Addendum
to
"The Health Consequences
of Sulfur Oxides: A Report
from CHESS, 1970-1971,"
Mav 1974
-------
This Page Intentionally Blank
-------
Contents
Part 1 : 3
The Environmental Protection Agency's Research Program with
Primary Emphasis on the Community Health and Environmental
Surveillanpe System (CHESS): An Investigative Report (the Brown
Committee Report), November 1976
The Environmental Research, Development, and Demonstration
Authorization Act of 1978 (ERDDAA) P^L 95-155,
November 1977.
Part 2 131
Research Outlook 1978 (Appendix 3). U.S. Environmental
Protection Agency, EPA-600/9-78-001', June 1978.
Research Outlook 1979 (Appendix E). U.S. Environmental
Protection Agency, EPA-600/9-79-005, February 1979.
PartS .....;.... 149
Report of the Health Effects Research Review Group (HERRG).
Science Advisory Board, U.S. Environmental Protection Agency,"'
February 1979.
-------
Introduction
This document is an addendum to the 1974 monograph published by the U.S.
Environmental Protection Agency, titled "Health Consequences of Sulfur Oxides: A
Report from CHESS, 197O1971" (EPA-650/1 -74-004). A key portion of the
addendum is a copy of a Congressional report which contained recommendations to
EPA regarding the 1974 monograph and the conduct of health research generally.
Also included are reports to Congress by both EPA and the Science Advisory Board on
the Agency's implementation of these recommendations. The Congressional report
and subsequent legislation which bears on the Agency's compliance with the report's
recommendations are described below.
In 1976, the Subcommittees on Oversight and on Environment and the Atmosphere of
the House Committee on Science and Technology released an evaluative report on
EPA air pollution research programs. The report, titled "The Environmental Protection
Agency's Research Program with Primary Emphasis on the Community Health and
Environmental Surveillance System (CHESS): An Investigative Report," is commonly
referred to as "the Brown Committee Report." The report sets forth in detail findings
and recommendations regarding epidemiologic investigations carried out by EPA into
the human health effects from air pollution., In particular, the report critically
addressed the 1974 monograph, which analyzed CHESS studies related to the health
consequences of sulfur oxides. Also addressed was the effectiveness of EPA's total
research and development program.
In 1977, Congress enacted the Environmental Research, Development, and
Demonstration Authorization Act of 1978 ("ERDDAA"), P.L. 95-155, November 8,
1977. Section 8(a) of ERDDAA provided a statutory basis for EPA's Science Advisory
Board ("SAB").* Section 8(c) directed that when the Administrator for EPA submits
his annual report to Congress on the Agency's five-year plan for research,
development, and demonstration, it be accompanied by the SAB's comments on the
plan. Section 8(d) further directed that the SAB independently report to Congress on
EPA's health effects research generally and on the findings and recommendations of
the Brown Committee Report in particular.
Section 10 of ERDDAA specifically required the Administrator of EPA to implement
the recommendations of the Brown Committee Report, unless he makes certain
determinations. Section 10 also requires the Administrator to report to Congress on
the status of implementation of the Brown Committee recommendations in each
annual report he makes on EPA's five-year plan for research and development.
Recommendation 3(c)of the Brown Committee Report was that EPA should publish an
addendum to the 1974 monograph, containing at a minimum chapters IV, V, and VI
and Appendix A of the Brown Report. This publication constitutes such an addendum
to-the monograph. As set forth below, the addendum contains various materials
which concern net only the 1974 monograph and the various CHESS studies, but also
EPA's research and development program generally.
*The SAB was originally established by EPA in 1974 in order to provide advice to the Administrator on
scientific and technical aspects of environmental regulation. By providing a statutory foundation.
Congress intended to enhance the status, scope and responsibilities of the SAB.
-------
Included as Part 1 of the addendum is the entire Brown Committee Report.
Part 2 contains appendices from EPA's Research Outlook reports for 1978 and 1979,
each of which include discussions of compliance with the Brown Committee
recommendations. The Research Outlook reports are EPA's annual reports to
Congress on the Agency's five-year plans for research and development. The
availability of these appendices as further information on the Sulfur Oxides
monograph was previously noticed in the Federal Register of March 7, 1979 (44FR
12490).
Finally, Part 3 contains a report made to Congress in February 1979, by the SAB's
Health Effects Research Review Group ("HERRG"). The HERRG report, as required by
Section 8(d) of ERDDAA, summarizes the nature of health effects research at EPA and
reports on steps taken by EPA to implement the recommendations of the Brown
Committee.
-------
Part 1
Congressional Report to EPA (1976) and subsequent
legislation which bears on the Agency's compliance
with the report's recommendations.
-------
This Page Intentionally Blank
-------
[COMMITTEE PEIHT]
THE EimRONMENTAL PROTECTION
AGENCY'S RESEARCH PROG-RAM
Wi'JL'JH PRIMARY EMPHASIS OJT THE
COMMUNITY HEALTH AND ENVIRONMENTAL
SURVEILLANCE SYSTEM (CHESS):
AN INVESTIGATIVE REPORT
REPORT :
PREPARED FOR THE !
SUBCOMMITTEE ON SPECIAL'STUDIES,
INVESTIGATIONS AND OVERSIGHT
AND THE i
SUBCOMMITTEE ON THE ENVIEONMENT
AND THE ATMOSPHERE
OS1 THE
COMMITTEE ON
SCIENCE AND TECHNOLOGY
U.S. HOUSE OF REPRESENTATIVES
NINETY-FOURTH CONGRESS
SECOND SESSION
Serial SS i
NOVEMBER 1976
Printed for the use of the Committee on Science and Technology
!
V.8. GOVERNMENT PRINTING OFFICE
77-S90 WASHINGTON : 1976 i
For eale by the Superintendent of Documents, H.S. Government Printing Office
Washington, B.C. 20402 - JPrlee $1.35
-------
COMMITTEE ON SCIENCE AND TECHNOLOGY
OLIN E. TEAGTJE, Texas, Chairman
KEN HECHLER, West Virginia CHARLES A. MOSHER, Ohio
THOMAS N. DOWNING, Virginia ALPHON20 BELL, California
EON FTJQTJA, Florida JOHN JAEMAN, Oklahoma
JAMES W. SYMINGTON, Missouri JOHN W.,WYDLER, Now York
WALTER FLOWERS, Alabama LARRY WINN, Ju., Eansas
ROBEET A. ROE, New Jersey LOUIS PREY, JR., Florida
MIKE McCORMACK, Washington BARRY Jr. GOLDWATER, Ja., California
GEORGE E. BROWN, JB., California MARVIN L. ESCH, Michigan
DALE MILFO ED, Texas JOHN B. CONLAN, Arizona
RAY THORNTON, Arkansas GARY A; MYERS, Pennsylvania
JAMES H. SCHEUER, New York DAVID F. EMERY, Maine
RICHARD L. OTTINGER, New York LARRY PRES3LER, South Dakota
HENRY A. WAXMAN, California
PHILIP H. HAYES, Indiana '
TOM HARKIN, Iowa
JIM LLOYD, California
JEROME A. AMBRO, New York
CHRISTOPHER J.'DODD, Connecticut
MICHAEL T. BLOtJIN, Iowa
TIM L. HALL, Illinois
ROBERT (BOB) KRTTEGER, Texas
MARILYN LLOYD, Tennessee
JAMES J. BLANCHARD, Michigan •:
TIMOTHY E. WIRTH, Colorado
SUBCOMMITTEE ON SPECIAL STUDIES, INVESTIGATIONS AND OVERSIGHT
OLIN E. TEAGTJE, Texas, 'Chairman
KEN HEOHLER, West Virginia CHARLES A. MOSHEE, Ohio
DON rUQTTA, Florida JOHN W.1 WYD'LER, Now York
JAMES W. SYMINGTON, Missouri
MIKE MoCORMACK, Washington i
SUBCOMMITTEE ON THE ENVEBONMENT AND THE ATMOSPHERE
GEORGE E. BROWN, JR., California, Cftsims-n
MIKE McCORMACK, Washington MARVIN' L. ESGH, Michigan
DALE MILFORD, Texas LARRY WINN, JR., Kansas
RICHARD L. OTTINGER, New York GARY A. MYERS, Pennsylvania
PHILIP H. HAYES, Indiana DAVID F. EMERY, Maine
JEEOME A. AMBRO,-New York
JAMES J. BLANCHARD, Michigan
JAMES H. SCHETIER, New York
fll)
-------
CONTENTS
Page
Letter of transmittal ; vn
I. Executive summary , ._• , 1
II. Introduction ~ 5
A. Background '. 5
B. History of air pollution health studies leading to CHESS
objectives i 7
III. Findings, conclusions, and recommendations 11
A. General findings 11
B. Specific CHESS findings and recommendations 12
C. R. & D. related finding^ 17
D. Recapitulation of guiding questions • 21
E. Closing remarks : 23
IV. CHESS aerometrio measurements _ 25
A. Introduction 25
B. Review of chemical and physical methods 27
1. The West-Gaeke method for measurement of ambient
S02 —_- . 27
2. Total suspended particulates : — 28
3. Suspended sulfate ', 28
4. Dustfall bucket, tape sampler, cascade impactor, and
cyclone sampler _ ; 29
C. Findings and evaluations of measurements and data reduc-
tion - 30
1. Sulfur dioxide : _ 31
a. Spillage of reagent, during'shipment 32
b. lime decay of the reagent SO3 complex 33
c. Concentration dependence of sampling
method .—! 34
d. Low flow correction _ ,— 34
e. Bubbler train leakage—: 34
2. General assessment of CHESS S02 data..., 35
3. Total suspendedpartioulate .. 37
4. Total suspended sulfate . 39
D. The CHAMP air monitoring program. -* 41
1. Introduction —. ; 41
2. System description _—; 42
3. Findings regarding the CHAMP program 43
4. Summary - ' ___, 44
V. Review of CHESS air quality analysis procedures and results 45
A. Introduction. , 1 45
B. Use of estimated data ' 45
1. Estimated sulfur dioxide concentration, 92 /tg/m*
(Magna) 46
2. Estimated sulfur dioxide concentration, 177 #g/ma
(Anaconda) __-_______-..__-_____-_„_ 47
3. Estimated suspended sulfate concentration, 15 #g/m3
(Magna) 49
4. Estimated suspended sulfate concentration, 7.2 jug/ms
(Anaconda) . ; 50
5. Estimates of suspended particulate, Salt Lake basin
study . 54
6. Estimates in the Chicago and New York studies-— 54
C. Use of mathematical dispersion models..: 54
(in)
-------
IV- •:
Page
VI. An analysis of the CHESS health effects studies 57
A. General problems of epidemiologic investigations of pollution
effects 1 . 57
B. Introduction and definitions . 58
C. Specific health effects studies ' 60
1. Chronic respiratory disease (CRD) prevalence 60
2, Retrospective • surveys for acute lower respiratory
disease (LRD) in children \ 68
3. Acute respiratory disease (ARD) in volunteer families 65
4. Asthma panel studies j 68
5. Cardiopulmonary symptoms in adult panels 71
6. Ventilatory function in schoolchildren 73
7. Other CHESS studies '. 75
D. Summary assessment of the population studies. , 75
E. CHESS current status \ 77
1. General i 77
2. Status of analysis , . 78
3. Recent planning directions of Population Studies
Division '„--. 80
VII. Current CHESS status and future programs— 81
A, Retrospection : 81
B, Current . , 82
C. Future ( 83
APPENDIX A—Recapitulation of the aerometric and meteorological
findings of the investigation as they relate to specific sections of the
CHESS Monograph , 85
A. Introduction '. 85
B. Critique •. : 85
1. Prevalence of chronic respiratory disease symptoms in adults:
1970 survey of Salt Lake Basin communities 85
2. Frequency of acute lower respiratory disease in children:
retrospective survey of Salt Lake Basin communities,
1967-70 ! 87
3. Aggravation of asthma by air pollutants: 1971 Salt Lake Basin
studies ' 87
4. Human 'exposure to air' pollutants in five Hooky Mountain
communities, 1940-70 ; 89
5. Prevalence of chronic respiratory disease symptoms in adults:
1970 survey of five Rocky Mountain,communities , . 90
6. Frequency of acute lower respiratory disease in children:
retrospective survey of five Rocky Mountain communities,
1967-70 90
7. Prevalence of chronic respiratory disease symptoms in military
recruits: Chicago induction center ' 92
8. Prospective surveys of acute respiratory disease in volunteer
families: Chicago nursery school study, 1969-70 93
9. Human exposure to air pollution in selected New York metro-
politan communities, 1944-71 _ 93
10. Prevalence of chronic respiratory disease symptoms in adults: .
1970 survey of New York communities 94
11. Prospective surveys of acute respiratory disease in volunteer
families: 1970-71 New York studies. 94
12. Aggravation of asthma by air poEutants: 1970-71 New York
studies ! 95
13. Frequency and severity of cardiopulmonary symptoms in
adult panels: 1970-71 New York studies 96
14. Ventilatory function in schoolchildren: 1970-71 New York
studies 97
15. Ventilatory function in schoolchildren: 1967-68 testing in
Cincinnati neighborhoods '. 98
C. Superficial and perfunctory treatment of meteorological informa-
tion : 99
D. Insufficient exploration of possible relationships between meteoro-
logical conditions and asthma attack rates 99
E. Failure to consider peak and episode concentrations 100
8
-------
PRJCC
P. Use of a single monitoring station to determine the exposure of a
community '. 100
G, Failure to establish, similarity of exposure and, stress factors be-
tween communities in the same study, excluding the exposure
to specific pollutants , 101
H, Impreciseness of monitoring station locations—, 101
I. Inexact locations of residences of individuals studied 101
APPENDIX B—Legislative history of the CHESS program 103
9
-------
-------
LETTER OF TRANSMITTAL
HOTTSE OF REPRESENTATIVES,
COMMITTEE ON SCIENCE AND TECHNOLOGY,
Washington, D.C., November 19,1976.
Hon. OLIN E. TEAGXTE, ;
Chairman, Committee on Science and Technology, U.S. House of
Representatives, Washington, D.C.
DEAR ME. CHAIRMAN: I am transmitting herewith a report entitled
"The Environmental Protection Agency's Research Program with
primary emphasis on the Community Health and Environmental
Surveillance System (CHESS): An Investigative Report." This report
provides the conclusions and reconunendations resulting from an
intensive evaluation of the scientific and technical adequacy of this
important air pollution-health effects surveillance system and related
research programs. The report indicates that there are a number of
serious issues in EPA's research programs which require resolution in
order to achieve more effective results.
The report was prepared by Committee staff members, Dr. R. B.
Dillaway, Science Consultant and Director, EPA CHESS Investi-
gation, -and Oversight Subcommittee on Special Studies, Investiga-
tions Task Team Leader, and Dr. Radford Byerly, Jr., Science Con-
sultant-, Subcommittee on the Environment and the Atmosphere, who
was detailed to assist in the investigation. Committee staff members
were assisted in the preparation of the report and in the conduct of
the oil-site investigation by a special team, of consultants. The Com-
mittee is grateful for the cooperation of the agencies which permitted
their participation and expresses appreciation to the consultants for
their assistance. The special consultants were: Mitchell H. Gail.
M.D., Medical Statistical Researcher, Biome'try Branch, National
Cancer Institute; Paul A. Humphrey, Senior Scientist (Retired),
Division of Meteorology, EPA/NOAA, National Oceanic and Atmos-
pheric Administration; Dr. James M. McCullough, Senior Specialist in
Science and Technology (Life Sciences), Science Policy Research
Division, Congressional Research Service, Library of Congress; Cdr.
John W. Poundstone, M.D., Chief, Occupational and Preventive
Medicine, Navy Regional Medical Center, Great Lakes; Dr. Harry L.
Rpojs, Chief, Activation Analysis Section, Analytical Chemistry
Division, National Bureau of Standards; and Charles C. VanValin,
Research Chemist, Atmospheric Physics and Chemistry Laboratory,
Environmental Research Laboratories, National. Oceanic and Atmos-
pheric Administration. In addition, the Committee sttiff solicited the
opinions and advice of a number of other scientists, statisticians,
and physicians in other agencies and institutes.
(VII)
11
-------
vni
In accordance with your direction, the staff in this investigation
reported^ to me. I have nothing but praise for the professional manner
with which they carried out the study. I have carefully reviewed this
report and am prepared to take full responsibility for its contents.
I further recommend its adoption as a Science and Technology
Committee report.
In closing, however, I would like to make an additional observation
that goes beyond this particular study. The need to regulate pollutants
in this industralized age is likely to grow,' not diminish. Regulating
knowledgeably is therefore essential. This report provides several
examples of the many instances in which the status of environmental
research and technology is not what it should be to assure adequate
and well founded protection of human health. While this investigative
report has much significance with regard to research management and
execution, it does not provide all the answers to the policj-maker who
is obliged to act in the face of less than certain proof. The policy
oriented questions of "acceptable risk" and "burden of proof" can be
assisted, but never fully answered by better research. I^vould hope that
those who read this report will not use it to undermine the minimal
research work that is now underway, but instead will use it as a reason
to improve and enhance the on-going programs.
Sincerely,
GEOEGB E. BEOWN, Jr., Chairman,
Subcommittee on the Environment and the Atmosphere,
12
-------
I. EXECUTIVE SUMMARY
The Staffs of the Subcommittee on Special Studies, Investigations
and Oversight and the Subcommittee on the Environment and the
Atmosphere have completed an evaluation of the Environmental
Protection Agency's research program related to the health effects
of air pollution, with primary emphasis on the OQTn.Tm.Tni.ty Health
and Environmental Surveillance System (CHESS). This investigation
was noted earlier in the report on joint hearings held by the House
Interstate and Foreign Commerce Committee and the House Science
and Technology Committee, April 9, 1976.1 While the joint hearings,
held to determine whether there had been any deliberate distortion
of data during the writing of the 1974 CHESS Monograph,2 produced
general agreement that no basis existed to question the integrity or
scientific honesty of ^he project leader for tie first CHESS Mono-
graph, the hearings did not resolvemany questions and supported the
need for a comprehensive evaluation of the scientific'and technical
aspects of the CHESS program and related Environmental Protection
Agency (EPA) research efforts.
The investigation reported herein was conducted by a team com-
posed of members of the staff of the House Science and Technology
Committee with the support of consulting specialists in aerometry
and analytical chemistry, meteorology, epidemiology, medicine, and
general environmental health. Conferences were held on site, in the
field, with individuals in EPA and in State and local government,
academic, and private institutions who had detailed knowledge of the
conduct of the CHESS programs. A number of findings and recom-
mendations are provided in the report including those highlighted
in the balance of this summary.
_ To some the findings may appear overly [Critical since CHESS
pioneered in a difficult research area. However, in part the report is
meant retroactively to complement the CHESS Monograph which
fails to fully appraise the reader of the errors, .deficiencies, and other
shortcomings in the information presented in, the Monograph. The
intent of the report is to aid in assessing the validity of the conclu-
sions presented in the Monograph and to assist researchers perform-
ing similar studies in avoiding similar problems. The endeavor was
•greatly assisted by hindsight and by^ the splendid cooperation and self-
examination by investigators both inside and outside of the Environ-
mental Protection Agency. The report also notes where problems
similar to those found in CHESS research still exist in the EPA
program.
i U.S. Congress. House Interstate and Foreign Commerce and House Science and Technology Committee.
94th Congress, 2nd Session. Report on Joint Hearings on the Conduct of the Environmental Protection
Agency's "Community Health and Environmental Surveillance System" (CHESS) Studies. Washington,
U.S. QPO April 9,1976. !
1 TJ.S. Environmental Protection Agency. Health Consequences of Sulfur Oxides: A Report from CHESS,
1970-1971. May 1974. EPA-650/l-74r004.
(1)
13
-------
The report identifies the need to reexamine the wisdom of legislative
and executive pressiires for rapid development of information for
pollution control standards, and the adverse impact of these require-
ments on the performance of research. Also identified were the prob-
lems which are occurring because of the Office of Management and
Budget (OMB) assuming operational control over details in the EPA
research programs.
The CHESS program, has historical value as a first attempt at a
broad based definitive study relating air quality to health effects in a
precise, quantitative manner. The program is a contribution to the
general field of air pollution epidemiological studies. However, many
problems critical to obtaining useful quantitive results were not
solved in the conduct of this research. This failure to solve procedural
problems became obvious when the studies were examined. There
were too many inconsistencies in the data, and too manj- technical
problems that resulted in large data uncertainties or errors associated
with the aerometrics for the results of this program to provide quanti-
tative support for policy decisions. _The 25 pg'™? lower sensitivity
limit of the method used for most SOa measurements coupled with
the overall large error band on all measurements (possibly exceeding
100%) and the apparent bias of SO2 concentration di
-------
Second, CHESS points out the many difficult problems involved in.
carrying out air pollution/health, effects research.
Of course, the first lesson had been learned long ago, but failures in
CHESS pi arming and execution precluded the development of new
quantitative information on the relationship of air quality to health.
Thus, CHESS cannot be called a major advance, only a confirmation
of previous advances. With respect to the second lesson, the value of
the CHESS experience lies mainly in the use'to be made of it in
correcting and improving operational procedures and research tools
(i.e., instrumentation, questionnaires) and in planning future health
effects studies. Finally, it seems that both these lessons could have
been, learned at much less cost in funds, elapsed time, EPA credibility,
and staff morale. ' .
EPA has only recently been reorganized to^modifj7- the National
Environmental Research Center concept. This reorganization was
examined in more detail in a previous report.3
The results of this present investigation suggest that the changes
made last year are less effective than anticipated. The issue of re-
search management should be continuously monitored to build the
Environmental Research Center, Research Triangle Park capability
and reputation. At the end of the 1977 fiscal year the situation should
be examined to determine further need for reorganization. Perhaps
consideration should be given now to some structure which provides
for a single point of review at both the Research Triangle Park and
Headquarters in order to alleviate the obvious lack of coordination of
multidisciplinary projects, without attempting a major reorganization.
There is considerable 'evidence from this investigation that some
critical disciplines, such as quality control, monitoring and sampling,
and statistical analysis continue to be inadequately addressed in pre-
project planning of health effects studies.
In the area of research management, the investigative team's
recommendations are critical of the control of research programs
within EPA. There is also concern with the fact that research is oeing
conducted by other EPA offices involving little coordination with the
Office of Research and Development. This functional overlap is
beginning to produce problems within the laboratories and needs to
be effectively resolved before these programs continue much further.
Related to this overlap and the need for the solution of the problem
of poor communication between the offices and laboratories, are the
needs to provide for a more effective input from the Science Advisory-
Board, to establish a permanent peer review system to insure coordina-
tion among laboratories, and to expedite the tamely publication of
research results.
Although the Administrator of EPA had :indicated during the
joint hearings that there was a five year plan for research on sulfates,
the investigative team found that only draft plans of this program
were available and that the development of this plan was being poorly
coordinated. Accordingly, no specific recommendations with regard
to this plan can be provided. However, the need for better information
on the species of chemicals, including sulf ates,'which are of primary
8 U.S. Congress, House Committee on Science and Technology. Organization and Management of EPA's
Office of Research and Development, 94th Congress, 2nd Session, June 1976. U.S. GPO. Washington, D.C.
Serial LL 40 pp.
15
-------
importance in health effects is emphasized and this recommendation
3ias direct relevance to the proposed sulfate program.
A number of other recommendations are imade with regard to the
•need for improvements in the air monitoring and quality control
programs, the need for additional strength in meteorology, and a
number of specific suggestions about improvements in the, type of
epidemiological studies reported in the CHESS document.
It should be noted that the investigative team received friendly
cooperation throughout the entire investigation and appreciation for
this help is acknowledged. A number 01 the recommendations for
change evolved from the suggestions and the very real concern of
EPA personnel for the improvement of their research program. The
report bears on the condition of a research a^rea which is the result of
many forces and restraints. There is no intent to impugn the integrity
or ability of any individual researcher, all of whom were found to
exhibit professionalism and dedication to their program.
Finally, a most important point must be made which grows from
but goes beyond the study. This report documents many cases of
deficiencies in EPA research both past and present. It does not discuss
many of the fine research efforts ongoing. The overriding purpose of
this report then is to point the way to improving and enhancing the
EPA research program, not to undermine the program. EPA can
regulate wisely only if its decisions are based'on information generated
in a sound and comprehensive research program.
16
-------
II. INTRODUCTION
A. BACKGBOTJND
The Environmental Protection Agency, in partial response to the
mandate of Congress under the Clean Air Act, conducts research to
study the adverse health effects of air pollution.1 A series of major air
pollution/health effects studies were carried out between 1967 and
1975 and were considered by many experts to be the most compre-
hensive of their kind. This program attempted to implement a co-
ordinated system of aerometric measurements, including measurements
of sulfur dioxide (SOz), particulates, total sulfate, carbon monoxide
(CO), ozone, hydrocarbons, nitrogen dioxide (NOg), and nitrate.
In areas where these measurements were made, the EPA initiated con-
current epidemiological studies embracing acute and chronic respira-
tory illness, ventilatory function of children, and the aggravation of
asthma symptoms. A portion of these studies eventually was given
the formal title of "Community Health and Environmental Sur-
veillance System" (CHESS).
Although data were gathered over the entire 1967-1975 period, the
monograph, Health Consequences of Sulfur Oxides, A Report from
CHESS, 1970-71, (EPA 650/1-74-5)04, May 1974), treats only some
of these data. Some controversy exists in the professional community
regarding the technical correctness of the results and conclusions as
presented in that monograph. The controversy was intensified follow-
ing a series of newspaper articles starting with' a February 29, 1976
article in the Los Angeles Times. In this article, scientists both within
and outside of EPA were reported to have alleged that the data and
its analysis were manipulated to more strongly support certain regula-
tory positions taken by EPA.
Following these allegations, the House Science and Technology
Committee and the House Interstate and Foreign Commerce Com-
mittee on April 9,1976 held a joint hearing on the alleged mishandling
of the CHESS program.1
While the hearing generally negated the idea of a planned manipu-
lation of the data, many questions were raised as to the reliability of
the d.ata, the technical soundness of its analysis, and the subsequent
-validity of the conclusions reached.
Further, the EPA Adminfetrator, Russell Train, testified that there
existed a five-year program plan for sulfate air pollution and health
effects R&D which would assure a sound data base for promulgating
air quality standards in the future. However, the plan itself was not
presented.
As a result of the questions raised by the April 9 hearings and other
sources, the staff of the House Science and Technology Oversight
1 U.S. Congress. House. House Interstate and Foreign Commerce Committee and the House Science and
Technology Committee. 94th Congress. 2d session. Report on Joint Hearings on the Conduct of the En-
vironmental Protection Agency's "Community Health and Environmental Surveillance System"
(CHESS) Studies. Washington. U.S. Govt. Print. Off. April 9,1976.25 pp.
17
-------
6 ;
•Subcommittee and the Subcommittee on the Environment and the
Atmosphere undertook a technical investigation of the entire CHESS
program, with particular emphasis on the work reported in the 1974
CHE5SS Monograph. The investigation also looked into the follow-on
CHAMP aerometric program and EPA. plans for research to establish
a sound technical basis for air pollutant health effects relationships.
As noted in the Report on Joint Hearings cited earlier:
". . . the Committee on Science and Technology is currently in-
vestigating EPA's research program. This technical investigation will
focus on CHESS and EPA programs planned to develop information
necessary to make a decision on a sulfate standard." 2
This investigation was launched dne to the importance of EPA
research to the Nation, not because of accusations raised in news-
. paper articles. It is simply that recognition of the importance of
environmental research as a basis for regulation has and will continue
to foster reviews of the EPA R&D program. We would like to em-
phasize that the team conducting this review worked with the goal of
helping to improve EPA'-s research for the benefit of the country.
In planning the investigation, the team formulated the following
13 questions as a means of setting the scope and direction to the in-
vestigation. The team did not expect necessarily to arrive at findings
in a format that would explicitly answer the ,13 questions in the form
they were set .down. The questions are as follows:
(1) Did the CHESS measurement system; utilize the best available
instrumentation and sound operational protocols and quality control
procedures?
(2) Were the precision and accuracy of the CHESS aerometric data
sufficient to quantify pollutant exposure differences between commu-
nities where health effects were studied? ;
(3) Were the aerometric data consistent with and accurately repre-
sentative of the exposures to pollutants received by the populations
used in the health effects studies?
(4) Were new knowledge, instruments, and procedures introduced
into the program as they became available and proven?
(5) Were the epidemiological populations adequately selected?
(6) Were the health measurement endpoints meaningful and reliable?
(7) Were the data reduction and analysis- accurately and correctly
carried out?
(8) What factors have retarded the analysis and reporting of the
CHESS data? .
(9) Did the health and monitoring data establish unambiguously
what concentrations of specific pollutants are associated with measure-
able adverse health effect? _ :
(10) If so, do the quantitative dose-response estimates in the
CHESS report have a firm empirical basis? : •
(11) Were the conclusions drawn from the CHESS program suffi-
ciently clear and unambiguous so as to form a sound basis for future
action?
(12) Does the new "CHAMP" 3 program provide a substantially
improved mechanism to provide aerometric data to support future
> U.S. Congress, House Interstate and Foreign Commerce Committee and House Science and Technology
•Committee. Report on Joint Hearings, op. cit. p. 25. .
3 Community Health Air Monitoring Program, a systematic program for monitoring air pollution in con*
nection with health effects studies.
18
-------
epidemiological research and is an epidemiological research program
planned to complement future aerometric moniiorings?
(13) What additional steps, if any, should EPA take to insure that
a sound technical basis will be available for future agency decisions
pertaining to SOa/sulf ates?
Answers to these 13 questions were developed from the specific
findings'and are given at the end of Chapter; III, Findings, Con-
clusions and Recommendations. :
The investigative team consisted of at least one person with tech-
nical knowledge in each of the scientific disciplines expressly involved
in CHESS, i.e., epidemiology, aerometric measurement, meteorology,
data analysis, and B&D planning and management.
The first phase of the investigation involved visiting field data
centers and measurement stations and interviewing the ^personnel
who ran monitoring systems and who collected health data. The
next phase involved a review of data and analysis procedures used in
the CHESS monograph and. intensive' interviews with the key scien-
tific and mana'gement'personiiel at the EPA laboratories at Research
Triangle Park, North Carolina. The third phase consisted of interviews
with EPA headcmarters personnel who had been involved in the
CHESS and CHAMP programs.
This report documents> the investigation which was focused, on
the CHESS program and its published output. It was no.t, however,
restricted to the CHESS program. The investigators looked into
related programs (e.g. current epidemiology projects), and into sup-
porting programs (e.g. development of measurement methods). As a
result, the scope of the report is broader than CHESS. Since the
CHESS program ended a year ago this report will have most value if
it is pertinent to existing and future EPA research. It is definitely
believed that many of the CHESS findings and recommendations are
indeed relevant to current research programs in EPA, and that the
CHESS mistakes might well be repeated if they, are not documented.
An Executive Summary is provided at the beginning of this Report
for the convenience of the Committee in quickly getting an overall
picture of the findings. The body of the report presents summarized
foldings and recommendations (Chapter III), followed by a detailed
discussion and critique of the aerometric measurement and epidemio-
logical parts of the CHESS program. ;
The field review naturally divided into air quality measurements and
epidemiological studies. The first part of the field review (Chapter IV)
presents the results of the evaluation of aerometric measurements,
including the next generation CHAMP program. Chapter V presents
an overview of the air Duality analysis procedures, Chapter VI pre-
sents an analysis of the health effects studies. Appendix A presents a
recapitulation of the aerometric and meteorological findings of the
investigation as they relate to specific health studies. Appendix B
presents a legislative history relative to the CHESS program.
B. HISTORY OF Am POLLUTION HEALTH EFFECTS STUDIES LEADING
TO CHESS OBJECTIVES
There has never been any question about the need for control of
severe air pollution. Undisputed historical data have clearly implicated
high levels of industrial pollutants, which invariably include oxides of
19
-------
8
I
sulfur and particulate matter, with increased incidences of health
effects—both moribidity and mortality. Those historical reports
clearly indicated the danger from such incidents and pointed the way
toward a need for additional information on the nature and extent of
the damage which such pollutants might produce. _ ;
In early December, 1930, a tliiek fog covered the industrial Meuse
Valley in Belgium. Frrket4 recounted that several hundreds were
sickened by suddenly appearing respiratory- symptoms, complicated
in a large number by gra,ve circulatory failure." lie notes that, "More
than sixty died on the 4th and 5th of December after only a few hours'
of sickness. A sizeable number of livestock;had to be slaughtered."
Having calculated that mortality rates were ten and one lialf times
average during this period, Firket commented that, should a similar
fog afflict London, "one would have 3,179 immediate deaths to
mourn." His words were prophetic. Four thousand excess deaths were
attributed to the foxtr day London Fog of December, 1952, and Logan5
added the following historical note to his mortality figures:
"The incident was a catastrophe of the first magnitude in which, for
a few days, death rates attained a level that has been exceeded only
rarely during the past hundred years—for example, at the height of the
cholera epidemic of 1854 and of the influenza epidemic of 1918-19."
Nor was the United States spared. Nearly half (42.7%) the inhabitants
of Donora,. Pennsylvania experienced symptoms during the "smog
episode" of October, 1948, and twenty died dining the final week of
October, a period which norm ally occasions but two or three deaths.6
A statistically significant increase in mortality was also documented
during a "pollution incident" in New York City in November,
1953/Thus, "air pollution," or "smog," or Yfog" was recognized as a
lethal menace, both in the United States and abroad.
These tragic episodes raised many questions and spawned a genera-
tion of research. Although various products £f sulfur combustion were
considered likely noxious agents in these episodes, the commentators
were duly cautious in suggesting which pollutant or combination of
pollutants might prove harmful. Recent reviews 8 910 describe popula-
tion studies, clinical experiments, and toxicological studies designed to
define which pollutant (s) pose risk, and a concommitant bod3r of
research sought to develop and refine techniques of pollution
measurement. ^
The above chronicled incidents and research logically led to the
question: If high levels of pollution can kill,;what can low levels do to
susceptible subpopulations, such as asthmatics, and what harm might
come to the general population if exposed tojlow levels of pollution for
< Fkket, M., Bulletin it L'Academic Regale de Medicine de Belgigue Series 5, Volume II, Number 10 (1931),
pp. 683-739.
B Logan, W.P.D., "Mortality in the London Fog Incident." Lancet Volume 264, Number 6755 (1953),
pp. 336-338.
• Shrank, H. H., Heimann, H., Clayton, O. D., Gafafer, W. M., Wexler, H., "Air Pollution in Donora,
Pennsylvania: Epidemiology of the Unusual Smog Episode of October 1848". Public Health Bullletin
Number 306 (1949). '
' Greenberg, L., Jaeongs. M. B., Drolette, B. M., Field,!'., Braverman, M. M., "Report olan Air Pollu-
tion Incident in New York City, November 1953." Public Health. Rsports Volume 77, Number 1 (1962),
pp. 7-16.
«RaU, D.R.. "Review of the Health ESeets of Sulfur Oxides." Environmental Health Perspectives
Volume 8 (1974), pp. 97-121. • :
«"Air Quality and Stationary Source Emission Control." National Academy of Sciences Report for the
Committee of Public Works, United States Senate Pursuant to S. Res. 133 Serial #94-4 (March 1975), pp.
58-189. ' f
« "A Critical Evaluation of Current Research Regarding Health Criteria for Sulfur Oxides." Technical
Report prepared for Federal Energy Administration By Tabershaw/Cooper Associates, Inc. (April 11,1973.)
20
-------
9
years? To answer such questions, epidemiologists began to measure sub-
lethal effects of pollution such as eye irritation, susceptibility to
chest colds, subtle changes in pulmonary function, and asthma attack
frequency (see references ll and lz.) The Community Health and
Environmental Surveillance System (CHESS) represented a natural
extension of efforts to gather concurrent morbidity and air pollution
data. In 1970, a description of CHESS1S stated: ;
The keystone of the CHESS program is the coupling of sensitive health indi-
cators to comprehensive environmental monitoring . . .; ,
. . . EPA health research needs are practical and problem oriented. CHESS
research is thus pragmatic and our goals are threefold: (1) to evaluate existing
environmental standards: (2) to quantitate pollutant burdens in exposed popula-
tions; and (3) to quantitate health benefits of pollutant cdntrol.
It was perhaps this last goal which led to title rapid expansion of
the CHESS system in 1968-70 and which necessitated the use of health
measurements and monitoring methods which in some instances had
not been adequately field-tested. Municipal regulation of industrial
and power combustion, space heating, and incineration resulted in
dramatic decreases in sulfur dioxide and total suspended particulates
in New York City, Chicago, and elsewhere,,14 These favorable pollution
trends were already evident in 1967. Thus, by 1970, it was urgent that
population studies begin immediately if corresponding improvements
in health indices were to be documented.. CHESS data was gathered
in New York City, Salt Lake City, Birmingham, Charlotte, and in
Chattanooga from 1970 through 1975 and Los Angeles between 1972
and 1975. The analysis of the data and compiling of conclusions was
begun aggressively as data became available. The facts relating to this
program history are discussed in the following chapters.
" Nelson, D.J., Shy, O.M., English, T., Sharp, C.R., Andleman, R., Truppl, L., Van Brugeen, J.,
"Family Surveys of Irritation Symptoms During Acute Air Pollution Exposures." Journal of the Air
Pollution Control Association Volume 23, Number'2 (1978), pp. 81-80.
" U.S. Environmental Protection Agency. Health Cousoq.-uen.ces of Sulfur Oxides: A Report From
OHESS.1970-1971. May 1874. EPA-650/1-7-HXH.
" Biggan, W. B., Hammer, D. L. Finklea, J. P., Hasselblacl.V. Sharp, C. R., Burton, E. M., Shy,
C. M., "CHESS, a Community Health and Environmental Surveillance System." Proa, of the Sixth
Berkeley Symposium on. Mathematical Statistics and Probabilitjr (1970).
u TJ.S. EPA. CHESS, op. ctt. ;
77-690—76 2
21
-------
-------
III. FINDINGS, CONCLUSIONS, AND KECOMMENDATIONS
A. GENBBAL FINDINGS \
The findings resulting from the CHESS program review have
extended into EPA's evolving task of developing a sulfate standard
as well as a general perspective on EPA's entire air pollution-health
effects research program.
The EPA five year plan for sulfates research cannot he commented
on at this time because the plan is still in the formative stage. Thus,
the recommendations deal primarily with what it is believed can be
accomplished to prevent repetition of the problems produced by or
related to the CHESS program.
The EPA CHESS program intended, as one of its prime goals,
to provide a sound basis for relating hximan health effects to precise
quantities of SO2 and other oxides of sulfur in air and.to provide a
reliable record of the improvement of the health of the U.S. popula-
tion as air pollution was controlled and reduced. The program, span-
ning 6 to 7 years, was to achieve these objectives by measuring and
analyzing 6 health indicators together with SO2, NO2, suspended
sulfates, total particulates, and oxidants in the air hi 6 representative
areas throughout the country. Only the 1969-1971 data were analyzed,
correlated and published by 1974 in the CHESS Monograph. Most of
the remainder of the data are in various states of analysis.
A critical review of the epidemiological and air monitoring programs
with conclusions for each subprogram is presented in the body of the
report. The results present a picture of a program pressured by EPA
management-imposed time constraints to meet legislated mandates
for promulgating new standards, hampered by inadequate mechanisms
to detect and correct technical problems, and handicapped by budg-
etary and management restrictions placed on the program after it
was well underway. • ;
In the drive for results, the program did not adhere to standards of
.quality control, validation of methods, cross-checking of data, ^nd
calibration of instruments required in such research. The decision
not to submit all results of on-going research for publication in pro-
fessional journals deprived the CHESS program of an important
source of peer review and evaluation at various stages, although the
Monograph was sent to many reviewers for comments. The simulta-
neous'publication of a large number of complex studies in a single
report—the 1974 CHESS Monograph—rather than as individual on-
going research reports may have been an administratively expedient
method of publishing and collating results, but it undoubtedly reduced
the effectiveness of peer review. The complexity of the document also
impeded the public from acquiring^ an understanding of the results
of the studies associated with policies of national importance.
(ID
23
-------
12 I
As noted in Chapters II and V, the investigative team is aware that
historic incidents and other clinical, toxicological, and specific epide-
miological studies support the general concept that health and welfare
will bo improved as air pollution is reduced. However, technical errors
in measurement, unresolved problems in statistical analysis, and
inconsistency in data in the 1974 CHESS Monograph renderlt useless
for determining what precise levels of specific pollutants represent a
health hazard. Nonetheless, the CHESS Monograph presents some
evidence of adverse health effects of air pollution, and, in particular,
slightly increase^ prevalence of chronic respiratory disease is found
in association with whatever levels of SO2 and.indeterminate particu-
late pollution were present in the air during the period the CHESS
health panels were active. CHESS also has historical merit for under-
taking the first massive air pollution health effects research program
in the Western World, It is important to record and remember the
lessons learned for future undertakings.
As pointed out in Chapter VII, the findings regarding CHESS
epidemiologies! work corroborate the issues raised in previous reviews.
This report collects the issues and goes .into depth as to effects of
errors, etc., and suggested corrective action. !
In addition, this review documents serious errors and omissions in
the aerometric data and analysis meteorology. Some of the more
important specific findings and related recommendations related to
the CHESS program are presented below, followed by ETP research
lab operational issues, by suggestions as to po'ssible improvements in
health effects programs, and finally, by a review of the 13 questions
set forth at the beginning of the investigation.
It should be noted that these findings do not contradict the con-
clusions in the report of the April 9 joint hearings.
B. SPECIFIC FINDINGS AND EECOMMENDATIONS
1. The data and analysisjn the_1974 CHESS Monograph do not
provide a reliable, quantitative basis for relating SO2 or sulfate levels
in the air to adverse health effects. j
The methods and instruments used had associated measurement
errors (inaccuracies) such that the threshold curves ("hockey-stick
curves") presented in the Monograph are withoxit basis in fact. The
indicated thresholds occurred at levels at which the measurement
methods (as used) were not reliable. Specifically in the case of SO2
measurements the method iised^had a threshold of 25/i/m3 and the
accumulative errors found in hindsight in handling, shipping, and
reading the instruments made SO2 reading between 50 and 200/i/m3
unuseable. There were similar problems with, other measurements as
pointed out in Chapters IV" ana VI.
2. The CHESS program did appear to demonstrate adverse health
effects in association with local air pollution of undefined level and
pollutant mix. Studies of chronic respiratory disease prevalence rates
Erovided the most consistent results. These results are not trans-
table to communities other than those measured (an objective of
the research) because the error uncertainties in the aerometric data
made it impossible to arrive at reliable quantitative relationships.
3. The CHESS monograph has been referred to either directly or
indirectly as a document supportive of regulatory and standards ac-
24
-------
13
tions. Specific examples were found such as its use to support EPA
positions on air pollution health effects in the following excerpt from
the Administrator's Eeport to Congress for 1974:
The CHESS studies have demonstrated [emphasis added] the benefits from im-
proved air quality with respect to the chronic respiratory disease experience of
subjects who have moved to communities having cleaner air. Also, the studies
have shown [emphasis added] that children living for '3 or more years in com-
munities having high level of air pollution have more! acute respiratory disease
episodes than recent immigrants to the community. . ... Data obtained from the
CHESS program indicate that adverse health effects are consistently [emphasis
added] associated with exposure to suspended sulfates, indeed, more so than to
SOs or total suspended particulates. This information has initiated further study
in the transport processes and control techniques for suspended swlfates.1
The CHESS study results were cited in an EPA document2 dated
May 1976. The preface of that document states "The report's objective
is to review results of research by and under the sponsorship of
ERC/BTP, related to criteria and hazardous pollutants since publica-
tion of the criteria and background documents." CHESS studies are
explicitly mentioned in the report (e.g., page 207, page 276) with no
references. Pages 303 through 308 seem to describe the CHESS studies
as reported in the CHESS Monograph, although no reference to a
document is given. Curiously, reference numbers 244 through 252 are
not found in the body of the text where they would be expected (be-
tween pages 301 and .308). Inspection of the references list at the end
of the pertinent chapter shows that the "missing" references, numbers
244 through 252, refer to the CHESS Monograph. Thus, it appears that
CHESS results are used without attribution, arid thus without caveat.
Recommendations:
(a) EPA should immediately publish and distribute a formal
announcement to the effect that the 1974 iCHESS Monograph is
at best a'preliminary document reporting research of varying
degrees of reliability and as such should be considered only as a
beginning study of larger problems. The notice should further
point out: (1) That many of the problems leading to the unre-
liability of the research are not reported in the Monograph; (2)
That no interpretations should be made based only on the data
in the Monograph; And (3) that other data available at EPA,
including data taken since the publication of the Monograph,
indicate the need for reexamination of the conclusions in the
Monograph. The notice should also announce the availability of
and describe how to acquire the addendum recommended below.
(6) EPA should not utilize the 1974 CHESS Monograph as a
source for specific quantitative data supportive of standards or
regulatory decisions without explicit qualification.
(c) EPA should publish an addendum to the CHESS Mono-
graph containing at least the materials in Chapters IV, V, and VI
and Appendix A of this report. EPA should take the initiative
to make _ sure this addendum is available to the public and
included in all future issues of the Monograph.,
4. If the EPA had drawn adequate attention to the many difficulties
being encountered in the CHESS program (which was under pressure
to provide data as fast as possible to support air quality standards
1 XJ.S. Environmental Protection Agency. Eeport to Congress. Progress in the Prevention and Control of
Air Pollution in 1974 (undated) p. 39.
* Scientific and Technical Data BBSS for Criteria and Hazardous Pollutants 1975 EEC/BTP Review.
Health Effects Eesearch Laboratory, U.S. E.P.A. (EPA-600/1-76-023) May 1976.
25
-------
14
under conditions set by the Congress, and on which OMB subsequently
placed restrictions as to the allocation of positions and funds) it- is
possible that the program could have been either terminated with
considerable savings or restricted in scope to permit a more satisfactory
research program to be completed. '
Recommendations:
(a) Legislation affecting the research responsibilities of the
EPA should be reexamined to insure that unrealistic procedures
and schedules are'not assigned.
(6) Research programs should be designed to gain information
and answer questions, not to support positions.
(c) If public policy requires expeditious research, every effort
should be made to insure that the OMB allows all necessary
funding and personnel. !
(d) In the event that budgetary restrictions evolve which will
have^an impact on the completion of; major projects, the EPA
Administrator should be required, to advise the Congress of the
potential for such impact. Consideration should be given to
requiring the Administrator to provide an annual analysis to the
Congress which estimates the probable impact of proposed budget
changes on major projects such as the CHESS program.
5. At the present time, OMB has the power to control the most
detailed aspects in a research program. For example, a delay in the
approval by OMB of health effects questionnaires submitted by
EPA prevents or significantly delays the continuation of epide-
miological studies associated with the establishment of air quality
standards. Thus EPA has the responsibility for its mission, but lacks
the necessary authority to carry it out. ;
Recommendation:
The OMB should be asked to develop procedures by which
questionnaires and similar documents .required for research pro-
grams can be given prompt review and approved without restric-
tive time limits for research programs of like nature.
6. The Health Effects Research Laboratory at Research Triangle
Park is currently proceeding with the processing of the remaining data
from the CHESS program. If adequate resources are made available
to complete both current and past data processing programs, then the
CHESS data may provide some additional insight into pollution health
relationships or at least identify pitfalls to avoid in future studies. The
loss of some key personnel involved in the CHESS program makes un-
available important insights provided by individuals who were involved
in gathering of the data. i
Recommendations: ",
(a) In view of the current unsettled conditions at the Health
Effects Laboratory and the severe restrictions on manpower and
budget, it is recommended that the unanalyzed epidemiological
and aeronietric data be examined critically by year (to determine
at what point improvements in measurement make the data
acceptable) and that analysis be carried out only on that data
that appears to have a higher degree of validity than the '69-71
data used as a basis of the 1974 CHESS Monograph.
(&) Steps should be taken to attempt to publish research in
traditional, referred, archival journals in a timely fashion.
26
-------
15
(c) The results of large or complex projects should nob be pub-
lished solely in. monograph form.. If EPA decides to publish such
results in monographs as well as journals, then EPA should take
care that, within the limits of reason, all pertinent data are
included.
($) Projects should not be initiated unless there is full assurance
that the data collected in such projects can be properly analyzed
and reports prepared in a realistic time frame for policy con-
sideration. If policy considerations dictate a time frame, then
research staff must have a part in deciding what can realistically
be accomplished in the allotted time.
7. While CHAMP looks more promising for providing reliable real
time measurements of air quality (absolute pollutant levels), the sys-
tem reliability and precision for SO2, sulfates, and N02 have yet to be
proven.
Recommendations: ;
(a) The aerometric and quality control programs in EPA should
be further, strengthened and improved.
(6) Every effort should be made to shorten the time between
the acquisition of data and quality assurance analysis of data so
that errors in instrumentation and collection may be corrected.
(c) The practice of employing development stage instruments
in field operations before qualification testing is satisfactorily
completed should be stopped. ;
(d) Field utilization of laboratory models of instruments is
not acceptable until the instrument has been thoroughly checked
out in the field and until all field personnel have been adequately
trained in its use.
(e) The opening of the CHAMP operation contract to competi-
tion now, before development is complete should be reexamined
by EPA policy management to see whether the merits of open
bidding at this time outweigh the problems of instability.
(/) Health effects research personnel need to coordinate closely
with air quality and monitornig personnel so that there is early
understanding as to the chemical species to be monitored as well
as the accuracy needed for support of health effects research.
8. Meteorological support to the epidemiologieal program has
been passive and not fully developed. >
Recommendations: :
Additional meteorological support is needed in the health
research-air pollution effects studies. Assuming that CHAMP
stations will continue in operation, the meteorological instrumenta-
tion should be uniform and complete for all stations.
9. The characterization of species of air pollutants and combinations
thereof which are of particular health or welfare significance needs to be
speeded up. The useful output from future large scale epidemiologieal
studies of general air pollution effects will be minimized if such char-
acterization is not available. Such characterization will require a
greater effort than is now possible with currently assigned resources.
The present Federal effort in this area is fragmented and unco-
ordinated.
27
-------
16
Recommendation:
The EPA Health Effects program as:well as interagency utiliza-
tion of all available Federal and extramural resources in the
health effects area should be examined with the objective of
significantly accelerating .research in this area.
10. The overall impression left with the review group was a general
awareness of many of the problems we found in the air quality health
effects research area but an apathy to really drive to correct them
quickly.
Recommendations:
(a) A. truly interdisciplinary task force led by an eminent
scientist should draw up a program plan for EPA to develop
a solid base of knowledge and procedures in aerometric instru-
mentation and measurements, meteorology, field data gathering,
quality control, epidemiology project design and testing and
panel planning. ;
The plan should be thoroughly reviewed with peer groups and
tnen funded for two years or as required at the expense of other
programs so that effective tools for meeting research needs will
be available for future programs.
(&) At (the very least, CHAMP should be pushed to verify
present instruments and protocols so that reliable data at
pollutant concentrations within instrument capabilities can be
achieved. New third generation instruments, such as the pulsed
fluorescence devices should be developed and proven for field
usage so that reliable values with uncertainties of less than ±20%
can be detected for pollutants at concentrations well below
current standards.
(c) The epidemiological questionnaires, panel selection criteria
and other items criticized in this report should be worked out and
approved by peer groups so that EPA will be prepared for the next
round of serial or episodes investigations with reliable tools and
plans. Specifically, questions such as the following must be
resolved for future epidemiological studies:
(1) How do CRD* questionnaire responses change on
serial administration in an area with unchanging pollution
patterns?
(2) What is the sensitivity of the self-administered CRD
questionnaire compared with its use in an interview?
(3) What is the nature of the statistical dependence of
ARD-^- attack rates, and what formal statistical methods are
appropriate to the analysis of relative attack rates?_
(4) What can be done to tighten the eligibility requirements
for asthma and cardiopulmonary panels?
(5) How can the statistical analysis of asthma and cardio-
pulmonary panels be unproved?
(6) What combination of CHESS health measurements is
most appropriate to long-term serial surveillance?
(7) What combination of CHESS health measurements is
appropriate to intensive studies of specific pollution hazards?
*CED=ohronlcrespiratory disease;
=acute respiratory disease.
28
-------
17
(in Magna, Utah. Such events
as this may allow one to disentangle the acute effects of
particular pollutants.
(7) Alternative populations should be considered, and
cohort studies of particular occupational or medical risk
groups may prove valuable.
(8) More attention might be given the need for individual-
ized dosimetry.
(9) Under what circumstances, if any, can epidemiologic
studies alone provide unambiguous dose-response data?
(10) More exotic studies might be considered. The health
effects of a new event could be observed, such as the con-
struction of a coal gasification plant. Changes in a town
which is known to plan selective modifications in the regu-
lation of a specific pollutant could be studied. Studying
vacationers from small towns who spend a week in a polluted
urban, center might be considered. Longer term migrant
studies might also be useful.
C. R&D RELATED FINDINGS
During the course of this investigation the extensive interviews
brought to light some issues which did not bear directly on the CHESS
program above but which are important to the effectiveness of the
29
-------
18
total EPA R. & D. program. Generally the Research Triangle Park
EPA Health /Effects Rose-arch Laboratory, and in fact all RTF
laboratories visited, appear to be inadequately staffed and suffering
from pooi- morale due to a number of factors including frequent
changes in leadership, poorly xmdcrstood organizational changes, im-
certaiiity about responsibilities, difficulty in recruiting high quality
personnel for leadership positions, poorly defined Ions-range task
objectives, and physical separation of working units. In addition, there
appear to be serious problems of coordination and communication
between divisions at RTF as well as at Headquarters, Washington.
These general findings are also borne out in the report on "The
Organization and Management of EPA's Office of Research and De-
velopment" prepared by the Subcommittee on the Environment and
the Atmosphere of the House Committee oil Science and Technology
in June, 1976.8
There1 follow several specific findings and [related recommendations.
11. The laboratories have not yet completely recovered from the
(inevitable) disruption associated" with the ,l3ecember, 1975 reor-
ganization of the Office of Research and Development.
Recommendation:
Since stability is an important ingredient in any good research
organization, the Center should be allowed-to function in its present
mode without further significant, reorganization until -the end of
Fiscal Year 1977 with all possible help it can use from Headquarters.
12. There is evidence of a need for improved coordination between
the four research laboratories at the Environmental Research Center,
RTF, particularly between the Health Effects Research Laboratory
and the Environmental Monitoring and Support laboratory. The
exchange of data and coordination of effort among the EPA/RTP
laboratories, and between EPA, contractors, and Federal laboratories
involved in similar programs is wholly inadequate.
A specific example of this lack of coordination is seen in. CHESS.
However, even after this learning experience the Environmental
Protection Agency is still not welf organized to undertake and con-
duct multidiseipliuary field projects and, in fact, certain of the recent
organizational changes are believed to have exacerbated this problem.
"When multi-disciplinary projects are organized and run by staff
having a single, particular discipline, important elements of ^the
problem ma3r be omitted or given insufficient attention. Objectives
need to be correlated more effectively with requirements and capa-
bilities. Statisticians and quality control experts need to play a larger
role in the planning of-such projects. Problems of coordination and
communication seem to aviso at project levels because of the Office
of Research and Development organizational structure in which
individual laboratories report directly to separate management offices
in EPA Headquarters. This is epitomized in the draft research plan
for sulfates, for example, wherein the research effort is divided into
four distinct segments at RTF but there is no program management
at RTF to insure coordination.. As a result, the four research elements
are not developed on consistent premises.
3 U.S. Congress, House Coaimittpp. on Science and Technology, Subcommittee on the Environment and
the Atmosphere, Orsanizalion and Management of EPA's Office of Research and Df velommcnt, G4th Con-
gress, 2nd Session, June 1976. U.S. Government 1'riuting Office, Washington, D.C. Serial LL, 40 pp.
30
-------
19
Recommendation:
(a) In accordance witli the recommendation under 11, above,
improved coordination should be sought at this 1 time without formal
organizational changes. This could be achieved through establish-
ment of an authoritative peer review panel to monitor all on-going
projects and insuie appropriate interface between the various dis-
ciplines. Such a review panel .should represent the highest level of
research skill and should assist not only in program planning and
review for scientific objectivity, but in peer review of major scientific
reports with policy implications. Such review should insure that
future reports are released in a timely fashion after complete scientific
assessment.
This need for coordination cannot be satisfactorily^ achieved by
ad hoc arrangements alone. Nevertheless, the formation of ad hoc
working groups, under the oversight of the continuing peer review
panel mentioned above, could be a useful mechanism for attacking
special problems, such as planning a coherent sulf ate research program.
In addition, the following additional steps might be taken without
affecting the stability of the laboratory:
(b) A stronger focus on management at the Environmental Research
Center, RTF, would aid in interf acing with the four Deputy Assistant
Administrator Offices at Washington to prevent separate tasking on
essentially identical problems.
(c) Consideration should be given to creation of a systems analysis/
operations research program review group at Headquarters of EPA
and perhaps at each technical center. This activity could insure that
proper attention is addressed to all scientific disciplines in program
plans and budgets, irrespective of which scientific discipline initiates
a program. This group might act as staff to the peer review panel.
(d) The Science Advisory Board's (SAB) charter should be re-
examined to improve the assistance rendered to the Health Effects
Research Laboratory and to Washington project directors. The SAB
should have direct informal access to the research programs and
opportunity to review and recommend actions relative to planning
and coordination of projects involving several divisions. The SAB
should have scientific review responsibility for programs, and its
participation should not be limited to public evaluations after some-
thing goes wrong. .
° -wroiw"—n1 i i I nil ••'! I nun m i in IIIUMIMIIIIIIIMBIIMI nn ill inu" r«r-| rn irr in
13. The EPA research programs in many cases were found to be
somewhat isolated. The research staff does not seem to avail itself
of a maximal technical exchange with scientific peers. It seems that
to a large extent outside contracts go to the, same groups again and
again.
Recommendations: ,
(a) EPA should actively seek cooperative research programs with a
variety of universities, other laboratories, other agencies. NO TE: This
is not a recommendation that more task-limited contracts be given,
but that truly cooperative programs be implemented in which the
outside laboratory can contribute its own ideas.
(b) EPA should arrange for research staff to spend six to _eighteen
month assignments working at other labs to reduce scientific isolation
and to promote career development (see below). Similarly, they should
31
-------
20
promote reciprocal visiting research assignments for outside scientists
to work at EPA laboratories.
(c) EPA should consider funding individual, worthwhile Ph. D.
thesis research as a means of broadening their technical community.
(d) The Science Advisory Board should be asked to help in devel-
oping such outreach programs.
14. There are examples of increasing overlap amon°; the Office of
Eesearch and Development (OED), the Office of Pesticides (OP), and
the Office of Toxic Substances (OTS), at Washington, in assignments
to the Environmental Research Center, ETP. Instances were noted
where both ORD and OTS have asked the RTP Center for research
on the same substances and where it was ;evident that neither the
divisions at the RTP Center nor the Offices at Washington had
coordinated the inputs to the Center during the development and
assignment of projects.
In addition it appears that research is being carried out in offices
other than ORD.
Recommendation:
The Administrator should clarify the role of the Office of Research
and Development in managing Agency research, and the role of the
laboratories in canying it out.
15. The various EPA groups at RTP are .now scattered geographi-
cally in such a way that the peer interaction needed to insure proper
coordination and exchange of ideas is physically difficult.
Recommendation:
Whatever final management concepts evolve, this separation of
facilities must, be resolved in an expeditious fashion if the EPA
program at RTP is expected to function- with any degree of co-
hesiveness. " ;
16. Poor staff morale seemed largely due to the superposition of a
ver.y difficult task (research in support of EPA's regulatory mission)
upon a very uncertain and fragile system for the support of research.
This would seem to cause or at least contribute to anxiety among the
professional research staff, especially since there appears to be no
plan of career development for this staff.
Recommendation:
EPA management should develop, implement, and defend a
professional career development plan for each professional employee
and for the staff as a whole. Developing such a plan will contribute to
better management by forcing a comparison of goals and resources.
An emplo3''ec secure in a peer-respected professional career will more
ha.ppily tolerate the exigencies of research in a regulatory agency.
Management of the ORD will sometimes have to defend its commit-
ment to such a plan against regulatory demands.
17. As stated above, major improvements in research execution are
needed. If, after the end of fiscal year 1977, the present structure is
judged inadequate to achieve this, a number of alternatives and
factors should be considered. The RTP laboratory directors are divided
as to the type of management structure needed "but there is a general
impression that the National Environmental Research Center
32
-------
21
(NERO) concept was preferred to the four-RTP-division/four
Washington-manager structure which currently exists.4 A major
problem is the lack of a single manager's office at the Center to insure
interdivisional review of multi-disciplinary projects and to secure
input from all divisions on such projects.
if success in recruiting key leadership and improving program
execution cannot be shown, even more fundamental steps may be
needed.
The steps to be considered might include (1),moving the existing
EPA research program into a new or existing agency; (2) establishing
a new agency to provide the basic^ long-term research support needed
(but not being done) and effectively dedicating the present EPA
laboratories to short-leash support of the agency regulatory programs;
(3) assigning long-term research support responsibilities to other
existing agencies.
Recommendation:
The Administrator must be prepared to answer hard and funda-
mental questions such as, should EPA conduct research?
D. RECAPlTTrLA.TIO]Sr OF GUIDING QUESTIONS
The following tabulation provides summary answers based on the
findings of the investigative team to the guiding questions formulated
at the, beginning of the investigation. ;
(1) Dm the CHESS measurement system utilise the best available
instrumentation and sound operational protocols and quality control
procedures?
The instrumentation was that of the state of the art at the time the
program started. However, it is quite clear that neither the protocols
nor quality control procedures were adequate to accomplish the objec-
tives of the program, and that the data usefulness could have been
greatly enhanced if well-known quality control procedures had been rig-
orously followed. For example, if the procedures described in the CHESS
Monograph (but not actually used) had been implemented, the data
would have been much more useful. If error bars nad been associated
with each data point the weakness of some of the cause-and-effect
relationships might have been evident. This, in turn, would probably
have prevented some of the over-interpretation of data.
(2) Were the precision and accuracy of the CHESS aerometric data
sufficient to quantify pollutant exposure differences between communities
where health effects were studied?
For many pairs of communities the answer is clearly no—both
because of inadequate procedures and also 'because at apparent
concentration levels in the lower-concentration areas the measurement
methods were incapable of detecting differences or changes in those
pollution concentrations.
(3) Were the aerometric data consistent with'and accurately repre-
sentative of the exposures to pollutants received by the populations used
in the health effects studies? L
« Under the NEED structure, the entire RTF laboratory reported to a NERC Director -who la tarn
reported to Washington Headquarters.
33
-------
22
This question cannot be answered, because there is no information
in CHESS on the relation between individual exposure and ambient
air quality. Thus, the CHESS researchers did not know this answer
either. In addition, the practice of having only one aerometry station
per community made it difficult to assess area-wide concentrations.
This illustrates the importance of personal dosimeters.
(4) Were new knowledge, instruments, and procedures, introduced into
the program as they became available and pro'ven?
It seems the introduction of new techniques was slow at best. For
example, as stated above even well-known quality control procedures
were introduced only slowly. The Committee understands the im-
portance of consistency in a project like ; CHESS, but this can be
assured with proper quality control techniques.
(5) Were the epidemiological populations adequately selected?
The findings raise doubts as to the validity of certain of the popula-
tion selections. ';
(6) Were the health measurement endpoints meaningful and reliable?
Again, the findings disclosed many shortcomings in this area. As
with the areometry, the health measurements could have been more
useful if validated, etc,
(7) Were the data reduction and a/nalys'is accurately and correctly
carried out? _ ^ ;
There, were instances of errors in both data reduction and analysis.
(8) What factors have retarded the analysis and reporting of the
CHESS data?
Several factors were uncovered. The most important appear to be
the forced change of computer and support contractors.
(9) Did the health and monitoring data establish unambiguously what
concentrations oi specific pollutants are associated with measurable
adverse health effects?
The findings clearly say no. ;
(10) If so, do the quantitative dose-response estimates in the CHESS
report have a firm empirical basis?
From (9) the answer must be no.
(11) Were the conclusions drawn from the CHESS program sufficiently
clear and unambiguous so as to form abound basis for future action?
Again, the answer is no. CHESS does indicate the need for a rigorous
program of health effects research. '
(12) Does the new CHAMP program provide a substantially im-
proved mechanism to provide aerometric data, to support future epi-
demiological research and is an epidemiological research program planned
to complement future aerometric monitorings?
The CHAMP system is clearly an improvement in aerometric
pollution measurement, although it is not yet fully validated. There
is no clearly denned, well-targeted, coordinated epidemiological
program to make use of CHAMP at this tune. There are no projects in
epidemiology making use of CHAMP data.
(13) What additional steps, if any, should EPA take to insure that a,
sound technical basis will be available for future agency decisions per-
taining to SOz/sulfates?
For the answer to this question, refer to all.the recommendations
in the sections B and C concerned with program improvements.
34
-------
23
E. CLOSING EEMABKS
The investigative team is aware that this is a very critical report.
The Environmental Protection Agency is a relatively new agency
which has been confronted with the need to grow rapidly and to assume
a large biirden of responsibility. The enforcement of these responsi-
bilities has increased in political significance with the growth in in-
tensity of the overall energy problem. All EPA'personnel consulted
in this study were generally cooperative and hopeful of being able to
provide constructive suggestions. Indeed, the recommendations pre-
sented here are, in many instances, suggestions and comments made by
EPA personnel. It was not difficult to identify many deficiencies
within the CHESS program or to relate these difficiilti.es to the need
for changes in organization and management within EPA. The
difficult task will be to resolve these issues arid construct a more
effective organization to meet the very real problems of pur society
which have been mandated to the Environmental Protection Agency
for resolution.
-------
-------
IV. CHESS AEBOMETKIC MEASUREMENTS
A. INTRODUCTION •
i •
As pointed put in the introduction, the attainment of precise, reli-
able, reproducible, and real time air Duality measurements in the field
(e.g., 562 and participates) was a critical element of the CHESS pro-
gram. This chapter provides a critical review of the aerometric
measurement aspects of CHESS.
However, before reporting on this review two facts about CHESS
aerometry should be mentioned. First, the methods used in CHESS,
especially in 1970-71, were probably as good as any available. Second,
quality control procedures were slowly introduced into the CHESS
program. EPA cannot be criticised, and is. not criticised in this report,
for using the best available methods. However, EPA can be criticized
for not pursuing a vigorous program of quality control throughout
CHESS. The review reported here showed that CHESS did not
employ well-established quality control measures. The quality control
program described in Appendix A of. the Monograph was not carried
out. A thorough quality control program would have discovered, for
example, the temperature effects on the method used to measure SC>2
(described below). It would also have placed bounds on the validity
of the data and precluded overinterpretations.
In the design and implementation of any measurement system, the
single most important consideration is the end user of the data pro-
duced by that measurement system. In the simplest of all measure-
ment processes, an individual scientist conducting his own research,
both measures the parameters _of interest and uses the resultant data
to draw conclusions about his experiment. In such a process the
individual involved has at his disposal all of the information contained
in the data, especially that concerned with the limitations of the data
and the constraints under which they should be used. In this type
of situation, few formal qualifications of the recorded data are neces-
sary since those'qualifications are implicit in the mind of the scientist.
In larger programs however, the measurement process and the
utilization process are quite often compartmentalized such that one
group of scientists is responsible for the collection, quality assessment
'and storage of the measurement data, and a second, usually nonrelated,
group of scientists is responsible for the synthesis of all pertinent
information into a final set of conclusions. In this type of systems
research, the determination of the fundamental quality 01 the measure-
ment data and transmittance of that quality assessment are the single
most important qualifier in the process of going from observation to
understanding.
The CHESS program, as designed and implemented by the Envi-
ronmental Protection. Agency, is a classic example of the large sys-
tems approach to research. The epidemiological measurements
were designed, conducted, and stored by one group of scientists; the
(25)
76 3
37
-------
26 \
aerometric measurements were designed, conducted and stored by a
second group of scientists. The desired end product, a correlation of
health effects with atmospheric pollution was then derived from these
two independent sets of data accumulated in a large data storage net-
work. It is important to reempbasize here that in such a research
program it is incumbent upon the measurement personnel to transmit
to the data user all of the information contained in the resultant data,
especially that relative to accuracy and precision. In order to under-
stand the problems encountered in a large research program such as
CHESS, it is necessary to understand the types of measurements that
were made.
The assessment of atmospheric pollution exposure received by a
defined population can be derived from.one of two broad classes of
measurement. The first is a measurement .that yields an "index" of
pollution. The second is a measurement that yields quantitative in-
formation about a specific pollutant as it is found in the atmosphere.
A pollution index is a measure of the relative level of pollution which
contains little or no information as to the specific chemical or physical
properties of that pollution.* These indices 'can be useful in assessing
short-term trends of atmospheric quality in well-defined and limited
geographic regions. They cannot be used to deduce information about
the source or chemical nature of the material being measured. They
also cannot be used to assess long-term trends of pollution burden
since gradual changes in pollution. sources will distort the quan-
titative aspect of the index. Most importantly, they cannot be used to
correlate atmospheric pollutant levels among diverse geographic
areas. Here again, the difference in chemical and physical makeup of
the pollutants being measured distort the quantitative aspect of the
index. - :
An example of a measurement that gives a pollution index is the
dustfall observations as applied in CHESS. In this method, an open
topped cylinder called a dust fall bucket is used to collect any par-
ticulate matter that falls out of the atmosphere. This collection is
carried out over a long time period, usually;one month; and the total
dry weight of material collected is used to estimate participate burden
of the atmosphere during that time period.1 A detailed description of
this process is given later in this Chapter. This measurement falls in
the index class because all solid material, regardless of its derivation
or chemical nature, is included in the final quantitative result.
The second class of pollution measurement is that which contains
information both on the specific species of pollutants and on the
atmospheric concentrations of those pollutants. In this type or
measurement the signal that is measured is derived from a process or
property which is specific to the pollutant of interest and which corre-
lates directly with the concentration of that pollutant in the atmos-
phere. An example of this type of method is the West-Gaeke proce-
dure for the measxircment of atmospheric sulfur dioxide. In this pro-
cedure, air is bubbled through an absorbing solution at a known rate.
The solution is specific for the absorption of S02 from the air. After a
known duration of sampling, the quantity of S02 which was absorbed
from the air is quantitatively determined by the formation of a
* N.B. This index is not the kind of "air quality index" often used popularly (in radio broadcasts, etc.)
to advise citizons of the relative air quality of a city. Such popular to quality indices are usually arrived at
by combining measurements of several pollutants.
38
-------
27
colored chemical complex of SO2. If carefully carried out, the procedure
gives an accurate value for the S02 concentration. The procedure is
described in detail later in this Chapter.
Measurements such as the West-Gaeke procedure, which are
specific and quantitative, m can be used _to compare atmospheric
pollutant burdens across diverse geographic areas and through long
tune periods. They can also be used to assess short-term variations in
pollutant levels provided that sufficient sensitivity exists in the
method to obtain a meaningful signal for the short time period used.
In conducting a program such as CHESS, where an attempt is made
to relate health effects to pollution burdens, only those measurements
that fall in the second class, specific and quantitative, can properly
be used to assess the relation between health % effects and pollutant
burden.
In this chapter, an attempt will be made to evaluate the method-
ology used to measure aerometric parameters and to assess the
validity of the resultant data. The review will encompass procedures
used in the field situation, the quality control exercised over the proce-
dures, and the data storage and retrieval network. Conclusions will
be drawn as to the adequacy of the measured pollution levels to assess
exposures received by specific CHESS population groups.
B. KEVIEW OF CHEMICAL AND PHYSICAL METHODS
1. THE WEST-GAEKE METHOD FOB THE MEASUEEMENT OF AMBIENT SO3
a. Description of the Method
The West-Gaeke colorimetric procedure for, S02 determination is
the designated Reference Method (Federal Register, 86, No. 84, 6168,
April 30,1971).* Atmospheric SO2 is collected by bubbling air through
a solution of potassium tetraehloromercurate (TCM). The product of
the reaction between SO2 and TCM is the nonvolatile dichloro-
sulfitomercurate that is then determined quantitatively by reaction
with formaldehyde and pararosaniline hydrochloride, followed by
photometric measurement of the resulting intensely colored para-
rosaniline methyl sulfonic acid.
b. Description of the Field Apparatus and Sample Collection
Outside air is drawn through a sample line:at the rate of 200 ml
min"1, then through a 6-inch long glass bubbler stem (tip diameter
of 0.025 in.) immersed in 35 rnl (50 ml after January, 1974) of
0.1 M TCM solution contained in a 32 mm diameter by 164 mm long
polypropylene> sample container. The exhaust air passed through a
glass wool moisture trap, then through a hypodermic needle used as a
critical orifice to control the flow, through another moisture trap, and
finally through a vacuum pump. A sample consisted of a 24-hour
collection. Collected samples were stoppered, and mailed to EPA/
RTP for analysis.
c. Validity as a Laboratory Procedure ,
A collaborative study by McKee et al. (H. C. McKee, R. E.
Childers, and O. Saenz, Southwest Research Institute, SWRI Project
21-2811, EPA contract CPA 70-40) indicates that "the method can-
•Alternately see CFR Title 40, Part 50, Appendix A.
39
-------
28
not detect a difference smaller than 10 percent between two observa-
tions by the same analyst in the range of 0 to 1000 jug m~3. A difference
of 20 percent or less may be detected above 300#g ra~3, and a difference
of less than 50 percent may be detected: above 100;*g m~3." For
analyses conducted by different laboratories on the same sample, "the
method.cannot detect a difference of less 'than 20 percent between
singlo-replicate observations of two laboratories in the range of 0 to
1000 jig m~3. At a level of 100 pg m~3, a difference of less than 100
percent is not detectable." The National Primary Ambient Air
Quality Standard for SOz is: For 24 hour average, 365 MgA*i3. For annual
average, 80 Mg/m3. Thus if the staticlard is iniet, most values will be
around or below SO fig/m?, no more than one will be above 365 M^/m3-
Regarding the lower limit of detection, the authors cited above
propose a value of 25 ng rn~3 as a practical figure. "A single determina-
tion less than this value is not significantly different from zero"
(Instrumentation for Environmental Monitoring, Air-SO3, Instru-
mentation, Lawrence Berkeley Laboratories, March 1972).
It is therefore evident that a single analysis is of little use, con-
sidering that the expected concentrations of SOa will usually be less
than the ambient air quality standard of SO ^g m~3. Results should
be regarded as valid only in terms of the mean of multiple determina-
tions, and only when the analytical method has been followed
rigorously by experienced analysts.
2. TOTAL SUSPENDED PABTICULATES
Total suspended partieulatps (TSP) were'measured using the EPA
Reference Method as specified in the Federal Register (S6 (84):
8191-8194, April 30, 1971J).
Total suspended particulates (TSP) were-measured by drawing air
through a prewieghed 8 x 10 inch glass fiber filter for a period of 24
hours. The apparatus used for this procedure was the standard High
Volume Sampler. At the end of the 24 hour time period, the filter was
reweighcd, and the TSP computed on the basis of total air flow. The
air flow rate was approximately 60 ftamin~1 at the start, and must be
not less than 40 ft'min"1 at the end for the measurement to be accept-
able. The average air flow rate was computed on the basis of a straight-
line interpolation between beginning and ending flow rates.
The National Primary Ambient Air Quality Standard for TSP is:
For 24 hour average, 260 Mg/in3. For annual geometric mean, 75 Mg/ni3.
3. SUSPENDED SULFATE
Suspended sulfate was analyzed, during the CHESS program, using
Portions of the TSP samples. From the 'beginning of CHESS to
eptember 1971 the turbidimetric method of analysis was used; then
the turbidimetric method was dropped in favor of the methylthymol
blue method, which was used throughout the remainder of the CHESS
program. '•.
Ihe turbidimetric method consists of the water extraction of soluble
sulfatcs on the TSP filter, the. addition of &' barium chloride prepara-
tion to the extract, and measurement of the resultant turbidity (from
Alli-rmiivriy sec CFB Title 40, Z'art 50, AppendixB.
40
-------
29
the formation of insohible barium, sulfate) with a spectrophotometer or
colorimeter. Accuracy of the method is affected by the kind and con-
centration of other ions present, as well as pH, conductance, tempera-
ture, and barium concentration in the test solution.
The methylthymol blue method also utilisr.es the water extraction of
soluble sulfates from the TSP. The filter extract isithen passed through
an ion-exchange bed to remove interfering ions, and barium chloride
is added under slightly acid conditions, forming barium sulfato. Then
the test mixture is made alkaline and methylthymol blue is added,
which forms a chelate with the excess barium. The uncomplexed
methythymol blue is equivalent to the amount of sulfate present, and
is measured spectrophotometrically. The methylthymol blue procedure
is automated (Technicon Auto analyzer) in ail steps following water
extraction of the TSP, and this part of the procedure is reproducible
within a range of 2 percent. Error in the determination of sulfate occurs
predominantly in the steps .preceding the methylthymol blue method.
4. DUSTFALL BUCKET, TAPE SAMPLER, CASCADE IMPACTOR,
AND CYCLONE SAMPLER :
In addition to TSP measurements using the Hi-Vol sampler, four
other means of estimating particulate concentrations were used at
various times. They are the dustfall bucket, the tape sampler, the
cascade impactor, and the cyclone sampler.
(a) The name "dustfall bucket" is adequately descriptive. It is
basically an open-topped cylinder, with some protection against wind
and rain loss, that is left out in the open, close to the ground or on a
rooftop, for a month. At the end of that time the dry matter collected is
weighed, and sometimes analyzed for trace metals. The dustfall bucket
method is very crude and misses almost completely the very significant
part of the aerosol, including the respirable aerosol, that does not settle
rapidly. It must be considered here, however, because dustfall measure-
ments were extrapolated to obtain estimates of suspended sulfates
and sulfur dioxide in New York City during the period 1949-58
((Table 5.2.1, CHESS Monograph), and intermittently in Chicago
(Table 4.1.A.3), CHESS Monograph). Dustfall measurements were
used as the basis for these extrapolations because there was no other
basis for such estimates, but it must be remembered that the relation-
ship between suspended sulfates and dustfall is unknown, and that
between sulfur dioxide and dustfall is another step removed from
reality.
(6) Coefficient of Haze (COH) is determined by the automatically
operating tape sampler. It is determined by measuring the optical
density; of an aerosol deposited on a filter tape. The aerosol deposit
is obtained by drawing air at a given flow rate through white filter
paper tape for a known period of time. If one could assume that the
composition and physical characteristics of the aerosol in a given
location did not change with time—that only atmospheric loadings
would change—then the COH would give a fairly; good approximation
of the variations of particulate loading and visibility.
However, this assumption is seldom justified, and even at a given
location the COH only roughly approximates the true particulate
loading. The COH method is worthless, or nearly so, for comparisons
between areas with dissimilar aerosols. For example, the aerosols
41
-------
30
collected at the Utah, sites are primarily the light-colored alumino-
silicate dust, whereas the aerosol collected within the inner core of
large cities has a predominantly sooty character. For a given par-
ticulate loading the Utah aerosol will often have as little as one-tenth
the optical density of the urban aerosol.
(c) The cascade impactor operates on the principle that particles
in an air stream will tend to follow a straight line when the air stream
is deflected, and thus can be impacted on a surface in their path.
The cascade impactor consists of a series of parallel plates separated
by precisely determined spaces. Alternate, plates contain a certain
number of holes of a size that is decreased as one goes throxigh the
series of plates from entrance to exit. Alternating with the plates
containing the calibrated holes are plates without holes. These may
be coated with a medium for the trapping ,of impinged particles. Air
is drawn through the apparatus at a known-rate, and the particles are
collected in decreasing size fractions related to the decreasing size of
the holes in the plates.
(d) The cyclone sampler is a device for the collection of the respi-
rable size fraction of an atmospheric particulnte loading. It operates
on the principle that the inertia of individual particles will tend to
keep the particles moving in a straight line when the air stream in
which they are carried is deflected. By this means the larger size
particles are removed by irnpaction and settling, while the respirable
particles are carried along with the air stream and are subsequently
collected on a filter.
C. FINDINGS AND EVALUATIONS OP MEASUREMENTS AND DATA
REDUCTION
It is important to profane this evaluation of the CHESS air moni-
toring program with a statement of the following facts. The inves-
tigative team looked backward at the program through a window in
time with all of the subsequent knowledg-e built up during that time.
More than ten years have passed since the initial planning of the
CHESS program and more than six years have passed since the first
data were collected. During that time there has boon a vast improve-
ment in the understanding of the methods used for pollution moni-
toring. Many of the procedures used in CHESS have subsequently
been found to contain serious errors. These problems were often
uncovered as a direct result of research and quality control programs
ongoing within EPA. It would thus be unjustified to lay criticism on
the principals in the CHESS program for using state of the art meas-
urement technology.
On the other hand, some serious oversights in scientific judgement
did occur. In the area of pollutant monitoring, these oversights could
have been completely avoided had .proper attention been paid to even
rudimentary quality control procedures. Throughout the program,
much more- emphasis was placed on the uninterrupted collection of
data than was placed on the systematic evaluation of data quality.
The field investigation stage of this review identified numerous prob-
lems that resulted in the propagation of unnecessarily large errors in
the aerometric data. These unevaluated errors persist even today in
the data as it is stored in the CHESS computer system. They could
have been avoided or easily discovered and quantified had a wel!-
42
-------
31
designed quality control procedure been applied to the CHESS aero-
metric monitoring program. This statement is contrary to the state-
ment of the quality control procedures in appendix A of the 1974
CHESS Monograph. Appendix A was not a manual provided to
CHESS data gatherers, but was written long after the data in the
1974 Monograph were collected. However, during the field investi-
gation of the CHESS monitoring contractors, it was found that the
quality control procedures as described in Appendix A of the CHESS
Monograph were routinely disregarded. In fact, for the first two years
of the program, virtually no EPA
-------
32
In an attempt to standardize the methodology and to eliminate
problems associated with, interlaboratory errors, a'CHESS policy was
instituted whereby all air sampling equipment was assembled and
tested at the central EPA research laboratory and then shipped to
the contractors for field use. Also, bubbler tubes were prefilled with
the appropriate absorber solution, shipped to the contractor for their
daily ^monitoring use, and shipped back to the central laboratory for
chemical anatysis. It was this long distance shipment of the chemical
solutions that led to the first of a series of field-use problems with the
procedure. These problem areas will be summarized below with an
attempt to evaluate their net effect on the resultant CHESS SO2 data.
Following this summary of individual problem areas, an assessment
of the overall S02 data quality will be given.
a. Spillage of Reagent During Shipment
The first field data were obtained in New York City and the Salt
Lake area (Utah) in November, 1970. By mid-1971, field personnel
at the Utah site reported to their CHESS field engineers that severe
spillage was occurring during shipment. Many bubbler tubes were
arriving partially filled with reagent and some were completely empty.
At the Salt Lake area an attempt was made to refill with solution
from extra tubes those tubes that were low. However, due to insuffi-
cient reagent, this was only partially successful. This problem was
not officially recognized until October, 1972, at which time an internal
EPA/CHESS memo was written outlining tlie problem and suggesting
corrective action. The magnitude of the problem can be best assessed
by quoting from the memo. "The present reagent tubes for S02 and
NO2 leak during shipment. . . . The S02 leakage rate (was foxmd to
be) 18% of the total volume, 50% of the time. ... It follows there-
fore, that the resultant pollution data are unreliable." Recommenda-
tions were made in this memo as to possible corrective measures. These
recommendations were not instituted until ;March, 1973.
During the subsequent years, many attempts were made to correct
this leakage problem. However, none were wholly successful and as
late as January 1975, another EPA memo described losses of solution
in SOs bubblers during shipment and suggesting appropriate corrective
action.
The effects of the reagent spillage problem on the SO2 data can
be only grossly estimated. Certainly, many samples were totally lost.
These lost samples were not the major problem. Of more significance
was the undetermined amount of daily S02 data that were in error
due to the loss of sample by spillage and yet included in the network
system.
If the reagent was partially lost during shipment to the sampling
site and used as received, an increased concentration of TCM-Sp2
complex would occur relative to normal sampling. This potential
positive bias would be corrected by for the analytical procedure used
(Page A-6 CHESS monograph—Analysis Procedure). "At the labora-
tory, the sample is brought back to its original volume by the addition
of distilled water to compensate for water loss during sampling."
If however, the reagent spillage occurred after sampling, the required
addition of water would result in data that were biased low in pro-
portion to the amount spilled relative to the total volume of solution.
44
-------
33
According to the EPA Memo of October, 1972, one half of all SO2
data taken between November, 1970 and March, 1973 are likely to
have been biased low by an average of 17%. This problem was cor-
rected after April, 1973.
6. Time Delay of the Reagent—S02 Complex
The Reference Method as originally described in the Federal
Register, was to be conducted at 20° C. There was a known error in
the method associated with tune delay between sampling and analysis
which was dependent on temperatures. This error was derived from
the spontaneous decomposition over time of the TCM-SO2 complex
as a function of temperature. The magnitude of the error and its
exact dependence on temperature was not known but a brief study
was conducted to determine its magnitude by scientists of the CHESS
monitoring group in November, 1971. As a result of this study, a
correction factor of rfl.5% per day was arithmetically applied to all
CHESS SO2 data to compensate for the time delay between sampling
and analysis.
A more recent and comprehensive study has been carried out within
the Quality Control Branch, Environmental Monitoring Laboratory
at EPA on the effect of temperature on "The Stability, of SO2 Samples
Collected by the Federal Reference Method." This study indicated a
much more severe problem than was estimated by the original CHESS
study. The evaluation was carried out over the range of 35 to 278
jug/m3 SOa concentration. The following findings were presented in
the report:
Over a normal range of temperature, the rate of decay of the
TMC-SO2 complex increases five-fold for every 10°C increase
in temperature, respectively.
The rate of decay is independent of S02 concentration.
At 20) 30, 40, and 50° C the following 80s losses were observed:
0.9, 5, 25, and 74% loss per day, respectively.
This study makes abundantly clear a second and even more severe
error associated with the SO2 measurements conducted by CHESS.
During the summer months, when the SO3 absorber solutions were
subjected to high and unknown temperatures between field sampling
and laboratory analysis, significant degradation of the samples did
occur. Estimates of time delay between sampling and analysis range
from 7 to 14 days. Estimates of summer temperature exposures range
from 25 to 40° C being most severe for the Utah CHESS sites. Thus,
CHESS SOs data can be estimated to be negatively biased, mainly
during the summer months. It would normally be difficult or impossible
to estimate the magnitude of the bias except to say that it is probably
large. However, simultaneous SO2 measurements were taken by the
New York City Department of Air Resources and by the Utah State
Division of Health. These results were obtained by an independent
method not susceptable to the temperature related error. A consistent
pattern emerged when side by side data are compared. From May to
October, the CHESS SO2 data were low with the largest error occurring
in the middle three summer months. The magnitude of the error varied
from month to month and year to year, but the CHESS data were
consistently low and represented only a portion of the true ambient SOz
concentration.
45
-------
34
c. Concentration Dependence of Sampling Method
The S02 reference method was subjected to a collaborative study
program in 1973. Four participating: laboratories tested the 24-hour
version of the Federal Reference Method. A previously unknown source
of error was documented that applies to the CHESS S02 data. It was
found that the 24-hour sampling method does have a concentration
dependent bias which becomes significant' at the high concentra-
tion levels (200jug/m3). Observed values tend to be lower thnn the
expected (known) SO2 concentration levels. This error source will yield
a negative bias on the daily CHESS S02 data when they exceed 200
jtg/m3 and on all monthly and yearly average data.
d. Low flow correction
The determination of atmospheric S02 concentration was dependent
on, among: other factors, the accurate measurement of air that passed
through the TCM solution. This flow was controlled by a critical flow
orifice in the form of a standard hypodermic needle. In practice, the
air flow throxigh the sampling system was measured at the start and
end of each 24-hour sampling, period. This, was done to detect low
flow due to needle blockage. The Federal Register Method (Reference
Method) calls for an air flow of 200 ± 20 ml/min. In field operation, the
CHESS procedure substantially broadened these tolerances. Replace-
ment needles were installed if the initial air flow was greater than
220 ml/min which is consistent with the Reference Method; however,
needles were not replaced nor were samples voided until the measured
flow dropped below 100 ml/min. Integrated'flows were calculated by
assuming a linear decrease in flow between the start and end of the
24-hour sampling period. If, however, the needle was partially blocked
near either the beginning or the end' of the sampling period, the
linear flow correction would be in error. Using the Reference Method
flow tolerance, only small errors would be introduced by this correction
(less than 10%). Using the CHESS procedure, however, errors as
large as 50% could be introduced and not detected. These errors
would be random (either positive or negative) depending on when dur-
ing the sampling period the needle blockage occurred. Thus a large
random error component was added to the SO2 daily data but this
component was somewhat damped statistically in the monthly or
yearly averages.
The modification of flow tolerance by ;the CHESS aerometric
group is a procedure that would not have withstood the critical review
of a competent quality assurance program.
e. Bubbler train leakage
The West-Gaeke method, as described in the Federal Register,
employs a vacuum .bubbler train. That is, the sampled air is drawn.
through the bubbler train by a vacuum pump rather than being
pushed through by a positive pressure pump. There are many ad-
vantages to the vacuum procedure, most important is that the air
does not come in contact with any internal pump mechanism. However,
there is a modest pressure differential between the atmosphere and the
internal bubbler; thus all fittings and joints must be gas tight. The
bubbler train used in the CHESS program had two points where frequent
air leak problems were encountered. One was around the rubber
stoppers for the bubbler tube and moisture trap and the other was the
46
-------
35
rubber tubing used to hold the glass assembly pieces together.. Field
operators reported consistent problems with leakage in the routine field
use of the bubbler train. In a severe leak situation, the samples were
voided due to out of tolerance (Low) flow rates. There were many
cases however, where small leaks occurred but the final flow was
within specifications so the sample was included as valid. In cases
where the leaks formed around the rubber stoppers, no significant
error would be introduced except due to the linear flow correction as
applied to instantaneously developing leaks. This error is similar in
nature to that discussed in the flow section. In the case of leaks up-
stream of the bubbler train, room air instead of outside air is drawn
through reagent. In normal situations, it has been observed that room
air is significantly less polluted than outside air. (See page 6-6,
CHESS Monograph—comparison of school air to outside air). This
effect may not be as large for the small buildings used to house CHESS
stations, but a somewhat decreased pollutant level would undoubtedly
be sampled. The absolute magnitude of this error cannot be adequately
assessed but it can be stated that the error would be in a negative
direction, that is, again to underestimate SO2 levels.
2. GENERAL ASSESSMENT OF CHESS SO2 DATA
The SO2 data, accumulated at "official" CHESS sites, followed a
remarkably uniform trend as the program progressed. The method
used was the EPA Reference Method which is specific for the chemical
species, SO2. Thus, regional changes in poEutant mix, i.e., the propor-
tion of other pollutant species relative to SO2, had minimal effect on.
the SO2 data. However, the sum effect of the errors detailed in this
section did have a profound effect on both the accuracy and the preci-
sion of the data.
Under normal circumstances, a retrospective evaluation of a
monitoring effort that occurred a number of years in the past and
which had been terminated, could yield only the broadest of estimates
of data quality. Fortunately for this review, two geographically
different locations with six different monitoring sites were involved
in the collection of simultaneous SO2 data. Further, the groups re-
sponsible for the two data sets were managed independently and the
methodology used was also independent. This fortunate circumstance
enabled the reviewers to acquire a quantitative understanding of
absolute.differences among data sets as well as correlations with respect
to time.
The locations where side by side data existed were the New York
City sites at Bronx and Queens and the Salt Lake Basin sites at Ogden,
Salt Lake City, Kearns, and Magna. In these locations, the local
environmental monitoring agencies nad sites located within 50 meters
of the CHESS sites and at similar elevations. At these sites, the local
agencies collected daily SO2 and TSP data for the entire life of the
CHESS program. The SO2 methodologies used by both State agencies
'were variations of the peroxide bubbler method in which twenty-four
1-hour samples were integrated to form a single 24-hour SO2 measure-
ment. In New York the samples were measured acidimetricly and in
Salt Lake City they were quantified conductiometrically. Neither
method is as specific for S02 as is the Reference Method, that is,
47
-------
36
pollutants that are in a significant concentration, relative to
and that also oxidize to form an acidic compound will be interpreted
as SO2. For this reason, when the NYC Department of Air Resources
initially brought to the attention of the CHESS Aerometric team the
large discrepancy between their respective data, the discrepancy was
dismissed as method bias on the part of the New York method. An
EPA memo dated November 3, 1971 described a limited study into the
Reference Method. The conclusion reached was "On the basis of
(this study) ... I feel there is no sound basis for discrediting the
EEvS (Environmental Exposure System) methodology."
No further attempt was made to uncover the cause of the discrep-
ancy in SOa data. Had the CHESS EES team obtained and compared
the Salt Lake Basin data, especially that from Magna site, a disturbing
similarity would have been immediately apparent. This data confirmed
in detail the discrepancies observed in New York. It is important that
the Magna site data were confirmatory since it was in a region of
single source pollution, that from the nearby copper smelter. In this
site very low levels of other pollutants existed relative to SO2, thus
the peroxide method was capable of giving reasonably reliable esti-
mates of the S02 concentration. Of equal importance the general pol-
lutant mix wasjvory different between this |rura1 smelter site and the
urban area of New York City. Despite these differences the compari-
son of side by side Federal-State data indicate the same discrepancies
in both trends and absolute concentrations. The following conclusions
as to SO2 data validity can thus be reasonably drawn from the review
of methodological errors and the comparison of existing side by side
data.
From November 1970 until December 1971 the SO2 data generated
from CHESS sites using the modified Reference method were biased
low by 50 to 100 percent in the High Exposure sites when compared
with existing State SO2 data. Thus, the 1971 annual average SOa
exposure estimates of 60jug/m3 as reported for Magna in the CHESS
monograph (page 2-24) are more likely in the vicinity of 100 Mg/m3.
Also, the same phenomenon occurred in New York and the reported
values are also in similar error.
A confirming fact is that during cool months after 1971 SO2 data
correlated well both in trends and absolute concentrations between
State and Federal analyses. It thus seems likely that the State data
were reasonably accurate throughout that time period. However, one
consideration must be applied here: namely, that due to the difference
between the independent methods an error oar of at least one hundred
percent must be applied to the data and explicitly correct data cannot be
drawn from these observations. In other words, where two or more
independent observations are in disagreement by a significant amount
it cannot be said by inference alone that one data set is more correct
than the other. It is reasonable to assume, however, from our review
of all State and Federal data in the time period of 1970 through 1971,
that the Federal SO2 data as collected in the CHESS program were
substantially low and went .through an abrupt upward transition in
concentration in December 1971 at all CHESS sites and Federal data
taken before that time may reasonably be expected to have a large,
unknown negative bias.
48
-------
37
In November 1971, the CHESS monthly mean SO2 data underwent
an abrupt change in the positive direction. The cause of this change is
not apparent. However, the result was prof (mud. From that time until
the conclusion of the CHESS program in. July of 1975, the fall-winter
data were in very good agreement with other existing data and very"
likely gave reliable estimates of S02 exposures.
Throughout the entire program, the CHESS SO2 data had an
associated negative bias during the summer months, becoming most
severe during the hottest periods of July and August. This error
usually reached a maximum of 60 to 80 percent underestimation of
exposures and was variable. As a result, even though wintertime
monthly SO2 averages appear valid from 1972-1975, annual averages
of the same data are biased low due to the inclusion of the summer
errors. The best estimate of error in the annual average data 1972-
1975 is approximately minus 15-20 percent relative.
The individual daily SO2 levels, when compared to city or State
data or to replicate CHESS measurements taken after 1973 had so
large a random error component that they are not useful to assess
daily SO2 exposure (as attempted in the asthma panels). The random
errors associated with the daily values were much larger than the
differences observed over time.
Due to inherent methodological errors, the following may be con-
sidered as minimum differences between High and Low SOg exposures
which may be considered "real." These are based on EPA's collab-
orative study of the reference method and used a 95 percent confidence
interval.
Below 100 Mg/m3 SO2, a difference of at least 50 /xg/nx3 is
necessary to be statistically significant.
Between 100 and 300 Mg/m.3 SO2, a difference of at least 60 ^g/m3
is necessary to be significant. :
Below 25 /ig/m3, a single determination is not significantly
different from zero. .
3. TOTAL SUSPENDED PABTICTLATE
The Reference Method for the determination of total suspended
particulate matter (TSP) is probably the simplest and most reliable
method used by CHESS. It has been well studied and most error
sources are known. However, it is a method that; measures an arbitrary
and poorly defined portion of the total atmospheric particulate burden
and the portion measured has unknown relevance to the human respir-
able portion. The size fraction measured is somewhat dependent on the
design of the shelter used for Hi-Volume sampler. The design and di-
mensions of the Reference Method shelter are specified in the Federal
Register, thus the portion of TSP that is collected by the method
is generally uniform. Best estimates of particle size range included
in the Reference Method are from 0.05 to 60 jon diameter. Above
60 ion diameter, the particle fall velocity is too great to navigate the
bend around the roof of the shelter. Below 0.05 pun the collection.
efficiency of the glass fiber filter used in the method diminishes.
A collaborative study was conducted on the Reference Method
using 12 different groups sampling ambient air at a common location.
The results of this study indicate the method is capable of reproducible
49
-------
38
measurements with less than 5 percent error at the 95 percent con-
fidence level. Also, the minimum detectable amount of TSP is approxi-
mately 2 Mg/nx3 for a 24-hour sampling period. This sensitivity is
more than sufficient for most 24-hour TSP measurements.
The TSP measurement method, as used in. CHESS, had one notable
difference from the laboratory procedure which was collaboratively
studied. The weighing procedure to determine TSP was performed at
EPA/RTP laboratory not by the CHESS contractors on site. This
necessitated the shipment of individual filter samples through the
mail and the subsequent storages of the samples at EPA. During
laboratory reorganizations at RTP, periods as long as 6 months
elapsed between actual field sampling and laboratory analysis.
The following is a summary of individual errors and an assessment
of overall TSP data quality.
Loss of partic-ulate matter before weighing >
In the TSP methodology there were field-related procedures that
resulted in partial loss of particulate matter from the Hi-Volume
filter samples. Due to the exposed location of the Hi-Vol TSP samplers,
wind and cold sometimes made it very difficult to remove the filter
paper from the apparatus without losing part of the sample. No
estimate has been made of loss due to.this problem; it would, of course
bias the reported results only in the direction of lower-than-actual
atmospheric loadings. This wus not a constant problem among; CHESS
sites. It was noted by field operators as being a particularly severe
problem in the Salt Luke City area during the winter months.
Two other error sources have been identified in the determination
of TSP, both of which would also produce a low-side bias: (1) the
shaking-oS of particles from the filter during transit from the field
site to EPA/RTP, uncl (2) the evaporation of organic substances. In an
attempt to quantify the mass loss during transit, David Hinton, EPA/
KTP, made a comparison of filters collected in. Utah, before and after
mailing from Salt Lake City to RTP (22). He found that there was a
average 4 percent loss. Carl Broadhcad, of the Utah Division of
Health, conducted a similar comparison.; however, he noted an
apparent loss of approximately 25%. This difference may, in part, be
due to the time of year the studies were conducted, During the dry
summer months in the Salt Lake City area, much of the TSP loading
is due to windborn crustal material (sand). This material is much more
easily lost in sample handling that is the finer anthropogenic particu-
late material.
A final error source, one more difficult, to assess, derives from wind
velocity versus collection efficiency. On days with relatively high wind
(>15 mph), the Hi-Voi sampler is more sxisceptible to the inclusion
of large diameter particulate material. To compound this problem,
the design of the shelter makes the magnitude of the error dependent
on the wind direction relative to the orientation of the shelter. The
main result of this problem is that two side by side Hi-Vol samplers,
oriented 90 degrees relative to each other, will produce dissimilar
measurements with the discrepancy increasing as the daily wind
velocity increases.
The overall effect of the summed errors -with the Hi-Vol TSP
measurement is a slight negative bias. This bias may be as small as
10% or may be as large as 30%. Side by side data from Now York
SO
-------
39
and Salt Lake indicate that this assessment is reasonable. These
data also indicate that the TSP data were by far the best quality
data taken in the CHESS monitoring program. Differences measured
between High and Low sites are probably reasonable estimates of
the differences of TSP exposures as received by populations within
these areas. Some local source variations undoubtedly did occur, but
average annual exposiires were reasonable.
In anyjoverall assessment of the CHESS TSP data it should be
noted that all of the sources of errors mentioned previously related
almost exclusively to the loss of large particulate matter and most
likely that matter is associated with crustal weathering. This material
is outside of the normal human respirable size fraction and by com-
position, it would be unlikely to be associated with aggravated health.
Thus, loss of that portion of the total material may not have di-
minished the quality of data for health effects studies. It may in fact
have rendered that data a closer estimate of the respirable TSP
exposure to which the CHESS population groups .were subjected.
It has been suggested by some environmental scientists that when-
ever Hi-Vol measurements are made for health related studies, the
filter pads should be "shaken out" much like a housewife does when
shaking crumbs from a used tablecloth. The resultant TSP exposure
estimates derived from such a procedure would then more closely
relate to the human respirable size fraction of the total atmospheric
particulate burden. Although never actually implemented, this
suggestion indicates the general level of dissatisfaction with the TSP
Hi-Vol measurement method.
4. TOTAL SUSPENDED STTLFATE
The determination of atmospheric ssulfate concentrations, as
carried out in the CHESS program, was a methodological extension
of the Hi-Vol TSP method. Thus, all errors associated with the TSP
method also affect the sulfate method. Subsamples were cut from the
exposed Hi-Vol filters and were analyzed for total water soluble sulfate.
Methods available for sulfate analysis at the time of CHESS deter-
mined all water-soluble sulfates as a class rather than distinguishing
them by chemical species. Two different methods were available for
total sulfate and both were used in CHESS. From November 1970
until September 1971, the manual turbidimetric method was em-
ployed. From September 1971 until July 1975, the methylthymol
blue (MTB) method was used. The methods are somewhat similar,
and are described in detail above.
The turbidimetric method is subject to interferences, many of them
being other common pollutants. In areas like the Salt Lake Basin
where the pollutants are dominated by a single source, the procedure
may be adequate. However, in urban areas like Cincinnati or New
York City, where the pollutant mix is derived from many independent
sources and is variable even within the city, the method is capable of
only the crudest estimates of sulfate levels. It should not be thought
of as an accurate measurement of atmospheric sulfate. Especiall3r,
small differences between High and Low exposure communities, such
as were reported in the Cincinnati Study hi the CHESS Monograph
(page 6-5) cannot be identified as real differences. When a realistic error
estimate is applied to the reported sulfate concentrations, the differ-
51
-------
40
i
ence becomes statistically insignificant. Any correlation of CHESS
health effects with sulfate levels where the sulfate data were obtained
using the turbidimetric method must be carefully qualified.
The MTB method is basically a better measurement method because
most of the aerometric interferences^ have been eliminated by its
revised methodology. The two remaining iriterferents, phosphate and
barium, are not normally found in atmospheric concentrations high
enough to cause inordinate problems. However, problems associated
with the sampling aspect of the method have been documented and
do impact on the general CHESS sulfate data quality.
First, problems associated with sulfate blanks (the level of sulfate
on the niter pad as manufactured) were reported to be high and
variable. In the 1971-1973 time period, problems of variable blanks
within the EPA NASN program were documented. The general
blank level was equivalent to an. atmospheric sulfate concentration
of 1-2/ug/m3. However, the major problem was variability of the
blank among manufactured lots of the filters. The blank level often
varied by more than 100 percent among lots so that routine and
continuous blank assessment should have been mandatory.
No evidence of routine sulfate blank determination was found in
the CHESS monitoring program until 1974. From that time period
on, adequate, blank assessment and correction were applied to the.data.
From 1971 until 1974 however, the blank contribution to the CHESS
sulfate data was not adequately assessed and consequently a positive
and highly variable bias of unknown magnitude was included in the
data.
Second, adsorption of atmospheric SO2 onto the fiberglass filter
material followed by spontaneous oxidation of the SO2 to sulfate had
been well documented. A 1966 publication by R. E. Lee and
J. Wagman provided results of their investigation of the problem. The
conversion was clearly documented with severe effects demonstrated
on four-hour samples. The conversion did appear to be an active-site
catalytic conversion that decreased in magnitude after an initial
saturation of sites. Thus, 24-hour samples were much less affected
by this problem than were those taken fpr shorter time intervals.
Even so, the paper by Lee and Wagman, presented data in which
routinely 0.5 to 1 Mg/m8 of the measured 'sulfate was derived from
SO2 conversion products. The maximum conversion presented was
2.1/itg/m3 derived from SO2; this constituted a 10 percent positive
bias of the sulfate data. A more realistic average bias is likely in the
5 percent range. However, there is clear evidence that in regions of
high levels of SO2, relative to sulfate, the positive measurement bias
becomes much more severe. This is probably the case in the Salt
Lake Basin area.
The third and most devastating problem associated with the
CHESS sulfate data occurred when the laboratory analysis of sulfates
was contracted to an outside firm. During this time period (October
1972-June 1974) the reported sulfate data underwent a sudden and
sustained decrease in apparent atmospheric sulfate level. Upon
investigation it was determined that the laboratory analysis of all
sulfate data from all CHESS sites were biased low by approximately
50 percent. The reason for this negative bias was and still is not
completely clear, but the continued dissemination of poor data was
clearly due to inadequate quality controls. An interim EPA report
52
-------
41
on a retrospective quality assurance evaluation of CHESS Sulfate
Data states:
A quality control protocol was designed for CHESS chemical analysis but has
not been implemented as per the contract .... The quality control protocol
should be implemented immediately.
In a series of following studies the magnitude of the affected data
and of the error were documented and an attempt was made to correct
and therefore recove-r the data. This type of procedure is difficult at
best and impossible in most cases. The validity of this data correction
was again assessed by the EPA Quality Assurance Branch. Their
finding was:
The basic question . . . is—How does one make bad data good? Whatever is
tried will be attacked for a multitude of (justifiable) reasons. Using the existing
data set for relative pollution level assessment will be acceptable, but statements
concerning absolute levels will not be. It would not be wise to submit these
data to the NADB,1 but rather answer all requests for these data internally.
Their statement gives a reasonable assessment of the CHESS
sulfate data between 1972 and 1974. The assessment of other year
CHESS sulfate data is more difficult. No comparative sulfate data
exists from the local agencies as it did for SQ2 and TSP. Based on
the intrinsic capabilities of the methods, and the error assessment of
the field use procedures, it -can -generallybe stated.that:
1. From 1970 to September 1971 the sulfate data were obtained
using the turbidimetric method. It should be used only as a sulfate
level indicator. Due to interferences, there will be severe problems if
an attempt is made to correlate sulfate levels in one part of the country
with sulfate levels in another.
2. From October 1971 until October 1972, the data are subject to
the following considerations:
a. The data are likely biased in the positive direction from
1-2 jug/m3. This bias may be more severe in areas of high 80s
concentration relative to sulfate.
b. The random error component of the measurement is probably
in the order of ±25% at an atmospheric concentration of
10 Mg/m3.
3. From October 1972 until June 1974, all CHESS sulfate data were
biased negatively by approximately 50% on an annual average basis
due to improper laboratory analysis by the contractor. These data
should be used only on an adjusted annual average basis to establish
local trends within site locations. The unknown cause of the bias
prohibits use of the data in shorter time structure (i.e., day, week,
month) increments.
4. From July 1974 until July 1975, CHESS sulfate data underwent
a marked improvement and was somewhat better than that collected
in the 1971-1972. era. The positive bias of the data is probably
similar to that of the earlier period but the random error component
was improved due to improved sulfate blanks on the TSP filters.
D. THE CHAMP AIR MONITORING PROGRAM
1. INTRODUCTION
Early in the execution of the CHESS program in 1969, a number of
staff members in the air quality measurements organization of EPA
1 National Aerometrlc Data Bank.
77-590—76 1
53
-------
42
decided it was desirable, indeed imperative, to improve the efficiency
and accuracy of short-terra air quality data monitoring coverage.
EPA coined the term CHAMP (Community Health Air Monitoring
Program) for this concept of a second generation automatic system of
air monitoring stations. Seven prototype stations were operated in
California from January, 1972 to February 1974. The manpower
ceiling placed on EPA resulted in a decision to contract for the devel-
opment, installation, and operation of the CHAMP S3rstem. A con-
tract for the development of the CHAMP system was awarded in
February, 1973. The developmental monitoring s^ystem was to contain
the newest technology in monitoring instrumentation. Accurate
measurement of all critical air and liquid flows in the system was
incorporated to enhance the accuracy of the system. The development
continued to mid 1974 when the first station systems were installed
in the Los Angeles area for field evaluation.
2. SYSTEM DESCRIPTION
The CHAMP air quality measurement system assembles the avail-
able discrete pollutant measurement devices and associated meteor-
ological instruments into a complete system in an air-conditioned
portable building. EPA specified the pollutants, to be measured and
selected the instruments with the advice of t.he CHAMP contractor.
All data are recorded digitally in a 'mini-computer integral to each
system. The data are checked and stored on tape at each CHAMP
site for transmittal to the EPA/RTP Laboratory at Durham, North
Carolina. S03 and NOj, and TSP measurements are also taken peri-
odically using older CHESS-type bubblers and Hi-Vol sampler
instruments described previously for backup and validation of the
CHAMP instruments. These bubbler and filter samples are sent to
the contractor's chemical laboratory in California for analysis.
All the CHAMP systems measure ozone, total gaseous sulphur
N0/N02, TSP,/KSP combinations, temperature, wind direction and
velocity, and humidity. Selected systems also incorporate CO and hy-
drocarbon sampling. The CHAMP system while automatic in principle,
requires periodic calibration and servicing by an operator to maintain
a high duty factor and an acceptable quality of data (less than 15%
error band). The operator repairs and adjusts instruments as required,
checks for failures, and does periodic calibrations and data verifica-
tions. A quality assurance specialist continually spot-monitors the
CHAMP sites carrying-out calibration and quality checks.
It should be noted that the instrumentation of the CHAMP
stations is not completely uniform. Some stations do not have wind and
pressure instruments; not all have CO and hydrocarbon instruments.
The manner in which meteorological data from the CHAMP stations
is being analyzed and used has not been investigated. This is a subject
of interest depending on the future of the CHAMP program.
CHAMP stations were visited in Thousand Oaks, California, and
Salt Lake City, Magna, and Kearns, Utah. The kind of meteorological
instruments in use appeared to be appropriate and they appear to be
well-located and properly maintained. Problems have occurred with
new dew-point measuring equipment that is now being replaced (this
has to do with humidity measurement. Except for occasional failures
of the sensing element of the dew-point apparatus, collecting meteoro-
logical data from the CHAMP stations should be routine.
54
-------
43
There are at present 18 CHAMP stations on. line at locations se-
lected by EPA; six in the, Los Angeles Basin, three in Birmingham,
Alabama, four in New York City, four in the Salt Lake Valley, and
one at the EPA Health Effects Research Laboratory at Research
Triangle Park, North Carolina. ;
S. FINDINGS REGARDING THE CHAMP PROGRAM
As in the CHESS program, all the instruments incorporated in the
CHAMP station were developed by the manufacturer for laboratory
use, In fact, some non-commercial instruments were selected by EPA
to try to use the most advanced technology. The CHESS experience
has demonstrated the need for validation in field use and ike con-
tractor appears to be attempting to do this.
There was apparently some attempt to standardize on one instru-
ment manufacturer for ease of maintenance, etc. Bendix ozone and
NOX instruments were employed. Flame photometric measurement
was selected for $>Q2 EPA apparently was interested in a pulsed
flourescence device but the equipment cost was too high for the budget.
The present instrument actually measures total gaseous sulphur and
it is assumed that this is S02. (The only other likely gaseous sulfur
compound HZS, does not seem to be widely present.) The rest of the
measurements appear to be well-validated. The backup measurement
with bubbler methods have validated NO2, to the extent possible.
The TSP/Hi-Vol measurements were apparently validated at the
beginning of the CHAMP program. However, because of the non-
linear calibration character of the flame photometric instrument in
the low concentration ranges of interest from Q to 50 jug/M3, calibra-
tion and range setting by the operator still ^results in 5% to 15%
range of uncertainty in the total sulphur readings. Further, while the
West-Gaeke bubblers used to check CHAMP SO2 are stored at 70°
F at the sites, they are shipped to the contractor's facilities for analysis
without temperature control and are subject to the unpredictable
temperature dependent decay of solutions prior to analysis. Thus, the
SC>2 validation in the CHAMP system may be in greater error than
EPA expects.
The execution of the CHAMP program has yielded validation
and quality control of field measurements better than CHESS. How-
ever, there are clearly numerous unresolved problems with the opera-
tion which have led to delays in validating the data bank and which
require high level attention for resolution oefore reliable quantitative
aerometric data can be obtained.
The data processing was 2,900 data-days behind at the time of this
investigation and no date agreed on for total backlog elimination.
Drift of zero setting and data span of instruments have invalidated
part of the earlier analyses. The data are only about 60 percent
machine validated. Field operator problems have arisen possibly due,
in part, to a lack of standardized operating procedures. Successful
operation of the CHAMP system requires well-trained instrument
technicians, and people of this high level of skill have not been em-
ployed in the past. Because of such circumstances, the SOs data
obtained through 1975 have been lost and apparently are not
recoverable.
55
-------
44
Some months ago EPA found that significant data were lost in
transmitting over "leased lines to the RTF laboratory. Thus, the
primary data source is the data tapes from the CHAMP site computer
which are mailed to ETP.
The CHAMP contract is up for renewal in November 1976 and the
bids are being solicited competitively. It is believed that at this time
competitive bidding could be a destabilizing step in this program and
could delay the achievement of reliable routine data gathering another
year. On the other hand there are obvious advantages to open com-
petitive bidding. When system development is more nearly complete,
it would certainly be appropriate for competitive bidding to be
adopted. The competition should include quality control considera-
tions. Unfortunately, the EPA quality assurance group was not
consulted on the renewal request for proposal, although that group did
participate in evaluating proposals received.
4. STJMMAEY
CHAMP appears to be an improvement in real time field measure-
ment of air pollutants in comparison with CHESS. However, the
system is still not completely validated and may not be ready for
routine use for 6 to 12 months. Data should not be stored in an ac-
cessible data bank until it is validated. ;
The present best estimate of expected accuracy is ± 15 to 20% on
the CHAMP measurements. However, this will be a significant im-
provement over previous CHESS aerometric network mesurement
systems when and if it is realized.
56
-------
V. EEVIEW OF CHESS AIE QUALITY ANALYSIS
PEOCEDURES AND EESULTS
A. INTEODTCTCTION '•
This chapter presents the results of the investigative team's critical
review of the utilization of aerometric data in the analysis and data
modeling presented in the CHESS Monograph. The citations to
pages, figures and paragraph numbers are to the 1974 CHESS Mono-
graph. The findings are highlighted in terms of examples wherein
it appears that estimates have been extended beyond the range of
credibility, models have been misused, or miscellaneous errors of
various types have occurred which lead to misinterpretation or over-
interpretation of data or results of analyses.
B. USE OP ESTIMATED DATA
A serious weakness in the CHESS study was acknowledged in the
last paragraph on page 7-9, which refers to the Salt Lake Basin study
and the Rocky Mountain study. It is in part:
Several factors should be remembered when interpreting tlie results of the
lower respiratory disease studies ... a majority of the pollution exposure
data in both studies were estimated from emissions data.
This statement applies to one of the most important and contro-
versial paragraphs in the CHESS report, also on page 7-9, which
follows:
It is interesting to note that larger increases in total lower respiratory disease
and two of its components were observed in the High pollution community of
the Salt Lake Basin study than in the corresponding communities in the Rocky
Mountain study. Also, the mean annual suspended sulfate concentration was
higher in the High pollution community in the Salt Lake Basin study than in
the Rocky Mountain study; the opposite was true for sulfur-dioxide. This suggests
that increases in lower respiratory disease frequency are probably associated
with suspended sulfates rather than sulfur dioxide.
The paragraph summarizes the argument that exposure to sus-
pended sulfates over a period of years produces significant adverse
health effects.
Analysis of the background material leading to the conclusion
shows that it is derived from an interpretation of the relationship of
four-numbers all of which are estimated values. The sulfur dioxide
values are estimated from smelter emissions and the sulfate values
are estimated from estimates of sulfur dioxide in one case and esti-
mates of suspended particulate based on smelter emissions in the
other, assuming no difference in the ratio of sulfate to suspended
particulate in the communities, Kellogg, Idaho,; Helena-East Helena
and Anaconda, Montana; and Magna, Utah.
The "High pollution community of the Salt Lake Basin" is Magna,
Utah. It is less clear what is meant by the words "than in the Eocky
(45)
57
-------
46
Mountain study". However, this paragraph refers to the preceding
paragraph of the CHESS report, which speaks of concentrations, "as
low as 7.2 jug/m3 in the Kocky Mountain Study".
From this it can be concluded that reference is being made to con-
centrations of sulfates in Anaconda, Montana.
A comparison is being made, therefore, between average sulfur
dioxide concentrations and average sulfate concentrations in Magna
and Anaconda. The period of the records being compared covers the
years 19GS-1970.
From the preceding paragraph the values being compared may be
obtained. They are as follows: :
(The concentration values are given in microzrams per cubic meter, written as fig/m*]
Sulfur •
dioxide Sulfates
Magna : 92 15.0
Anaconda 177 7.2
Because of the methods xised for making estimates, the absolute
values of these concentrations are questionable. The next four sections
discuss these estimates.
1. ESTIMATED SULFUR DIOXIDE CONCENTHATION, 92 ;iG/M3 (MAGNA) .
The concentration value 92 jug/m3 for Magna can be obtained from
Table 2.1.A.14 or Table 2.1.A.16. It is based on the following estimated
values for three years: ;
Year: . • wr/w*
1970 ; 84
1969 _ 103
1968 __„ _ 90
Average : 92
These estimates of annual sulfur dioxide exposures were derived by
multiplying the yearly smelter emission for sulfur dioxide by the ratio
of the 1971 measured annual average sulfur dioxide concentration
(61.8 /ug/m3)'to the same year's sulfur dioxide emission rate (193 tons/
day). The last chapter established that these data could be off by
100 percent, probably on the low side.
61.8/193 =.320 (jug/m3)/(tons/day)
The emission rates used were as follows (page 2-37):l
Tons!
tr AW
Year: (soi)
1970 1 261
1969 322
1968 _ 281
In order to obtain the estimated sulfur dioxide concentrations, it
must be first assumed that the meteorological conditions for each of
the years 196S, 1969 and 1970, were identical to those conditions hi
> These rates of emission ere off by a factor of two. Tons of sulfur, not tons of sulfur dioxide, are listed.
Tliese values corrected should be 522,644 and 562 tons/day. However, this does not change the estimates of
sulfur dioxide concentrations, which depend on a ratio between measured 1971 concentrations and 1971
emissions, whatever they might be. Doubling the emission rate also doubles concentations estimated by the
application of a mathematical diffusion model (Page 2-23).
-------
47
1971. There was no presentation in the Monograph of the use of
climatological data to show that 1971 was similar to the other years,
an average year, or a generally representative year. Even if the
meteorological conditions for all four years had been identical, there
is still a problem because the year 1971, on which the estimates are
based is" not a normal year for smelter operations. Emissions were
zero, or practically zero for two weeks during July, and nearly zero
for sis weeks in July and August. Therefore, the emission/concentra-
tion ratio is deficient in showing the effects of the summer season,
when wind direction frequencies from the smelter to Magna might
have been less than during the remainder of the year. This suggests
that the average concentration of sulfur dioxide in Magna is likely to
have been slightly over-estimated, but it supports rather than changes
the conclusion that average concentrations of sulfur dioxide are less
in Magna than in Anaconda. Primarily this estimate is criticized
because it is not supported by climatological information.
Also it should be realized that the method used for estimating the
annual average concentration can result in ah incorrect estimate if
there is a significant background of sulfur dioxide from a source or
sources other than the smelter. Multiplying the emission rate of the
smelter by a factor assumes that all individual observational values
that make up the annual average can be multiplied by this same factor,
when actually only those values totally resulting from the smelter
emissions would be effected. The Salt Lake^City airport wind rose
(Figure 2.1.2) is probably not representative for estimating the
percentage of time that Magna is downwind from the smelter because
the smelter stack is at the base of the Oquirrh mountain range.
However, the frequency of west northwest and northwest winds at
the airport suggest that Magna is only downwind about 5%
o'f the time. Allowing for the effect of calm and variable winds, it
seems unlikely that Magna would be under the influence of the
smelter more than 10% of the time. It follows then, sulfur dioxide
values for only these hours would be affected. On the other hand,
if the smelter is the only significant source of sulfur dioxide, as may
be the case, then multiplying individual observation values of zero
concentration would yield only zero, and the procedure for estimating
yields a true result, assuming no change in meteorological or emission
conditions. Since the sulfur dioxide background in Magna is not
known, the error that could be produced by background concentra-
tions cannot be determined. Probably most of the sulfur dioxide
does come from the smelter, so this source of error is not significant.
2. ESTIMATED STTLPX7B DIOXIDE CONCENTRATION, 177 Mg/m3 (ANACONDA)
A paragraph in the right hand column of page 3-12 explains how the
average concentration of 177 jug/m3 for sulfur dioxide was estimated
for Anaconda for the period 1968-70 using sulfation plate data and
emission rates. However, the explanation is incomplete, because it
requires the 1971 emission rate of the smelter, which has been omitted
from the Monograph. Thus, the validity of the entire procedure is im-
possible to verify. Table 3.1.2., which fists the emission rates by year
begins with the year 1970. The ratio of 0.343±.253 (Mg/m8)/(ton/day)
was obtained by a very dubious procedure. To begin with, sulfation
plate data are of somewhat uncertain nature. The document "Air
-------
48
Quality Criteria. for Sulfur Oxides", U.S. Department of Health, Educa-
tion and Welfare, Public Health. Service, National Air Pollution Con-
trol Administration, Washington, B.C., January, 1969, pp 24-25
says that sulfatipn "candles" (and plates) give only "an empirical
estimate of the average concentration". It also sajrs '^results are influ-
enced by -wind movement and humidity" and that "the lead peroxide
candle provides intelligence on the oxidizable sulfur compounds in
the _ atmosphere which seldom can be directly related to sulfur
dioxide".
The CHESS Monograph paragraphs refer to sulfation plate data for
1965. The sulfation plate is a variation of the lead peroxide candle.
Developmental work on the plate was reported in the following
reference: Huey, N.A. "The Lead Peroxide Estimation of Sulfur
Dioxide Pollution" J. Air Pollution Control Association, Vol. 18,
pp 610-611, Sept. 1968. Consequently it is unlikely that sulfation
plates were in use in Anaconda in 1965.*
ndioxide~ from.
g^ate an empirical relationship must be used. For example, in the
elena Valley, Montana, Area Environmental Study, (EPA, Office
of Air Programs, Research Triangle Park, North Carolina, January
1972) the sulfation. values were converted to sulfur dioxide values by
means of the relationship: 1 mg SO3 per 100 cm2 per day is equivalent
to 0.035 ppm SO2. In the history of the use of lead peroxide devices,
there has , not been general agreement as to what ratio should be
used, and a belief prevails that sulfation candle or plate data are
conservative, i.e., that sulfur dioxide concentrations are sometimes
higher than indicated. Further, more information is needed concerning
the location of the station, or stations, in the Anaconda area, where
the sulfation data were obtained. In order to validate the Anaconda
sulfur dioxide data further work needs to be done.
In 1965 the annual average concentration of sulfur dioxide was
reported to be 80 Mg/m3 with an emission rate of 609 tons/day. Since
the 1971 emission rate is omitted from the report it cannot be compared
with the corresponding concentration of 286 Mg/m3. Assuming that
the 1971 emission rate is also on the order of 600-700 tons/day, then
there seems to be too great a difference between the 80 /xg/m3 con-
centration and the 286 Mg/na3 concentration. (Center paragraph, right
hand side, page 3-12.) J
The ratio 0.343 ±.253 has a large- error factor. The range is from
.090 to .597. If the low value is multiplied by the emissions for the
years 1968-1970, the following concentrations are obtained:
iTons per day]
Table 3.1.2 New value
Year (SOZ) Montana SDES
1971 ...... _____ . . . „ . . ........ __________________ Omitted 638
1970 ____ ...... ~ •• 635 856
M69 ......... ______________________ ........... _____________________ . 545 824
1968 ____________________________________________________ • ___ . _____ „ 367 C62
1967 _____________________________________________________________ 346 459
NOTE.— The omission of the 1971 emission rates maies it impossible to check the effect of using the
new value for 1971 on the estimated emission rates.
*Tbe chemical reaction for "candles" and "plates" Is the same.
{According to information recently received from the Montana State Department of Health and Environ-
mental Sciences, the emission rates listed for the Anaconda smelter are low.
60
-------
49
Estimated average
SOi emissions concentrations
Year (tons par day) (pg/m')
1970 „ „ 635 57
1969" • " - . '' 545 49
19631 I" U , 367 33
The average of these values is 46 jugM3. This concentration's
considerably less than the 92/tg/m3 value at Magna. However, in-
formation received from a representative of the Montana State
Department of Health and Environmental Services, suggests that
the 80 Mg/mS value and the 286 jug/m3 value -were measured in two
different locations in Anaconda, and that the 80 jug/m3^ value is too
low. This indicates that the estimated values of sulfur dioxide in the
table comparing Anaconda and Magna values are somewhat too^ low.
The estimates are further weakened by the fact that an assumption is
made that meteorological conditions during all of the year is identical
for all years. No supporting climatologieai information is presented.
Also, note that the Table 3.1.7 lists a sulfur dioxide concentration
of 177 Mg/m3 far 1971 instead of the 286 ftg/m? value obtained from
the Montana State Department of Health.
The procedure for estimating sulfur dioxide concentrations in
Anaconda seems unnecessarily crude, making the average concentra-
tion value for the years 1968-1970 uncertain. However, since the
reported 1971 values for Anaconda and Magna are 286 jtg/m3 and
61.8 Mg/ro3* and these values are the basis for estimates, it would
appear that it was fairly certain that there was more sulfur dioxide
present in Anaconda, than at Magna during the '68 to '70 period of
the CHESS studies.
8. ESTIMATED SUSPENDED SULFATE CONCSJNTEATION, 15 /Jg/ms (MAGNA)
The 15 Mg/m3 estimate is a double estimate since the sulfur dioxide
concentration data on which it is based is also estimated. The sulfate
value seems to be an average for the years 1968-1970. It is obtained
by using the following regression equation, which is found on page
2-39.
Magna—SS=0.09(SO2)+6.66
This equation is based on 1971 conditions.
It is of interest to note that with a zero concentration of sulfur
dioxide there would still be 6.66 MgM3 of sulfate, or approximately
half the average annual value reported on 1971, which was 12.4 Mg/m3
Further, 44% of the 15 jug/m3 of interest for ,the years 1968-1970 is
unrelated to sulfur dioxide concentrations. The Figures 2.4.2 and
2.4.4 suggest some lack of complete correlation between sulfur dioxide
and sulfate concentrations.
During the strike with zero sulfur dioxide concentrations, there
still is an appreciable amount of suspended sulfate. Also, a peak
value of sulfate occurred during the third week that does not corre-
spond with sulfur dioxide value behavior during the same p_eriod.
Similarly, tihe very large rise in sulfur dioxide that peaked in the
ninth week hardly shows in the sulfate values. Consequently, the
regression equation can be questioned because the reason for the
61
-------
50
sulfate values is not understood. What is the physical source of the
sulfates?
Since the sulfur dioxide concentrations used in the regression equa-
tion are themselves estimated, uncertainties in the sulfur dioxide
estimates are compounded in the sulfate estimates. Further, since the
source of a considerable amount of the sullate seems to bo not associ-
ated with the sulfur dioxide, it is not clear what effect the strike period
has on the estimates.
'The CHESS report lists the suspended sulfate concentration as
12.4 Mg/m3 hi 1971 and this is the basis for the estimate of 15 pg/'m3
for the 1968-1970 period. Observations of sulfate in Magna area
subseqxient to 1971 support the argtiment that average annual con-
centrations are in the neighborhood of 15 jig/ms, or that they are sig-
nificantly higher than reported for Anaconda.
On page 2-79, in Table 2.4.1, it may be noted that suspended sulfate
values for the High community do not follow the sullur dioxide con-
centrations, particularly for the Spring and Summer. This raises a
question about using sulfur dioxide as an indicator of sulfate, as was
done with the regression equation on page 2-39. (Median values for
the High community are: Sulfur dioxide, Spring 64, Summer 9, whereas
for suspended sulfate they are 8 and 7, respectively.)
Wind blowing from the smelter stack ito Magna would generally
cross a portion of the Great Salt Lake and, therefore, might carry
more moisture, thereby facilitating the conversion of sulfur dioxide
to sulfate. Perhaps this mechanism helps to account for the high
sulfato concentrations observed in Magna.
4. ESTIMATED SUSPENDED SULFATE CONCENTRATION, 7.2 /ig/m3
(ANACONDA)
The 7.2 Mg/m3 suspended sulfate value can be obtained from
Table 3.1,7, page 3-12, by taking an average of sulfate values for
three years, as follows:
Year:
1970 ;
1969 ,_ _
196S
Average 7; 2
These sulfate values are estimates, based on estimates of total sus-
pended particulate and an estimate of the ratio of suspended sulfate
concentration to total suspended particulate concentration, based on
results from East Helena and Helena, Montana, and Magna, Utah.
The same procedure was used for Kellogg, Idaho.
On page 3-11, in an attempt to explain how the suspended sulfate
estimates were made for Kellogg, it is stated that "Data observed for
Magna during the period January 1971-June 1972 indicated an average
ratio of suspended sulfate concentration to total suspended par-
ticulate of 0.159." Following this is the reference number "22," re-
ferring to National Air Pollution Control Administration Publication
No. AP-61, "Characteristics of Particulate Patterns 1957-1966."
This publication presents graphs of suspended particulate concentra-
tions for various cities over a ten year period. In it, suspended sulfates
are not mentioned, the time period is wrong, and there are no data
62
-------
51
for Magna; therefore, it must be concluded that the reference is an
error.
An obvious reference for this paper would have been the paper by
Marvin B. Hertz, et al., "Human Exposure to Air Pollution in Salt
Lake Communities, 1940-1971," however, it is not referenced. Perhaps
this was titie reference intended. Even so, the ratio 0.159 cannot be
obtained from the Hertz paper.
In the Hertz paper, page 2-11, Table 2.1.2, which gives CHESS
1971 Annual Averages for Magna, the suspended sulfate concentra-
tion is 9.6 Mg/in3 as.d the total suspended particulate concentration
is 53.9, which gives a ratio of 0.178. In Tables 2.1.5 and 2.1.A.16, the
following concentrations are given: TSP, 66 MgA&3, SS, 12.4 /ug/m3.
Here the ratio ia 0.188. Other ratios can be determined for various
time periods from Tables 2.1.A.4 and 2.1.A.5, but none of these is
0.159.
Note (page 3-11) that the unexplained ratio 0.159 for Magna is
used with the 0.063 ratio for East Helena to obtain the ratio 0.111
plus or minus 0.057 that is used to estimate suspended sulfate con-
centrations for Kellogg, and the 0.11 plus or minus 0.06 ratio for
Anaconda (page 3-13).
(Pages 3-8 and 3-9) Particulate emissions for East Helena are given
in two tables on pages that face each other. The headings of the'second
column in Table 3.1.4 should be "Emissions, Tons/year," not "Emis-
sions, Tons/day." i
On page 3-7 it is stated that estimates of stack emissions for both
particulate and sulfur dioxide for East Helena for the years 1941-1970
were provided by Asarco. Presumably the data in Table 3.1.3 tare
Asarco data. The source of the data in Table 3.1.4 is not stated.
The Offi.ce of Air Programs Publication No. AP-91, Helena Valley,
Montana, Area Environmental Pollution Study, gives more informa-
tion about the industrial complex at East Helena. This study was
conducted during the period June 1969 through June 1970. The
table below is from this study. ;
EMISSIONS FROM EAST HELENA INDUSTRIAL COMPLEX
[Tons per day]
Emissions
Company and operation
SOs production
Participates production
Seduced
Normal Maximum Reduced Normal Maximum
Asarco:
Sintering .
Smelting.... . .. . ..
Miscellaneous
Subtotal
Anaconda:
Fuming
Miscellaneous... . .....
Subtotal
American Chemet: Pigment production.
184.6
8.4
0)
193.0
13.0
13.0
0)
315.6
14.6
0)
330.2
13.0
13.0
0)
355.1
23.2
378.3
13.0
(*)•
13.0
C)
0.8
(0
w
.3
1 0
1.0
0.5
0)
w
.5
1 0
1.0
0.5+
<")
.5+
1.0
1.0
Total...
208.0
343.2
391.3
1.3
1.5
1.5+
< Negligible.
> The outside storage of concentrates contributes a significant but undetermined amount of participates.
' Emissions also occur during the slag charging and tha coal mill, but ro estimates have been made.
< Emissions occur when slag is dumped, but no estimate of their quantity has been made.
* Emissions are controlled by cyclones and bag filters with high collection efficiencies.
63
-------
52
It may be noted that ASARCO is only one of several particu-
late sources for the East Helena area. Fuming and other sing
processing activities of the Anaconda Co. are estimated to produce
1.0 tons per day of ^articulates, resulting in a normal total of 1.5 tons
per day, not a rate in the neighborhood of 0.3 tons per day as Table
3.1.3 suggests. Further, the _ total normal sulfur dioxide emission
rate in the preceding table is 343.2 tons per day, a considerably
higher rate than is given in Table 3.1.2. (i.e.3 1969: 221 tons/day;
1970: 239 tons/day).
On page 3-7, right hand side, is given an explanation of how the
data in Table 3.1.4 were used to obtain a ratio of total suspended
p articulate concentration to tons of particulate emitted per day for
Bast Helena. However, after giving this explanation, the estimates
of TSP in Table 3.1.5, that were used to make the suspended sulfato
estimates were not obtained by means of this ratio. They seem to
have been obtained'from the-particulate emission data in .Table 3.1.3,
.using-fee-fo&tei^-3^3-.^—(Trg^^)/{-to^
factor is not explained. The ratio that is explained never seems to
have been used. The suspended sulfate estimates are obtained by
multiplying the total suspended particxilate concentrations by the
factor 0.063, which is explained on page 3-8.
Both observed and estimated suspended particulate concentrations
are given in Ta.ble 3.1.4 and 3.1.5. It may be noted that the estimated
TSr values are used to estimate the suspended sulfate concentrations
and not the observed values for the years 1966 through 1969. In
1966, the observed value was 87 fig/in3, whereas the estimated value
is 114.2 Mg/m3. No explanation is given for rejecting the observed
values.
Data for Magna during the period January 1971-June 1972 indi-
cated an average ratio of suspended sulfate concentration to total
suspended particulate of 0.159. The available data for East Helena
indicated a suspended sulfate to total suspended particulate ratio
of 0.063±0.022 jug/m3. For Kellogg, the assumption has been made
that the ratio of suspended sulfate to total siispended particulate
is the average of these values, or 0.111 ±0.057. For Anaconda, this
value was rounded to 0.11 ±0.06. It is rmiltiplied by the estimated
concentrations of total suspended partieulate listed in Table 3.1.7,
to obtain the suspended sulfate values for each year.
The following table has been prepared from the Helena Valley
study, June through October 1969.
Station
1
2
3. ..
4 ..: ,
Average „
Location
Degrees
34
105
112
274
i
Miles
0.8
2.5
.4
4.5
particulate
108
74
59
62
76
• sulfate
3.5
3.7
4.4
2.9
3.6
Ratio
0.032
.05
.069
.047
.050
»With respect to the smelter stack,
64
-------
53
The data from stations 1 and 3, the stations nearest the stack, were
used to obtain a ratio range (0.037, pages 3-8), but for some curious
reason the available ratios from the Helena Valley study were not
used. The average ratio for stations 1 and 3 is 0.051.
The ratio chosen for East Helena, 0.063 plus or minus 0.022 (jttg/m3)/
Gig/m3), is not significantly different from that which might have been
obtained had more use. been made of the Helena Valley study, but
there is no basis for the assumption that the ratio of suspended sulfate
to suspended participate is similar in Magna, East Helena, Helena,
and Anaconda.
The dubious nature of using sxispended particulate concentrations
to estimate suspended sulf ate can be seen by comparing Figures 2.4.3
and 2.4.4. In the Low Exposure Community, the sulf ate level remains
low and nearly constant while the suspended particulate concentra-
tions fluctuate.
In_the High Exposure Community, the highest concentration of
susperide'd parfic'ulaBe "occurred on" thTfourth" week "whereas th~e~pea"k
sulf ate value occurred on the third week. On the fourth week, sulf ate
levels dropped. A corresponding drop in the sulfate levels does not
occur until the fifth week. Only during the last seven or eight weeks
do suspended particulate and suspended sulfate concentrations
fluctuate together. There may be some situations where suspended
particulate and suspended sulfate concentrations are well correlated.
Justification for assuming correlation in the Salt Lake Basin and the
Rocky Mountain communities is inadequately supported by scientific
evidence presented in the CHESS Monograph.
Further, the 7.2 Mg/m3 suspended sulfate estimate for Anaconda is
based on an estimate that comes from another estimate of suspended
partieulate values based on rates of emission from the smelter, touring
the period 1961—1962, the annual total suspended particulate concen-
tration was found to be 84.5 /*g/m3. In 1971, the average suspended
particulate level was observed _to be 52 jug/m3. By comparing the
observed total suspended particulate concentration with the par-
ticulate emitted from the Anaconda plant, a ratio of 9.1 ±2.3 (jug/m3)/
(ton/day) was determined. This ratio was multiplied by the particu-
late emission for Anaconda shown in Table 3.1.3 to estimate the total
suspended particulate concentrations for the years 1940-1970. This
ratio cannot be actually obtained from the data presented in the report
because particulate emissions for the year 1971 are not given, i.e.,
they are not listed in Table 3.1.3.
The basis for this ratio is unfounded since there are sources for the
suspended particulate other than the smelter emissions.
Although there are no actual sulfate observations from the Ana-
conda area included in the CHESS report there are some actual
observations of suspended sulfate versus total suspended particulate
available for the year 1971, that were obtained from the Montana
State Department of Health and Environmental Sciences. These
suggest that annual average suspended sulfate levels in Anaconda are
in the neighborhood of 4 or 5 MgM3, even less than the estimated
value (7.2 /ig/m3).
65
-------
54
There are also pronounced seasonal effects, with much higher values
in winter than in summer. The months pf February and April had
values of 7 and 9 M!/m3 whereas the months of July and August have
values of less than 1 ug/m3. Local heating emissions and relative humid-
ity may be significant factors determining the measured concentra-
tion as well as the smelter emissions.
5. ESTIMATES OF SUSPENDED PARTICT7LATE, SALT LAKE BASIN STUDY
On page 2-23 it is stated that "the number of sulfric acid plants
utilizing sulfur recovered from emissions have increased from one in
1940 to seven in 1971, and that air pollution control devices in the
form of baghouses, scrubbers, cyclones, and mist eliminators have
been installed. Such changes in the smelter operations would greatly
effect the ratio of suspended particulate to tons of copper produced.
Therefore, aside from the fact that there would be differences from
year to year because of meteorology, the procedure described in the
first paragraph, right hand column, page 2-24, for estimating sus-
pended particulate from copper production in tons for 1971, is highly
questionable.
6. ESTIMATES IN THE CHIGACO AND NEW TORK STUDIES
In the Chicago and New York studies suspended sulfate concen-
trations were estimated from suspended particulate concentrations.
In Chicago, the estimates were used to fill in data for some years when
no data were available. In the New York study measured values for
suspended sulfates for 1956-1970 were available from the Manhattan
121st Street station, and these values were used as citywide values.
The observed annual ratios of suspended sulfate to dustfall for New
York City were used to estimate the suspended sulfate levels in Queens
and Bronx. In Table 5.3.1 suspended sulfate levels for the Low Com-
munity (Riverhead) are listed as about 10 jug/m3 for the years 1961
through 1970. The basis for this estimate is not given, although it was
probably determined from the 1971 concentration, which was 10.2
In summary, it appears that some values, on which are based
important conclusions that sulfates may ;be harmful to health, are
estimated values.
C. USE OF MATHEMATICAL DISPERSION MODELS
The dispersion model shown in Figure 2.1.16 is incorrectly applied.
It was used in the Salt Lake Basin study to determine sulfur dioxide
contours around the smelter source and to >show that annual exposure
estimates obtained from the ratio of 1971 observed air quality to
1971 emissions were not unreasonably high or low. First, the contours
are incorrect because the model used does not take into account the
elevation of the terrain and the wind direction frequencies for the
Salt Lake City airport, which were used are different from those
affecting the smelter plume, which originates at the base of the
Oquirrh Mountains. Second, a^dispersion model is based on numerous
assumptions and applied in this way might be off by a factor of two,
or more. It does not make sense to use a model to check observations.
66
-------
55
The usual application, is to apply observational data to calibrate,
or verify, a model. A model such as the one used might have been
applied to show some sort of relative distribution of concentrations
across the Salt Lake Valley, however, it should not have been used
to justify estimates of concentrations over the period 1940-1970.
(See Tables 2.1 .A. 14 and 2.1.A.16). Further, during this review of the
CHESS report it was discovered that smelter emissions used for the
model estimates were tons of sulfur, not tons of sulfur dioxide. There-
fore, the model estimate is only half what it should have been. Doub-
ling the emission rate and reducing the wind direction frequency
somewhat with respect to Magna might result in an estimated con-
centration near that measured, which was 61 Mg/m3-
Apparently the dispersion model was run only once and then the
ratio between the emission at the smelter for 1971 and the calculated
concentration was applied to emission values for the other years in
order to obtain the other listed concentrations in the column headed
"Diffusion Model". No account is taken of the fact that meteorological
conditions, or perhaps stack conditions, were not the same for all
years. More information should have been included in this report on
exactly what meteorological data were used in the model. The model
requires the use of the STAB program, which is obtained from the
National Climate Center. Frequently the results of running this
program are based on data for the year 1964, which is the only year
when wind directions were punched on data cards to the nearest 10
degrees each hour rather than each 3-hours. Therefore, the model is
likely to have incorporated meteorological data for some year other
than 1971, the year of the emission data. No. at tempts is made to
show that the year (or period) of the meteorological data is average,
good or bad. Similarly there is no attempt to show that 1971 was an
average year, yet all of the estimates are based on this assumption.
Considering how the model estimates for the years 1940-1970 were
obtained it is misleading to include them in the table, and they serve
little purpose since the ratio for the year 1970 is repeated throughout.
On page 2-43, bottom of right hand column, the following state-
ments appear: "Estimates of sulfur dioxide, total suspended paitic-
ulates, and suspended sulfate concentrations in the High exposure
community for 1940-1970 and the Intermediate II exposure com-
munity for 1950-1970 were obtained by a mathematical dispersion
model, which utilized^ emissions from the industrial source and exten-
sive local meteorological data, and by observed relationships among
pollutants. Observed suspended particulate, suspended sulfate, and
sulfur dioxide concentrations for 1970-1971 were used to calibrate the
models used to estimate exposure levels for previous years." This is an
overstatement. The estimates were obtained from simple ratios and
the application of a regression equation. See page 2-39. The model
was only applied once to demonstrate that annual exposure estimates
obtained from a ratio were not unreasonably high or low.
In the Chicago study, another attempt was made to apply a dis-
persion model (Figure 4.1.10). This^ model gives a false picture of
pollution conditions that prevailed in the study area because it is
based only on pollution sources within the city limits of Chicago,
omitting effects of adjoining large industrial sources in Indiana and of
some suburban communities to the southwest of the Loop area, which
have considerable air pollution.
67
-------
56
Maps recently published by the Chicago Department of Environ-
ment Control, for the years 1970 and 1975 clearly show that pollution
concentrations are not simply concentric around the urban core as
the model indicates.
On page 4-8, it is stated Measured data from the City network,
from which the exposure estimates were made, were best supported
by the -Mitre model. It is not clear why a greater use was not made of
the available actual measurements instead of the model estimates.
Also, it is not sufficiently clear why the model happens to be for the
year 1968.
68
-------
VI. AN ANALYSIS OF THE CHESS HEALTH
EFFECTS STUDIES
A. GENEHAL PROBLEMS OF EPIDEMIOI.OGIC INVESTIGATIONS OP
POLLUTION EFFECTS
Before discussing health effects problems specific to CHESS, some
discussion of general difficulties inherent to pollution epidemiology
may be helpful.
Exposure to suspect pollutants is not controlled in population
studies. Indeed with, current technologies, it is not possible to be sure
that the correct pollutant is even being measured. Combinations
of pollutants may be more harmful than any single pollutant, and the
number of studies needed to investigate such synergisms (interactions)
increases rapidly with the number of pollutants under consideration.
The analysis of synergisms is often impractical since sites with the
needed configurations of pollutants are seldom at hand.
Not" only "is "exposure uncbTItS'olled; it isr often "difficult to measurer
Even when aerometric measurements are valid, special meteorologic
conditions or personal habits may cause a given subject to experience
pollution levels very different from those measured at a nearby fixed
monitoring station. These problems are exacerbated in long term
studies during which the quality of aerometric data has bee^x variable
and individuals have changed jobs and residences. Aerometric methods
for measuring hourly or daily pollution levels are often less reliable
than required for studies associating pollution levels with short term
health effects.
The health measurements are often subjective responses to a ques-
tionnaire or interview. An individual may give one answer on a self-
administered questionnaire and another to a friendly interviewer.
Other factors, such as the public announcement of a pollution alert,
can also influence subjective health measurements. Some health
measurements, such as pulmonary function tests or,blood analyses, are
less influenced by jpoorly defined conditions surrounding the meaure-
ments and are said to be objective. However, even objective end-
points respond to uncontrolled events like an undetected influenza
epidemic or high pollen count.
Whether the health measurement is subjective or objective, the
response is often affected by factors (covariates) associated with the
subject studied and unrelated to pollutant exposure. Whether tihe
individual smokes or is subjected to cigarette smoke at home or work
is a covariate of dominant importance in pollution studies. Educational
attainment may affect responses to questions about phlegm or pneu-
monia. Occupation, age, sex, race, immunity to influenza, allergy,
access to air-conditioning and countless other covariates complicate
the interpretation of epidemiologic data. Epidemiologists treat
covariates in two ways. They try to choose study populations which
(57) !
77-590—76 5
69
-------
58
have similar covariate characteristics so that health differences
between such populations can be ascribed to pollution effects. Al-
ternatively, they make mathematical adjustments to nullify the
effects of covariate imbalances. Both strategies have weaknesses,
and neither works if the investigator is 'unaware of an important
covariate or has failed to measure it.
The epidemiologist has little control over the subjects studied. He
cannot assign them at random to reside in polluted communities of
interest. Thus, a clean town may contain many asthmatics because
asthmatics have wisely chosen to live there rather than in a more
polluted community. This fundamental problem of self-selection must
qualify any conclusions obtained from non-randomized population
studies: it may be possible to demonstrate temporal or spatial associa-
tions between health and pollution measurements, but a causal rela-
tionship cannot be inferred on the basis of a single epidemiologic study.
Students of'pollution counter these weaknesses in several ways. One
strategy is to replicate an epidemiologic 'study in a variety of cir-
cumstances and serially in time. If a consistent association between
pollution and health measurement is .observed, it is held to be reliable
since covariate imbalances and problems of self-selection are unlikely
to affect all sites and to persist over time. Clinical stiidies, in which
healthy volunteers are subjected to controlled pollution exposures,
and toxicological studies, in which animals are subjected to various
combinations and doses of pollutants, complement information ob-
tained from epidemiologic studies. This body of information from
clinical and toxicological studies and from several epidemiologic
studies may substantiate an interesting association suggested by the
health and pollution measurements of a single epidemiologic study.
In addition to these general issues, several questions directly perti-
nent to the CHESS health measurements;were examined, namely:
(1) Was the health measurement a reliable and meaningful
indicator of public health?
(2) Was the statistical analysis sound and impartial?
(3) Were the methods used to ascribe specific health effects to
specific pollutants and to establish dose-response relationships
logically compelling?
The sources of information used in this assessment include:
(1) The CHESS Monograph cited previously which contains
data gathered in 1970-1971 as well as some earlier studies.
(2) Preliminary internal EPA drafts of 1971-1972 studies.
(3) External peer review documents.
(4) Interviews and progress reports of contractors and EPA
personnel who gathered health and aerometric data.
(5) Interviews with EPA personnel who gathered and analyzed
the data.
Time limitations prohibited a complete reanalysis of the primary
health and aerometric data.
B. INTRODUCTION AND DEFINITIONS
What follows is a detailed discussion of each type of study found in
the CHESS Monograph. Each health measurement is considered
together with_ special problems associated with particular CHESS
sites. Comparisons across CHESS sites and comparisons with sub-
70
-------
59
sequent CHESS studies are made (when possible). A relatively non-
technical summery follows the discussion of each health measurement.
Unfortunately, much of the terminology is specialized. A short glossary
of statistical terms is given below for those who wish to read the
detailed assessments. This glossary is intended to convey the purpose
and use of certain statistical terms rather than to give precise defini-
tions, which are available from standard statistics texts. The defini-
tions are particularized to pollution applications.
Adjustment Procedure.—This is a mathematical method, based on
an assumed model, to compare populations exposed to various pollu-
tion levels when other factors which might affect health measurements,
such as age, sex, or race also differ among communities.
Autocorrelation.—This term describes the way in which the asthma
attack rate on a particular day depends on the attack rate on other
days. Pollution levels, which are also time series, also have auto-
correlation structure, since a pollutant which is elevated above usual
levels on one day is likely to be elevated on the next day also.
Correlation.—Correlation is a measure of the strength of the linear
relationship between two quantities. Positively correlated quantities
tend to rise and fall together, whereas negatively correlated quantities
tend to rise and fall out of phase.
Goodness-o/-Fit Statistic.—This quantity is used to gauge how well
a mathematical model fits a set of data. !
Least Squares Method.—This is a method of estimating the param-
eters of a mathematical model so that the final estimates bring the
model into closest possible congruence with the data.
Linear Model for Categorical Data.—Linear models are models in
which the average value of the observation (dependent variable) is
assumed to be a linear function of independent variables, such as
pollution category, age or sex. If the expected frequency of a cell in a
contingency table is assumed to be a linear function of such inde-
pendent variables, one has a linear model for categorical data. The
value of such a model, if it fits the data, is that it enables one to
estimate the relative contributions of the various independent
variables.
Multiple Regression Analysis.—Multiple regression analysis is a
method for relating an observation (dependent variable) such as
asthma attack rate to several independent variables, such as sus-
pended sulfate levels, sulfur dioxide and perhaps other pollutants.
Usual applications require that the expected value of the observation
is a linear function of the'independent variables, and that observa-
tions are statistically independent with constant variability about the
expected value. A multiple regression model can be used to describe
a given set of data and to predict the outcome of future observations
for known values of the independent variables. However, the _fact
that a multiple regression model fits a given data set well or yields
good predictions of future observations need not imply that the inde-
pendent variables determine the observations or bear any causal
relationship to the observations.
Pooled.—Pooled data is data obtained by [combining data from
two or more samples and then ignoring the fact that the combined
data came from various sources. •
Residuals.—Kesiduals are differences between observed data and
values expected under a mathematical model., Kesiduals are helpful
in assessing the appropriateness of the model.
71
-------
eo ;
Ridge Regression.—If independent variables are highly correlated,
estimates of the effect of a particular independent variable on the
dependent variable can be greatly influenced by what other inde-
pendent variables are included in the multiple regression model.
Ridge regression is a method of determining the importance of a
particular independent variable in the presence of other highly cor-
related independent variables.
Statistical Significance.—A statistical hypothesis is tested by com-
puting how probable a set of data is under that hypothesis. This
level of probability is called the significance level and a level p<.05
means mat the chances are less than 5 in 100 that the observed
outcome could have occurred if the hypothesis were true. The signifi-
cance level is often taken as a measure of the strength of evidence
against the hypothesis, with smaller significance levels providing
more evidence against the hypothesis, buppose, for example, the
hypothesis being tested is that two populations have the same risk of
chronic respiratory disease. This is called a "null" hypothesis—that is,
there is no difference. If, in view of the data, this hypothesis is found
to be highly improbable (that is, to have low probability of occur-
rence) one rejects the hypothesis (of no difference in risk of disease
between populations) and assumes instead that there is a difference in
risk of chronic respiratory disease between the two populations.
C. SPECIFIC HEALTH EFFECTS STUDIES
1. CHRONIC RESPIRATORY DISEASE (CUD) PREVALENCE
CHESS estimates adult chronic respiratory disease prevalence by
means of a self-administered questionnaire which inquires whether
the subject coughs and produces phlegm for at least three months of
the previous year. These studies compare CUD prevalences in ad-
jacent communities with different pollution levels, often designated
"High," "Intermediate" or "Low." The Monograph reports on data
gathered in the New York City and Salt Lake City areas in 1970-1971
and also includes studies from the Chickgo area and five Rocky
Mountain communities. CHESS monitoring stations were never
present in Chicago or the Rocky Mountain communities, and ex-
posure estimates were based on local data sources and theoretical
extrapolations in these areas (see Chapters IV and V). CHESS
monitoring stations onlv became operational in Salt Lake City in
December, 1970. Attention in this investigation was confined to the
four surveys, which wer.e reviewed by Chapman et al,1 and to New
York City CHESS follow-up data 2 from 1971-1972.
The CHESS CRD questionnaire was adapted for self-administra-
tion from an interview-administered questionnaire used by the British
Medical Research Council.3 Although a similar questionnaire had
been validated for self-administration in a 1971 Japanese study,* the
i Chapman, E. S., Shy, C. M., Finklea, J. F. House, D. E., Goldberg, H. E., Hayes, C, G., "Chronic
Respiratory Disease in Military Inductees and Parents of Schoolebildren.'T.4rcAise» of Environmental Health. ,
Volume 27, Number 3 (1973), pp. 138-142.
> Galke, W. A., House, D., "Prevalence of Chronic Respiratory Disease Symptoms in New York Area
Adults, 1972." EPA In-House Technical Report (June 7,1976).
s "Standards for Epidomio'.ogic Surreys in Chronic Respiratory Disease." National Tuberculosis and
Respiratory Disease Association. (1969).
4 Tsunetosht, Y,, Shimiza, T., Takanashi, E., Ichinosava, A., TJeda, M., Nakayama, N., Yagamata, Y.,
"Epidamtological Study of Chronic Bronchitis with Special Reference to Effect of Air Pollution." Int
Arch Arbeitsmed. Volume 29 (1971), pp. 1-27.
72
-------
61
Monograph does not contain validating data for U.S. surveys. An
internal EPA report8 dated 1973 showed that the self-administered
questionnaire detected only three of nine cases of chronic bronchitis
found through interview. Survey supervisors noted that the word
"phlegm" was not understood by some subjects, suggesting^ that
educational attainment affects CRD response. However, even if the
CHESS questionnaire underestimates CRD prevalence in a given
community, it may give a reliable indication of the difference in CRD
prevalence between two communities of similar educational
attainment.
The Bronx, Queens and Riverhead, Long Island represented two
Intermediate and one Low level of urban pollution in the New York
City area. The Intermediate regions< exceeded1 Riverhead in total
suspended particulate (TSP), sulfur dioxide (SOz), suspended sulfate
(SS) and suspended nitrate (SN). Generally,' questionnaires were
given to school children who gave them to their parents to fill out.
Thus, the sample is not representative of adults in general but only
of parents of school children. Only 73% of Riverhead parents re-
sponded to the questionnaire which probably reflects the fact that a
large number of Riverhead questionnaires were mailed to the parents
rather than distributed through the children. Only white respondents
who provided sufficient questionnaire data and' who had stable
residence histories were included in the analysis. Queens was shown
to have the highest income and educational level and unpublished
data showed Queens to have a larger proportion of Jews than Bronx.
These facts may account for the higher CRD! prevalence rates re-
ported for the Queens than Bronx. No formal adjustments were
applied for these socio-economic covariates. While smokers and non-
smokers were treated separately, no covariate adjustments were
made for exogenous exposure to cigarette smoke at home or work or
for occupational exposure (which affected less than 1% of respond-
ents). The statistical methods were sound. Sex- and smoking-specific
prevalence rates were consistently lower in. Riverhead than in Queens
or Bronx, the difference being 5% typically. For comparison, smokers
had prevalence jrates nearly 10% above non-smokers in the same
community. A linear model for categorical data 6 was used to make
smoking-, age-, and sex-adjusted tests for community differences in
prevalence' rates, and the differences were statistically significant.
A parallel analysis using severity scores for; CRD confirmed the
analysis of prevalence. While the statistical methods were appropriate,
additional information would be helpful. In 'particular, confidence
intervals on the prevalence rates in ^Table 5.2.7 (unless otherwise
indicated, figure, page and Table citations refer to the CHESS
Monograph) and observed and expected rates under the linear model
of Table 5.2.8. would allow the reader to make specific prevalence
comparisons and to verify that the linear model held for all
subcategories. '.
Communities in the Salt Lake region provided an opportunity to
study the effects of sulfur oxide (SS, SO2) smelter emissions. Other
urban pollutants, including SN and TSP, were moderate or low.
Magna was exposed to intermittent smelter fumigations of SO2 and
' House, D., "Eaaabflity of the CHESS School and FamUy Health Questionnaire." EPA Haman Studies
Laboratory Internal Kaport (April 12,1973).
1 Grizzle, I.E., Starmer, C.F., Koch, G.G,, "Analysis of Categorical Data by Linear Models."
Volume 25, Number 3 (September 1969), pp. 489-504.
73
-------
62
had the highest average S02 and SS levels of these pollutants, and
Ogden was designated Low. Aerometric copper data was not available.
The qxiestionnaire was distributed through elementary school children
and mailed to parents of high school students. Response rates of
85%.and 35% were found for child-carried and mailed questionnaires
respectively. Although the 65% nonresponse rate to mailed question-
naires admits the possibility of serious reporting bias, the authors
assure the reader that inter-community CRD differences (presumably
similar) were observed for both sets of parents. Respondents were
excluded for incomplete questionnaires, for >a residential change within
the previous two years, and for occupational exposure to irritants
such as coal dust, cutting oils, asbestos, jnine dust, smelter fumes,
cotton dust and foundry dust. Subsequent analysis showed that the
occupational exclusion gives a conservative estimate of. effects
attributable to pollution. All races were included, but the proportion
of black respondents was trivial. No covariate measurements were
made to assess religion, exposure to exogenous cigarette smoke at
home or work, or racial composition, although Salt Lake City has
proportionately fewer Mormons, and Magna more Spanish Americans.
Educational attainment was comparable; in the four communities.
CRD prevalence rates reflected pollution levels faithfully, and rates
in Magna (High) exceeded those in Ogden (Low) by 2-7%, depending
on sex and smoking status. These differences were found to be
statistically significant using the sex, smoking and age adjusted
linear model.7 For comparison, differences in CRD prevalence
attributable to smoking and to occupational exposure were 10-20%
and 2-8% respectively. Thus, air pollution and occupational exposure
were associated with comparable increases in CRD prevalence, and
personal smoking habits seemed to be a more important determinant.
The exposure data from the Rocky Mountain communities were
least adequate. Two smelter communities, Kellogg and Anaconda, had
relatively ^ high levels of SO2. East Helena and two non-smelter
communities, Bozeman and Helena, were classified as Low. Ambient
zinc, copper and lead were not measured. Nearly 85% of the ques-
tionnaires distributed by elementary school children were returned.
Respondents were excluded for occupational exposure, but no
residence duration requirement is mentioned. Over 97% of respondents
were white. Low communities were better educated than High
communities, and most occupational exclusions were from the latter,
but these factors tend to reduce the apparent pollution effect.
Conversely, the increased residential crowding in High communities
might increase apparent pollution effects. Sex-, education-, smoking-
and age-adjusted prevalence rates were statistically significantly
higher in High than Low communities (Table 3.2.6). High com-
munities had CRD prevalence rates 1-3% higher than found in
Low communities (Table 3.2.5). Corresponding differences associated
with smoking and occupational exposure were 10-16% and 0.1-3.6%
respectively.
The Chicago studies compared urban (High), suburban (Inter-
mediate) and clean surrounding areas of Illinois and Indiana (Low).
Local SOz, SS and TSP measurements were used, and other urban
pollutants were not measured. Gary, Indiana was included in urban
' Grizzle N. E., at al., op. clt.
74
-------
63
and may represent a special industrial pollution hazard. A CUD
questionnaire designed by Julius Goldberg was administered to
recruits reporting to the Chicago induction center. This population
obviously differs from the previous^ three and contains only young
male adults. Some with serious respiratory problems may have been
exempted prior to induction. To be analyzed, the questionnaire had
to be adequately filled out and the respondent had to have lived at
'his present address for at least three years. Statistically significant
age-, race-, smoking-, and education-adjusted community differences
in prevalence rates were found for black smokers and white non-
smokers (Table 4.2.8). However, the significance levels for black
smokers and the adjusted rates in Table 4.2.7 are suspect since the
linear model does not fit the data well for black and white smokers.
(The "adjusted" rates in Table 4.2.7 are really expected rates
under the linear model, which is a useage unfamiliar to epidemi-
ologists.) Race and smoking specific prevalence rates tend to be
0.4-3.6 percent higher among urban than Low communities, while
smoking" effects are 0.3-3.5 percent for blacks and 12.4-15.1 percent
for whites.
To summarize, these four studies demonstrate that higher adult
CRD prevalence rates are associated with pollution in two urban
sites and in two smelter-exposed sites with high sulfur oxide emissions.
The magnitude of these differences is comparable to that associated
with occupational exposure but smaller than that associated with
smoking. Such differences were not seen in a 1970 CHESS study in
Chattanooga, comparing communities with different nitrogen dioxide
levels.8 New York CHESS studies in 1971-1972 showed no significant
differences in CRD prevalence among Riverhead, Queens Bronx and
Sheepshead Bay,9 and prevalence rates werejdramatically reduced
compared to 1970-1971 values. Perhaps subjects tire of the CRD
questionnaire and report less symptomatology on repeated sampling.
This would limit the usefulness of the CRD questionnaire as a health.
surveillance measurement. An alternate explanation is that decreasing
CRD prevalence reflects decreasing pollution in New York.10
These studies demonstrates that small CRD prevalence increases
are associated with pollution levels which are: moderate by historic
standards in urban and smelter sites. The Monograph^ contains no
formal methods for associating these effects with a specific pollutant
or for establishing a dose-response relationship. Any such inferences
are tenuous, especially since the aerometric data is of dubious quality
and completeness (See Chapters IV and V). The authors are to be
complimented for applying newly developed and appropriate statisti-
cal methods.
2. EETHOSPECTIVE StTBVEYS FOE ACUTE LOWER EESPIBATOEY DISEASE
(LED) IN CHILDEEN
The Monograph contains data from questionnaires asking mothers
to recall how many times each child under age twelve had had pneu-
monia, croup, or bronchitis during the previous three years. Data was
also gathered on related hospitalizations and doctor visits.
• Chapman, E.S., et. al., Prevalence of Chronic Respiratory Disease In Chattagnooga: ESeet of Com-
munity Exoosors to Nitrogen Oxides. EPA Internal Draft. July, 1973.
• Galfce, W.A. and D. House. PrsTalence of Chronic Eespiratory Symptoms in New York Area Adults-
1972. EPA Internal Draft. Jane 7,1976.
10 Galke, W.A. and D. House, op. cit.
75
-------
64
These questionnaires were distributed through, school children
together with the 1970 Salt Lake and Rocky Mountain CRD ques-
tionnaires. Children less than one year old, those with incomplete
questionnaires, and asthmatics were excluded from analysis. This
last exclusion tended to minimize estimates of pollution affects (See
Table 2.3.4). The sample is not a random selection from all children
aged 1-12, but represents only schoolchildren and- their siblings
Because the data is retrospective, validation is crucial. Nurses
examined medical records to document LRD (any croup, bronchitis
or pneumonia) detected by questionnaire. In Ogden (Low), 70%
of questionnaire-detected cases were confirmed, whereas 78% were
documented in Kearns (Intermediate) and Magna (High). These
differences would tend to reduce estimates of pollution effects. The
corresponding figures were SS% and 75% for pooled Low and pooled
High JRocky Mountain communities, which would tend to increase
estimates of pollution effects.
Parents smoked less in Magna than in other Salt Lake communities
(Table 2.3.3), which would minimize pollution effects. The Mono-
fiph presents a higher percentage of parental smoking in pooled-
gh Rocky Mountain communities (Table 3.3.2), but argues no
adjustment is needed.
The monograph does not specify how age-, sex-, and socioeconomic-
"adjusted" rates are computed. Since the epidemiolpgic interpretation
is predicated on these adjusted rates, some clarification, obtained from
EPA interviews, is given.
A linear modelfl containing community, age, sex, and socio-
economic main effects and no interactions was fitted separately for
the three residence designations hi Table 2:3.12. The "adjusted" rates
hi each column of this table are simply the sum of the estimated
overall mean-and community effect. Differences of rates within each
such column are estimated community effects under the linear model.
These comments may explain the anomaly that adjusted rates exceed
unadjusted rates hi every community, both for Salt Lake and Rocky
Mountain studies. A more reasonable "adjusted" rate would be
obtained by applying the total estimated linear model, including^ all
main effects, to the actual cpvariate composition of each community.
The "adjusted" rates given in the Monograph still provide reasonable
estimates of community effects, provided' the linear model used fits
the data well. The Salt Lake studies give no goodness-of-fit statistic
or analysis of residuals, and the model is riot even specified by Table
2.3.9. Thus the Salt Lake "adjusted" rates and significance levels must
be regarded as provisional. Although Rocky Mountain studies give
no analysis of residuals, the goodness-of-fit>statistics suggest the linear
model is valid.
If one assumes the validity of the linear model used, the Salt Lake
studies demonstrated statistically significant community differences
in three-year croup, bronchitis, and LTJD incidence rates after adjust-
ment for sex, age, and socioeconomic status (actually educational
attainment of the mother). Magna residents of three years duration
had LRD, croup and bronchitis rates which exceed corresponding
rates in Ogden by 10.9 percent, 9.5 percent, and 7.1 percent respec-
ively. However, these rates did not increase hi strict accordance with
11 Grizzle, J. E. et si., op. tit.
76
-------
65 .
pollution levels and Magna residents of fewer than three years actually
experienced the lowest croup, bronchitis, and LED rates. No signifi-
cant community differences were found for pneumonia or hospitaliza-
tion rates. ;
Pooled High Rocky Moxmtain communities had significantly higher
adjusted croup rates than pooled Low communities. No significant
differences (at p<.05) were found for LED, bronchitis, pneumonia,
or hospitalization. Those residents in. pooled Low communities for
fewer than three years experienced higher adjusted croup, bronchitis,
and LED rates than corresponding inhabitants of pooled High
communities.
The basic observation that children of families resident in sraelter-
exposed communities for three years or more experience higher croup
rates than those hi nearby communities may be valid. However, these
rates do not increase consistently with increasing pollution (Tables
3.3.A.2, 2.3.8, and 2.3.A.2), and recent migrants'to High areas have
lower croup rates than migrants to Low areas (Tables 2.3.8, 2.3.A.2,
3.3.A.2, and 3.3.8). These inconsistencies deserve further elucidation.
The documented insensitivity of the LED questionaire reminds us
that this endpoint is no better than a mother's ability to recall a
three-year illness history for each of her children and to remember
specific diagnoses, such as croup and bronchitis. Finally, the conclusion-
in the Monograph (p. 7-9) that "increases hi lower respiratory disease
frequency are probably associated with suspended sulf ates rather than
sulfur dioxide" is tenuous at best, since no formal methods of relating
health effects to a specific pollutant are described and since suspended
sulfate measurements are not available in Salt Lake and Eocky
Mountain communities for the years 1967-1970 surveyed by these
questionnaires.
3. ACUTE BESFIRATORY DISEASE (ABD) IN VOLUNTEER FAMILIES
The Monograph contains New York AED studies from 1970-1971.
Telephone interviewers made biweekly calls to mothers of families
enrolled in the study to inquire whether any family member had
developed upper or lower respiratory illness in the preceding two weeks,
and, if so, whether a doctor had been consulted and how many days of
restricted activity had eventuated. If an individual was reported to
have both upper and lower respiratory symptoms, his illness was
classified as lower respiratory disease. Thus, the least ambiguous
diagnostic category is "all respiratory illness." A 10 percent subsample
of families who had cooperated on the CED questionnaire was re-
cruited for the AED study. An eligible family had to have resided at
least the past year within 1.5 miles of a CHESS monitoring site, to
be white, to have one or more children age 12 or less, and to have a
working phone. Priority was given to families with many preschool
siblings, and in a subsequent protocol, such "priorities" were specified
in writing. These priority schedules introduce ambiguity into the
eligibility requirements. 'The major response variables were the num-
ber of respiratory illnesses per hundred person-weeks exposure (the
attack rate) and the severity score, which reflected physician visits,
fever, and restricted activity. The severity score scale lias been criti-
cized as arbitrary, and inversions for scores II and III in the upper
77
-------
66
respiratory column of Table 5.3.9 support this view. Table 5.3.4
(which has obvious misprints for Bronx data) shows that only 750
of 1,136 families in Riverhead participated for the full 32 weeks of
study, white 887 of 1,153 families did so for Queens. Selected inter-
views were repeated a few days after the initial call and concordant
results wero obtained in over 90 percent of 'those previously reporting
ARD and 93 percent of those previously denying ARD, This valida-
tion shows the interviews were reproducible, but it does not verify
the medical content.
The authors point out that the pooled Bronx-Queens community
has a higher proportion of adult smokers and of children exposed to
smoke at home than Riverhead. More cooking gas was also used in
Bronx-Queens. But no formal adjustments were made for these co-
variates. The percent of crowded families was 15 percent in Riverhead
and 17 percent ia Bronx-Queens, a difference the. authors judged to be
inconsequential. (Table, 5.3.14).
_ The analysis of these panel studies is complicated by the fact that
biweekly attack rates are statistically dependent, since the same
people are repeatedly surveyed. The same comment applies to severity
scores. Significance levels or confidence limits cannot be estimated
without an analysis of such dependence, which the Monograph does
not provide. In particular the Monograph does not specify the methods
used for computing significance levels in Tables 5.3.5, 5.3*7, and 5.3.15,
and the quoted significance levels are suspect. The formal statistical
properties of "attack rates" and "relative risks" are insufficiently
ciefined to allow us to calculate confidence intervals and perform
significance tests, even though these quantities may be useful de-
scriptive statistics.
Figure 5.3.1 shows that Riverhead attack rates are usually lower.
than either Queens or Bronx rates. It is inconsistent with the pollution
hypothesis that Queens has higher rates than Bronx for 13 of the 15
surveys shown. Perhaps this reflects socioeconomic differences men-
tioned in the CRD discussion.
Tabla 5.3.6 contains ratios of attack rates with the Low community
rate as denominator.
Bronx and Queens rates usually exceed Riverhead rates for lower
respiratory illness and all respiratory illness, but the relation is in-
constant for upper respiratory illness.- This difference may be an
artifact of classifying all those with upper and lower respiratory
symptoms as lower. For the unambiguous category of "all acute
respiratory disease," Table 5.3.6 shows the pooled Queens-Bronx rates
to exceed Riverhead rates for fathers, mothers, school children, and.
preschool children. Table 5.3.13 divides the populations shown in
Table 5.3.6 into those who have lived five or more years at their
present residence (stable) and those who have moved (mobile). It is
surprising that stable fathers and mothers actually experience lower
attack rates in pooled Queens-Bronx than Riverhead, and this in-
consistency warrants special consideration. Likewise confusing is the
result (Table 5.3.14) that crowded families have lower attack rates
than less crowded families.
The 1969-1970 Chicago ARD study compared Intermediate, High,
and Highest Chicago neighborhoods. Twenty monitoring sites run
by the City of Chicago were ranked in order of increasing 1960 TSP
78
-------
67
values. The lowest 10 were designated as Intermediate, the next five
as High, and ^ the five with highest TSP values as Highest. Such
designations did not correspond to SOj ranking (Table 4.3,1) and SS
was not measured for these sites. Families with 2-5 year old children
in day care centers located near monitoring stations were recruited.
The family was assigned the exposure classification of a monitoring
station in the same census tract, if any, but the Monograph is unclear
about exposure designations for families in unmpnitored census tracts.
Over 500 of the families initially enrolled participated throughout the
study. Biweekly telephone interviews were used to gather data for
attack rates and severity scores as in New York. Review of doctors'
records documented 95.3% of telephone-survey-idetected ARD.
The Highest exposure category contains more smokers and more
professionals than the two other categories. The Monograph gives
smoking-adjusted relative attack rates (Table 4.3.15) which show
this covariate does not alter conclusions, but the method of adjust-
ment is not presented. Although families of professionals appear at
higher risk than those of non-professionals (Table 4.3.13), adjustment
for this covariate was said to produce little change. The racial com-
position of the exposure categories is not given.
Figure 4.3.1 shows that Intermediate families experienced the
lowest attack rates during 20_of 25 biweekly surveys, but the largest
differences were evident during the three month influenza season.
One wonders why the first three surveys were omitted hi Chicago but
not New York. Severity score rankings show inversions in days of
restricted activity (Table 4.3.7) as in the New York study, which
argues against this scoring system. The relative risk data in Table
4.3.5 shows higher ABD attack rates in Highest than in Intermediate
exposure families for fathers, mothers, older siblings, nursery school
children, and younger siblings. This result holds both for mobile and
, stable (three year residency) families (Table 4.3.11), in contrast to the
New York study. The Monograph does not define Table 4.3.12
adequately. If the analyses were performed on attack rates, the
significance levels quo ted ^ are untrustworthy. If by "frequency" the
authors mean the proportion of families ever experiencing ABD dur-
ing the study period, then use of the linear model is justified, but more
explication is needed.
These two ARD studies have the advantage that health data were
gathered prospectively by telephone interview, so the recall of the
mother was not taxed as in the LRD studies. The quality of the
health data are probably adequate, and descriptive statistics, such as
"attack rates" and "relative risks" are no doubt useful guides. How-
ever, some mathematical statistical investigations to compute ^confi-
dence intervals and significance tests would be a valuable addition to
this methodology, especially since these problems are non-trivial. It is
likely true .that families exposed to high levels of urban pollution
experience higher ARD attack rates than those less exposed. However,
two surprising findings require further study, namely: (1) Why do
residentially stable fathers and mothers experience lower attack
rates in Bronx-Queens than in Riverhead? (2) Why do crowded
families in New York and less educated families in Chicago appear at
lower risk?
Summary statements linking specific SO2, SS and TSP levels to
excesses hi ARD attack rates must be viewed with customary caution.
79
-------
68
4. ASTHMA PANEL STUDIES
The 1970-1971 Salt Lake asthma panel studies attempt to corre-
late the daily proportion of panelists suffering an asthmatic attack
with daily pollutant levels. The studies were complicated by practical
and theoretical problems. Many relevant factors, including; medica-
tion (steroids), humidity, exercise, daily temperature changes,
nitrogen dioxide levels, copper levels, and exposure to smokers at
home or work were not evaluated, and the,study focused on TSP, SS,
SO2, SN (suspended nitrates), and temperature. Daily SO2 measure-
ments may be too unreliable for meaningful assessment (Chapter 3),
and Awhile midnight to midnight daily asthma rates were used, air
monitoring samples were obtained from 10:00 A.M. one day to 10:00
A.M. the next, introducing a lag.
Potential panelists, recruited from families who had responded to
the CRD questionnaire, or suggested by doctors or known asthmatics,
were interviewed to obtain covariate and eligibility data. To be eli-
gible, the asthmatic had to have been previously diagnosed by a
physician and had to give a history of two or more episodes of wheezing
and shortness-of-breath during the previous year. In addition he had
to live within two miles of a monitoring site. This residency require-
ment was violated from time to time, and1 in 1975, 38 of 222 asthma
panelists resided outside a 2.5 -mile-radius.12
Since a physician was not available to make the diagnosis of asthma
for prospective candidates, certain individuals with unlikely asthma
histories (such as elderly smokers with a sudden onset of "asthma")
were eliminated. While this second level of panelist selection may
have reduced the number of non-asthmatics, it introduces uncertainty
as to what the exact eligibility requirements were. The monograph
states only: "Highest priority for selection; was given to non-smokers
over sixteen years of age." In _1974,_ EPA staff suggested that ^an
improved initial interview qiiestionnaire was needed " for separating
asthma from other cardiorespiratory disease, but these suggestions
were not implemented. Asthma panelists were mailed weekly ques-
tionnaires and asked to record day by day whether no attack, one
attack, or more than one attack had d'ccurred. Some failed to return
diaries, and some filled them out for the wrong week. Those who con-
sistently failed to return diaries were eliminated from the panel, as
were those who never reported an attack and those who reported
an attack every dav.
The 1970-1971 "Utah CHESS Progress BeportM gives summary
asthma panel data indicating that mailed diaries are inconvenient.
During week eight, 183 'diaries were mailed out. Of these, 126 were
returned on time, and 30 additional diaries were returned after
telephone prompting. Ten calls were required to obtain or correct
particular entries. Of the 183 diaries sent out, 156 provided "useable"
data for this typical week. The mailed diary was discontinued in the
1973-1974 protocol and replaced by a weekly telephone interview
to gather diary data kept by panelists at home.
w Back, M. M. and P. Carl, Bureau of Environmental Epidemiology Teohnlcal Eeport Number 3. TTtab
State Department of Social Services. Division of Health. 1975.
» Calanore, D. Memorandum to CHESS Field Epidemiology Director. HSL., and Attachments, (Janu-
ary SO.1974).
« Utah CHESS Progress Eeport. October 22,1970 to Dacember^l, 1971.
80
-------
69
The Monograph deompliasizes covariates since no intercommunity
comparisons are made. Some community differences in age, sex,
education, and smoking habits are noted (Table 2.4.2).
Dropout rates varied by community. Magna, Kearns, Salt Lake
City, and Ogden had respective dropout rates of 13/48, 24/55, 20/46
and 14/46 during 43 weeks of study.13
The measured response was the daily proportion of panelists re-
porting an asthma attack. The daily composition of the panels (age,
sex, severity or illness) changes as panelists drop-out and are replaced.
Since some panelists continue throughout the study, daily asthma
attack rates are statistically dependent, which complicated the
analysis.
Figures 2.4.1-2,4.4 relate mean weekly attack rates for each panel
to weekly minimum temperature, weekly TSP, weekly SC>2, and
weekly SS. By superimposing these graphs one can see that Ogden
(Low) often has higher attack rates than Magna (High). The Mono-
graph makes no such comparison. Conspicuously absent are corre-
sponding graphs for Salt Lake City and Kearns.;
Sample correlations for daily attack rate with daily pollutant
levels are shown in Table 2.4.3. Sulfur dioxide has negligible positive
correlations with daily attack rate. Correlations for TSP are positive
in all sites' and negative for SN and imfiimiinxt-eniperatttre in all sites..
Suspended sulf ates are negatively correlated with daily attack rate in
three of four sites. These correlations indicate that SS, SN, and mini-
mum temperature are usually low when daily attack rates are high.
Quite the opposite was observed in New York'(Table 5.4.3.) where
SOa shows negative correlations in two of three sites, SN is positively
correlated in all sites, and SS is positively correlated in two of three
sites. While these conflicting correlations may have descriptive value,
the significance levels in Tables 2.4.3. and 5.4.3. are untrustworthy
since daily attack rates are statistically dependent.
This remark applies to the multiple regression analysis (Tables 2.4.4.
and 5.4.4.). Researchers at EPA and collaborators have recognized
that both the daily attack rates and pollution levels are time series
with autocorrelation structure, and they have begun to explore
alternative analyses. Stebbings 16 noted that Autocorrelations affect
multiple regression parameter estimates and mentioned the need for
methods which could detect lagged relationships between pollutant
and attack rate.
Bloomfield" responded by suggesting an analysis of spectral
coherence to relate attack rate and pollutant level time series
Hasselblad 18 recently suggested methods based on first order Markov
structure, and French, et. al.19 recently used an informal but sensible
analysis to relate pollutant combinations to attack rate.
Even if daily attack rates were statistically independent, so that
the significance levels in Table 2.4.4 had meaning, one would need to
exercise caution when inferring that particular pollutants are
associated with daily attack rate. This is because daily pollutant
11 Utah CHESS Progress Report, op. clt.
« stebMngs, J. H. Some Problems In the Design and Analysis of Panel Stadlesln Epidemiology. Seminar
at Princeton s Department of Statistics. October 18,1975.
" Blnomfielcl, P. "Spectrum Analysis of Epidemiological Data." Updated reprint from the Department
of Statistics, Princeton University.
i« Hasselblad, V., "Analysis of Panel Studios." EPA Internal Draft (June tBTff).
« French, J. G., Hasselblad, V., Sharp, O., Truppi, I/., "A Study of Asthma la the Los Angeles Basin:
1972-1973." EPA Internal Dra,}t (June 10,197S). ;
81
-------
70
levels are correlated (Table 2.4.3.)- Ridge regression methods20
applied to the Salt Lake data to investigate this "multicollinearity"
of pollutants show that temperature adjusted SS effects are stable.
This result tends to support the hypothesis that SS may be important,
but formal statements of statistical significance are not justified.
To compute the relative risks in Table 2.4.5. (and 5.4.5.), 'the data
from all days with minimum temperature between 30-50° F and
integrated SO2 less than or equal to 60 pg/M3 were pooled, and the
total number of attacks was divided by fthe total number of per?ons-
days to obtain the rate 16.9 (Table 2.4.5.). Ratios of such numbers
were termed "relative risks." The ^statistical properties of these
relative risks require further study, since the days contribtite statis-
tically dependent data. Also no consideration is given special com-
binations of pollutants since the same days are used (except for days
with missing pollutant data) for each pollutant. The only large in-
creases in relative risk occurs on days with high SS and high minimum
temperatures, but these relative risks are based on numerators with
few person-days exposure^ and the corresponding increases in New
York (Table 5.4.5.) are unimpressive.
The "hockey stick" model21 used to determine threshold is. a three
parameter non-linear model (attack rate intercept, threshold, and
suprathreshold slope) which the authors fit by least squares. Con-
fidence limits are not given for the estimated slopes and thresholds;
nor are these appropriate since the daily attack rates are dependent
and possibly heteroscedastic (i.e., with different variances). The
Monograph gives no analysis of residuals, nor is the original data
plotted on the fitted graphs (Figs. 2.4i5.-2.4.7.). Thus, the reader
cannot assess whether the "hockey Stick" model fits the data or
whether the data suggest that a threshold even exists. One such
scatterplot of S02 data does not support the concept of a threshold 22
and it is likely that uncritical use of the hockey stick function has
led to estimates of non-existent thresholds on occasion.
t The New York asthma studies had similar practical and theoretical
difficulties as documented in the 1970 Biannual Report.23 Figure 12 of
that Report shows 74 of 80 mailed diaries were useable, ofwhich 14
required telephone prompting or interviewing. Eleven of 43 Bronx
Eanelists dropped but during the study whereas the rates for River-
ead (8/38) and Queens (11/52) were smaller (Figure 10 of the Bian-
nual Report24). Superimposition of the Figures 5.4.1.-5.4.4. shows
that. Riverhead often has lower weekly attack rates than Queens, but
such data is not given for Bronx. Preliminary analysis of 1971-1972
New York asthma panels using the same statistical methods as in
the Monograph suggest that SN is a more important risk factor
than SS. The Monograph gives no relative risk calculations for SN,
and SN was discounted in correlation and multiple regression, analyses
in 1970-1971 Salt Lake and New York studies. Relative risk tables
from the 1971-1972 New York study confirm the 1970-1971 associa-
» Smith, E. J. C., "A Ridge Regression Analysis of the CHESS Studies; Salt Late Asthma." Undated
EPA Internal Draft.
» Hassalblad, V., Creason, T. P., Nelson, W. C., "Regression tJsing Hockey Stick Functions." EPA-
600/1-76-024 (June 1978).
» Deinill, K. M., Wride, W. P., "Technical Memorandum,76-2." Kennecott Copper Corporation, Opera-
tions Research. Salt Lake City.
a Prsvda, M., "Community Health Effects Surveillance Studies: New York City." Internal Biannual
Kaport to EPA (December 1970).
"Ibid.
82
-------
71
tion of SS with daily attack rate, but the effects of SN cannot be
disentangled since SN is highly correlated with TSP, ESP (respirable
difficulty of associating a specific pollutant with asthma attack rate,
but both studies show attack rate is associated with daily urban
"pollution." ;
These panel studies suffered from several practical and theoretical
deficiencies. Panelists were not examined by a physician and eligi-
bility requirements were somewhat obscured by a "priority" system
of secondary selection. Numerous factors known1 to precipitate asthma
attacks were ignored. The analyses failed to consider autocorrelation
in the asthma attack rate and daily pollution levels, and the authors
did not present spectral or lag analyses. The significance levels
presented are untrustworthy, since the attack rates are dependent.
The statistics used have descriptive value, but the patterns of cor-
relation between pollutants and attack rate are strikingly discordant
in Salt Lake and New York. It is disturbing that Ogden (Low) often
has higher attack rates than Magna (High), and this observation re-
quires further investigation. Although SS may well precipitate asthma
attacks, the hockey stick model used to estimate thresholds was
not shown to fit the data. Tentative findings of the 1971-1972 New
York asthma study highlight the difficulty of relating asthma attack
rate to a specific pollutant. . :
This commentarytemphasizes limitations of the analytical method-
ology and problems in data collection because these topics are scarcely
mentioned in the Monograph. The panel strategy is nonetheless an
appropriate tool for studying the effects of pollution on selected high
risk populations, and the statistical methods used provided consider-
able insight in exploring possible relationships between pollutants
and asthma attack rates. The fact that time series methods were not
used or that theory could not provide formal inference (without
further research) for statistics such as the "relative risk" should not
obscure the descriptive and exploratory merits of the methods used.
5. CARDIOPULMONAKT SYMPTOMS IN ADULT PANELS
These 1970-1971 New York studies attempted to associate symptom
aggravation in high risk elderly panels with daily pollutant levels.
Since these studies were similar to the asthma panel studies, they
shared similar practical and methodological problems. Figure 14 of
the Biannual Eeport25 shows that recruitment varied by community.
Eiverhead contacts came from trailer parks and from recreation and
golden age clubs whereas Queens and Bronx participants were con-
tacted on park benches or through the New York Housing Authority.
It is hardly surprising, therefore, that Eiverhead (Low) participants
were generally younger and healthier than those from Queens and
Bronx (Table 5.5.A.I.) Candidates were interviewed and question-
naire information was used to classify them as well, heart, lung, or
heart-lung panelists. No previous or concurrent physician diagnosis
was specified in the 1971-1975 protocols (although the ^Monogpaph
summary on page 5-104 mentions physician diagnosis). EHgible
candidates had to reside within 1.5 miles of a1 monitoring site and to
_____
83
-------
72
be 60 years of age or more. In later studies, younger panelists were
accepted and some candidates were found outside the 1.5 mile moni-
toring radius, especially in Biverhead.
Substantial numbers of enrollees dropped out during the study,
many because they were too sick, had difficulty filling out the weekly
diary,, or just found the study bothersome (Figure 16 of Biannual
Report26).
The Monograph states that dropout rates exceeded 50% in Bronx
and Queens and approached 30% in Riverhead.
Many panelists failed to return weekly diaries. Figure 15 of the
Biannual Report 2r> shows that only 49 of 86 Riverhead panelists
provided useable diaries for the week November 22-28, and corre-
sponding fractions for Bronx and Queens were 38/99, and 9/59. The
fraction of useable diaries for the heart, lung and heart-lung panels
was usually even smaller. Thus Bronx, Queens and Riverhead yielded
respectively 24/47, 18/40 and 10/13 heart-lung diaries, 14/31, 6/21,
and 20/26 lung diaries and 26/59, 19/52, and 27/39 heart diaries.
Mailed diaries were replaced by telephone interview in the 1973-1974
protocol.
The high dropout rates and poor weekly diary response rates limit
the generality of these studies since the data was generated by a self-
selected population. This fact also complicates community comparison.
Panelists were asked to specify daily whether symptoms were
"worse," "much worse," "the same," "better," or "never present"
and daily aggravation rates were computed as the sum of "worse"
plus "much worse" divided by all participants that day. (The "much
worse" category was eliminated in subsequent protocols.) Table 5.5.2
shows that Bronx and Queens aggravation rates are usually higher
than Riverhead rates, but the authors do not ascribe these differences
to pollution since Riverhead is known to contain younger, healthier
panelists.
Figures 5.5.3-5.5.6 relate composite weekly symptoms aggravation
rates to weekly S02, TSP, SS, and SN for the Riverhead and Queens
heart-lung panels. The aggravation rate seems to rise and fall with
SS levels. No such graphs are given for Bronx nor for well, heart and
lung panels.
Table 5.5.4 shows that SS is positively correlated with daily symp-
tom aggravation in the well and heart-lung panels, but the stated
significance levels are untrustworthy since daily aggravation rates
and pollution levels are each autocorrelated. Multiple correlation
analysis confirms that temperature adjusted SS is associated with
symptom aggravation, especially in the heart-lung panel (Table
5.5.6), but these significance levels too can only be taken as qualita-
tive indicators. Because the daily pollutant levels are highly corre-
lated (Table 5.5.5), specification of a particular pollutant for particular
concern is speculative. However, it is impressive that other pollutants
have little explanatory power after temperature and SS adjustment
whereas SS still explains substantial variation in daily aggravation
rate after adjustment for temperature, SN, TSP, and SO2. Table
5.5.3 shows that daily SS was strongly positively correlated with
shortness of breath aggravation in six panels and strongly negatively
correlated in none. These figures are derived from Table 5.5.A.3
"Ibid.
"IbiO.
84
-------
73
which shown that, the. correlation was strongly positive iu. all three
heart-lung paaols, in two of three heart panels and in one lung paael
(Queens). Temperature1 specific relative risk calculations (Tables
5.5.7-5.5.9) also suggest that RS is associated .with increased symp-
tom aggravation, rates, especially in the heart-lung panel on warm
days. Such a table is not presented for SN. All these methods suggest
an' association between SS and symptom aggravation rate in heart-
lung panels. It would be interesting to know if the SS data indeed
exhibited a threshold effect and fit the. hocko.y-stu'fc model in Figure
5.5.S.
Tiic severe dropout rates and low weekly diary returns (especially
for the heart-lung, heart and lung panels)" raise important questions
as to the type of respondent and the meaning of the response "wor.se."
The statistical melhods have the limitations -[Mentioned i'or asthma
panels. Moreover, a subsequent New York 1971-1972 study28 failed
to confirm the dominant importance of SS but did find that SS, SN,
SO2 and RSP were associated with symptom aggravation of the well
panels. (This study did not use the same analysis as the Monograph
and, in particular, did not include temperature adjusted multiple
regression or temperature specific relative risk calculations.) None-
theless, the 1970-1971 data suggest an association between symptom
aggravation rates and various pollutants, especially suspended sulfates.
6. VB.NTIIiA.TOHY KUNCiTIOiV IN SCHOot OHILDRBN
The Cincinnati study of ventilafcory function in school children
preceded CHESS. Children from schools in an industrial valley of
.Cincinnati were, compared with children from schools in a non-
industrial river valley on the. east sido of the. metropolitan area. Two
upper-middle white, lower-middle, white, arid lower-middle black
schools were selected from each valley. Air monitoring stations Avithin
three blocks of the schools showed that seven month average TSP
values were from .18 to 32 /ag/m3 higher in the. industrial valley than
in the non-industrial valley, but corresponding differences for SS, SN
and SO2 ranged from 0.1 to 1.1, 0.1 lo 0.8, and 0.6 to 10.4 respectively
(Figure. 6.1.1). Thus, the industrial valley had more TSP thaii tlie
non-industrial Vi'illoy, but its levels of SS, SN, and S0s> exceeded
those iu the non-industrial valley by veiy small margins. Ventilatory
function was measured as the forced expiratory volume at thre'e
quarters second (FEV .75), and height, sex, and race were noted to
make adjusted FEV .75 comparisons.
The study was confined to 394 second graders who participated in
weekly measurements during November 1967, February 1968 and
May 1968. These students represented 93 percent of second graders
in the classrooms selected. Mothers were interviewed to obtain socio-
economic data. The educational attainment of, fathers was similar for
corresponding schools in tho industrial and non-industrial valleys
(Figure 6.1.2).
Figure 6.1.3 shows that average height adjusted FEV .75 in "clean"
schools exceeded that in "polluted" schools in all three months for
lower-middle whites and in two of three months for upper-middle
n Stebbinss, J. H., Hayes, C. U.. "Panel Studies of Acute Health.Efforts of Air Pollution I. Cardio-
pulmonary Symptoms tn Adults," Environmental Research Volume 11 (1978), pp. S9-1U.
77-580—76 6
85
-------
74
whiles. Blacks hud consistently lower FJEV .75 values, and a "pollu-
tion" effect was only seen among blacks during: one of throe months.
The absolute differences in average FEV .75 were roughly 40-120
milliliters (or less than 10%) in most cases. Table? 0.1.5 presents n
multivaritittt analysis of variance which allows one to test for com-
munity effects adjusted for a possible month effect and for the co-
variatcs height, sex, race, and social class. The dependent variable
for each child was his vector of three monthly average FEV .75 values.
No test for a community effect (industrial valley versus non-industrial
valley) was performed, but the significance, lovels given suggest that
community differences exist. The conclusion that "suspended sulfates
exerted the strongest influence" (pages 6-7) is unwarranted in view of
the high correlations commonly Found among these- pollutants. None-
theless, these studies support the notion that FEV .75 is 3-10 percent
less among white second graders in. the industrial valley than among
those in the non-industrial valley.
The 1970-1971 New York ventilatory study included children ages
5 to 13. Schools were situated within 1.5 miles\>f CHESS air monitors,
and Riverhead, Bronx, and Queens were represented by three schools
each. Only white children were eligible for analysis.
A new electronic spirometer, the National Gas Cylinder (NGO),
exhibited serious drift (perhaps 350 railmiters). This problem was
detected by field teams who had difficulty reproducing measurements
and who made several trips to the manufacturer for adjustments.
The June 1071 Progress Report28 gives an account of difficulties with
the spirometer, arid the Assistant Commissioner of the New York
Cjty Department of Health expressed serious reservations about the
reliability of those measurements in. a January 18, 1972 letter to the
Acting Chief of the EPA Epidemiology Section.30 It is possible that
drift of this magnitude biased the results, since one community may
have been systematically subjected to a spirometer with extreme
drift (especially if drift varied in phase with the rotation of spirometers
through ^communities). Even if rotation of the spirometers among
communities; distributed drift equally, the variability of the observa-
tions is greatly increased by random distribution of drift, since the
community effects (60 milliliters or less) are much smaller than the
drift. The NGC spirometer was replaced in late 1971 by a bellows-
type spirometer described in the May 1972 CHESS protocol. How-
ever, the_NGC spirometer was used throughout the 1970-197,1 study
reported in the Monograph.
The 1970-1971 protocol instructs investigators to "deduct one inch
for stacked heels for those girls wearing such shoes." In the. .1972-1973
protocol this height adjustment was reduced to % inch.
No data were presented to compare families of children studied in.
Riverhead, Queens, and Bronx, but the ago distribution and parental
smoking habits were said to be similar. Income and educational
attainment were said to decrease in the order Queens, Riverhead,
Bronx. The Monograph contains no comparison of children's smoking
habits in these communities; this covariate may be relevant since
.statistically significant differences were found only for older children.
» "New York City CHESS Progress Saport." (Juna 1971).
*> Bergner, Dr. L. (Assistant Commissioner for the Department of Health, N&w York City), Letter to
Dr. Douglas L. Hammer, Acting Chief, Epidemiology Section, EPA (January 18,1972).
86
-------
75
Figure 5.6.1 shows thai male* height and age adjusted FEV .75
values from Riverhead wore intermediate between. Queens and Bronx
values for throes of four test periods. For females, the lliverhead
values exceeded Bronx and Queens values in each tost period, hut.
the differences were usually less than 50 mUlilitors. Table 5.(3.5 show
that Biverhead height adjusted FEV .75 values were- largest during
one of four test, periods for young males, three, of four for older mates,
one of four for young females, and three of four for older females.
Thus these differences were inconstant, and even for older females,
the average difference between Riverhead and Queens was only 42
milliliters (about 2.4%). Table 5.6.6 is milling since, it includes a
degree of freedom for "ethnic, differences'5 not doso.ri.bed in the. text.
Perhaps blacks were included .in this analysis. The imiltivaria-to
analysis of variance used in Table 6.1.5 is mbr.e appropriate than
averaging over all test periods as in Table 5.0.6. However, the analyses
for individual test periods in Table 5.6.6 show statistically significant
differences for older males and females.
The Cincinnati study demonstrated small (less than 10%) differences
ii\ FEV.75 between white- children in schools located in a non-hid list-rial
valley and those in schools in an industrial valley. For New York,
small, inconstant differences were fount! between liiverhoad and.
Queens-Bronx only for older (9-13 years old) children. Preliminary
analysis of 197J--1972 New York CHESS winter and spring data
obtained with a bellows-type spirometer shows, that older Riverhead
children had statistically significantly higher FEV .75 values than
Bronx children in winter and statistically significantly lower values
iu spring.31 In view of these findings, the problem of spirometer drift,
and the possibility that older children in Bronx and Queens srnoke
more than in Riverhead, one must be caution's in linking New York
pollution effects with ventilatory differences. .
7. OTHKll CHKSK STUDTKS'
CHESS has conducted two types of studies not found in the- Mono-
graph. The. "Episode. Study" measures rates of discomfort (cough,
shortness of breath, restricted activity, and |»ye, throat, and chest
discomfort) during control periods and during]air pollution episodes.
A 1973 report32 shows higher rates of cough, chest discomfort and
restricted activity during episodes of high air pollution. "Pollutant-
Burden" studies are designed to quantitate the levels of metals and
other pollutants found in human tissues and to relate these levels to
exposure history. Scalp hair was collected from families participating
in ARD studios, maternal hair, blood, cord blood, and placenta!
samples were obtained from willing mothers, and various tissues were.
obtained at autopsy from pathology laboratories.
.1). SUMMARY ASSESSMENT OF THE POPULATION STUDIES
This assessment is made from the point of view of an epidemiol-
ogist. More global judgments, made in consideration, of the quality
of both epidemiologic and aerometric data, are found in Chapter IT I.
at Chapman. R. S., Hasselhlad, V., Burton, E., Williams, .T.. "Air Pollution and Vcntllatory Function."
EPA Draft (April 8, W75).
32tNeIson, C. J., Shy, C. M., English, T., Sharp, C. R,, Atidleman, E., Truppi, L., Van Brngpcen, J.,
"Family Surveys of Irritation Symptoms During Acute Air Pollution Exposures." Journal of the Air
Pollution Control Association, Volume 23, Number 'I (Fn.bnmry 1073), pp. 31^00,
87
-------
76
No formal methods arc used to link specific pollutants with specific
health effects in the, CRD, LED, ventilation, and ARD studies; If a
demonstrated health difference between communities cannot be
explained in terms of imbalances in known coyariatcs, it- is generally-
ascribed to pollution. The CRD studies convincingly demonstrated
small increases in the prevalence of productive cough in polluted
urban and smelter sites. The retrospective LRD studies may have.
demonstrated higher croup and bronchitis rates in cliildren in polluted
smelter communities, and the prospective ARD are rather more con-
vincing in. demonstrating an increase in acute respiratory disease in
polluted urban settings. The Cincinnati ventilation study demonstrated
small differences in FEV .75 between white second graders attending
schools in industrial and non-industrial urban valleys. If the aerometric
measurements had been entirely accurate and reproducible, during
these studies, it might "be safe to conclude that health effects had been
demonstrated in association with known patterns of pollution. Unfor-
tunately, this was not the case. It is not possible to know which, specific
pollutants, if any, or what concentrations of any suspect pollutants,
were responsible for the health effects. The health effects data provide
at most a rough guide for making general judgments about probable
health effects in other communities with similar pollutant sources,
meteorology and population composition. However, even though the
aerometric data are of poor quality, the studies appear to demonstrate-
some health effects of air pollution at concentrations which might
possibly be moderate (lower) compared with some high urban levels
of the early 1960's. In this light, the data in the CHESS study provide
no support for those who would significantly relax current standards;
oil the other hand, neither do they support a need for more stringent
standards. Thus, the CHESS studies have proved to be valueless for
assessing or supporting setting of quantitative regulatory standards
which was a goal of the program.
The methodology used in the panel studies (asthma and cardio-
pulmonary) attempts to disentangle the effects of the several pol-
lutants. The multiple regression and relative risk calculations are
interpreted as implicating suspended sulfates, particularly in the
Salt Lake asthma and New York cardiopulmonary studies. While this
interpretation appeal's reasonable, these formal methods do not pro-
vide logically compelling evidence that: SS, or indeed any of the
measured pollutants is of dominant/ importance. Despite the many
practical and methodological difficulties of these studies, they support
the belief that asthma attacks and cardiovascular symptoms occur
with greater frequency when the air is polluted.
These findings may appear overly critical especially since CHESS
pioneered many research techniques. However, this report comple-
ments the CHESS Monograph, which often fails to apprise the reader
of practical or theoretical problems encountered in this research.
These remarks are meant to aid in the assessment of the validity of the
conclusions presented in the Monograph and to assist researchers
performing similar studies and encountering similar difficulties. This
ondoavor was greatly assisted by hindsight and by the splendid
cooperation and .self-examination of investigators both inside and out-
side the Environmental Protection Agency.
88
-------
77
E. CHESS CURRENT STATUS
1. GENERAL
la the process of interviewing EPA personnel concerning CHESS
background, a good deal of information ivas obtained as to present
and future plans for tho program. Many other program and manage-
ment- ideas were also offered by tho Ef'A personnel.
Tho CHESS program uo longer officially exists within EPA except
os a historical rofere.ii.co point. Administrator Train on April 9th com-
mitted to have all datw. from the program analyzed by mid-1977.
However, the effort within the Health Effects Research Laboratory
directed at processing; the. several years of data remaining (1972-75)
unanalyjsed from the CHESS studies is progressing slowly. A forced
change of computers (see below) caused major delay. In addition,
the investigative team sensed ambivalence as to how much effort
is justified. Discussions with EPA Headquarters jjersonnel indi-
cated that, in keeping with the Administrator's commitment, the
analysis is of sufficient importance in its own right as well as to con-
tinuing programs to warrant early completion. Yet, from discussions
at Research* Triangle Pai'k, it can only be concluded that many would
rather spend resources initiating new projects or refining CHESS
methodology. Perhaps this is just as well, since there is serious doubt
that the analysis even when completed will ever be sufficiently credible
to support the stated objectives of the program. Any results from the
(inal CHESS analysis may be. examined with suspicion since the doubt
concerning the reliability of aorometric measurements and the accuracy
a.nd acceptability of health measurements which evolved following
the first CHESS report will affect any further reports from this same
program. The past loss of experienced" personnel has impaired the long
iprm research effort required by CHESS objectives and has hurt
morale among the. remaining researchers.
The investigative team found that the disorganized state- of affairs
described above is compounded by the iact that the Health Effects
Research Laboratory still lacks key personnel in the health research
Held. The Laboratory Director has been actively seeking persons to
fill several key positions, but without success;
Several personnel problems continue to frustrate members of the
Population Studies Division, of the Health Effects Research. Labora-
tory. An impending retirement leaves the Division without leadership
in a period when long range plans arc to be reformulated. Members
of the Division wonder, in view of recent investigations including this
one, whether yet another reorganization is soon to come, and many
regard this organizational instability as an impediment to sound
work and career development. Most believe that Population Studies
staffing is inadequate in relation to program funds. This disproportion
requires that most real work be done by contractors. Not only are
ETP staff deprived of the incentive of planning and executing their
own studies, but they lack time and travel funds to monitor the con-
tracts satisfactorily. The Population Studies Division responds to
special regulatory needs (e.g. sulfate studies) and to unforseen events
such as the demand for assessment of Kepone effects in Hopewell,
Virginia. While these situations and demands might form the basis of a
muled program in the future, at present they are regarded as "fire-
fighting" distractions.
89
-------
78
The team found strong evidence of a continuing lack of coordination
between field and Headquarters and insufficient coordination among
regional programs, Environmental Research Center Laboratories, con-
tractors, other Federal agencies studying health effects of air
pollutants, and Headquarters, not only in total program objectives
but also in program coordination of quality control, sample collection
and analysis, and standardization of health effects data collection
tools.
2. STATUS. OF ANALYSIS
Schedules 3IJ for analysis of CHESS data show that CHESS health
data were collected from 1970 through 1975 for asthma panels,
cardiopulmonary panels, and yentdlatory function panels. The last
surveys for acute respiratory disease (ARD), and chronic respiratory
disease (CRD), and acute episodes took place in 1973, 1974, and 1975
respectively. CHESS surveys were completed for 25 asthma panels,
12 cardiopulmonary panels, 28 ventilatory function panels, 15 ARD,
12 CRD, 8 episodes and 8 acute lower respiratory disease panels
during this period. The Monograph reported on only eight of these
panels conducted in the 1970-1971 New York and Salt Lake studies.
(The Rocky Mountain, Chicago, and Cincinnati studies contained in
the Monograph were not CHESS studies.) Thus, CHESS generated
only a portion of the data analyzed and reported in the 1974 Mono-
graph, and the Monograph reports on only a small fraction of CHESS
data.
The present analysis of the studies proceeds hi steps. First, the
survey data is edited for computer processing and subjected to pre-
liminary statistical analyses. Then an epidemiologist writes a draft
on the basis of the preliminary statistical 'analyses of health and aero-
metric data. Other analyses may be done if judged necessary by the
epidemiologist or statistician. KPA staff have tabulated the fraction
of surveys subjected to preliminary statistical analysis and the fraction
for which a preliminary draft has been written.34 Drafts exist for 6 of
8 episode and 7 of 8 LRD panel studies. Drafts have been written for
7 of 15 ARD and 8 of 12 CRD studies, but no CRD drafts have been
written for survey years 1973 or 1974, By comparison, asthma, cardio-
ptihnonary and ventilatory function analyses are far behind schedule.
Of 28 ventilatory surveys, 10 have undergone preliminary analysis,
and 8 drafts have been written (as of June 28, 1976). Of'25 asthma-
surveys, nine have undergone preliminary analysis and nine drafts
have been written, and of 12 cardiopulmonary panels, two have been
analyzed and two drafts written. These figures substantiate the degree
to which the CHESS program has fallen far behind in analyzing and
writing up survey data (61 data sets are yet to be written up).
The figures quoted above indicate that a major bottleneck in the
analysis of survey data has been the preliminary editing and statistical
processing. It should be noted that EPA had little or no control over
two disruptive events which retarded, preliminary analyses by a year or
more. GSA ordered a change of computer from IBM to Univao in
August 1974 as part of a general government policy related to com-
petitive bids. Both hardware and software conversions were difficult,
and most programs had lo be completely rewritten. Compounding
a Unpublished tables giving dates of CHESS data collection prepared by Dr. Dorothy Calafiore, EPA/
RTF.
» Unpublished tables showing chronology of survey data analysis and draft •writing prepared by Drs.
Carl Hayes and/or Katbryn McClain (June 28,1070), EPA/RTP.
90
-------
79
this problem was the need to rely increasingly on outside data process-
ing contractors, since increases in EPA staff needed to meet the
increased data load never materialized even though verbal com-
mitments had been given to the laboratory directors. 6SA contracting
procedures required two changes in data processing contractors
between "1972 and 1974. Other problems impeded the preliminary
statistical analyses. The loss of key personnel imposed new scientific
and administrative burdens on remaining statisticians and epidemi-
ologists. Moreover, responsibilities of the Statistics and Data Manage-
ment Office were broadened during this period to include projects
with the Toxicology said Clinical Studies Divisions as well as special
tasks, such as preparing data tapes in reponse to outside requests for
EPA data. CHESS analyses (Population Studies Division) now ac-
count for about half the Statistics and Data Management Office work
load. However, the most important delays in the preliminary analysis
are attributable to the forced computer change and to problems with
outside data processing contractors.
The epidemiologists charged with writing a draft monograph must
await receipt of aerometry computer output and biometry output.
Epidemiologists interviewed attributed delays to the failure of these
preliminary analyses to arrive and to subsequent changes in aero-
metric values used in the manuscript. Other problems retard the
epidemiologists once preliminary analyses are in hand. To begin with,
the Population Studies Division has lost many of the important early
authors or researchers within CHESS. It is hard for an epidemiologist,
not familiar with the background and history of the data collection
and faced with two computer outputs, to. assess the quality of the
data and to obtain specific scientific guidance.
Morale is low, and this may be due in part to the loss of key per-
sonnel, to investigations of the CHESS program, and to organizational
instability affecting Population Studies. Furthermore, during the
past year, epidemiologists have been encouraged to pursue competing
scientific interests, and they have had to shoulder administrative
burdens associated with increased reliance on external contractors
and to meet special tasking such as the Keppne'Task Force.
It was interesting to find that several epidemiologists interviewed
complained that the current team structure for analyzing survey data
does not function well without a strong authority to resolve scientific
and scheduling conflicts between epidemiologists and statisticians.
While these factors and lack of staff have not seriously impeded the
drafting of CHESS reports until now, they may become pressing
problems as the backlog of preliminary analyses'clears the new com-
puter system in the near future.
The continued delays in analyzing CHESS data have had several
adverse effects. Contractors gathering survey data and panelists have
been embittered and demoralized by lack of data and feedback from
EPA. The failure to provide contractors with timely aerometric and
health data summaries stifled local initiative in analyzing the data.
Thus, the CHESS program did not benefit fully from local field level
insights as to local peculariti.es of the survey and populations or from
comparisons with independent local air pollution measurements and
data analysis.
91
-------
80
3. RKCBNTT PfiANNTXtr DIIUSCIIONS OF THE POPULATION
STUDIES DIVISION
The following comments ui'e largely speculations and impressions
based on interviews with EPA staff.
CHESS data acquisition ended in 1974-1975 when OMB did not
approve questionnaires needed for the study. Had OMB approved
the forms, however, it is likely that 6 asthma, 6 ventilatory function,
2 LED and 2 CED surveys 'would have taken place in 1975-1976."
Instead the Populations Studies Division worked on old CHESS
data and made several proposals for noii-CIIESS studies to be run
on contract. One of these is a cohort study of ventilatory function
in. adolescents which uses a CHESS enclpoint (measured variable)
but does not use a cross-sectional CHESS replicate 'population.
Another uses the CHESS AED enclpoint to study the effects ,of
f>ak NO2 values. One study entitled "A study to Determine the
ealth Effects Associated with Emissions from Coal Combustion.
and Coal Gasification" uses the CHESS AED and FEV .75 ondpoints
to study the impact of coal combustion and gasification in local
populations (Hair, blood, and urine samples are also obtained to
assay for products of combustion and gasification). Other proposed
studies do not use CHESS endpoints. These include: (1) a study of
mortality patterns in Montana, where, excessive can'cor mortality
has been noted, (2) a general household health survey given to mem-
bers of a community located near a new sewage treatment plant,
(3) a study of trace metal burdens found in the hair, blood, and urine.
of people Jiving near non-ferrous smelters, and (4) an evaluation of
hyperactivity and cognitive function in children with asymptomatic
lead burdens. These proposals build on CHESS experience, but they
do not embody any continuation of CHESS surveillance. However,
only one study plan and questionnaire was approved by OMB at.
the time of this investigation. The contained resistance by OMB to
approve new study forms and questionnaires is a great frustration
to the researchers. There was a strong feeling evident that this re-
sistance stems from a misguided assessment of scientific merit as
much as from a Presidential concern for the rights of privacy.
The Scientific Advisory Board (SAB) has fallen short of EPA staff
expectations. The staff members had'hoped to work closely with
SAB to formulate and refine new proposals. Instead, the SAB has
provided only formal criticism, expressed in public meetings attended
by press and industry. This is a time when the EPA staff needs new
leadership from a Director of Population Studies and from close
informal working relations with the SAB.
» Table of proposed CHESS studies for 1W5-1976 prepared by Dr. Carl Hayes, KPA/UTP.
92
-------
VII. CURRENT CHESS STATUS AND FUTURE PROGRAMS
A. RETBOSPECTION
It is always easier to look at the past and detect mistakes than.
it is to look forward and avoid them. There are lessons which can be
learned from the research which produced the CHESS Monograph
as well as from the efforts to process the balance of the CHESS data
which remain unanalyzed and unpublished.
This investigative report supports much of the criticism of the
CHESS Monograph as developed by other earlier and independent
examinations. For example, a report prepared for the Federal Energy
Administration in 1975 f describes the use of the unvalidated version
of the British Medical Research Council questionnaire as limiting
the reliability of the health effects analysis in programs, utilizing this
health indicator. The Subcommittee's investigation, no ted this same
defect. Questions regarding the analysis of the data on asthma attacks
indicated serious discrepancies. In the report to FEA, however, the
significance of the errors in the aerometric data which are detailed
in the Subcommittee'sinvestigation was not fully recognized. In
their review of the CHESS Monograph, the EPA Science Advisory
Board (SAB) identified many of the same problems noted by the
investigative team, again primarily in the epidemiologies! areas.
In the SAB, or so called Whittenberger, Report 2 other deficiencies
included: limited population sampling control; shortcomings in the
methods used for assessment of past air pollution levels; variations
in sample populations which limit comparison of communities; high
drop-out rates in asthma panels; inadequate consideration of tempera-
ture as an important variable in consideration of asthma attacks; and
serious concern about the problem of gaining information on in-
dividual exposure to pollutant levels. All of these points and others
are confirmed by the Subcommittee's investigation.
As noted in the legislative history of this report (Appendix B), other
independent investigators also had offered criticism of the CHESS
Monograph during the brief period of peer review which occurred
prior to publication and this criticism also appears to be well founded
for the most part. In separate talks with researchers who had either
participated in some phase of CHESS or had reason to follow the
progress of the CHESS investigations and review of data samples,
the Subcommittee investigators had occasion tp gain separate con-
firmation of many of these same concerns about the validity of the
instruments, questionnaires, and methods used for collecting data as
well as the methods of analysis.
1 Preliminary Report. A Critical Evaluation of Current Research Regarding Health Criteria for Sulfur
Oxides. Technical Seport Prepared tor the Federal Energy Administration. Tabershaw/Cooper Asso., too.
April 11,1975.
1 Review of the CHE S3 Program. A Report of a Review Panel of the Science Advisory Board—Executive
Committee, March 14,1975.
77-890—78
93
-------
82
Thus, the conclusions contained within Chapter Til of this report are
not exceptional in their general tenor. However, the serious errors in
the aerometric data and the verified significant effect of the poor
quality control during the early stages of the CHESS air monitoring,
measurements and analyses efforts are identified more precisely for
the first time by an outside investigation.
B. CURRENT
Moving from the historical perspective, an appropriate question to
be asked is whether the errors suggested in other critiques and con-
firmed in greater detail by the investigative team are being adequately
addressed in the current and planned Environmental Protection
Agency's air pollution/health effects programs.
It was of interest, first, to determine the approximate status of the
CHESS data which remain unanalyzed and to ascertain whether
.these data will suffer from the same or other discrepancies as the data
presented in the 1974 CHESS Monograph. As is documented at the
end of the last chapter (VI), in discussions with EPA personnel at
Research Triangle Park, the investigative team determined that there
still remains a considerable amount of analysis to be completed. The
impressions gained from these interviews was that it would be at
least until the summer of 1977 before the final assembly of data and
examination for potential analysis and publication would be possible.
However, as noted earlier in the report (Chapter VI), a number of
draft studies were completed at the time of the visits to ETP. For
example, draft reports have been prepared for 6 of the 8 remaining
episode type health studies and for seven of the 8 remaining; lower
respiratory disease studies. Seven of the remaining acute respiratory
disease studies have been prepared and 8 of the remaining 12 chronic
respiratory disease studies are in hand. No draft of the data on chronic
respiratory disease studies appears to be ready for the survey years
1973 or 1974. Analyses of a number of other health effects studies
have not been initiated.
The general impression is that there is still considerable effort
required for the basic transfer of field data to computer tapes for
analysis. Unfortunately, serious reservations are retained by the
investigative team about some of these data since some of the same
criticisms about the techniques used to collect the health effects
information in the CHESS Monograph continue to apply to some of
the data accumulated after 1971.
The establishment of a more vigorous quality control system appears
to have improved the collection of aerometric data after 1974. The
aerometric review (Chapter IV) points out that CHESS instrumenta-
tion, as used in the CHESS studies, did not consistently^ generate
valid, reliable data in the range of the National Ambient Air Quality
Standards. Thus, it is likely that no amount of effort on analysis of
CHESS data would achieve the desired objectives. There is doubt that
adequate meteorological support is being provided. Further, as noted
in other sections of this report, there does not yet appear to be an
adequate systematic planning effort to insure the full participation of
all research components.
94
-------
83 i
C. FtTTtTBE !
Since sulfates were identified witMn the CHESS monograph as
being a pollutant requiring an immediate increased priority of investi-
gation, the investigative team attempted to determine the status of
work on this pollutant. Mr. Train had indicated during testimony
before the joint hearing on April 9, 1976, that a five year plan for
study of .sulfates was available. Actually, it was found during the
investigation in the summer of 1976 that the sulfate_plan was still
in draft form although passing through a second; iteration. There was
some indication that even, now there is not a full appreciation of the
need for planning input with regard to quality control, ak monitoring,
level of cooperative preplanning does not scemj to be present. In a
letter to the chairman of the Subcommitfcee on f;he Eiivironment and
the Atmosphere (undated) in May, 1976, Mr. Train summarized
EPA's projected research program on sulfates. '
In measurement and monitoring, the emphasis in the draft plan is
placed upon the development of reliable analytical and sampling
techniques and instrumentation to support laboratory and field
measurements pf sulfates and sulfur compounds'; the development of
quality assurance programs; data audits, and providing technical,
support. The ongoing research and near-term [research is aimed at
increased data collection on sulfates with emphasis on measurement
methodology, instrument development and field! tests. Emphasis is to
be placed also on the development of reliable models to predict sulf ate
concentrations and precursor emissions over long distances. The plan
indicates a recognition of the need for instrument and field tests.
A continuation of the concept that more data on the health effects
of sulfates are needed is reflected in an intention^ to expand emphasis
on the toxicological, epidemiologies! and clinical studies on the
effects of sulfates. The plan includes a recognition by EPA of the
difficulty of conducting epidemiological studies, particularly in
correlating ambient pollutant (sulfates) concentrations with health
effects.
With the exception of the particular emphasis on sulfates, these
objectives are essentially the same as the objectives cited as necessary
upon the initiation of the original CHESS studies. If these objectives
are to be achieved in the near future, it will be necessary for EPA
to avoid the errors of CHESS. It is not clear either from the proposed
programs or from discussion with individuals who will be implementing
these programs that some of the basic errors committed during the
CHESS studies will be corrected before such field, air monitoring, and
epidemiological studies are continued with • different pollutants.
Certainly, the manpower problems have not been resolved.
In the aerometric area, the CHAMP aerometric measurement and
analysis is entering a new phase of development. It is essential that
the aerometric measurement and analysis errors detected during the
CHESS time period be eliminated as the CHAMP effort is expanded
and greater emphasis is placed upon this program. The CHAMP
instruments, when verified, may be capable of identifying reliable
differences in absolute air nollutant levels but, the instruments will
55
-------
84
be pushed to their limits of sensitivity and will require constant
attention by skilled technicians. 'New instruments just developed
may be able to meet the ultimate requirements of the air pollution.
health effects program. ,
A sound program of development and testing (verifying) question-
naires and other approaches to epidemiologieal research also must be
mounted and completed before another ;CHESS-type program is
initiated. Completion of both phases of such a program to provide
the tools and techniques for a sound health effects program might
require up to 2 years and several million dollars. If this work is
not carried out, progress in the area is doubtful.
96
-------
APPENDIX A
A EBCAPITDLATION OF THE AEROMETRIC AND METEOROLOGICAL
FINDINGS OF THE INVESTIGATION AS THEY DELATE TO SPECIFIC
SECTIONS OP THE CHESS MONOGRAPH AND THE HEALTH FINDINGS
I
A. INTRODUCTION
This section contains citations of errors and omissions found in a
careful review of the CHESS Monograph which ishow that the use of
aerornetrie and meteorological data in correlation with health effects
end point measurements can easily mislead the reader of the CHESS
' document into inferences which are not wholly or even partially
supported by the data in the report. Page, paragraph, and figure
references are to the 1974 CHESS Monograph, i
Since an important application of the aerometrie data is to deter-
mine correlations with, health effects, any errprs or overusage of
aerometrie data based upon estimates or improper measurements will
obviously reduce or negate the value of any health effects correlations-
which are attempted. This misusage or overusage of aerometrie data
will be particularly damaging as the extension of the conclusions is
made in an attempt to discover possible threshold effects.
B. CRITIQUE
1. Prevalence of Chronic Resmratory Disease Symptoms in Adults:
1970 Survey of Scdt Lake Basin Communities
Observed concentrations for only one year;have been used to
crudely estimate concentrations of sulfur dioxide and suspended
sulfates relating to a 4-7 year exposure. The. 1971 observed annual
average concentration of sulfur dioxide was used with the 1971
emission rate from the smelter to obtain a ratio that was then mul-
tiplied by emission rates for other years to estimate concentrations
for the other years. The estimated sulfur dipxide concentrations
were then used in a regression equation based on a 1971 relationship
to estimate suspended sulfate concentrations. Possible changes in
meteorological conditions and mode of smelier operations were
neglected. Acknowledgment is not given in the1 discussion and sum-
mary that the critical concentrations relating to health effects are
nothing more than estimated concentrations. ,
It is questionable whether or not long-term exposures should have
been attempted for Magna, based on only one year's record of ob-
servations mat are abnormal because of the smelter strike. It would
certainly have been appropriate to have mentioned that only es-
timated long-term data were available and indicated their degree of
uncertainty in the discussion and summary.
(85)
97
-------
86
Further, we find many errors on Page 2-37, Table 2.1.A.14. It seems
that this table should have never been included in the report. Aside
from the misuse of the diffusion model (discussed in Chapter IV) this
table lists suspended sulfate values for Magna for the years 1940-1970,
that are not the same as listed in Table 2.1. A. 16, on page 2-39. The
values_ are estimated by a simple ratio froih the smelter emission rates,
but this is not explained. On page 2-39 a regression equation is used for
thesame purpose. All of the sulfate concentrations under the heading
CHESS are estimated observations except those for the year 1971.
This has not been properly indicated, e.g. by the use of parentheses.
On pages 2-37 emission rates are not sulfur dioxide rates as indicated
but emission rates in tons of sulfur per day. This means that the sulfur
dioxide emissions. were twice the values listed. It also means that the
dispersion model estimates are incorrect. However, the listed estimated
•concentrations in Magna and Kearns, which are based on a simple
ratio between observed concentrations in 1971 and some emission rate
for 1971, whatever it might be, are not changed.
Note that the regression equation for suspended sulfates, Salt
Lake City, (pages 2-39) which is:
SS=0.101(TSP)-J3.65
i
is quite different than that which can be' obtained from Table 2.1.4,
SS=0.065(TSP)-}-1.93
exposures were derived by multiplying the yearly smelter
emission of SO2 by the ratio of the 19711 measured annual average
SO2 concentration to the 1971 SO2 emission rate (193 tons/day).
Estimates of suspended sulfates were derived from the estimates of
SO2, using the following regression equation for 1971 :
88=0,09(302) +6.66.
The annual TSP exposures were derived by multiplying the yearly
smelter production of copper by the ratio jof the 1971 measured annual
arithmetic mean TSP concentration to the 1971 copper production
rate (260,000 tons/year).
Smelter emissions of sulfur dioxide in the early 1940's were roughly
three times greater than they were after 1956 although copper pro-
duction has remained more or less constanib. The method for estimating
are estimated lower in 1940 than in 1971. The procedure used produced
very high ratios between SS and TSP for the earlier years. For example,
the 1940ratio (34.6/63) is 0.55. This ratio is so large that it is obviously
questionable.
The audacity of the estimates can be; seen in Figure 2.1.17. The
lowest value, which occured in 1971, is extrapolated aE the way back to
1940, reaching unusually high annual average concentrations of more
than one part per million. Considering the effects of wind direction,
which would result in low concentrations much of the time because the
Bmelter stack plume would not be blowing toward the town, such an
annual average would result in short-period concentrations many tunes
98
-------
87
'the annual average. It is questionable thai; such, high concentrations
ever occurred. If they did, they_ would be well-remembered, and living
conditions in MagJmwxmW be different than hi 1971. Such unreasonably
high estimates should have been further investigated before being
presented.
The.grossness of the estimates made overrides other shortcomings
in this study pertaining to exposure that might |be mentioned. How-
ever, more carefully made estimates would have required considerably
more work, including obtaining meteorological records and details of
smelter operations affecting plume behavior over the period of years
studied. Such a large effort may not have been worthwhile considering
the inexactness of some of the other aspects of-the study. Nevertheless,
a study of this nature seems to call for actual observations, more ac-
curate estimates, or considerably less exactitude in its conclusions.
2. Frequency of Acute Lower Respiratory Disease in Children: Retro-
spective Survey of Salt Lake Basin Communities, 1967—1970
The same comments apply to this study as for the preceding study
on the prevalence of disease symptoms in adults. Inadequate recogni-
tion is given to the fact that only estimated concentration data are
being used in the discussion and summary. ;
3. Aggravation of Asthma by Air Pollutants: 1971 Salt Lake Basin
Studies
In this study, daily entries in a dairy were used to determine weekly
asthma attack rates. A statistical relationship was then determined
between the attack rates (weekly) and observed air pollution concen-
trations (averaged weekly). Participants lived within a 2-mile radius
of air monitoring stations.
Daily exposure of asthmatics in a community such as Magna, which
is close to the smelter, are poorly characterized by a single monitoring
station. On a given day, one side of the community could be much more
affected by the smelter stack plume than the other, and high concen-
trations from looping or fumigation might affect one neighborhood
but not others. The study inadequately assesses 'the effects of peak ex-
posures and episodes.
This report does not make clear that the minimum temperatures
used were from the Salt Lake City airport. The assumption seems to
have been made that temperature was uniform'over the entire study
area. This is not true because of the differences in elevation and the
effects of the mountains, and the lake. Perhaps the differences were
not important, but they should have been considered. It is not clear
why days were stratified by mmimum rather than mean temperature.
Minimum temperatures occur during the early morning when peo-
ple are generally indoors and perhaps in bed. When temperatures are
low, windows are generally closed. Also, lower mmimum temperatures
are correlated with other meteorological phenomena that could also
affect asthma attack rates, e.g., lower humidity land lower wind speed.
Further there may be a correlation with wind direction. A lower^than
average minimum temperature probably is also associated with a
strong temperature inversion which would be; conducive to lofting
the smelter stack plume. Because of the many questions raised, the
findings pertaining to temperature merely suggest further study and
have no general application. !
99
-------
Near the middle of the left hand column, page 2-89, the following
sentence appears. "The shut-down of operations by the strike was
accompanied by a pronounced improvement in air quality and a reduc-
tion in asthma attack rates that occurred sooner and were larger than
seasonal reductions observed in the more 'distant study communities
some 2 weeks later." Here there is a lack of appreciation of the natural
climatic differences that exit in the Salt Lake Basin. Some effects of
summer weather could easily be delayed 'two weeks before reaching
Ogden. The average date for the last killing frost in Ogden is about
May 6, whereas the average date of the last killing frost at Saltair
(the climatic station nearest to Magna with a long record) is about
April 12.
On Page 2-76 (near middle of page, right hand column) the smelter
is not "5 miles north of Magna." I
On page 2-81 the first graph in Figure 2.4.1 is incorrectly drafted.
After the 17th week the broken line should be solid and the solid line
broken. The temperature curve should appear as hi the graph for
the high exposure community.
Figure 2.4.2, page 2-81, shows a weakness in the argument that the
sulfur dioxide concentrations are responsible for the asthma attack
rate. In the High Exposure Community the attack rate starts up at
the 18th week as the sulfur dioxide concentrations approach zero, or
near zero, and remain very low for about six weeks. It is noted that
this same graph shows the highest S02 peak occurring at the 9th week,
which seems to begin about May 9. The graph on page 2-16 seems to
show the peak in April. i
In Figure 2.4.4, page 2-82, with respect to tlie High Exposure
Community, it may be noted that the sulfate concentrations are not
particularly well-correlated with the sulfur dioxide concentrations
plotted in Figure 2.4.2, on the preceding' page. The highest sulfate
reading occurs in the 3rd week, whereas the sulfur dioxide levels
build up to a peak in the 9th week.
On page 2-87, left hand column, it is stated that a threshold concen-
tration of 1.4 Mg/m3 was calculated for suspended sulfates for the
higher temperature range. In Figure 2.4.4 all of the plotted concentra-
tions are greater than this value. Considering the background of
suspended sulfates generally observed, thislow threshold value seems
to have no practical significance.
The third paragraph that appears in the right hand column, page
2-89, probably applies to Magna, however, this is not made clear.
There is a possibility that the paragraph could be given broader
interpretation than actually intended since the last three sentences
seem to refer to conditions in urban areas generally. The paragraph
g'obably should have been divided into two separate paragraphs.
owever, the main fault with the paragraph is that important con-
clusions are drawn that are not supported by information presented
elsewhere in the report. It says "excess asthma attributable to sulfur
dioxide might be expected 5 to 10 percent of summer days", "total sus-
pended particulates could occur on up to 5 percent of summer days and
30 percent of fall and winter days", and 'i'excesses due to suspended
sulfates are likely to occur on 10 percent of fall and winter days and
90 percent of summer days." Assuming that the stated relationships
100
-------
between concentrations and temperature are true, the report does
not explain how the percentages of days were obtained. The study
covered only 26 weeks, but these conclusions apply to an entire year.
The percentages given seem to be rough estimates since they appear to
be given only to the nearest 5 or 10 percent. The percentages might
have been obtained from daily values for the minimum temperature,
pollutant concentrations and asthma attack rate:; but it is not clear
how they were obtained.
Presumably daily average concentration levels of specific pollutants
were used in the construction of the "hockey stick" curves shown on
pages 2-86 and 2-88. The discussion implies that "24-hour levels"
were used, but the precise nature of the air quality data used in the
threshold analyses is nofe made clear. j
There could be various reasons not explored by the study why the
thresholds for asthma attacks were lower on w)armer days. One of
these is that there may be more plume looping on warmer days. This
might result in localized, short-period, high concentrations, but
relatively low average concentrations. '
The validity of scientific work can be tested by the repeatability of
results. In this and the other CHESS studies there were factors
affecting asthma attack rates that were not cpnsidered and whose
effects are unknown, Such factors are: time spent Outdoors, percentage
of time windows are open, temperature change, relative humidity, etc.
The incompleteness of the study and the lack of 'Understanding of the
causes of the asthma attacks suggest that it might be repeated with
significantly different results. !
Short-term exposures to concentrations much higher than average
annual or weekly concentrations could have occurred in the commu-
nities studied that were near large sources of air pollution such as
smelters. There exists the possibility that asthma attacks could be
triggered by brief-duration high concentrations. Such exposures could
have been determined only inadequately by the procedures used in
the study. The report does not make clear why more attention was
not devoted to peak concentrations. i
4, Human Exposure to Air Pollutants in Five Rocky Mountain Com-
munities, 1940-1970 '
On pages 3-7 through 3-12 beginning with the second columnj
paragraph near middle of page, which begins "By comparing . . .".
There is not a simple relationship between average daily pollutant
emissions and average annual pollutant concentrations because the
receptor area is often now downwind. Also, some; consideration should
have been given to determining if the years for which data are available
were representative meteorologically.
(Page 3-11) Second paragraph, left hand side ;of page. Information
obtained during this investigation indicates that the ratio 1.63 ±0.21
should be 1.42±0.21. (The value 1.63 is the upper limit of this ratio.)
(Page 3-12) Emission ratios of particulate and sulfur dioxide for
1971 are omitted from this report. Therefore, it is not possible to
verify the ratios given here. :
(Page 3-12) According to information obtained during this investi-
gation, the two values for TSP listed as 99.5 for the years 1971-70,
should be 98.1 for both years.
1O1
-------
90
S. Prevalence of Chronic Respiratory Disease Symptoms in Adults:
1970 Survey of Five Rocky Mountain (Communities
Four comTn.TinJ.ties were studied: Bozeman, Helena (Low), Helena
(High), Anaconda, and Kellogg. These communities are all in moun-
tainous areas and subject to terrain effects, some much more than
others. Average pollution patterns are irregular so that exposures in
a particular community will vary considerably. It is particularly
important whether the community is somewhere downwind of ^a
source, such as Magna, or on either side of it, as might be the case in
Helena (High), or perhaps Kellogg. In these communities assump-
tions should not have been made about exposure without some effort
of sampling and/or meteorological modeling to determine what the
actual exposure might be. I
In this study long-period concentrations of sulfur dioxide and sus-
pended sulf ates are estimated. Also, the mbst important finding, which
was that excess bronchitis occurred with 2-3 year exposure to sulfur
dioxide concentrations of 177 to 374 /^/m3 and suspended sulfate
concentrations of 7.2 to 19.9 jug/m3 (in !the presence of low total
suspended participates), covers such a wide range of concentrations
it has hardly any practical significance. '
On page 3-31, in the Summary is stated i "Metallic sulf ates may well
have accounted for the findings of excess bronchitis." The nature of
the sulfates is nowhere mentioned in the -report. The smelter is ap-
parently the source of only some of the sulfates, and the nature of none
of the sulfates has been determined.
6. Frequency of Acute Lower Respiratory Disease in Children: Retro-
spective Survey of Five Rocky Mountain Communities, 1967-1970
Sulfur dioxide levels are crudely estimated from sulfation plate
(or candle) data and smelter emission rates. Suspended sulfates are
estimated from suspended particulate data using a ratio found in
East Helena, Helena and Magna. It was assumed that the estimated
values prevailed throughout the community. This is unlikely to have
been true considering the topographic effects in the Rocky Moxtntain
Communities. A major shortcoming in this paper is a failure to make
clear the inaccuracies that might be associated with estimating the
pollution levels.
There is an irregular distribution of annual average sulfur dioxide
concentrations in. the vicinity of a smelter. These differences affect
concentrations in East Helena and can be! confirmed by examination
of a report on the Helena Valley Envirqnmental Pollution Study.1
During 1965-66, 66 measurements of suspended sulfate indicated
an average concentration of 7.9 ng/ia3. The observed total suspended
particulate during this period of time was 96 /tg/m3. Hence the ratio
of suspended suMate observed to particulate during this period was
.082. A series of 25 measurements made during the time period April
to August 1968 indicated a suspended sulfate to total suspended
particulate ratio of 0.057.
Then to determine a ratio range, measurements made during the
period June through October at two stations of the Helena Valley
study were used. The original information is as follows:
i Helena Valley, Montana, Area Environmental Pollution Study, Environmental Protection Agency
Office of Air Programs, Research Triangle Park, North Carolina,' January 1972.
702
-------
91
Location from stack Suspended Suspended
paniculate- sulfate
Station De& Miles Oig/m:) g/in3. It may also be noted that the ratio for Magna is much*
larger than for other locations. No explanation is given for including-
the Magna value in the average. From the applications of these ratios
to estimates of suspended particulate concentrations came the con-
clusions pertaining to the health effects of sulfates. It is stated that.
"we conclude that excessive acute lower respiratory illnesses can be
expected among asthmatic and nonasthinatic children who are exposed
for longer than 2 years to elevated annual average sulfur dioxide
levels (177 Mg/m3) accompanied by elevated annual average suspensed
sulfate levels (7.2 /xg/m8) in the presence of low levels of suspended
particulates (65 /ug/m3)." This statement refers to findings in Ana-
conda. No mention is made that concentrations of suspended sxilf ate
might range from 3.51 yg/m3 to 10.92 ng/m? because of the uncertain.
nature of the ratio alone. ;
The following statement from the Helena Valley Study (Page 45) is
of interest. "These results (i.e., average concentrations ranging from
2.9 to 4.5 Mg/m3 for the period of the study) are considerably below the
national average and serve to point out that there is no problem from.
sulfates. It should be noted that the sulfate content is evenly distrib-
ute^ among the five sampling stations, which: indicates that there is
no significant increase in concentration in the vicinity of the East
Helena industry." ;
Suspended sulfate observations made by tjae Montana State De-
partment of Health and Environmental Sciences suggest that average
annual concentrations of 4 or 5 Mg/m3 occur! in Anaconda [See dis-
cussion of Estimate Suspended t Sulfate Concentration, 7.2 jtg/m3
(Anaconda), Chapter II]. Assuming that suspended sulfate concen-
trations of about 7 Mg/ua3 did occur in the Rocky Mountain communi-
* Referenced publication contains no Information pertaining to Magna.
1O3
-------
92 ;
ties where certain health effects were observed, the source of the
suspended sulfate is inadequately determined. The study findings are
much too incomplete to call for the stringent control of suspended
sulfates as has been done on page 3-51.
7. Prevalence of Chronic Respiratory Disease Symptoms in Military
Becruits: Chicago Induction Center (Paragraph 4-®)
Exposure estimates in this study are extremely crude. In tiie sum-
mary the following statement is made, "Available evidence indicates
that exposures lasting 12 years or morei to ambient air pollution
characterized by elevated annual averagfe levels of sulfur dioxide
(96 to 217 jug/m3), suspended particulates (103 to 155 pg/m3) .and
suspended sulfates (14 Mg/m3) were accompanied by significant in-
•creases in the frequency of chronic respiratory disease symptoms."
The 96 Mg/m.3 value is the average urban core value for 1969-70,
which ranges from 54 to 138 Mg/m3, whereas the 217 pg/ia.3 is an average
•value for five suburban communities for tlxe year 1969. Going back
12 years concentrations were much higher. During the period 1960
through 1965, the lowest value was 222, and there was a high of 344
in 1964. For the five suburban communities there was data only for
one other year. It averaged 183 jug/na3. The 14 pg/m3 concentration for
sulfates is for a period of 7 years, not 12 as' stated. It basically repre-
sents data for the Chicago core area, with some scattered observations
from East Chicago and Hammond, Incl. The average concentrations
for the city should be somewhat less than in the core area. Use of the
core area value would generally result in an overestimate.
It is difficult to characterize exposures lasting 12 years for the entire
Chicago area. Either this should have been done in very general
terms, nonquantifcatively, or a greater effort should have been made
to present more representative estimates, j
The assumption is being made that sulfate observations made at a
central urban location in Chicago, averaged with a few observations
from East Chicago and Hammond, Ind. are generally representative of
the entire Chicago area. : - •
(Page 4-8) Referring to the Chicago area the following statement is
made: "Each sampler location, identified by a station name in
Figure 4.1.2, represents the central business-commercial district of
that particular area." This statement is not true. Practically all, if
not all the samplers are located on the roofs of school buildings in an
effort to obtain representative community values. They were not
located deliberately in business commercial districts and do not
slightly overestimate area-wide concentrations as suggested.
(Page 4-23) In reading this paper abouti the prevalence of chronic
respiratory disease symptoms in military recruits, questions arise about
the actual locations from which the men came and the local pollution
levels to which they might have been exposed. Some rural occupations
result in high exposures to dusts, plant allergens, etc.
(Page 4-35) (Summary) The 12-year value for suspended sulfates
should be 16 mierograms per cubic meter, not 14, as stated. Also, it
appears that the concentrations of sulfur dioxide and suspended par-
ticulate are for only the period 1969-1970 and not for 12 years as
is stated. (See Table 4.1.A.6)
104
-------
93
8, Prospective Surveys of Acute Respiratory Disea.sf in Volunteer Fam-
ilies: Chicago Nursery School Study, 1969—1970
On page 4-41, in Table 4.3.1, it is not clear where the sulfur dioxide
data for the years 1959-63 come from. The Chicago network, which
would have provided community data, was not operating effectively
until 1964. :
On the same page the suspended stilfate data; are probably repre-
sentative for the core, area but are high to be used !as an average for the
city as a- whole. ... i '
A serious weakness in this sttidy is that the communities are ranked
Intermediate, High, and Highest according to a ranking that was
determined by suspended particulate values, whereas the most im-
portant finding pertains to sulfur dioxide. Referring back to Table
4.1.A.I, it can be seen that a considerably different ranking would have
resulted if the communities had been ranked i according to sulfur
dioxide concentrations. In Table 4.3.1, it may be noted that during.
the study that the "High" community had the lowest concentration
of sulfur dioxide.
Also note in Table- 4.I.A.I that the Highest communities include
GSA, which happens to be on the south edge qf the Chicago Loop
area. This station probably contributed considerably to the high
concentration of sulfur dio?cide attributed to the. Highest community
during 1969-1970, yet it is very npnrepresentative of a nursery school.
Also, note that the Highest stations include Carver, which for some
reason ranks highest because of suspended participate concentrations .
whereas the sulfur dioxide concentrations arc relatively low.
Sulfates are not considered in the summary of ,this study, which
seems to focus on sxilfur dioxide without quantitative considerations of
suspended sulf ate levels.
(Page 4-54) In the first paragraph of the Summary, the following
statement appears: "It is also possible that more recent lower air
pollution levels contributed to increased respiratory illness." On page
4-51 the following statement is found. "Acute respiratory morbidity
was significantly lower among families living in neighborhoods where
sulfur dioxide levels had been substantially depreased." These two
statements are contradictory and require clarification. The first
statement is remarkable. It can be interpreted tp mean that some air
pollution is good for you. Did the authors intend to say this? Such an
important finding is inadequately supported by the contents of the
report.
9. Human Exposure to Air Pollution in Selected New York Metro-
politan Communities, 19^4—WH1
An overusage of estimated data can be found on page 5-19. The
following two statements appear: (Left hand column, middle para-
graph) "Measured values for suspended sulfates for 1956-1970 were
available from the Manhattan 121st Street Station, and these values
were used for citywide values." (Last paragraph on page) "The
observed annual ratios of suspended sulfate to dpstfall for New York
City were used to estimate the suspended sulfate levels in Queens and
Bronx."
105
-------
94
10. Prevalence of Chronic Respiratory Disease Symptoms in Adults:
1970 Survey of New York Communities
Three communities were compared: Rivethead, Long Island, a low
exposure community, Queens, an intermediate exposure community,
and the Bronx, a high exposure community. Parents of all children
attending certain elementary schools located within 1.5 miles of an
air monitoring station in each community were asked to participate
in the study. JEach child was given a questionnaire to be filled out by
his parents and returned. !
Regarding exposure, were the concentrations measured at the
monitoring stations generally representative? Assuming a person
remains reasonably near the station, in this case within lj| miles, and
breathes the outside air, the station measurements would be generally
representative for long-term average exposure. Maps or annual
concentrations which are for sulfur dioxide and suspended particulate
matter, show reasonably uniform concentrations across the study
areas. However, as has been mentioned in the report (5.1) Human
Exposure to Air Pollution Selected New York Metropolitan Com-
munities, 1944-1971, by Thomas D. English, et al., the Queens Com-
munity lies about 1 mile west of the John F. Kennedy International
Airport. The effect of this airport and the various other possible
sources of -air pollution that could have; affected, particular local
areas were not determined.
The fact that the CHESS monitoring si^es were the same as used
in the city air pollution control programs suggests that the sites
were picked and are being used because they seem to be generally
representative. \
More important than the representativeness of the monitoring
site locations in this study is the proper interpretation of the effects
of the greatly reduced pollution levels during the period 1969-1971.
It is not meaningful to draw conclusions from sulfur dioxide exposures
ranging from 144 to 404 Mg/m8 and sulfat'e exposures ranging from
9-24 jig/m3, as was done in this study. The| implication is that health -
effects can be caused by the lowest concentrations mentioned, and this
is not shown in the study. Also, it is stated that annual sulfur dioxide
levels of 50 to 60 Mg/m3 (accompanied by annual average suspended
sulfate levels of about 14 Mg/m3 and annual arithmetic mean total
suspended particulate levels of about 60 to 105 ptg/m3) could be assoc-
ciated with such effects. These are levels that were measured in 1971,
whereas in the study there seems to have been no way to have differ-
entiated between the effects of pollution in 1971, or that might have
occurred during some earlier time. It is not reasonable to infer that
lower pollution levels are responsible for the observed health effects.
11. Prospective Surveys of Acute Respiratory Disease in Volunteer
Families: 1970-1971 New York Studies
In this study families were telephoned once every two weeks and
questioned about possible health effects. The families resided within
1 to 1.5 miles of the air monitoring stations.
In the discussion it is stated that acute lower respiratory disease
morbidity can be attributed to exposures! to 2 to 3 years involving
annual average sulfur dioxide levels of 256 to 321 jug/m3 (accompanied
by elevated annual average levels of total suspended particulate of
1O6
-------
95
97 to 123 Atg/m3 and animal average suspended sulfate levels of
10 to 15 fig/nr). These values are average values for the period 1966—
1970, a five year period, and not period of 2 to'3 years as indicated.
Also, they are the averages for the Bronx and Queens, respectively,
and therefore do not represent a range of concentrations that would
have occurred in any particular community, as implied. For example,
the sulfur dioxide concentrations in the Bronx; ranged from 184 to
472 Mg/ni3 and in the Queens from 131 to 420 jug/m3, during the five
year period. Three year averages are 174 to 247 Mg/m8? aild two year
averages, lower still.
On page 5-16 the dustfall concentrations shirwn in Figure 5.1.21
seem to be greater than would be obtained fromithe data presented in
Figure 5.1.16. ;
On page 5-36 (Table 5.2.1) the values in this table seem to come
from Table 5.1.A.8, The values in the column headed 1949-58 are,
except for dustfall, for shorter time periods. For example, the values
for Queens come from data for the years 1956-58.
On page 5-45 (Summary) we find that since the concentration data
base comes from Table 5.1.A.8, the long term exposure values repre-
sent a period of less than 20 years. '.
Further, it is stated that there is a distinct possibility that in-
creased suscep'tibility to" scute lower respiratoryillness is maintained
or induced by exposures involving annual average sulfur dioxide
levels of 51 to 63 Mg/EQ8 (accompanied by annual average total sus-
pended particulate levels of 63 to 104 jig/in3 and annual average sus-
pended sulfate levels of 13 to 14 Mg/m3). The 51 io 63 MgM3, is a range
resulting from two different analyses of samples (see page 5-53).
It represents uncertainty in measurement techniques rather than a
range of exposure as would be interpreted. These concentrations
and the suspended sulfate concentrations of 13 to 14 Mg/m3 happen to
have occurred in the Intermediate I and the Intermediate II com-
munities during 1971. This particular study as conducted could not
have differentiated between the effects of these levels of pollution
and the effects of higher levels that occurred earlier.
Only average annual concentrations were considered and not peak
or episode concentrations. !
IS. Aggravation of Asthma by Air Pollutants: 1970-1971 New York
Panelists who lived within a 1.5 mile radius of three monitoring
stations in communities identified as Low, Intermediate I, and Inter-
mediate II, because of their average air pollution concentrations,
recorded asthma attacks each day in a diary for a period lasting 32
weeks, October 1970-May 1971. From a statistical association between
asthma attack rates, 24-hour average concentrations from the moni-
toring stations, and daily minimum temperatures from airports near
the study communities, it was concluded that 24-hour suspended
sulfate levels of 12 Mg/m3 on cooler days (Tmll, equal to 30 to 50°)
and 7.3 Mg/m8 on warmer days (Tmm greater than 50°F) were thresh-
olds for the induction of excessive asthma attacks. No firm evidence
could be found to associate elevations in sulfur dioxide (100 to 180
jug/m3 on 10 percent of days) with excessive asthma attack rates on
either cold or warmer days.
1O7
-------
Regarding exposure levels, there is much! less assurance that daily
average levels throughout a community would be more or less uni-
form, than would be the case with annual average levels. More moni-
toring stations might have been operated, or mobile stations used,
to determine how pollution exposure varied'from location to location.
The determination of such differences in air pollution concentrations
might have been important, but probably more important is that the
other factors (in addition to the observed air pollutants) that could
have caused or contributed to the asthma attacks were not examined.
It would not be worthwhile to refine the information on the distri-
bution of the air pollutants studied, unless a greater effort were made
to study all of the various possible causes of the asthma attacks more
thoroughly.
The study focused on the effects of minimum temperature. The
possible effects of other meteorological variables could also have been
explored. Of particular interest would be the effects of sudden, large
temperature changes. .
It is not made clear why minimum instead of average, or even
maximum, temperatures were picked for correlation. Generally
there would be less actual exposure to minimum temperature, which
usually occurs about sunrise, than to warpaer temperatures. Asth-
matics would generally be expected to protect themselves from colder
temperatures, staying indoors and keeping;windows closed, whereas
on warmer days they might be more subject to exposure to outdoor
air with its assortment of possible allergens. There are diverse reasons
why temperature might be an important factor determining asthma
attack rates. No attempt was made Jin., the study to provide an
explanation. ; '
It is expected that there would be noticeable temperature differences
between Biverhead (the Low community); and Queens (the Inter-
mediate I, community). Although it is stated that the temperatures
come from nearby airports, the temperature curves plotted in Figure
5.4.1 seem to be identical for both communities. It may be noted
that a different curve is plotted for the low!community Figure 5.5.2.
(Figure 5.4.4) Although at a glance it appears that for the Inter-
mediate community that the "Attack Kate" and the "Suspended
Sulfate" curves are similar, close inspection shows that more often
than not, they are out of phase. Between the 2nd and 3rd week the
attack rate (AB) curve continues down as the suspended sulf ate (SS)
curve starts up, between the 10th and llth week the AB-curve
continues down after the SS-curve starts up, between the 14th and
16th week the AB-curve goes up while the SS-curve continues down,
between the 19th and 20th week the AB-cur,ve starts up while the SS-
curve continues down, and again on the 27th week the AB-curve
rises a week before an increase hi the suspended sulf ate concentrations.
In all, three of the five increases in attack rate precede, rather than
follow, increases in suspended sulf ate concentrations.
IS, Frequency and Severity of Card'iopulmo'nary Symptoms in Adult
Panels. 1970-1971 New York Studies (Paragraph B.ff).
Symptom diaries were maintained daily, for the 32-week period
October 8, 1970 through May 22, 1971, by four panels, depending
on state of health. The panelists were distributed in three communities
1O8
-------
97
i
and lived within 1.5 miles of air pollution monitoring stations. 16
was concluded that elderly panelists in the low iexposure community
reported higher symptom rates on days when sulfate levels exceeded
10 jug/m3. There seemed to be good evidence bf a threshold effect
between 6 and 10 fig/m?, with a greater morbidity excess on warmer
days. •
Since suspended sulfates seem to be more uniformly distributed
than a pollutant such as sulfur dioxide, the concentrations determined
by monitoring should be generally representative of outdoor exposure
ard in most cases indoor and outdoor average exposures would be
expected to be similar. The question not answered by this study is
whether or not the panelists are also being exposed to some other
causative agent, or stress factor, that might happen to correlate with
the sulfate concentrations. It, and not the suspended sulfate con-
centrations, might be the cause of the observed health effects.
(Page 5-91) (Figure 5.5.3) The low value of sulfur dioxide that
began at the 19th week and continued until the 24th week are sus-
pected of not being true values. Near the end of the last paragraph
on the preceding page it is suggested that meteorological conditions
may have been responsible. A careful study of the meteorological
conditions and fuel usage would be necessary to determine if these
might have caused the persistent low concentrations. However, a
scanning of the daily local climatological data shows no obvious
reason for the reported low values. ',
Furthermore, the minimum temperature curve for the Low com-
munity in Figure 5.5.2 is not the same as given in Figure 5.4.1.
The New York Department of Air Resources also reported a large
drop in concentrations following the mid-winter peak at the Queens
(Intermediate I) monitoring station, but reported values^ were never
as low, and a period of low values was not followed by a rise as shown
in the Figure. Further, the low values shown, which are about 25
jug/m8, or .01 ppm or less, are quite low for the New York metro-
politan area. Average weekly low values two or three times this value
would generally be expected for a comparable period.
14- Ventilatory Function in School Children: 1970-1971 New York
Studies (Paragraph 6.6).
Pulmonary tests were made in three elementary schools in com-
munities with different air pollution levels, and there were four rounds
of testing, November-December 1970, January 1971, February-March
1971, and April 1971. The children lived within 1J5 miles of a particular
air monitoring station. The Queens monitoring station is on top_of a
school where the testing was done. However, the Bronx station is on
top of a "court house hi the center of a busy commercial area" (page
5-6) and may not be close to the school. For the Riverhead com-
munity it is not made clear whether or not-the school and the monitor-
ing station are at the^same location or near each other. It is assumed
that the schools in Riverhead and the Bronx were within 1J^ miles of
the monitoring stations, but this is not actually, stated.
It was concluded that 9 or more years exposure to annual sulfur
dioxide levels of an estimated concentration of 131 to 435 jug/m3
(accompanied by suspended particulate levels of about 75 to 200
Pg/m3) and suspended sulfate levels of about 5 to 25 jig/m3 can be
i
77-590—76 8 •
709
-------
98
associated with a small but significant impairment in ventilator
function. These values are from Table 5.6.2,' and are the extreme big
and low values listed. There is an implication here that the low con-
centrations, 131 jug/m3 for sulfur dioxide and 5 ^g/m3 for suspended
sulfates represent threshold values. Actually they are only annual
average concentrations for the years 1969 and 1970. The observed
health effects may have been the result of exposure to much higher
concentrations in other years, or to some other cause.
16. VentHatory Function in School Children: 1967-68 Testing in Cin-
cinnati Neighborhoods (Paragraph 6.1}.
This study included a pair of public elementary schools in each of
six neighborhoods differing in socioeconomic level, race, or pollution
exposure. All children in one or two classrooms of the second grade of
the elementary schools were asked to participate in the study to achieve
sample sizes of 60 to 75 children in each of the six study sectors.
Ventilatory performance as measured by a spirometer was obtained 12
tunes from each child: once weekly in the months of November 1967
and February and May 1968. The tests were; administered on Tuesday
and Wednesday mornings. ;
Air monitoring stations were placed in locations within three blocks
of each school to provide samples representative of the air quality in
the neighborhood served by the school. No information is reported
on the distances of the homes of the children from the school. Ap-
parently it was assumed that the home environment and the school
environment were the same. Indoor soiling, index and sulfur dioxide
observations were taken in the schools, but results are not reported.
It is reported that it was determined that indoor and outdoor sulfur
dioxide, soiling index, and suspended particulate levels measured
over ^the 24-hour or 4-hour period directly preceding pulmonary
function tests did not consistently correlate with the test values.
Details of this lack of correlation are not given, but it was concluded
that 'Ventilatory performance of children thus did not appear to be
acutely affected by variations in pollutant;levels on the day of the
test." Possible exposures over intermediate periods, say three days or
one week, prior to testing were not considered. Conclusions seem to be
based on possible long-period exposures, probably over a lifetime.
Concentrations of sulfur dioxide were low1 (less than 52 jag/m3) in all
areas, so health effects were attributed to particulate pollutants
independent of atmospheric levels of gaseous sulfur dioxide.
Average sulfate levels during the period of the study were observed
to be between 8.9 and 10.1 Mg/m3, in the polluted lower middle white
community, but previous average exposure .was estimated to be 10.7
to 12.1 jug/m8, based on the National Air Surveillance Network station.
The average suspended sulfate level in the clean white sectors was 8.3
/ig/m.3, a relative difference of I3;percent. (The largest differences in
area exposure were in the concentrations of suspended particulates.
Levels of total suspended particulates were 131 jtg/m3 in polluted sec-
tors and 61 to 92 jug/m3 in clean sectors.
In reading this paper one wonders about the psychological inter-
action between the children and the team members administering the
tests, who could anticipate the outcome of the experiment. The curves
for the black children in Figure 6.1.3, are particularly interesting.
110
-------
99 ,
There was only 0.1 jig/m? difference between the average sulfate levels
between the Clean Lower Middle Black and the Polluted Lower
Middle Black, yet there was a large difference in ventilatory function
that disappeared as the study progressed. The discussion acknowledges
a culturally determined response of these children to the white
interviewers. The effect of the interviewers on all of the children may
have been greater than was realized.
It was concluded that the observed area effect on ventilatory per-
formance was probably attributable mostly tp suspended sulfates.
There is no particular reason to challenge the estimated average ex-
posure of the children to suspended sulfate concentrations. However,
it should be noted that the sulfate level in the (f clean" school #4, was
9.1 Mg/na3> more than the 8.9 MgM3 concentration in the polluted
schools #8, #11, and #12. The sulfate concentration differences be-
tween the schools is acually very slight. It seems far-fetched that the
observed average differences in ventilatory performances were caused
by a difference in suspended sulfate concentrations of only 1.2
C. STJPEBFICIAL AND PEBFT7NCTORY TBEATMENT OF METEOBOLOGICAIi
INFOKMATION *
A discussion of meteorology is included in the CHESS report on the
studies in the Salt Lake Basin, Chicago, and 'New York. The treat-
ment is so superficial and perfunctory that there is hardly any useful-
ness. Little attempt has been made to tie the meteorological informa-
tion with information subsequently presented!, or to use it to make
clearer the possible reasons for the study findings.
The air now over the 'Salt Lake Basin is much more complicated
than implied by the wind information included on pages 2-4 and 2-5.
Winds in the vicinity of the Oquirrh Mountains, where the smelter
is located are significantly different from those at the Salt Lake
City airport. Also, the wind rose for Hill Air: Force Base shows the
effect of a nearby canyon and is not indicative of air _ trajectories
affecting the Ogden area generally. The meteorological discussion on
these pages could have made clearer the fact that Ogden is rarely,
if ever, affected directly by the smelter plume, and that Salt Lake
City is infrequently affected.
Similarly, Chicago winds are inadequately explained by the Midway
and O'Hare Airport wind rose; and the reader remains unsure of the
wind conditions at Riverhead even though John F. Kennedy and
La Guardia wind roses are given.
Even so, the wind rose data, no matter how carefully given, is of
practically no value for interpreting the CHESS studies because
source-receptor orientations are generally unknown and were not a
factor considered in developing the CHESS conclusions.
D. INSUFFICIENT EXPLORATION OF POSSIBLE RELATIONSHIPS BETWEEN
METEOBOLOaiCAL CONDITIONS AND ASTHMA ATTACK BATES
With regard to meteorological correlations; the CHESS studies for
the most part only explored a possible relationship between asthma
attack rates and minimum temperature. In designing studies of
asthma attack rates it would seem that steps would have been taken
111
-------
100
to investigate other meteorological factors also. Of particular interest
would be the effect of large and sudden temperature changes. Some
other factors that could affect asthma attack rates are relative
humidity, frontal passages with their change of air mass, wind
direction and speed, precipitation, type and amount, frozen or wet-
ground, and possibly atmospheric pressure.
A community study of asthma attack rates would also be expected
to take note of any air pollution episodes resulting from atmospheric
stagnation that occurred during the course of the study. If there were
none, notice would be taken of this also.
B. FAILURE TO GONSIDEE PEAK AND EPISODE CONCENTRATIONS
The CHESS studies do not adequately consider the short-period.
exposures to high concentrations that would occur in the smelter
communities. It is conceivable, that two different communities might
one of them might be subjected much more to short-period high level
concentrations that could trigger health effects, asthma attacks in
particular. Three causes of much higher than average concentrations
are plume looping, plume fumigation associated with the dissipation
of a surface-based temperature inversion, .and episode conditions-
caused by a shallow mixing layer and light winds, perhaps with fog.
In order for the CHESS studies to be complete much more needs to-
be known about the short-period exposures to high air pollution
concentrations in the smelter communities. :
P. tTSE OF A SINGLE MONITORING STATION TO DETERMINE THE EXPOSURE:
OF A COMMUNITY ,
Although it may be reasonable to assume an average concentration
based on a single monitoring station, in a large metropolitan area
without any well-defined sources of air pollution, such an assumption
may not be reasonable in a mountain community with a single large
source such as a smelter and possible topographic effects. For example,
air pollution concentrations in East Helena, j Anaconda, and Kellogg
are likely to be rather non-uniform. There is no assurance that the
air Duality values being used by CHESS are average, or that some
families were not subjected to much higher or -lower levels of pollution.
One would expect hi community studies such as conducted under
CHESS that more effort would be made to determine if a single
monitoring station is sufficiently representative, particularly in areas
where there is but one major source of air pollution. More use could
have been made of meteorological technique^ in the CHESS studies
to estimate exposures and the representativeness of (monitoring
stations, including the wider application of simple dispersion models.
Where there were particular sources, wind direction frequency infor-
mation should have been examined. Generally, a visit by a meteor-
ologist ^ to a monitoring station location would have been helpful in
determining its suitability. Temporary or mobile monitoring stations
might have been used to determine if there were significant concen-
tration differences from one location to another within a community.
112
-------
101
G. FAILURE TO ESTABLISH SIMILARITY OF EXPOSURE AND STRESS
FACTORS BETWEEN COMMUNITIES IN THE SAME STUDY, EXCLUDING
THE EXPOSURE TO SPECIFIC POLLUTANTS
I
No account seems to have been taken in the CHESS studies of
actual exposure inside of buildings with various kinds of ventilating
•or air conditioning systems, or from air at various elevations above
•ground. Neither was there any effort to determine if there were any
significant differences in vegetation. Any of these things might have
-caused the actual exposure to air pollutants and' allergens to be quite
different from that estimated on the basis of usual ambient air moni-
toring. Further, the general stress factors affecting health are no doubt
different in the Bronx than they are in Riverhead, or perhaps between
Ogden and Magna, where, for example, homelife might nave been
upset by the smelter strike. I
Still another shortcoming^was that no information was collected on
possible exposures of individuals when tfaey"weTrtroTrteidtnjf1fae study
community. i
The CHESS reports do not carefully distinguish between outdoor
concentrations such as might be measured at >a monitoring station
and true exposure, which might be significantly different. Throughout
the report the concept of dosage is never used, Whole communities
were assumed to have the same exposure and presumably were con-
sidered to have received the same dosage. A more careful study might
have considered the ^actual community dosages of air pollutants,
which might be significantly different, even though the potential
exposures are the same.
H. IMPRECISENESS OF MONITORING STATION LOCATIONS
Monitoring stations used in the CHESS studies are not precisely
located within the. community with which they are identified. In some
instances it may not be necessary to know the exact geographic
location, but it is not ^true for a community such as Anaconda. The
Rocky Mountain Studies report leaves completely indefinite where the
Anaconda, station was located with respect to the smelter,, or the com*
munity itself. It may even be possible that more than one station is
involved in the determination of the exposure' estimates. Similarly
the reader of this particular report has no idea where the monitoring
stations in the two Helena areas were located. Perhaps the general
nature of the epidemiological studies is an excus0 for omitting details,
b\it a follow-up effort to verify some of the GJHESS results would
require the precise location of certain monitoring stations from which
exposures were determined.
I. INEXACT LOCATIONS OF RESIDENCES OF INDIVIDUALS STUDIED
Assuming that an attempt was made to
•apply meteorological
•* * " • -- .vwpv&e-v **• r~t
modeling techniques to better determine erppsure in the CHESS
studies, there would be a severe limitation on what could be accom-
plished because the locations of residences of participating individuals
within the study area are never given except!to say that they are
within a certain radius of a monitoring station.,
113
-------
102
There .are two problems. First, one cannot ,be sure that the popu-
lation, was all within the stated radius as claimed. Inquiries made in.
New York and elsewhere indicate that some of the children studied
rode school buses for greater distances and; that for various reasons
some other participants in studies lived outside of what'might be
considered the effective radius of an air monitoring station. Second,
and probably much more important, are the uncertainties with respect
to populations in the vicinity of single large sources such as the
smelters in the Rocky Mountain communitie^. Insufficient information
has been given about the distance and direction that participants
lived from the smelter stack, and perhaps where they live with respect
to significant topographic features. When determining possible expo-
sure, it is particularly important to know if participants lived in
different directions from the stack because of the effects of wind
direction. ;
In general, the conclusions from the CHESS studies are weakened
because there is a lack of assurance that the .populations were actually
located as claimed, and in the smelter communities that the popula-
tions were so located with respect to a stack, or stacks, that it is
feasible to represent exposure with a single monitoring station^ per
community. The presentation of maps that showed, at least in a
general way, the location of the residences of participating individuals
would have helped to engender more confidence in the estimated
exposures.
These commentSj many of which are minor and have various orders
of significance in total implication, are an indication of the many
observations which cumulatively support the need for careful peer
review of each research project. In the limited time available for
analysis of historical research of this nature, it is difficult to determine
the net impact of these errors on the final conclusions in the study but
such errors certainly tend to reduce the credibility of the research
results as published. ;
114
-------
APPENDIX B :
LEGISLATIVE HISTOBY OF THE CHESS PKOGBAM
i
The programs now referred to as CHESS studies evolved during:
a period of increasing concern about air pollution. This same period
produced significant changes in government organization to cope
•with new air quality legislation and substantial 'increases occurred in
the funding levels in support of this legislation. In order to place this
evolution in some legislative perspective, the following brief history
has been prepared. The information is summarized and citations are
primarily (except as noted) from House Appropriations Hearings
held during the period 1964-1975.
By 1964, weU before the 1974 publication of'the 1970-71 CHESS
studies, air pollution health effects had been under investigation
within the Department of Health, Education and Welfare's (DHEW)
Bureau of Disease Prevention and Control, National Center for Air
Pollution Control. The Center was'conducting clinical and laboratory
studies in humans and experimental animals. The plans of the Center
included a comprehensive program of research on the health effects of
exposure to oxides of sulfur alone and in combination with particulates.
In addition to clinical and laboratory studies, epidemiological studies
were planned for,the St. Louis-Philadelphia areas as well as a con-
tinuation of a statewide study in Alabama to assess the relationships
of air pollution to the prevalence of pulmonary emphysema. The
budget proposals in 1964 were the first to be specifically rela-ted
to the Clean Air Act which had been adopted in 1963. At that time
also, there was a concern within the scientific community about the
conversion of SC>2 to sulfate and the potential health effects of such
sulfates. (Estimated funding levels for 1965 were $2.1 million for
the medical and biological studies and $1.5 milliqn for the epidemiolog-
ical studies). ;
In 1965, a spokesman for the Air Pollution Control Center in-
dicated during hearings that the 1963 Clean'Air Act had opened
"a new era of opportunity to cope with the growing national problem
of air pollution.. . . There is overwhelming evidence that air pollution
is a serious threat to public health and welfare as well as an economic
burden costing the Nation several billion dollars annually." In the
request for appropriations at that time, the, Center indicated an
intent to continue the same programs supported in the previous year.
Again, studies of the acute and chronic effects of inhalation of oxides
of sulfur were specifically cited as receiving! support. Human ex-
periments to study the effects of these pollutants under controlled
laboratory conditions which had been accelerated in 1965 were
designated for expansion in 1966. Additional!, surveys to identify
field sites with varying levels of sulfur oxides and to clarify the
effects on human health of single pollutants or single-source pollutants
(103)
115
-------
104 !
were identified as priority fields for investigation. A small increase
to support these additional studies was requested.
The field studies identified for continuation in the 1967 program
included the Alabama air pollution-emphysema studies and the
detection of asthma epidemics; the epidemiology of air pollution of
asthma in New Orleans, and the relationships between air pollution
and absenteeism. It was also indicated that in 1967 there would be
a retrospective survey of the effect of air pollution on human health
through the use of Social Security disability records, Veterans' Ad-
ministration records, and^ insurance records. A study of nursing
home residents as a high-risk group was to ibe initiated and a study
to survey air pollution in San Antonio was \planned to place special
emphasis on the synergistic effect of man-toade air pollution and
aeroallergens. A mortality study in large cities to determine if excess
mortality could be correlated with air pollution was considered for
support.
Studies scheduled for continuation in 1968 included the Alabama
studies mentioned previously; the New Orleans studies; the continua-
tion of the excess mortality studies; and thje designation of Chicago
as one area for studying the correlation of disease with air pollution.
Plans for 1969 again emphasized the need for continuation of studies
on several air pollutants including sulfur oxides. By .this time, the
testimony being provided to the Appropriations Committees was
citing as justification for some of the work the new requirements
specified in the 1967 Air Quality Act. The need for data to develop air
quality criteria was being emphasized as a prime goal for the continua-
tion of many of the epidemiological studies. The criteria for sulfur
oxides, had been published. By July 1, 1968, the reorganization plans
of the Administration had resulted in the establishment of the Con-
sumer Protection and Environmental Health Service, and the re-
sponsibility for air pollution programs was assigned to the National
Air Pollution Control Administration (NAPCA). Objectives of this
reorganization included increased autonomy and a higher visibility
within DHEW for the air pollution control programs. Progress cited
during the • 1969 hearings included the publication in the Federal
Register of February 11, 1969 of the air quality criteria and related
information on control technologies for sulfur oxides and particulate
matter. The discussions also directed attention to the increased level
of activity required to generate data on health and economic effects
of pollutants. Plans for leasing facilities at1 Research Triangle Park
(RTF), North Carolina, for NAPCA were under way and an increase
of almost $11 million for this transfer was requested. Among the many
justifications offered in support for the concentration of air pollution
research at RTP and the move to that location of personnel from
DHEW Washington and Cincinnati facilities was the planned con-
struction of the new National Institute of Environmental Health
Sciences at RTP. It was believed that a consolidation of NIEHS and
NAPCA at RTP would facilitate general planning and centralization
of related research programs. A continued expansion of the general
air pollution monitoring network was supported.
Hearings in 1970 indicated that further reorganization of the
Environmental Health Service had occurred iwith the splitting out of
the Food and Drug Administration. Fiscal Year 1971 was designated
116
-------
105
as a key year for developing Federal-State machinery for effective
na tional control of air pollution. A majority of NAPCA personnel were
located at RTF and close liaison with. NI'BHS was cited as real and
viable. By this tame, NAPCA was referring to the epidemiplogical
studies which were being correlated with air monitoring collection as a
health surveillance network (HESN). The effort was described as a
"medical research effort destined to assess changes hi the health of
the population as a result of air pollution control program activities.
It is also specifically directed toward conducting studies in unique air
pollution types so we can have the kind of data we require for the
proper development of air quality criteria." A Chattanooga, Ten-
nessee study of nitrogen dioxide emissions was described as a specific
example of the type of study intended to provide data for further air
quality criteria development. At this time the! HESN program was
being 'highlighted among the several types of epidemiological studies
and funding levels for ITY 1970 were estimated for this program at
about $400,000 with an expected growth in 1976; to about $2.3 million.
Other epidemiological. studies had been initialed or were planned
including studies of air pollution in New York and Seattle, and in
several university grants. In general, the HESN program involved
cities purportedly selected for high to low exposure gradients to specific
pollutants to correlate with appropriate health indicators, as a means
to measure health effects. The HESN program had as its objective the
construction of a .framework of information in \ which selected health
parameters such as pulmonary functions, acute respiratory disease and
chronic respiratory disease could be evaluated at one or more points in
time. The network was to be initiated on a nine-city basis in fiscal
year 1970 and was to be expanded to a 12-city basis in fiscal 1971.
By appropriations hearing time ha April 1971, a major Executive
Reorganization had occurred (Plan No. 3 of 1970—December 2,1970)
and the air pollution program had been transferred from DHEW
NAPCA to the newly created Environmental Protection Agency.
At this time, Mr. Euckelshaus, the first Administrator of EPA to
appear before an appropriations committee, prdvided some summary
data about the health effects surveillance studies which had now been
given the new acronym CHESS (Community Health Effects Surveil-
lance Studies). Despite this change in title, it1 was obvious that the
CHESS studies were simply a further expansion and formalization
of the HESN programs which in turn had evolved from earlier effort
in epidemiology. Several of the comments offered by Mr. Ruckelshaus
are of interest as they relate to the priorities associated with CHESS
at that time: ;
CHESS studies have already associated environmental pollution with a number
of health, problems. Air pollution was associated with significant increases in
acute respiratory disesases among schoolchildren and their parents who also
reported chronic respiratory disease symptoms more frequently. Both parents
and children residing in polluted areas exhibited significant decreases in lung
function not accounted for by other factors. Asthmatics were also adversely
affected, reporting more attacks on days with increased air pollution.
The CHESS concept is anchored to extensive environmental and health
monitoring in sets of communities demonstrating an exposure gradient for specified
pollutants. Three years of intense effort have been devoted to the design and
field testing of the health impact indicators now a part pf CHESS. The air quality
aspects of CHESS will not be fully operational until 1973 and will evaluate the
health effects of specific pollutant sets. In other words', the air quality aspects of
117
-------
106
CHESS are designed to obtain the maximum amount of health effects information
about the most important air pollutants. These effects win be continually moni-
tored as pollution is controlled and can thus document the health benefits of
abatement. Extending intensive CHESS air monitoring with limited health
monitoring to a larger number of cities would probably be counterproductive in
terms of health information. '
First, to evaluate existing environmental pollutant standards as these relate
to health. Second, to measure pollutant burdens in exposed populations. And
third, to determine the health benefits of environmental pollution control; that
is decreased adverse health effects as a result of improved environmental quality.
Presently, major CHESS studies are being conducted in the following areas:
New York City; Chattanooga, Tenn.; Birmingham, 'Ala.; Salt Lake City, Utah;
and Los Angeles, Calif.
CHESS operations essentially involve four basic: integrated functions; data
collection; bioenvironmental measurement; information synthesis ', and research
and development.
Testimony at the time indicated that about $6.7 million was
programed lor the CHESS program for fiscal year 1973 with opera-
tions anticipated in 27-30 communities in fiv& geographical areas. To
add additional areas was estimated to require about $1 million and 15
staff persons per area. CHAMP (Continuous' Health Air Monitoring
Program) was described for the first time during these appropriation
hearings as an integral portion of the CHESS program which would
eventually be the total air measurement part of CHESS. As noted
by Dr. Greenfield: ;
Comprehensive monitoring is required to relate changes hi sensitive health
indicators to existing environmental pollutant levels, i From these data, pollutant
effects that occur from very short exposure, that is, from 10-minute peak to 24-hour
average values, can be documented'and available for ambient air quality standard
setting. To date, stations [CHAMP] in the Los Angeles Basin have been selected,
partially equipped, and data is being transmitted to North Carolina [RTF], The
cost for initiating the more comprehensive CHAMP , program, to have 27-30
completely equipped and installed stations, is $2,400;000. The in-house costs are
$750,000 and the contract costs are $1,650,000. The contract includes $465,000
for operating the stations in Los Angeles for this year. The yearly operational
costs for the 27 to 30 stations will be $1,240,000; $465,000 is in contract and
$775,000 of in-house funds.
During the 1973 hearings, the question was raised as to whether
the health effects being measured in the CHESS studies were actually
based on medical diagnosis and not on individual reports of health on
a questionnaire which "then some research clerk .codes as a specific
disease." Mr. Euckelshaus replied:
To the fullest extent possible, the medical response to EPA health questionnaires
are validated by appropriate techniques. It would 'not be possible to have a
physician visit the home of thousands of respondents to ascertain if their responses
to medical questions were correct. However, we can; assure the committee that
the questionnaires we employ are a representative subset of the responses received
and nave been appropriately validated against physician records. Further, the
questionnaires used in the CHESS program were previously well validated with
clinical examinations by competent investigators. :
In more detailed hearings held in September 1973 (Committee on
Interstate and Foreign Commerce) it was already being reported tihat
the CHESS program had provided information that adverse health
effects might be more closely associated with suspended sulfates than
with SO2 or total suspended particulates. The tune period estimated
118
-------
107
at that time to permit expansion, and modification of health effects
research on sulfuric acid and suspended sulfates in the ambient
atmosphere was 5-10 years. A very extensive expansion of the study
of oxides of sulfur with emphasis on sulphates, including sxilfuric acid.,
was being recommended. The FY 74 program rwas estimated to be
about $1.2 million with about $780,000 devoted specifically to the
CHESS studies, the remainder directed toward supportive biomedical
research, methods development, atmospheric chemistry and moni-
toring. Reference was made to the fact that the data from CHESS
suggesting significant health effects from sulfates had been presented
in'a scientific conference as early as December; 1972. The total air-
health research program was summarized as having expanded from
about $4 million in FY 69 to about $11 million in FY 74. During this
hearing, EPA personnel pointed out that the National Academy of
Sciences was reviewing a number of data bases on air pollution health
research and would be providing reports. >
At this time, the program at RTF was being affected by manage-
ment adjustments involving the transfer of temporary personnel to
permanent positions with some total reductions in personnel as a
result of insufficient permanent positions to accommodate all temporary
personnel. Thus, although the air-health effects program at RTF
showed some growth in general trends,--there actually appeared to be
some slowing of growth and some adjustments'in programs because
of the impact of inflation and fewer permanent positions than had been
anticipated. With regard to the National Academy of Sciences' re-
view of the EPA research program on sulfates, it is of interest to
note that shortly after these congressional hearings, the implications
of the potential dangers^f sulfates were discussed and acknowledged
by several investigators at an Academy conference. At this same con-
ference, however, Higgins and Ferris noted with regard to the EPA
CHESS studies: !
This is not the place to criticize these studies in detail. A few general points
only will be made. The need for a great deal of information in the shortest possible
time has meant that the E.P.A. has been forced to attefaipt too much too super-
ficially. A more deliberate approach with outside consultation might have led to
a more solid body of knowledge. As it is, the studies have a number of
deficiencies which make evaluation difficult. The samples studied, the response
rates in certain categories, the methods and procedures which have been used and
the analysis of the results can all be criticized. It is particularly disquieting, in
view of these deficiencies, that there has been a tendency to select findings which
point to an effect of pollution on health and ignore those which do not. Perhaps
the wisest conclusion which can be drawn from these studies in relation to current
standards is that taken in the aggregate they do not appear to justify any lower-
ing of the current standards for SOs and particulates. But on the other hand
neither do they suggest that any relaxation ot these standards is justified.1
It is of interest for purposes of this report that the Academy in
the final report to the Senate on the results of a special review of the
health effects of ah- pollutants noted also with regard to the CHESS
report:
The 1970-1971 CHESS'studies have received a good deal of criticism, although
it must be noted that many of the potential problems >are discussed in detail in
the CHESS document (EPA 1974). Specific problems with the two asthma studies
1 Higgins, Tan T, T. and Benjamin Q. Ferris. Epidemiology of Sulphur Oxides and Particles. In:
Proceeding of the Conference On Health Effects of Air Pollutants. TT.Ss. Senate. Committee on Public
Works. Proceedings of the Conference on Health Effects Of Air Pollutants. 93d Congress, 1st session.
November 1973. Serial No. 93-15. Washington, D.C., Govt. Print. Off. p. 247.
779
-------
108 :
described above include the relatively poor response rate and high turnover,
which may be inherent in diary studies of patients with this disorder; the presence
of anomalous data, such as a decrease in the temperature-specific risk of asthma
attacks associated with increasing sulfate concentrations on days when the
minimum temperature was 30-50° F in the Salt Lake area; the failure to consider
temperature change, rather than absolute temperatures; the lack of information
on atmospheric allergens, which conceivably vary with pollution concentrations;
and the tendency toward overanalysis and overinterpretation of the available
data. Reasonable preliminary conclusions from these studies are that they do
provide support for the association of air pollution with asthma attacks and that
they further implicate suspended sulf ates as an important deleterious component
of polluted air. It does not appear to be necessary or reasonable at this time to
draw firm conclusions concerning the 1970-1971 CHESS asthma studies, inasmuch
as data have been collected and analyzed for later.years and the reports will be
forthcoming soon. If the more recent studies are able to replicate the 1970-1971
findings, this will provide much firmer support for the interpretations advanced
by the CHESS investigators.3 ;
In April, 1974 Mr. Train, now the Administrator of EPA,pointed
out that he considered 1975 to be a critical year for the EPA air
programs. He emphasized the need for considerable more research in:
Better defining the health effects of air pollutants,' determining the causes and
effects of atmospheric concentrations of sulfates and developing improved in-
strumentation and analytical methods for monitoring air pollutants.
Our proposed 1975 program shows a substantial increase of approximately
$9 million over the funding levels of this area during' 1975.
The overwhelming bulk of this increase, some $7 million, will be used to con-
siderably accelerate our programs dealing with the health effects of air pollutants.
During detailed hearings in July 1975 'held by the Subcommittee
on the Environment and the Atmosphere, Committee on Science and
Technology, Dr. Greenfield, as a former administrator of EPA re-
search which included the CHESS programs,, provided comment on
this program from-his new perspective as a private consultant:
If you take the report from CHESS for 1974, the thing that is forgotten is that
this relates to the data that was collected in 1970-71. Where are the data for 1971-
72 and 1972—73. These have never been released. Yet, if you go back and look
at these data there are interesting ambiguities relative to the correlations between
sulfates and these health effects and whether or not they show up again the fol-
lowing year. There are interesting ambiguities, for example, in the number of
chronic bronchities shown in the 1970-71 period thatiactually dropped out in the
1971-72, as if you have suddenly removed chronic bronchitis, and they don't
remove from the population in that way, which raises serious questions as to
whether or not the sampling techniques were really correct.
Finally, we come to the question of health effects,; and potentially dangerous
pollutants. I do not want to spend a great deal of time describing the EPA Com-
munity Health Environmental Surveillance System, the CHESS program.
I'm sure that's been described here previously. But these studies have given rise
to the most recent sulfur oxides health effects data which have implicated sul-
fates. Suffice it to say that this was a massive epidemioiogical program, probably
the largest ever attempted, but suffers from many of'. the deficiencies inherent in
all such studies. These deficiencies identified by the CHESS investigators, them-
selves, have included an [in] ability to adequately quantify the exposure of in-
dividuals to various pollutants simultaneously present. These deficiencies pre-
clude the ability to unilaterally ascribe the observed adverse health effects to any
single pollutant. It should be noted that an even more fundamental problem
pervades those data that purport to demonstrate adverse health effects due to
sulfates. The sulfates measured were total water soluble sulfates. No attempt
was made to separate the various sulfate, and sulfite, compounds so as to de-
termine which forms may be innocuous and which |forms may be deleterious.
Neutral sulfate compounds such as sodium and magnesium sulfate occur naturally.
* 7.8. Senate. Committee on Public Works. Air Quality aad Stationary Source Emission Control. A.
Report By the Commission on Natural Resources, National Academy of Sciences, National Academy'of
Engineering, National Research Council. March 1975. Serial No. 94-4.94tb Congress, 1st session. Washing-
ton, B.C., Govt. Print. OS. p. 81.
720
-------
109
Are thfise compounds harmful? If not, what fraction of the measured sulfates
represent a potentially harmful concentration? The problem is further compound-
ed by the fact that nitrates may correlate better with the selected health indicators
than sulfates, and this, incidentally, is shown more clearly in the 1971-72 data
than in the 1970—71 data. That was what I mentioned earlier. In fact, in one
analysis made of the data, the most consistent correlation occurred with sus-
pended particulates including irritant respirable particulates.
This does not necessarily mean that some oxides of sulfur are not deleterious
to health. Rather, it implies that the health effects observed are probably due to
a complex, multipollutant mix of irritant respirable pa.rticulates consisting of sul-
fates, nitrates, organic and inorganic acids, et cetera, the sources of which are
as complex as the mixture itself. There is no question that we just do not under-
stand the nature of this complex mixture pollutant well enough today to either
assign a safe threshold level or ascribe a cause/effect relationship to any or all
components of the mixture. In the absence of such ability it is almost impossible
to stipulate an effective control strategy for these pollutants.
The March 1976 House Appropriation hearings duplicated much of
the information which had been provided the previous year. The 1976
planned program was modified slightly to indicate that the objectives
on air health research were intended to focus studies toward selected
poptilation subgroups to describe health effects of exposure to sulfur
oxides, respirable participates and other pollutants as well as refining
the data on acid sulfate aerosols. The general: impression is one of
reduction and refinement rather than expansion of surveillance pro-
grams. This impression is strengthened by examination of the pro-
posed 1977 program which does include a number of objectives re-
lated to clarification of sulfate air chemistry and health effects. Again,
no information is provided with regard to anticipated completion of
final analysis ol the remaining three years of diata from the 1971-74
CHESS data.
To briefly recapitulate this legislative history, the CHESS pro-
grams evolved from smaller sets of epidemiological studies initiated
well before the establishment of EPA. The concept of longitudinal
health studies hi selected cities to try to correlate health effects with
air pollutants was a well established requirement prior to 1970. The
phasing together of health indicators with real time measurements
of air pollutants received more attention as the air monitoring cap-
ability became more available. The history does show that the execu-
tion of such epidemiological studies requires considerable preplanning
and coordination; some of the studies have persisted for considerably
longer periods of time than had been estimated for the data gathering,
analysis, publication and utilization in policy planning. Associated
with the evolution of the surveillance type health studies^ was a
sequence of reorganizations and establishment of new organizations
which undoubtedly produced significant management problems in
maintaining continuity for projects requiring long term commitment
of manpower and management attention. The CHESS program thus
was essentially a continuous program, including many specific studies
not originally contemplated as a part of a total system of programs.
Reorganizations within EPA, such as the consolidation of laboratories
at B.TP under a different management system and significant re-
structuring of data processing systems are identified as other tem-
porary obstacles to effective progress. It now appears that the
epidemiological studies of the health effects of air pollutants have
been somewhat restructured and reduced in scope to focus on specific
subsets of populations for more specific objectives.
121
-------
110
i
It is evident that the February 1976 Los! Angeles Times series of
articles on the 1974 CHESS report were preceded by a number of
public discussions of the total problem of the'health effects of sulfur
oxides and including criticisms of the significance of the analyses in-
these reports. The 1974 CHESS monograph is actually a simultaneous.
publication of a number of studies of health! effects of sulfur oxides,.
including some not actually a part of the heialth surveillance system
known as CHESS. Critical evaluation of these analyses has been
difficult as compared with the usual type of scientific peer review.
Access to the data had been somewhat limiied, and there was little-
time provided for review outside the writing team. Unlike most
reports in scientific literature, the CHESS papers were not referreed
in the usual manner nor^ are the epidemiological experiments readily-
repeated by individual independent researchers as is the case with
more controlled experiments. Rather significant resources in air
monitoring and data processing as well as in statistical analysis and
collection of health effects information are required and for this reason,
the studies represent a unique contribution to jbhe literature. Neverthe-
less, experts in the field had drawn attention' to several of their own
misgivings about the conclusions of the CHESS reports. As discussed
herein in more detail, the conclusions from! epidemiological studies
frBquently.-ar.e-diffieult-4o~suppor4^
in the context of many other variables. ;
The task of evaluating the CHESS program then must include an
understanding of the historical issues which identified the oxides of
sulfur as a potential hazard to health, the problems of conducting
epidemiological studies, and then an examinsition of the various sci-
entific processes and management structures 'which- were brought to
bear upon the problem during the past 6-8 years. As noted in the
Introduction to the Monograph, the purpose of the program was "to
evaluate existing environmental standards, obtain health intelligence
for new standards, and document the health benefits of air pollution
control." It is this last task of defining benefits that requires a high
degree of precision and for which maximum assurance must be present
that the data being considered are indeed valid and not overintei'—
preted. '
o
122
-------
PUBLIC LAW 95-155—NOV. 8, 197?
91 STAT. 1257
Public Law 95-155
95th Congress
An Act
To authorize appropriations for activities of the Environmental Protection
Agency, and for other purposes.
Be it enacted by the Senate and House of Representatives of the
United States of America in Congress assembled, That th1 is Act may
bo cited as the "Environmental Research, Development, and Demon-
stration Authorization Act of 1978", ;
SEC. 2. (a) There are authorized to be appropriated to tjhe Environ-
mental Protection Agency for environmental research, development,
and demonstration activities for fiscal year 1978— - i •
(1) $92,500,000 for water quality activities authorized under
the Federal Water Pollution Control Act of which—
(A) $25,200,000 is for the Health and Ecological Effects
program; ;
(B) $9,300,000 is for the Industrial Processes program;
(C) $6,069,000 is for the Monitoring and Technical Support
program;
(D) $'22,300,000 is for the Public Sector Activities pro-
-gr.am-;.-and-..
(E) $29,631,000 is f or the Energy program.
(2) $10,800,000 for activities authorized under the Federal
Insecticide, Fungicide, and Rodenticide Act, in the! Health txnd
Ecological Effects program.
(3) $16,000,000 for water supply activities authorized under
the Safe Drinking Water Act, in the Public Sector program.
(•t) $8,200,000 for toxic substance control activities'authorized
under the Toxic Substances Control Act, in the Health and Eco-
logical Effects program.
(5) $830,000 for radiation activities authorized tinder the Public
Health Act, in the Health and Ecological Effects program.
(6) $35,000,000 for air quality activities authorized under the
Clean Air Act, which shall be in addition to funds previously
authorized in the Clean Air Act Amendments of -1977 i ( Public Law
95-95), so that the total amount authorized for such activities in
fiscal year 1978 is $155,000,000, of which—
(A) $36,000,000 is for the Health and Ecological Effects
program;
(B) $11,000,000 is for the Monitoring and Technical Sup-
port program;
(C) $7,000,000 is for the Industrial Processes program;
and
(D) $101,000,000 is for the Energy program.;
(7) $31,273,000 for interdisciplinary activities, of which—
(A) $9,230,000 is for the Health and Ecological Effects
program; !
(B) $6,066,000 is for the Industrial Processes program;
(C) $1,599,000 is for the Public Sector Activities program;
and ;
(D) $14,378,000 is for the Monitoring and Technical Sup-
port program.
Nov. 8.1977
[H.R. 5101]
Environmental
Research,
Development,
and
Demonstration
Authorization Act
of 1978.
33 USC 1251
note.
7 USC 136 note.
42 USC 300f
note.
15 USC 2601
note.
42 USC 201 note.
42 USC 1857
note.
Ante, p. 685.
29-139 O - 77 (170)
723
-------
91 STAT. 1258
PUBLIC LAW 95-155—NOV. 8, 1977
Ante, p. 687.
Appropriation
authorization.
Transfer of funds,
restriction.
(b) In addition to any other sums authorized by this section or by
other provisions of law— ; - •
(1) there are authorized to be appropriated to the Adminis-
trator of the Environmental Protection Agency for fiscal year
1978, $10,000,000 for long-term, research and development in
accordance with section 6 of this Act;
(2) there are authorized to be appropriated to the Adminis-
trator, for fiscal year 1978, $2,000,000 for training of health sci-
entists needed for environmental research and development in
fields where there arc national shortages of trained personnel;
and ',
(3) there are authorized to be. appropriated to the Administra-
tor, for fiscal year 1078, $3,000,000 to implement the study author-
ized in section 103 (d) of the Clean Air Act Amendments of 1977
(Public Law 95-93).
(c) There is authorized to be appropriated to the Administrator
$19,000,000 for fiscal year 1978 for program management and support
related to environmental research and development.
(d) No funds may be transferred from any particular category
listed in subsection (a) or (b) to any-other category or categories
lis.ted in either such subsection if the total of the funds so transferred
from that particular category would: exceed 10 per.centum thereof,
and no funds may he transferred to any particular category listed in
subsection (a) or (b) from any other category or categories listed in
either such subsection if fhe. total of the funds so transferred to that
particular category would exceed 10 per centum thereof, unless—
(1) a peribd of thirty legislative days ha.s passed after the
Administrator of the Environmental Protection Agency or his
designee has transmitted to the Speaker of the House of Kepre-
sentatives and to the President of the Senate a written report
containing a full and complete' statement concerning the nature
of the transfer and the reason therefor, or
('2) each committee of the House of Representatives and the
Senate having jurisdiction over the subject- matter involved,
before the expiration of such period, has transmitted to the Admin-
istrator written notice rw the effect that such committee has no
objection to the. proposed action.
SBC. 3. Appropriations made pursuant to the authority provided
in section 2 of this Act shall remain available for obligation for
expenditure, or for obligation and expenditure, for suoh period or
Budget
projections.
42 USC 4361a.
42 USC 4361.
Public sector
agencies, grants.
42USC300j-3a.
gress under section 3 of Public Law S)4-4ff>, .shall include budget pro-
jections for a "no-growth"- budget, for a "moderate-growth" budget,
and for a '"high-growth" -budget. In addition, each such annual revi-
sion shall include a detailed explanation of the relationship of each
budget projection to the existing laws which authorize the Adminis-
tration's environmental research, development, and demonstration
programs.
SEC. 5. (a) The Administrator of; the Environmental Protection
Agency shall offer grants to public sector agencies for the purposes
of—
(1) assisting in the development and demonstration (includ-
ing^ construction) of any project!which will demonstrate a new
or improved method, approach, or technology for providing a
dependably safe supply of drinking water to the public; and
124
-------
PUBLIC LAW 95-155—NOV. 8, 1977
91 STAT. 1259
(2) assisting in -the development and demonstration (includ-
ing construction) of any project which will investigate and dem-
onstrate health and conservation implications involved in the
reclamation, recycling, and reuse of wastewaters for drink-
ing and the processes and methods for the preparation of safe
and acceptable drinking water. :
(b) Grants made by the Administrator under this section shall be
subject to the following limitations:
(1) Grants under this section shall not exceed 66% per centum
of the total cost of construction of any facility and 75 per centum
of any other costs, as determined by the Administrator.
(2) Grants under this section shall not- be made for any proj-
ect involving the> construction or modification of any facilities
for any public water system in a State unless such project has
been approved by the Jstate agency charged with :the responsi-
bility for safety of drinking water (or if there is no such agency
in a State, by the State health authority).
(3) Grants under this section shall not be made -for any proj-
ect unless the Administrator determines, after consultation, that
such project will serve a' useful purpose relating to the develop-
ment, and demonstration of new or improved techniques, methods,
or technologies for the provision of safe water to the public for
drinking.
(c) There are- authorized to be appropriated for the purposes of this
section $25.000.000 for fiscal year 1978.
SpjC. 6. (a) The Administrator of the Environmental Protection
Agency shall establish a separately identified program to conduct eon-
timiing and long-term environmental research and : development.
Unless otherwise specified by law, at least 15 per centum of any funds
appropriated to the Administrator for environmental research and
development under section 2(a) pf this Act or under any other Act
shall be allocated for long-term environmental research and devel-
opment under this section.
(b) The Administrator, after consultation with the Science Advisory
Board, shall submit to the President and the Congress; a report con-
cerning the desirability and feasibility of establishing a national
environmental laboratory, or a system of such laboratories, to assume
or supplement the long-term environmental research .functions cre-
ated by subsection (a) of this section. Such report shall be submitted
on or before March 31, 1978, and shall include findings and recom-
mendations concerning— " , . ...
(1) specific types of research to be carried out by such labora-
tory or laboratories; !
(2) the coordination and integration of research to be con-
ducted by such laboratory or laboratories with research conducted
by existing Federal or other research facilities; ;
(3) methods for assuring continuing long-range funding for
such laboratory or laboratories; and •%;-.
(4-) other administrative or legislative actions necessaiy to
facilitate the establishment of such laboratory or laboratories.
SEC. 7. (a) The Administrator of the .Environmental .Protection
Agency shall assure that the expenditure of any funds appropriated
. piirsuant to this Act or any other provision of law f orienvironmon^l
research and development related to regulatory program aefiv>Hes
shall be coordinated with and reflect the research needs and priorities
Grants,
limitations.
Research and
development
program.
42 USC 4363.
Report to
President and
Congress.
Contents.
42 USC 4364.
725
-------
91 STAT. 1260
PUBLIC LAW 95-155—NOV. 8, 1977
Program offices.
Report to
President and
Congress.
Science Advisory
Board.
Establishment.
42 USC 4365.
Membership.
42 USC 4361.
Report to
Administrator,
President, and
Congress:
of the program offices, as well as the overall research needs and priori-
ties of the Agency, including those defined in the five-year research
plan. . ;
(b) For purposes of subsection '(a), the appropriate program
offices are— •
(1) the Office of Air and Waste Management, for air quality
activities; .1
(2) the Office of Water and Hazardous Materials, for water
quality activities and water supply activities;
(3)(the Office of Pesticides,; for environmental effects of
pesticides;
(4.) the Office of Solid Waste, for solid waste activities;
( 5 ) the Office of Toxic Substances, for toxic substance activities;
(6) the Office of Radiation Programs, for radiation activities;
and
(7) the Office of Noise Abatement and Control, for noise
activities.
(c) The Administrator shall submit to the President and the Con-
gress a report concerning the most appropriate means of assuring, on
a continuing basis, that the research efforts of the Agency reflect the
needs and priorities of the regulatory program offices, while main-
taining a high level of scientific quality ..Such report shall be submitted
on or before March 31,1978. '
SEC. 8. (a) The Administrator of the Environmental Protection
Agency shall establish a Science Advisory Board which shall provide
such scientific" advice as the Administrator requests.
(b) Such Board shall be composed 6f at least nine members, one of
whom shall be designated Chairman, and shall meet at such times and
places as may be designated by the Chairman of the Board in consul-
tation with the Administrator. Each ^member of the Board shall be
qualified by education, training, and Experience to evaluate scientific
and technical information on matters referred to the Board under this
section. ;
(c) In addition to providing scientific advice when requested by the
Administrator under subsection (a)j the Board shall review and
comment on the Administration's five-year plan for environmental
research, development, and demonstration provided for by section 5
of Public Law 94-4.75 and on each annual revision 'thereof. Such
review and comment shall be transmitted to the Congress by the
Administrator, together with his comments thereon, at the time of the
transmission to the Congress of the annual revision involved.
(d) The Board shall conduct a review of and submit a report to the
Administrator, the President, and the Congress, not later than
October 1,1978, concerning—
(1) the health effects research authorized by this Act and other
laws; ;
(2) the procedures generally used in the conduct of such
research; ,;
(.3) the internal and external "reporting of the results of such
research;
(4) the review procedures for such research and results;
(5) the procedures by which such results are used in internal
and external recommendations on policy, regulations, and legisla-
tion ; and
(6) the findings and recommendations of the report to the
House Committee on Science and Technology entitled "The
126
-------
PUBLIC LAW 95-155—NOV. 8, 1977
91 STAT. 1261
Environmental Protection Agency's Research Program with
primary emphasis on the Community Health and Environmental
Surveillance System (CHESS) : An Investigative Report".
The review shall focus special attention on the procedural safeguards
required to preserve the scientific integrity of such research and to
insure reporting and use of the results of such research in subseqxtent
recommendations. The report shall include specific recommendations
on the results of the. review to ensure scientific integrity throughout
the Agency's health effects research, review, reporting, and recom-
mendation process.
(
-------
91 STAT. 1262
PUBLIC LAW 95-15S—NOV. 8, 1977
Report.
42 USG 4361.
Report to
President and
Congress.
Legislative
recommenda-
tions.
Presidential
report to
Congress.
42 USC 4361b.
Personnel
positions,
increase.
mentj which may need to bo more effectively coordinated in order
to minimize unnecessary duplication of programs, projects, and
research facilities; '
(2) to determine the steps which might be taken under existing
law, by him and by the heads of such other agencies, to accomplish
or promote such coordination, and to provide for or encourage the
taking of such steps; and
(3) to determine the additional legislative actions which would
be needed to assure such coordination to the maximum extent
possible. ...
The Administrator shall include in each annual revision of the five-
year plan provided for by section 5 of Public Law 94-475 a full and
complete report on the actions taken and determinations made during
the preceding year under this subsection, and may submit interim
reports on such actions and determinations at such other times as he
deems appropriate.
(b) The Administrator of the Environmental Protection Agency
shall coordinate environmental research, development, and demon-
stration programs of such Agency with the heads of other Federal
agencies in order to minimize unnecessary duplication of programs,
projects, and research facilities.
(c)(l) In order to promote the coordination of environmental
research and development activities, aiul to assure that the action taken
and methods used (under subsection (a) and otherwise) to bring about
s"ch rpordination will be as effective 'as possible for that purpose, the
Council on Environmental Quality iii consultation with the Office of
Science and Technology Policy shall promptly undertake and carry
out a joint study of all aspects of th£ coordination of environmental
research and development. The Chairman of the Council shall pre-
, pare a report on the results-of such study, together with such recom-
mendations (including legislative recommendations) as he deems
appropriate, and shall submit such report to the President and the
Congress not later than May 31,1978.
(2) Not later than September 30, 1978, the. President shall report
to the Congress on steps ne has taken to implement the recommenda-
tions included in the report under paragraph (1), including any rec-
ommendations he may have for legislation.
SEC. 10. The Administrator of the Environmental Protection Agency
shall implement the recommendations of the report prepared for the
House Committee on Science and Technology entitled "The Environ-
mental Protection Agency Research Program with primary emphasis
on the Community Health and Environmental Surveillance System
(CHESS): An Investigative Report", unless for any specific rec-
ommendation he determines (1) that such recommendation has been
implemented, (2) that implementation ;of such recommendation would
not enhance the quality of the research, or (3) that implementation of
such recommendation will require funding which is not available.
Where such funding is not available, jthe Administrator shall request
the required authorization or appropriation for such implementation.
The Administrator shall report the'status of such implementation
in each annual revision of the five-year plan transmitted to the Con-
greps under section 5 of Public Law 94-475.
SEC. 11. The Administrator of the Environmental Protection Agency
shall increase the number of personnel positions in the Health and
Ecological Effects program to 862 positions for fiscal year 1978.
725
-------
PUBLIC LAW 95-155—NOV. 8, 1977
91 STAT. 1263
SEC, 12. (a) Each officer or employee of the Environmental Protec-
tion Agency who—
(1) performs any function or duty under this Act; and
(2) has any known financial interest in any person who applies
for or receives grants, contracts, or other forms of financial assist-
ance under this Act, ;
shall, beginning on February 1, 1978, annually file with the Admin-
istrator a written statement concerning all such interests held by such
officer or employee during the preceding calendar year. Such statement
shall be available to the public.
(b) The Administrator shall—
(1) act within ninety days after the date of enactment of this
Act—
(A) to define the term "known financial interest" for pur-
poses of subsection (a) of this section; and
(B) to establish the methods by which the requirement to
file written statements specified in subsection (a) '.of this sec-
tion will be monitored and enforced, including appropriate
provision for the filing by such officers and employees of such
statements and the review by the Administrator of such state-
ments; and
(2) report to the Congress on June I of each calendar year with
respect to such disclosures and the actions taken in regard thereto -
during the preceding calendar year.
(c) In the rules prescribed under subsection (b) of this section, the
Administrator may identify specific positions of a nonpolicymaking
natiire within the Administration and provide that officers or employ-
ees occupying such positions shall be exempt from the requirements of
this section. ......
(d) Any officer or employee who is subject to, and knowingly vio-
lates, this section, shall be fined not more than $2,500 or | imprisoned
not more than one year, or both.
SEC. 13. It is the national policy that to the maximum extent possible
the procedures utilized for implementation of this Act shall encourage
the drastic minimization of paperwork.
Approved November 8, 1977.
Annual
statement, filing.
42 USC 4367.
Report to
Congress.
Violation,
penalty.
Paperwork
minimization,
encouragement.
LEGISLATIVE HISTORY:
HOUSE REPORTS: No. 95-157 (Comm. on Science and Technology) and No. 95-722
(Comm. of Conference). i . . .
SENATE REPORT No. 95-188 accompanying S. 1417 (Comm. on Environment and
Public Works).
CONGRESSIONAL RECORD, Vol. 123 (1977):
Apr. 19, considered and passed House. ;
May 27, considered and passed Senate, amended, in lieu of S. 1417.
Oct. 20, Senate agreed to conference report.
Oct. 25, House agreed to conference report.
o
129
-------
-------
Part 2
EPA's annual reports to Congress on the Agency's
five-year plans for research and development, as
required by Section 10 of ERDDAA.
737
-------
-------
APPENDIX E
CHESS--The Community
Health and
Environmental Surveillance
System-- ;
Congressional
Recommendations-
Status Report U •
!
Background: A controversy about the
scientific credibility of results from the
CHESS study prompted a series of con-
gressional hearings in 1976. Subsequently,
17 major recommendations were made by
Congress to EPA regarding methods and
means to upgrade its environmental •
research. These recommendations.;
covered a wide spectrum of subjects.
Public Law 95-155, the Environmental
Research, Development and Demonstra- \
tion Authorization Act of 1978 specified ,
that EPA annually report on the im- ;
plementation status of each recommended \
action. This report is the second status <
report; the first was in Research Outlook
1978. ;
Because last year's report was the first :
status report, it included descriptions of \
all the recommendations and subsequent •
EPA actions. The reader is referred to ,
that report and documents in the
bibliography for detailed information _
about the recommendations. Given below
is a status update of only those EPA ac-;
tions taken last year. ;
Recommendation ;
3--CHESS Monograph
This recommendation concerns thei
public access to and understanding of the.
limitations of the CHESS monograph. To,
comply, we have sent an appropriate
cover letter and copies of the Research]
Outlook 1978 (which contains the first,
status report on the CHESS recommenda-
tions) to locatable holders of the CHESS
monograph. Any additional copies of the
CHESS monograph that are distributed
will be accompanied by copies of the
cover letter and the Research Outlook
1978. Finally, a notice is planned for the
Federal Register informing the public of
the availability of further information on
the CHESS monograph. These actions,
we believe, satisfy the intent of recom-
mendation 3.
Recommendations 10(a),
10(c) and 12(a)
These recommendations concern peer
review of EPA's scientific research. The
subject has received increased attention in
various quarters of EPA in the past year.
To improve the quality of research
throughout ORD, the EPA Assistant Ad-
ministrator for Research and Develop-
ment has directed establishment of peer
review mechanisms at the laboratory
level. He has also directed the head-
quarters line managers, to whom the
laboratories report, to structure a head-
quarters level peer review. Submission of
ORD research results to referred scientific
journals is also being stressed.
The EPA Science Advisory Board
(SAB) is a prime source for peer review
of research and research program plan-
ning advice. Its various components
regularly examine elements of ORD's
research program.
Recommendation 10(a) specifically
concerns establishment of an inter-
disciplinary task force to draw up a plan
for EPA to develop "a solid base of
knowledge and procedures in aerometric
instrumentation and measurements, mete-
orology, field data gathering, quality con-
trol, epidemiology project design, and
testing and panel planning." Last year's
status report indicated that this recom-
mendation would be discussed with the
EPA's Science Advisory Board. The
Board's Health Effects Research Com-
mittee was directed by Public Law 95-155,
the Research, Development and
Demonstration Act of 1978, to review
EPA's health effects research, including
the recommendations of the CHESS In-
vestigative Report. The Committee is in
the process of completing its report, We
await that report, and the recommenda-
733
-------
RESOLUTION OF INVESTIGATIVE REPORT RECOMMENDATIONS
Number
3
3(c)
*M
4
5
6
12(b)
12(c)
12
CHESS dale analyses should be carried out only OB data with high validity potential '
EPA should publish research in referred journals in a timely fashion
EPA should not publish large projects solely in monograph form
EPA should not initiate projects for policy consideration unless they can be completed ,
in a realistic time frame ',
EPA should strengthen the CHAMP aerometric and quality control programs ;
EPA should shorten the time between data acquisition and quality assurance analysis
of data
EPA should stop employing development stage instruments before qualification testing •
EPA should not use laboratory models of instruments In the field until they have been
field checked and operating personnel trained
EPA should reevaluale the opening of the CHAMP operations contract to competition •
EPA research and monitoring personnel should closely coordinate regarding
chemical species '
EPA should have additional meteorological support for air pollution health effects
research studies '
EPA should examine accelerating research In pollutant characterization
An interdisciplinary task force should draw up an integrated air epidemiology-exposure ;
assessment program plan for EPA
CHAMP should verify instruments and protocols so that reliable data can be achieved
EPA should have epidemiologies! questionnaires and panel selection criteria approved
by peer groups
EPA should review research concepts obtained from learn interviews
The Environmental Resarch Center at Research Triangle Park (RTF) should not be '
reorganized until the end of FY 77 ;
EPA should establish authoritative peer review panels (o assist In improving j
research coordination
EPA should have a stronger focus on management at the Environmental Resarch Center, RTF '
EPA should create a systems analysis-operations research program review group
The Science Advisory Board's charter should be expanded
EPA should seek cooperative research programs with universities and other
laboratories and agencies
EPA should promote the exchange of scientists within and outside the Agency
EPA should fund individual Ph.D. thesis research
Action
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Implemented
Under
consideration
734
-------
(CONT.)
Number Summary of Recommendations Action
13{d)
14
IS
K
17
The Science Advisory Board should develop outreach programs
The Administrator should clarify the role of the [Office of Research and Development
and its laboratories
EPA should resolve the separation of facilities at RTF
EPA should develop a professional caner development program for each
professional employee
. The Administration should determine if EPA should conduct research under its present
organizational configuration
Under
consideration
Shall be
implemented
Under
consideration
Implemented
Implemented
Source: Research Outlook 1978
tions of the Committee's Subcommittee ;
on Epidemiologic Studies. We will take '
appropriate action based on the recom-
mendations.
Recommendation 10(c) concerns peer <
review and approval of epidemiological i
questionnaires and panel selection
criteria. Appropriate review and approval ,
are part of the review of epidemiology '•
studies in EPA's research program, con-
ducted by the SAB's Health Effects ;
Research Review Committee. It should be
noted that the Interagency Regulatory
Liaison Group (IRLG), composed of
EPA, OSHA, CPSC, and FDA has a j
working group on epidemiology. A
subgroup addresses standards which •
would apply to epidemiological studies in '.
order to assure scientific validity for use
as court evidence.
Recommendation 12(a) refers to peer :
review panels for increased coordination !
of research. The review responsibilities of i
the Committees mentioned in recommen-
dations 10(a) and 10(c) fulfill this recom-
mendation. In addition, improved peer
review of EPA research was incorporated
within a new research planning system
established within the past year to im- ,
prove the responsiveness of ORD to EPA ,
program offices. This subject is discussed ;
in two reports to Congress: "The Plan-:
ning and Management of Research and
Development Activities Within EPA,"
June 1978, and a follow-up status report
in December 1978.
These reports describe research plan-
ning via specially formed research commit-
tees. Each committee will consist of
representatives from ORD and EPA Pro-
gram Offices, and will plan research
specific to those offices. The new research
planning system calls for incorporation of
peer review mechanisms throughout the
planning and management process in
order to improve research quality. Several
research committees have been operating
successfully on a pilot basis, and more are
planned. While systematic peer review has
not yet been implemented in the pilot
committees, EPA hopes to do so within
the coming months.
Recommendations 13(a),
13(c) and 13(d)
These recommendations address a
perceived isolation of EPA research. They
concern technical information exchange
and interaction of EPA scientists with
peers outside EPA, particularly with the
university community.
Last year's status report described
EPA's ties with the university com-
munity, mentioning, most notably, EPA
extramural research which involves grants
735
-------
with universities. In FY 1977, EPA
awarded 618 grants to 355 academic in-
stitutions. Each grant is monitored by an
EPA project officer, an arrangement
enabling our people to work closely with
researchers outside EPA.
Additionally, the Intergovernmental
Personnel Act has allowed an exchange of
researchers between EPA and state and
local governments and universities.
A report to Congress, "Laboratories
Needed to Support Long-Term Research
in EPA," April 1978, further explores
possible ties between ORD and univer-
sities. The report recommends a selec-
tively expanded program of long-range
research with both intramural and ex-
tramural components. An important
means of assuring coordination of these
components is scientisMo-scientist con-
tact between EPA and the academic com-
munity. The extramural portion of the
program would initially be a series of
small centers for long-range research at
universities and other institutions
dedicated to specific research problems.
In FY 1979, ORD will propose three such
centers, one each for advanced control
technology, epidemiology, and ground-
water research. These centers would serve
as bridges to the academic community
and should provide ORD with a reservoir
of talented scientists.
ORD's Minority Institutions Research
Support (MIRS) program also serves the
spirit of recommendation 13. MIRS was
established in 1972 to help minority in-
stitutions develop the potential for con-
ducting environmental research and thus
become more competitive for federal
funds. The EPA's MIRS staff maintains
continual liaison between university
researchers and the ORD scientific staff
to develop relevant research proposals.
Some expansion of the program is being
considered.
Recommendation 13(c) refers
specifically to EPA programs to fund in-
dividual PhD theses. EPA's workforce
training program, recently placed under
ORD's aegis, includes both academic
training grants given to various institu-
tions and fellowships to individuals.
EPA's role in this type of program is
unique since many whose work is crucial
to achieve environmental • goals are not
directly employed by EPA. This non-
federal workforce includes wastewater
treatment operators, state/environmental
employees, and other professionals. EPA
currently is working with the Department
of Labor and the Office of Education to
address various options for further work-
force development. :
Recommendation 13(d) concerns
Science Advisory Board ; assistance to
EPA to develop an outreach program. We
have actively sought SAB counsel for
many activities related to our connections
with the academic community. For exam-
ple, the SAB helped prepare the report
cited above, "Laboratories Needed to
Support Long-Term Research in EPA."
Additionally, the EPA Assistant Ad-
ministrator for Research {and Develop-
ment made a formal presentation to SAB
concerning ORD's university relations.
As a result of this presentation and
previous discussion, the SAB will develop
outreach program suggestions for further
consideration. ,
Recommendations 14 and
17 ;
These recommendations -state that the
EPA Administrator should clarify the
role of the EPA Office of Research and
Development and determine if research
should be conducted in its present
organizational configuration.
The Environmental Research, Develop-
ment and Demonstration !Act of 1978
(Public Law 95-155) directed the EPA
Administrator to report to ithe President
and Congress the most appropriate means
of assuring that EPA's research efforts
reflect the needs and priorities of the EPA
regulatory program offices. The EPA Ad-
ministrator fulfilled that mandate with
the distribution of the report "The Plan-
ning and Management of Research and
Development Activities within EPA"
(June 1978). The information in this
736
-------
report satisfies both recommendations.
Concerning Recommendation 14—the;
clarification of the role of ORD and its,
laboratories—the report provided a.
mechanism for improved coordination;
between ORD and program offices forj
planning research. This mechanism,,
described above in connection with peer;
review, is a series of research committees, j
established for each research planning
unit. Five pilot committees have already:
been established for key research areas;
and more are planned. Each research
committee is chaired by a research,
manager (designated by ORD) and has'
representatives from relevant program of-j
fices and EPA Regions. At a hearing
before the House Science and Technology
Committee, Subcommittee on the En-
vironment and the Atmosphere, the EPA
Administrator and Assistant Administra-
tors attested to the success of this ap-
proach. A formal status report on the
pilot activities was provided to ORD's
authorizing committees in December, 1978.
Further exploration of ORD's role in
the Agency was provided by the report
mentioned under Recommendation 13.'
This report, "Laboratories Needed to
Support Long-Term Research in EPA"
(April 1978) examines alternative ar>i
preaches for conducting long-term en-
vironmental research and presents find-
ings and recommendations. The report
also reviews the history of ORD labora-
tories, describes representative research
areas that could benefit from enhanced
long-term support, describes mechanisms
used by other federal agencies for carry-
ing out this kind of research, and presents
options for long-term research within
EPA. ,
Concerning Recommendation 17—that
EPA determine whether it should conduct
research under its present organizational
configuration—a study group found that
"the Agency's R&D problems have no't
resulted from the way the R&D program
is organized." To reach this conclusion,
the group reviewed a number of similat
organizational structures in various
federal agencies. Possible use of some of
those organizational structures was re-
jected and, instead, the study group pro-
posed for ORD the management system
changes described above.
Recommendation 15
This recommendation directs EPA to
resolve the separation of facilities at
Research Triangle Park (RTP), North
Carolina. EPA's Office of Research and
Development (ORD) is presently prepar-
ing a program of requirements for a new
research and development facility at
Research Triangle Park. Additionally, the
Agency has set up an EPA long-range
plan for special purpose facilities. This
long-range plan would be agency-wide
and would include the EPA regional of-
fices' facilities as well as the Office of
Research and Development laboratories.
137
-------
CHESS :
Health Consequences of Sulfur Oxides: A
Report from CHESS, 1970-1971. United
States Environmental Protection Agency.
EPA 650/1-74-004. May, 1974.
Laboratories Needed to Support Long-
Term Research in EPA: A| Report to the
President and the Congress,. United States
Environmental Protection Agency. EPA
600/8-78-003. April, 1978.
Pilot Study of the Revised Planning and_
Management System for Research and
Development in the Environmental Protec-
tion Agency: A Status Report to the Con-
gress. United States Environmental Protec-
tion Agency. November, 1978.
Report of the Task Force to Review
CHESS. United States Environmental Pro-
tection Agency. April 7, 1976.
Report on Joint Hearings on the Conduct
of the Environmental! Protection Agency's
Community Health and Environmental
Surveillance System' (CHESS) Studies.
United States Congress. House Interstate
and Foreign Commerce and House Science
and Technology Committee. 94th Con-
gress, 2nd Session. Washington, United
States Government Printing Office, April,
1976.
Report on the Environmental Protection
Agency's Research Program with Primary
Emphasis on the Community Health and
Environmental Surveillance System
(CHESS). United States Congress. House
Committee on Science and Technology
Committee. 94th Congress; 2nd Session.
Washington, United States Government
Printing Office, November, 1976.
Research Outlook 1978. United States En-
vironmental Protection Agency. EPA
600/9 78-001, June 1978. I
Rood, W.B. EPA Study-r-The Findings
Got Changed. Los Angeles Times,
February 29, 1976. i
The Planning and Management of Research
and Development Within EPA: A Report
to the President and the Congress. United
States Environmental. Protection Agency.
June 30, 1978.
138
it US. GOVERNMENT PRINTING OFFICE 1973 (3—291-132
-------
Appendix 3
Community health and
environmental surveillance system
Introduction
On November 24, 1976, the House Sub-
committee on the Environment and Atmo-
sphere of the Committee on Science and
Technology released a report titled, "The
Environmental Protection Agency's Research
Program with Primary Emphasis on the
Community Health and Environmental Sur-
veillance System (CHESS): An Investigative
Report" (Ref. 1). The Environmental Re-
search, Development, and Demonstration
Authorization Act of 1978 (P.L. 95-155)
specifies that EPA shall report the implemen-
tation status of the Investigative Report's rec-
ommendations in each annual revision of its
five-year plan. This is the first EPA im-
plementation status report.
Background
CHESS emerged as a major program in
1970. A discussion of goals and objectives is in
"Environmental Science and Technology"
(Ref. 2). CHESS was designed, as the name
indicates, to monitor the health status of the
United States population with respect to vary-
ing environmental conditions. For the most
part, the environmental considerations were
limited to meteorologic conditions and pollu-
tion levels. The program included studies of
health groups and potentially susceptible
groups such as asthmatics. CHESS data,
gathered over five years, have been analyzed
for relationships between health effects and
exposure to such pollutants as sulfur oxides,
nitrogen oxides, paniculate matter, and
oxidants.
In May 1974, EPA published the "Health
Consequences of Sulfur Oxides: A Report
from CHESS, 1970-1971" (often referred to
as the CHESS Monograph) that included sev-
eral of the early CHESS aerometric and
health studies (Ref. 3).
On February 29,1976 (he Los Angeles Times
published the first of several articles implying
that studies hi the CHESS Monograph on die
health effects of ambient sulfur oxides were
distorted (Ref. 4). Basically, die Times articles
made three allegations: (1) the analysis of the
CHESS data shows a stronger than actual
correlation of adverse health effects with; in-
creased levels of ambient sulfate; (2) Dr. John
F. Finklea was responsible for the distortion,
with the passive assistance of his subordi-
nates; and (3) the EPA regulatory program
for sulfur oxides rests solely on the CHESS
program.
On April 7,1976., an EPA investigative task
force appointed to review the entire matter
reported its findings (Ref. 5). This group
interviewed'most of the EPA employees who
participated in the CHESS data analysis.
From these interviews, it became apparent
that comments of EPA personnel made to the
Times reporter referred to the 1972 draft
version of the report, and not to the final
publication of 1974. The group's unanimous
opinion concerning both the draft and final
versions was "that there is no evidence of dis-
honesty or deliberate distortion of data by Dr.
John F. Finklea or members of his staff who
worked on the Monograph. On the contrary,
there is evidence of an honest and aggressive
effort to publish the sulfur oxide findings
from the CHESS studies so that they would be
available in a timely fashion for use by the
Agency and die public at large."
On April 9, 1976, the allegations concern-
ing the CHESS report and the Times articles
were the subject of a Congressional hearing
convened by two House Committees: Science
and Technology and Interstate and Foreign
Committee (Ref. 6, pp. 23-24). Three of the
Committees' conclusions, which directly re-
late to the Los Angeles Times allegation, are:
—"There was agreement that the CHESS
studies confirm an association between
sulfur oxides emissions and adverse
health effects."
—"There was no evidence that Dr. Finklea
tampered with, distorted, or withheld
data."
—"The National Ambient Air Quality
Standards (NAAQS) for sulfur dioxide
were set before CHESS, and were based
on odier data."
An examination of the Times articles and
the Congressional hearing is published in Sci-
ence and the Environmental Health Letter (Ref.
7,8).
The House Committee on Science and
Technology started an investigation concern-
ing technical issues relating to CHESS in
April of 1976. The Committee released the
Investigative Report on November 24, 1976
(Ref. 1).
Investigative report
recommendations
The recommendations in the Investigative
Report concern two major topics: the scien-
tific assessment of the CHESS program, and
the present and future management of EPA
research. EPA agrees with many of the Inves-
tigative Report's statements and recommen-
dations regarding the quantitative limitations
of the CHESS results. Also, EPA concurs with
the majority of the recommendations for the
139
-------
improvement of research. The implementa-
tion status for each recommendation is dis-
cussed in the following paragraphs.
Recommendation 3
CHESS Monograph
Recommendations 3(a), (b), and (c) in the
Investigative Report concern the CHESS
Monograph. Recommendation 3(b) directs
that the Monograph should not be used with-
out explicit qualifications. Recommendations
3(a) and 3(c) state that EPA should publish an
announcement regarding the limitations of
the Monograph and publish an addendum to
the Monograph (including at least Chapters
IV, V, VI, and Appendix A of the Investiga-
tive Report). Priorto presenting our response
to these recommendations, the EPA air
health effects research program, the CHESS
Monograph, and EPA's regulatory respon-
sibilities will be placed in perspective.
The air health effects research program
.uses a combination of research approaches:
human epidemiological studies, human clini-
cal studies,, and toxicological studies on ani-
mal models (Ref. 9). Integrating the capabili-
ties and advantages of these approaches
provides the best overall scientific strategy
for informed regulatory decisions. Concern-
ing the CHESS program, epidemiologic in-
vestigations offer the advantages of studying
the biological responses of people, including
vulnerable groups, under ambient condi-
tions.
The major problems are related to quan-
tifying the exposure, dealing with many typi-
cally unknown covariates, and interpreting
association vs. causation. Several publica-
tions, including the Investigative Report,
contain detailed explanations of the strengths
and weaknesses of epidemiology (Ref. 1, pp.
57-58; Ref. 10, pp. 16-17; Ref. 11, pp. 16-17).
EPA acknowledges the limitations of in-
dividual epidemiologic studies. For example,
the Summary and Conclusions of the CHESS
Monograph states:
"The findings summarized in this paper
must be substantiated by replicated obser-
vations in different years and under dif-
ferent circumstances. Weil controlled
human and animal studies are required to
isolate several of the important intervening
variables that are inherent to studies of free
living populations, and to elucidate the
precise nature of the pollutant-disease rela-
tionship. Hence, the conclusions put forth
at this time cannot be definitive, but are of-
fered in the sense of developing more re-
fined quantitative and scientific hypotheses
concerning pollutant-health effect associa-
CHESS Study Areas
Rocky Mountain Cities—5
(Sulfur oxides)
Salt Lake Basin—4
(Sulfur oxides)
New York City Metropolitan Area—
(Sulfur oxides and paniculate matter
Los Angeles Basin—7
(Photochemical oxidants)
and nitrogen oxides)
Chattanooga—3
(Nitrogen oxides)
Southeastern Cities—9
(Paniculate matter)
Note: Numbers indicate number of neighborhoods studied per area.
140
-------
Appendix 3
dons in a real life environment." (Ref. 3, p.
7-4).
Most epidemic logic studies are open to crit-
icism and this was the reason spatial and tem-
poral replication was fundamental to the
CHESS designs. Further, in practically ail
areas of epidemiology, conclusions rest on
the weight of evidence from many studies,
not on individual studies. Thus, there is
ample justification to cite CHESS studies as
they bear on existing EPA standards. An im-
portant feature of several of the CHESS
studies reported in the Monograph is their
general consistency with the majority of
epidemiologic, clinical, and toxicologic
studies previously published in the sulfur
oxides and particular matter literature. The
CHESS studies tend to support the rea-
sonableness of existing ambient air quality
standards for sulfur oxides and paniculate
matter. However, EPA agrees that there is far
too much uncertainty and lack of qualifica-
tion in findings contained in the Monograph
to support any new or modified air quality
standards.
Finally, the Monograph assessments of cur-
rent pollution exposure were among the most
complete that had ever been performed
within the then existing state-of-the-art.
These epidemiologic findings, although hav-
ing a limited ability to affect EPA regulatory
policy, have materially advanced our knowl-
edge concerning the general distribution and
behavior of the exposure-response variables
employed.
Misunderstandings still exist, however,
over the CHESS Monograph and EPA's regu-
latory posture on sulfur oxides. Therefore,
this report to the Congress shall be widely cir-
culated and sent to all holders of the CHESS
Monograph with an appropriate cover letter.
With this action, EPA believes that the intent
of Recommendation 3 will be adequately im-
plemented.
Recommendation 4
research responsibilities
Recommendation 4 addresses research re-
sponsibilities and resources. Recommenda-
tion 4(a) directs thai: legislation should be re-
examined regarding unrealistic procedures
and schedules. Legislative mandates are the
most important considerations in the annual
program planning process. EPA maintains an
in-house research capability and expertise to
respond to short deadlines. However, when
procedures or schedules are unrealistic, the
Congress and the O ffice of Management and
Budget (OMB) are informed by either the
normal budget submission process, during
oversight hearings, or by other appropriate
mechanisms. Recommendation 4(b) specifies
that research be designed to gain information
and not support positions. The Office of Re-
search and Development is organizationally
separated from offices having regulatory re-
sponsibilities. Therefore, scientists conduct
research to gain accurate information and are
riot under pressure to support existing or
preconceived positions held by the regulatory
offices. The Science Advisory Board, an in-
dependent advisory body, has established a
Subcommittee on Epidemiological Studies to
independently review EPA's epidemiology
(Ref. 12). Recommendations 4(c) and (d) con-
cern the Office of Management and Budget
allowing all necessary funding for expedi-
tious research and advising the Congress of
budgetary restrictions affecting completion
of major projects. Through the normal
budgetary process, the Office and the Con-
gress are advised for EPA resource require-
ments and which programs are affected by
budgetary restrictions.
Recommendation 5
{questionnaires
Recommendation 5 advises that the OMB
should be asked to develop procedures for
prompt review of questionnaires. OMB and
EPA's research managers and scientists have
.discussed this matter. These discussions have
i expedited questionnaire clearances. How-
: ever, the total number of questionnaires allo-
! cated to EPA is small and therefore limits the
number of epidemiologic studies that can be
performed (Ref. 13). The control of ques-
. tionnaires by the Federal Government was in-
tended to reduce involuntary solicitations
from the private sector. Only selected volun-
teers participate in EPA's epidemiologic
studies. Therefore, we believe that EPA's
voluntary questionnaires should be free of al-
location limitations.
Recommendation 6
i CHESS data
Recommendation 6 concerns the process-
i ing and publication of the remaining CHESS
1 data. Recommendation 6(a) directs that un-
: analyzed data be examined and that analyses
be carried out on those data that appear to
have a higher degree of validity than the
CHESS Monograph data base. In general,
the quality of the CHESS data improved as
, experience was gained. Therefore, a plan has
. been developed for validation of 61 of the 65
, data sets for which reports have not yet been
r published (Ref. 14). Four episode studies will
! not be validated because their usefulness is
141
-------
questionable. Recommendations 6(fa) and (c)
concern publishing research in traditional,
refereed, archival, journals and not publish-
ing solely in monograph form. EPA endorses
this policy. Independent university scientists
are being used to analyze, interpret, and re-
port on appropriate CHESS data. Manu-
scripts of research investigations are being
submitted for publication in the scientific lit-
erature as relevant studies are completed. As
of December .1, 1977, there have been 28
CHESS publications in scientific journals
(Ref. 15). Monographs are, and have mainly
been, used as a vehicle to present all pertinent
data that would be inappropriate for publica-
tion hi scientific journals. Recommendation
6(d) states that projects for policy consid-
erations should not be initiated unless they
can be completed in a realistic time frame and
unless the research staff can be involved in
the process. Several mechanisms have been
incorporated to develop achievable program
plans. These include a joint program plan-
ning process where the staffs of the research
laboratories and headquarters, and the pro-
gram offices participate. In addition, labora-
tory program reviews are conducted and
problems associated with the implementation
of investigations are discussed and resolved.
Recommendations 7,8,
and 10(b)—CHAMP
Recommendations 7, 8, and 10(b) are di-
rected toward CHAMP, EPA's Community
Health Air Monitoring Program. Recom-
mendation 7(a) states that the aero metric and
quality control programs should be further
strengthened and improved. An expanded
quality control program is being im-
plemented in EPA (Ref. 16). Specifically for
CHAMP, comparisons for instruments,
techniques, and standards are being con-
ducted among the EPA laboratories measur-
ing air quality. In addition, a contractor is
providing additional quality control audits of
die CHAMP field and laboratory systems
(Ref. 17). This includes the use of National
Bureau of .Standards flow and measurement
standards as well as gas mixture standards.
Recommendation 7(b) directs a shortening of
time between data acquisition and quality as*
surance analysis of data. The new CHAMP
contractor has been given: technical direction
to minimize the time between data collection
and validation (Ref. 18). To accomplish this,
software is being developed and maintenance
practices have been revised.
Recommendations 7(c) and (d) specify that
development stage instruments should not be
employed before qualification testing has
been done and that laboratory models of in-
struments should not be used in the field until
they have been field checked and operating
personnel have been trained. The present
CHAMP policy calls for complete checkout,
acceptance testing, and personnel training
before field placement of any developmental
stage instruments. Recommendation 7(e) di-
rects that the opening of the CHAMP opera-
tions contract to competition should be reex-
amined to see whether the merits of open
bidding outweigh the problems of instability.
This revaluation took place and it was de-
termined that the merits of open bidding with
the possibility of improved performance
outweighed any problems of instability. In
due course, the contract was awarded to a
new contractor after competitive bidding.
The transition period between the old and
new contractors disclosed several major defi-
cient areas. This finding confirmed the wis-
dom of the decision to reopen the CHAMP
operations contract. Recommendation 7(f)
concerns health effects personnel closely
coordinating with air quality and monitoring
personnel to understand chemical species to
be monitored. The CHAMP staff is cooperat-
ing closely with the epidemiologists and other
health scientists in the design and protocol
development for epidemiologic studies. This
cooperation includes reaching mutual
agreement on chemical species to be moni-
tored. ;
Recommendation 8 concerns additional
meteorological support for health research-
air pollution effects studies. This recommen-
dation also directs that meteorological in-
strumentation be uniform: and complete for
all stations. There is one full-time
meteorologist assigned to the epidemiologic
program. He works >with both the
epidemiologists and CHAMP personnel to
assure that the appropriate and uniform
meteorological measurements are made. Be-
cause of the reduction in the number of
epidemiologic studies since the termination
of CHESS field studies, the meteorological
support to this activity is now at the proper
level. Additional meteorological support will
be seriously considered if the number of
epidemiologic studies is significantly in-
creased. !
742
-------
Appendix 3
Recommendation 10(b) directs that in-
struments and protocols used in CHAMP be
verified to ensure: reliable data. EPA is
testing all instruments in present use for
precision and accuracy. The present system
of continuous air monitors appears to have a
precision and accuracy such that errors are
less than plus or minus 15 percent. Third
generation instruments are being evaluated
when obtained to improve the present sys-
tem.
Recommendation 9
pollutant characterization
Recommendation 9 directs that the EPA
Health Effects program as well as interagency
utilization of all available Federal and ex-
tramural resources in the health effects area
should be examined with the objective of sig-
nifscandy accelerating research in pollutant
characterization. Liaison is maintained with
other Federal agencies and the non-Federal
sector regarding air pollution characteriza-
tion. In accord with the 1977 Clean Air Act
Amendments, EPA has organized an intera-
gency task force to determine the effects of
environmental pollutants on cancer, heart,
lung, and other chironic diseases (Ref. 19). In
addition, on a case by case basis we are exam-
ining the proper l>alance between pollutant
characterization and health effects research.
Adjustments are being made as deemed ap-
propriate. We fully appreciate, for example, :
the necessity of having adequate pollutant ,
characterization data prior to the beginning i
of laboratory toxicoiogic experiments. The
same principle clearly applies to '
epidemiologic research. >
Recommendations 10(a), 1Q(c), <
12(a) and 12(d)—peer review
Recommendations 10(a), 10(c), 12(a) and i
12(d) concern peer review. Recommendation '
12(d) concerns expanding the Science Advi- ;
sory Board charter. The charter has been
modified in accord with the Environmental:
Research, Development, and Demonstration '
Act of 1978. (Ref. 20). Under the previous
and new charter, die Board is authorized to ;
.conduct peer reviews. The Board, an inde-
pendent advisory body, decides how it will re-
spond to requests for assistance. The Office ,
of Research and Development encourages:
die Board to conduct such reviews. However,
it is impossible for die Board to review all re-
search programs because of the limited time'
Board members can devote to EPA activities, j
TheOfficeof Research and Development has
started discussions with the Board to establish <
a more effective internal peer review process.
Recommendation 10(a) states that a truly
interdisciplinary task force led by an eminent
scientist should draw up a program plan for
EPA to develop a solid base for knowledge
and procedures in aerometric instrumenta-
tion and measurements, meteorology, field
data gathering, quality control, epidemiology
project design and testing, and panel plan-
ning. This recommendation will be discussed
with the Board. The activities of several
groups are pertinent to the recommendation.
The Board's Subcommittee on Epidemiolog-
ical Studies provides advice and assistance in
the review and evaluation of. proposed or
existing programs of epidemiologic studies
relating to the health effects of environmen-
tal pollutants (Ref. 13). Interactions between
our scientists and this subcommittee are con-
tinuing. The Environmental Measurements
Advisory Committee has visited several EPA
laboratories and evaluated current analytical
methods and instrumentation research (Ref.
21).
Recommendation 10(c) directs that EPA
should have epidemiological questionnaires
and panel selection criteria approved by peer
groups before the next round of investiga-
tions. Specific questions that must be resolved
are identified. Regarding the latter, all ques-
tions are presently being addressed either by
the in-house staff or through contracts. As
mentioned previously, the matter of peer re-
views is being discussed with the Board.
Recommendation 12(a) directs EPA to es-
tablish authoritative peer review panels to as-
sist in improving research coordination. This
is being discussed with the Board.
Recommendation 10(d)—ideas
Recommendation 10(d) instructs that EPA
should review several ideas raised in the in-
vestigation team interviews. These have been
reviewed and some have been incorporated
into the epidemiology program.
143
-------
Recommendation 11
reorganizations
Recommendation 11 states that no signifi-
cant reorganization should occur at Research
Triangle Park's Environmental Research
Center until the end of Fiscal Year 1977, No
significant reorganizations have occurred
during that specified time period.
Recommendation 12(b)
management
Recommendation 12(b) instructs that EPA
should have a stronger focus on management
at the Environmental Research Center, Re-
search Triangle Park, North Carolina. In its
laboratory reorganization, the Office of Re-
search and Development established a line
structure with accountable managers direct-
ing the. research programs at each laboratory.
In turn, each laboratory has programs as-
signed on the basis of scientific areas that are
carried out by its complement of scientists
and engineers. In addition to the line man-
agement of each Laboratory, the Research
Triangle Park includes an office of the senior
research and development official. Essen-
tially, this official (who is, also a Laboratory
director) is responsible to the Assistant Ad-
ministrator for Research and Development to.
assure effective operation and administration
in the laboratories. '
Recommendation 12(c)
systems analysis
Recommendation 12(c) instructs that EPA
should create a systems analysis-operations
research program review {group. Using sys-
FederaJ Health Facilities in
Research Triangle Park, North Carolina
Durham A
Duke University J
10 Kilometers
EPA Interim
— Facilities
Chapel Hill
University of
North Carolina
16 Kilometers
National
Center for
jjjfl Health
National Institute
for Environmental
Health Sciences
EPA Interim
facility
National
Environmental
Health Sciences
Center
Permanent
Federal Site
Raleigh >
North Carolina
State University
144
-------
Appendix 3
terns analysis and operations research as
tools, significant program reviews are carried
out by the Office of Planning and Manage-
ment in its Program Evaluation Division. As
needed, the Division:
—Assembles and evaluates scientific,
. technological, cost, benefits, and institu-
tional data to critique existing program
activities, and recommends alternatives.
—Develops a long-range policy framework :
for EPA goals and objectives in consulta-
tion with other Agency offices; identifies
strategies for accomplishing these goals; ;
and assures that program activities are '
evaluated in relation to such strategies. :
—Conducts and coordinates analyses and •
evaluations of Agency-wide programs,
' including those crossing EPA orgaraiza- :
tional lines.
Recommendation 13
technical exchange
Recommendation 13 relates to the EPA re- ,
search program and maximal technical ex-
change. Recommendation 13(a) directs that j
EPA should seek cooperative research pro-
grams with universities and other laborato-
ries and agencies. The research program in
EPA has a large extramural component in- :
volving research grants with universities and i
interagency agreements with other Federal
organizations. For example, the overall coor-
dination and detailed planning of the Inter-
agency Energy/Environment Program is the
responsibility of the EPA (Ref. 22). Research .
and development activities under this pro* •
gram are performed by several agencies in
addition to EPA. Also, most of the EPA re-
search laboratories are located on university :
campuses, or in research parks developed by
universities, and have close working relation*
ships with nearby institutions. Recommenda-
tion 13(b) directs that EPA should promote
the exchange of scientists both within and j
outside the Agency, Over the last few years,;
the Office of Research and Development has
used the Intergovernmental Personnel Act,
mobility program. This program authorizes
the temporary exchange of career employees
between the Federal Government and state:
and local governments, institutions of higher,
education, and Indian tribal governments.
Currently, 66 individuals are participating in
the Office's mobility program.
Recommendations 13(c) and (d) concern;
EPA of funding individual Ph.D. thesis re-
search and the Science Advisory Board de-,
velopment of outreach programs. The Office,
of Research and Development will discuss
these recommendations widi the Science Ad-1
visory Board and-specifically seek their assis-
tance in developing effective outreach pro-
grams. Board members have provided gen-
eral comments but have not yet undertaken
any formal actions related to outreach pro-
grams.
Recommendations 14 and 17
research rote
Recommendations 14 and 17 state that the
EPA Administrator should clarify the role of
the Office of Research and Development and
determine if research should be conducted in
its present organizational configuration. EPA
is preparing a report to the Congress on
planning and management of the Agency's
research and development activities. This re-
port will address the most appropriate means
of assuring, on a continuing basis, that re-
search in the Agency reflects the needs and
priorities of the regulatory program.
Recommendation 15—facilities
Recommendation 15 directs EPA to resolve
the separation of facilities at Research Trian-
gle Park. This is the largest EPA field facility
and is located in North Carolina within the
geographical triangle bounded by the North
Carolina cities of Raleigh, Durham, and
Chapel Hill. In this area, three major EPA
components with a total of 1,857 employees
and contractors occupy leased space in nine
buildings. In 1967, a 509-acre tract at the
Park was donated to the Federal Government
for the construction of the National En-
vironmental Health Science Center. An over-
all master site plan was completed in 1971
and EPA was assigned a 44-acre site for con-
struction of a permanent facility (Ref. 23).
Recently EPA developed a long-range space
plan for its activities in the area (Ref. 24). This
plan is presently under consideration,
Recommendation 16
career development
Recommendation 16 states that EPA man-
agement should develop, implement, and de-
fend a professional career development pro-
gram for each professional. It is the policy of
EPA to plan and provide for the training, de-
velopment, and necessary career planning
for employees (Ref. 25). In July 1977, the
Agency strengthened existing mechanisms to
insure adequate career development (Ref.
26). As part of their annual performance
evaluation, supervisors are required to de-
velop a yearly training plan for each
employee.
145
-------
Resolution of Investigative Report Recommendations
Number
3(a)
3(b)
3(c)
4(a)
4(b)
4(c)
4(d)
S
6(a)
6(b)
6(0)
6(d)
7(a)
7(b)
7(c)
7(d)
7(8)
7(f)
8
Summary of Recommendations
EPA should publish an announcement regarding the
limitations of the CHESS Monograph,
EPA should not use the CHESS Monograph without
explicit qualification.
EPA should publish an addendum to the CHESS
Monograph including most of the Investigative Report.
Legislation should be reexamined regarding unrealistic
procedures and schedules.
EPA should design research to gain information and not
support positions.
OMB should allow all necessary resources if public
policy requires expeditious research.
EPA should advise Congress if budgetary restrictions will
impact completion of major projects.
OMB should be asked to develop procedures for prompt
review of questionnaire.
CHESS date analyses should be carried out only on data
with high validity potential.
EPA should publish research in refereed journals in a
timely fashion. ,
EPA should not publish large projects solely in
monograph form.
EPA should not initiate projects for policy consideration
unless they can be completed in a realistic time frame.
EPA should strengthen the CHAMP aierometric and
quality control programs.
EPA should shorten the time between data acquisition
and quality assurance analysis of data.
EPA should stop employing development stage
instruments before qualification testing.
EPA should not use laboratory modete of instruments in
the field until they have been field checked and
operating personnel trained.
EPA should reevaluate the opening of the CHAMP
operations contract to competition.
EPA research and monitoring personnel should closely
coordinate regarding chemical species.
EPA should have additional meteorological support for
air pollution health effects research studies
Action I
Shall ba
implemented
Shall be
implemented
Shall be
implemented
Implemented
t
Implemented
!
Implemented
i
Implemented
Implemented
Implemented
;
Implemented
Implemented
1 . 1
Implemented
' ,
Implemented
;
Implemented
i
Implemented
!
Implemented
Implemented
;
Implemented
Implemented
146
-------
Appendix 3
Number
9
10(8)
10(b)
1
-------
-------
Part 3
Report to Congress in February 1979, as required by
Section 8(d) of ERDDAA,
149
-------
-------
Report of the
Health Effects
Research Review
Group
U.S. Environmental Protection Agency
Science Advisory Board I
February 1979 \
75?
-------
EPA NOTICE
This report has been written as a part of the activities of
the Agency's Science Advisory Board, a public advisory group
providing extramural scientific information to the Admin-
istrator and ot her ' of fj_c1_aj j>__ of _th.eJE nv irojun e_nt_aj _Rr_oi JBJ: tl o. n. _______
tre~"B^aTrcr i s structured to provide a balanced expert
assessment of scientific: matters related to problems facing the
Agency. This report has not been reviewed for approval by the
Agency, and hence its contents do not necessarily represent the
views and policies of the Environmental Protection Agency.
752
-------
TABLE OF CONTENTS
Page
I. Introduction , 154
II. Summary and Recommendations 156
III. Committee Membership, Approaches, and Procedures ........ 164
IV. Research in a Regulatory Agency: The Conflict Defined 167
V. Observations of Current EPA Research and Development 172
VI. Utilization of ORD Results . ... . . . . . . . . ...... .189
VII. Status of Implementation of Two Sets ofNational Academy of
Sciences (NAS) Recommendations to EPA 190
VIII. Community Health and Environmental Surveillance-System"""
(CHESS): An Investigative Report . . . V~. . . . . ... . . . 205
APPENDICES
i
A. Charge to Committee and Authorization for Charge in
Public Law 95-155 211
B. Committee Membership and Consultants 219
C. Chronology of Committee Visits to Facilities and
Meetings 221
D. List of Principals Interviewed and/or Supplying
Information ........ ..".":.. .......... .223
E. Agenda of Major Regulations Being Considered by EPA .... 239
153
-------
I. INTRODUCTION
The Congress required an evaluation of the health effects
research efforts of the U.S. Environmental Protection Agency in
section 8(d) of Public Law 95-155, enacted November 8, 1977.*
Subsequent to the passage of the Act, EPA's Science
Advisory Board formed a special committee to perform the
mandated evaluation. This Committee, named the Health Effects
Research Review Group (HERRG) and composed of experienced
scientists and research managers, began their task i.n May 1978.
The Act stated that the evaluation include the following:
1) The health effects research authorized by this
Act and other laws;
2) The procedures generally used in the conduct of
such research;
3) The internal and external reporting of the results
of such research; I
4) The review procedures for such research and
results; j
5} The procedures by which such .results are used in
internal and external recommendations on policy,
regulations, and legislation;i and
6) The findings and recommendations of the report to
the House Committee on Science and Technology
entitled "The Environmental Protection Agency's
-Research Program with Primary! Emphasis on the
Community Health and Environmental Surveillance
System (CHESS): An Investigative Report."
The Act further stated that :
"the review shall focus special attention on the
procedural safegards required to preserve the scien-
tific integrity of such research and to insure
reporting and use of the results of such research
in subsequent recommendations. The report shall
include specific recommendations on -the results of the
review to ensure scientific,integrity throughout the
Agency's health effects research, review,^ reporting,
and recommendation process." \
. ^
The word "research" takes on a broad .meaning in a regula-
tory agency. For the purpose of this evaluation, health
effects research will be defined as requested by Mr. Costle in
his letter of June 17, 1978, to the Chairman of the Science
Advisory Board. A quotation from that-letter foil ows. ...
*Section 8(d) of this Act requires that a special evaluation
of EPA's health effects research be prepared by the Science
Advisory Board (SAB) and the report be submitted to the
Administrator, the President and the Congress.
154
-------
"To delineate the Congress1 charge more sharply, I
urge the Study Group to define health effects research
to include all planned activities, collection and
analyses of data done within the Agency for the purpose
of adding to the scientific basis for understanding
the effects of environmental factors on human health.
This definition would include those activities within
the Agency which may be used to assess human risk, and
which support standard setting and; regul atory deci-
sion and any activity which gathers new knowledge
about human health, or improves oujr understanding of
human health either directly or which can be used to
extrapolate to human health impacts."
In view of the limited time available to the Committee,
this study focused on the collection and analysis of data
primarily to add new knowledge. The analysis of existing
information and data, which already satisfies generally
acceptable criteria for scientific adequacy, was not considered
to be within the scope of the charge to; the Committee. Some
requested data were unavailable or not provided to the
Committee, therefore,the evaluation is ;not as complete as
initially anticipated or desired.
755 .
-------
II. SUMMARY AND RECOMMENDATIONS
A. Summary
The purpose of this report is to summarize the nature of
health effects research in a regulatory agency, to describe the
current status of that function in EPA, and to present conclu-
sions and recommendations. Supporting data and reports
relating to individual ORD facilities are available but are not
included.
The Committee visited (either as a full or partial
committee) all EPA laboratories performing health effects
research. Interviews were conducted with senior laboratory
staff, managers, and bench scientists as well as with senior
managers in the Office of Research and Development (ORD) and in
the Program Offices. For the purposes of this report, a
"Program Office" refers mainly to the Offices of Water and
Waste Management; Air, Noise, and Radiation; and Toxic
Substances, as these are the offices responsible for developing
regulations and setting standards or tolerances in response to
specific legislative acts. A 1ist of the faci1 ities visited,
Committee members visiting each facility, a:nd those EPA
employees interviewed or providing information can be found in
Appendices C and D.
The Committee also utilized the services of SAB members,
other scientists, and research managers on .an ad hoc basis
(Appendix B).
Programs and facilities were evaluated using a number of
criteria relating to the objectives of the .research and the
quality of facilities, staff and results. Among these criteria
were responsiveness of the research functio-n, research
influence in the decision making process, coherence of planning
and goal-setting between ORD and the Program Offices, and
quality assurance through peer review and publications.
The Committee interviewed many competent and dedicated
people with a real desire to work in a more effective,
efficient and involved way«
Research and development in a regulatory agency is a
complex task, one requiring research targeted to regulatory
requirements usually having short (six month to two year) time
frames. Research and development must be related to specific
regulatory needs. Identification of gaps in data and needed
research effort necessitates cooperative planning between
156
-------
program managers*, often unfamiliar with; research, and research
managers, who are often insensitive to regulatory pressures and
requirements. Researchers, as professionals, may have
difficulty in identifying results which will satisfy regulatory
needs when these results are not in theiir scientific
specialties. Constantly altering budgetary allocations to adapt
to rapidly changing regulatory needs aggravates research-
program staff relations. For these and ;other reasons, ORD has
frequently been viewed as unresponsive by many program
managers, who do not, in general, depend; upon ORD to support
their regulatory efforts. The Committee concluded that it
would require far greater joint planning and coordination of
ORD and Program Office staffs if ORD outputs, useful to
regulation, were to be commensurate with the funds allocated.
At present, it is not an effective or an efficient system. The
dilemma of research in a regulatory agency is further treated
in Chapter IV. . <
The most successful and useful research programs were
found where there was a close working relationship and
understanding between scientists in the 11aboratories and their
counterparts in the Program Offices. Such communications are
essential to an understanding of priorities, quality demands,
timing and what was truly needed to back up the regulatory
process in the short and long terms. Poor results were .seen
all too often, however, because close relationships did not
exist. "".
i
Pilot research committees have helped to establish
essential communications between those who have direct and
indirect responsibilities. Where successful, the resulting
agreements, e.g., Drinking Water and Pesticides, have helped to
make research more responsive and have icut across juris-
dictional barriers to establish objectives, goals and plans.
The pilot research committees are one rrveans to an end, but
shorter and more direct communications flines are needed between
data generators and data users. !
Beyond a committee approach, therje seemed to be little
consideration of organizational structures designed to
streamline decision making. Hopelessness was expressed many
times by those concerned when faced with the seemingly obdurate
character of the civil service system and the highly placed,
inflexible, and sometimes less than adequate individuals who
occupy unessential positions. Inflexibility makes it difficult,
indeed, to place people properly and to transfer or get rid of
people not performing up to expectations in their jobs.
*A program manager is defined as ;that person in the
Program Office who is responsible for developing the regulatory
or standard-setting activity for a specific program as mandated
by legislation. A research manager is;that person in ORD who
is responsible for formulating, planning, and executing specific
research programs. I
157
-------
Recent changes in the civil service laws were not seen as
adequate to effect much improvement. Desirable changes can
occur, but they will require enormous ieffort, training in, and
application of the principles of management by objective and
job performance evaluation to establish a clear understanding
of what is expected of each employee.
B« Recommendations ;
The Committee recommends that:
(1) ORD and Program Office
leadership take immediate sjteps to
coordinate all research plajnning and
activities in the Agency. Joint
planning to identify information needs
must begin as soon as a decision is
reached to prepare a regulatory
proposal.
!
Immediately following a program decision to develop a
regulatory proposal, Program Office and ORD staff should be
assigned to review existing information nee"ds. This group
should be given authority to organize Program Office-ORD staff
to identify regulatory needs for specific proposals and outline
the required research to fill the gaps";
(2) ORD continue to use
appropriate research committees, but
they should no.t be ORD's exclusive
planning mechanism.
Research committees, initiated on a pilot scale in 1978 to
help ORD plan and coordinate its research activities with the
Program Offices, should be used sparingly. These research
committees, really task forces, will be most useful when
research needs relate to multiple Program Offices and
laboratories.
*
The research committees should beiused for identification
and prioritization of needs. These committees should not be
involved with research implementation.
Key managers within ORD should devise mechanisms to
develop well understood objectives, goals, plans and measures of
performance for how research should be conducted.
The Committee does not believe 'that it will be possible for
ORD to fulfill its function without extensive agreement by key
personnel on objectives, goals, plans,,:and measures of
performance. It might be helpful for ORD to hire experienced
management specialists, as consultants, to help address some of
the difficult managerial problems which currently exist.
158
-------
ORD leadership must take steps as sopn as possible to work
out an understanding with Assistant Administrators in the
Program Offices to simplify and shorten lihes of communication
and to cut to a minimum the reprocessing of decisions by the
Washington ORD staff.
(3) The scientific staff ojf ORD
identify subject areas and establish
active investigatory groups to pursue
long term research essential to
regulatory needs. (Implementatilon of
recommendation 1 will ensure th;at long
term research efforts remain
relevant.) i
There should be a long-term ORD investment in researchers
and facilities to develop highly active and productive groups
in those research areas which are central !to large segments of
the Agency's regulatory activity. Allocation of a specific
percentage, at least 10%, of the ORD budget for relevant
research in case subject areas seems to be reasonable.
" - • "(4) The incorporation of ORD
research results into-criteria,
standards, and regulations be
strengthened. ;
ORD must stress, at all levels, the importance of producing
results and assisting with their incorporation into regulations
and standards. ORD has neither fully recognized or accepted this
criterion for judging its efficiency, nor1 developed mechanisms
for efficient utilization of research results by Program
Offices. ORD does not maintain records of; results which have
been incorporated into regulations.
The formation of the Environmental Assessment Groups is a
step in the right direction. Part of the* responsibility of
these groups -shoul d be the documenting o.fL which research results
have been utilized, the continuing audit iof the usefulness of
ORD results to regulations and standard setting, and getting
feedback from the Program Offices about the research and
research planning activities. The Committee found the model,
outlined on page 9 of Volume 11,1 of.. "Research and Development in
the Environmental Protection Agency," toibe still relevant for
Agency use. -..-- ,. - '
1
(5) Responsibility and ;
authority for implementation;of
research and reporting of research
be vested in the laboratory \
directors and the staff scientists,
after agreement on research p'la'ns.
159
-------
The Committee feels that too many specific directions
regarding research implementation come, from headquarters. This
prevents the scientists from using their talents and diminishes
the scientific climate for innovative research.
(6) After agreement ori
responsibilities for research
implementation, laboratory i
directors and their scientific
staff be permitted to performed
their assigned tasks. (See
recommendation 5.)
Laboratory staff need protection against unwarranted
mandates, incursions into allotted time for research, and
reorganizations and spurious changes iin policies that occur with
the all-too-frequent changes in leadership. The scientists also
need a sense of the Agency's long range commitment to its stated
goals. ;
(7) An expansion of the
Interagency Regulatory Group (IRLG)
activities be carried out. i The
- - excellent planning initiatives of
IRLG should be "extended to include
environmental health research.
-The IRLG is seen as an excellent beginning with the
potential of reducing duplication and confusion among agencies.
This effort should be extended to strengthen coordination of
research planning by all agencies conducting environmental
health research. [
(8) A simple, easily under-
stood accounting system be
established for planning, assigning
and monitoring use of funds'and
personnel relative to ORD's'
intramural and extramural programs.
Effective use of limited funds and personnel requires that
they be carefully managed. The accounting systems now in use
are inadequate. At the present time, analyses are not performed
to,place in perspective salaries, equipment costs, services,
etc. Those cost breakdowns are necessary to give ORD
information about responsive and nonresponsive work at the
different laboratories performing health effects research.
(9) Standard procedures for
awarding contracts, grants, land
cooperative agreements, and
monitoring extramural research be
simplified and enforced. ;
160
-------
Current elaborate rules for contract ^and grant awards
should be reviewed and revised to promote^efficiency and
timeliness of extramural awards. All personnel must adhere to
these new procedures. This would end the durrent abuses of the
extramural award system. Procedures should be adopted to ensure
adherence to the new requirements after revisions are made.
The monitoring procedures should indicate methods for
evaluating the performance of contractors and grantees during
and after completion of their work. Furthermore, the extramural
research results should be published in peer reviewed scientific
journals. EPA-published reports are no substitute for open
literature publications.
Adequate travel funds should be allocated for proper site
visits and for monitoring of extramural work. Presently, there
is no routine, operational audit of the quality of extramural
research. :
Responsibility for extramural research (planning, awards,
and monitoring) should be made according to the staff's
.capabilities: ,t.o_ effectively plan and monitor such research. This
should take into account the amount of independent in-house
research .expected from"the staff scientists. Extramural
monitoring assignments should only be mad£ to scientists who
have demonstrated professional competence; and are thoroughly
fajniliar with, bow research is conducted i|n the field being
monitored.
(10) Scientific peer rev;iew of
proposals, programs, and intramural and
extramural research be greatly
intensified.
i
Scientific credibility and def ensi bi'l ity of research done
in support of regulations are key elements of the success and
acceptance of the Agency's role by the public. The Committee
feels th'at, to the maximum extent practical, scientific peer
review mechanisms should be utilized to improve the quality of
final research results. i
All programs and organizational units should be
periodically subjected to peer review byiqualified scientists
from outside the Agen;cy.' All proposals and completed research
should be reviewed by peer scientists withi n the Agency, and
representative items should be reviewed by scientists outside
the Agency.
The quality of research in EPA is important not only
because any research should meet standards acceptable to the
scientific community but also^-for reasons derived from the
regulatory nature of the Agency. ;
161
-------
To ensure acceptability of research results, the studies
must be reviewed by one's scientific peers and published in
reputable scientific journals. Failure to so treat results of
research investigations involves the rjis'k that review will occur
at a later date, in a adversary situation, with possible
refutation of results and embarrassment to the Agency.
(11) A dual-ladder promotional
system be implemented for qualified
scientists to advance in grade and
salary without having to undertake
supervisory or managerial !
responsibilities. ;
Presently EPA has a promotion ladder inadequate to allow
scientists to remain in the laboratory and be promoted strictly
on the basis of their scientific excellence. EPA suffers from a
poor reputation as far as the scientific quality of its health
effects research is concerned. This reputation is not totally
deserved. There does need to be a system whereby both qualified
scientists and qualified managers can ieach advance and be
rewarded in their own fields.
1
Well qualified personnel are the key ingredient to the
conduct of a scientifically sound research program. At the
present time, there are both formal and informal procedures that
encourage scientists seeking promotion's to accept supervisory
and administrative responsibilities, thereby reducing the amount
•of time they have to spend on laboratory research.
When personnel are assigned to senior management positions,
primary consideration should be given ;to individuals who have
demonstrated scientific and managerial! capabilities; an
understanding of how research is plann.ed, conducted and
reported; and the ability to communicate research information
and needs to both scientists and non-s;cientists.
(12) Research management give
immediate-attenti on to instituting,
in the laboratories, a variety of
procedures to create an atmosphere
conducive to scientific excellence.
i
Even though the laboratories are located on or near
university campuses or other research institutions, EPA
scientists were somewhat outside the mainstream of scientific
events. The Committee, therefore, urges management to regularly
schedule seminars in which both outside scientists and Agency
scientists participate, invite outside scientists to spend time
in EPA laboratories (in addition to use of the Interagency
Personnel Agreement--IPAs), encourage :EPA scientists to spend
time in outside laboratories (an exchange program), sponsor
workshops and symposia, and generally .institute a closer
interaction with geographically close institutions.
162
-------
(13) ORD and senior Program Office staff
rotate assignments, preferably on the basis of
those ORD and Progam organizational; units
which consistently interact. '.
It is essential for effective performance that Program
Office and ORD managers understand the problems and capabilities
in each organization. Program managers are often unfamiliar with
research planning, laboratory work and this inherent time
constraints. Likewise, research managers are often unaware and
insensitive to regulatory pressures and rjequirements and with
the dilemma of how to present data in a fiorm useful to the
Programs. \
(14) The research progra!m
using the clinical inhalation ;
exposure facility at Chapel Hi;ll,
North Carolina, be fully staffed and
a sound research program implemented
as soon as possible. ;
i
The clinical inhalation facility at Chapel Hill is a unique
facility,, engineered to deliver the desirjed exposure levels;
however, the scientific program, staffing, and plans to utilize
the facility are: total ly inadequate — a ve^ry conspicuous waste,
as it now stands. j
!
ORD should immediately assess the fu:ture need for and use
of this facility, establish goals and support for the facility,
and assure that the facility is not wasted—even if EPA has to
make it available to outside groups. This facility was designed
for long range studies to accurately asse!ss and predict the
potential adverse effects of selected environmental chemical
agents. !
The inhalation program, once developed, should be
scientifically peer reviewed and approved.
163
-------
III. COMMITTEE .MEMBERSHIP, APPROACHES AND PROCEDURES
A. Committee Membership ; .
The Health Effects Research Review Group (HERR6) consisted
of core members and consultants selected for their scientific
expertise and research management skillis. The consultants
supplemented core members and were used; to provide specific
expertise for the evaluation of individual laboratory programs
or special topics of research. A list iof Committee members and
consultants is Appendix B.
T
B. Approach to the Assessment of R&D and Procedures
Used
It was apparent from the outset thjat the Committee needed a
clear understanding of the mission of health effects research as
seen from the viewpoints of the personnel in both the various
Program Offices and ORD. Responsiveness; of the research function
to the pressing (often mandated) needs ;of the Program Offices
has been inadequate in the past; this problem has been clearly
described in a report by a committee of the National Academy of
Sciences, Analytical Studies of the U.S. Environmental
Protection Agency, Vol ume III: "Rese'arch and Development in the
Environmental Protection Agency," 1977.•
Of necessity, the Committee had to; subdivide much of its
-i nvesti gat ion . into-smal 1 study group activities. A common
approach was taken to make it easier 'to- analyze and assemble, the
findings of the various study groups into an integrated final
report. Thus, the research function of the Agency was to be
analysed 1n'the1 context of the regulatory responsibilities of
the Agency, which in turn requires a reliable and defensible
data base for decision making. The Committee agreed that
research can only be understood if the {reciprocal relationship
.between the users of the information (the Program Offices) and
the generators of the information (ORD) was examined. The
perceptions of both/the generators and ithe users were,
therefore, to be probed to determine if: there were shared goals
and a shared understanding of what is known, what is unknown,
and what needs to be known. It was also necessary to determine
whether there was a shared understanding of the time frame
necessary to generate or assemble the needed data. These
perceptions -were to be examined at several hierarchical levels
to determine if the intentions of the supervisors were accepted
in a way that motivated the respective Organizational units
regardless of location or attitudinal preferences.
While conducting interviews and fact-finding sessions,
Committee members tried to use some of :the following checkpoints
as they were appropriate for the various situa-tions. These
points were- the basis for the formulati'on of this report.
164
-------
a. Checkpoints relating to the mission of health-
related research as it supports short-tefm and long-term
Agency needs:
1. Responsiveness of the res'earch function (as
defined at the outset) '•
2. Sense of urgency and commitment of the
research function ;
3. Research influence on judgments made on the
decision making process (jle^el of influence
and dependence by the program offices)
4. Coherence of planning and goal setting be-
tween the Program Offices, and ORD (Are
budgets really reconciled and supported
by both the Program Offices and ORD?)
5. Examples of good and poor! responses by ORD
6. How and by whom is the decision made to
initiate and conduct specific research
investigations? s
7. How are information gaps 'identified? How
are long-term trends with potential
environmental impacts identified? How are
: long-term research needs jdefined and planned
to assure budgetary support"?"" "
8. Beyond the Program Offices and the ORD
functional organizations, what other factors
help influence what research is to be done?
b. Checkpoints relating to the qua!ity of health
effects research as it supports short-term and long-term
Agency needs:
1. Quality assurance:
a) Good laboratory practices
b) How is quality assurance implemented to
improve the defensibi1ity of results?
c) Evidence of attention to detail and
carefulness (facilities, work flow,
" " "' housekeeping, attitude, safety program)
d) Personal scientific integrity,
including quality of planning and
experimental design:, rigor of analysis,
courage to disprove;one's hypotheses
(or hypotheses of a; superior), and
•--•"•• acceptance of opinions of qualified
•"•"•"" peers
e) Can the most qualified people be
quickly identified?'
•• f) Is the- civil service system seen as- a
positive factor in the encouragement of
a good research program within EPA?
165 I
-------
2. Publication of results (reporting)
a) In journals requiring scientific peer
review, internal government
publications, journals or meetings not
requiring scientific peer review
b) Methods for approving manuscripts
before release or publication
c) Is publication seen as helpful to
career development?
With these checkpoints in mind, the Committee conducted
its assessment through a series of fact-finding sessions and
public meetings in Washington and in va'rious EPA laboratories
(see Appendix C). The Committee chairman and co-chairman first
discussed the charge arid the plans f or 'accompl i shi ng the
evaluation with the appropriate Congressmen and their staffs.
Subsequently, the Committee met with the Administrator, the
Assistant Administrators and other senior EPA policy and
management staff in various Program Offjices, and with
representatives from the regions, laboratory directors, senior
science managers, and individual laboratory scientists
(Appendix D). The Committee members reviewed legislative
mandates, various EPA documents, and othei—p-apers and memoranda
relating to the Committee's charge. ,
166
-------
IV. RESEARCH IN A REGULATORY AGENCY: THE CONFLICT DEFINED
A. Present and Future Agency Needs; for Data
Volumes have been written on regulatory agency
research needs in general and on EPA research needs in
particular. Therefore, the Committee approached the subject of
the research and development needs of EPA with trepidation and
elected initially to describe the pressures and constraints
imposed generally upon a research and development group in a
regulatory agency and those imposed upon EPA in particular.
Program administrators in regulatory agencies are captives
of the calendar deadlines imposed for regulation by the specific
statutes they enforce. These agencies rqutinely deal with
Congress, irate constituents, citizen groups, the media, and
others. The professional skills which contribute to their
success and/or survival are all devoted to integrating immediate
pressures and existing knowledge into a s.et of regulations
acceptable to all. This is a difficult situation, one requiring
sensitivity to human behavior and appreciation for the relevant
available data base. Regulations are usually compromises, their
political socio-economic impact and whet her,..they., can be
enforced. The scientific and technical bases for a regulation
will be put to rigorous test if, and only if, the regulation is
challenged. Judicial review will incorporate and consider all
relevant data; an administrative "gamble" made in the absence of
sufficient data to support regulation will very likely lead to
remanding the rule to the Agency. Development, promulgation and
enforcement of regulations, particularly jin an area as
underdeveloped and evolutionary as environment, is a difficult
exercise. ;
i
The formal challenges to regulat 1 on(are cyclical. Because
of inflationary pressures on regulatees since 1976, there has
been an increasing trend toward challenging environmental
regulatory promulgations,. The courts have been sympathetic to
the innovative promulgations of EPA, but'the economic impacts of
EPA administrative interpretations of enabling statutes have led
to regulatee demands for more complete substantiating data for
promulgated rules; those demands will increase in the future.
Even those sympathetic to prudent Federal environmental
regulations are demanding higher standards of proof during this
highly inflationary period of increasingly demanding and varied
Federal regulation. Because environmental rules are still
perceived by many as a luxury affordableionly by a prosperous
private sector, EPA must anticipate continuous, more
sophisticated private sector challenges because of inflationary
pressures.1 i" :r
167
-------
These challenges will be overcome only by convincing
arguments for regulation, arguments drawing upon defensible
data. These data will have to relate:specifically to
improvements in human health if EPA is to fulfill its mandate as
an Agency. In the future EPA will increasingly have to document
health gains anticipated from allocation and expenditure of
large sums of money for regulation and control of environmental
pollution.
B. Investigatory Time Frames '
Specific statutes include timetables for regulation
assigned by Congress. The Agency has;formulated a table of
regulations scheduled or in progress (Appendix E). Program
administrators will formulate these regulations with whatever
data are available prior to and until|the scheduled completion
date. In general, schedules for EPA to write regulations are
short; 6-12 months is normal, while 18 months is considered
long. These are short time frames for: generation of new
information in. the laboratory or in the field. EPA Research and
Development Office (ORD) personnel have had enormous difficulty
responding within-the t-ime al 1 otted. |t_ jjs essential that ORD
and Program Office personnel carefully evaluate information
needs critical to implementation of scheduled regulations. This
must be done as soon as a statute is {assigned to EPA for
enforcement. In this way, ORD will be able to utilize the
maximum available time to generate needed data -for--regulation.
We did not perceive that research needs are routinely approached
in this manner. ;
i
C. Investigator and Program Staff Interactions
The perceived needs of program managers are usually
very specific and often conflict withjneeds perceived by
researchers. For example, researchers may regard experiments
requiring toxicity data from animal exposure to pollutant agents
at concentrations far in excess of those likely to occur under
normal exposure as of little relevance to scientific
understanding. Program personnel, howiever, may regard
demonstrated toxicity data, even at unrealistically high
exposure levels, as a rationale for regulation. Sorting out
these differing perceptions requires personal interchange if ORD
is to respond in a timely and meaningful manner. Too often in
the past the Program Offices have perceived ORD as unresponsive
because results were of a kind different from what had been
anticipated and because research time; frames were too long to
allow the Program Offices to use the data produced. Under these
circumstances, program administrators did not look to.ORD for
solutions to their problems. ;
168
-------
Principal Program Office and ORD adm(i ni strators are located
in Washington, D.C. ORD investigators are located in laboratory
facilities throughout the nation. Specific administrative
mechanisms are required to ensure that communications occur
between Program Office administrators and; ORD investigators as
research in support of specific regulations progresses. In 1978
five research committees were initiated on a pilot basis to help
ORD plan and coordinate its research activities and become more
responsive to the needs of designated Program Offices. These
pilot research committees have helped to provide an essential
communication function; furthermore, they have helped to
establish understanding and commitment to: objectives, goals, and
plans. Carefully selected research commijttees are seen as a
means to an end, although a cumbersome onie, because their
meetings help to educate those who need to know. In the long
run, however, the functions served by the pilot research
committees need to be institutionalized so that laboratory
directors are not excluded from key roles in leadership or from
maintaining a high level of competence in their respective
laboratories.
Program administrators frequently have their primary
training in the legal or engineering professions; they are often
not familiar with the state-of-the-art of ORD scientific
research. ORD utilizes scientifi-cal ly trained personnel, at all
levels of the organization, those working at science on a daily
basis. One can draw flow diagrams of the decision making
processes in a regulatory agency, diagrams illustrating ORD and
Program Office personnel interactions. However, in the final
analysis, exchange of information and resolution of issues is
required of persons with essentially different bases of
understanding. There will be a major built-in obstacle to
communications between ORD and Programs Offices as long as ORD
relies entirely on scientific managers and the Program Offices
on managers who pride themselves on their pragmatic approach,
managers grounded in law and/or engineering sciences. By one
mechanism or another (rotation of assignments, creation of new
positions for complementary professionals in each Program Office
and ORD), .there must be' promotion of ORD-Program Office
communication by ensuring that senior managers have a common
1anguage(s).
• D• Evaluating the Responsiveness lo'f OR'D'
- The responsiveness of-ORD-.i-s-judged by a variety of groups
and individuals, including EPA program managers, Congress,
citizen groups, and the media, to name a' few. The Committee
probed primarily EPA program managers' perceptions of ORD's
responsiveness to their needs... .Senior program managers have.
indicated that there have been recent improvements, but much
169
-------
remains to be done. In the past, many Program Offices did not
participate in ORD planning. Recent joint ORD-Program Office
research planning exercises, such as the; pilot research
committees, have caused Program Offices to be more favorably
disposed toward ORD activities.
Ultimately, ORD's response to the Program Offices will be
more stringently judged by how effectively the research results
meet the specific needs of the regulators in a timely and
scientifically rigorous fashion. The current auspicious climate
for ORD pilot research committee planning must not be confused
with future ORD outputs necessary to satiisfy hard-pressed Agency
program managers. For this reason, the 'major ingredients of ORD
research that would allow ORD to be considered "responsive" to
regulatory program needs will be briefly discussed. Following
this discussion will be comments on the ;current EPA research
process from the planning stages to the 'final utilization of
results by Agency Program Office staffs.:
The timing of the delivery of research results to a Program
Office is a major factor contributing to the perception of ORD's
responsiveness to Agency needs. Regardless of the quality of
research results, they are viewed as only.marginally. useful if
available after statutory deadlines have; passed. One can argue
that in the long .run "late" results will b.e-integrated into
environmental programs, but this does not engender Program Office
staff confidence in or support for ORD. >
The scientific and technical soundness 'of-ORD'-'-result's is
crucial if EPA Program Offices are to sustain their regulatory
positions. Transfer of weak results by iORD will lead either to
rejection of these resul'ts by administrators or to utilization
with subsequent public embarrassment upon disclosure of a weakly
supported position and/or reversal of the Agency position by the
Courts. ;
In addition to being scientifically defeasible, research
results must be targeted to meet Program Office needs.Needs
must be commonly perceived and agreed upon-by researchers and
program administrators. Dictation of needs by regulatory staff
to researchers can result in untimely and fruitless
investigations; likewise, researchers with inadequate
understanding of program needs may pursue scientifically sound
studies,which are irrelevant to the Programs.
f h e understanding of ORD results b'y potential users i s
probably a major ingredient of the perception of responsiveness.
ORD must not only deliver sound results in a timely manner, but
must also trans!ate_these_results into terms and ...concepts
understandable to the users, i.e., the Program Offices. ORD has
a responsibility to assist its users in ^understanding the
strengths, weak/nesses and full significance of those research
results transmitted for Agency use. :
17O
-------
The above ingredients of "responsiveness" relate to the
research function as it serves regulatory needs. Each ingredient
must be carefully developed and nurtured, literally on a project
basis, if expectations of ORD efforts are [to be fulfilled.
With this brief introduction to the idemands placed upon
ORD, specific aspects of performance of he'alth effects research
and development in the Agency will now be idiscussed.
E. What is an Investigatory Product in a
Regulatory Agency? ;
The investigatory product in a regulatory agency is that
body of scientific information and data base which is either
available to or resides with the scientific staff. The product
must be provided to the Program Office in;a form that is useful,
understandable, and defensible in setting^reasonable standards
and for writing regulations. ;
I
This scientific information can be provided to the Program
Offices in many ways. The best way would^undoubtedly be to have
the research described and published in professionally peer
reviewed journals, but information can also be provided-through
monographs, letters and verbal presentations. The key to the
desired investigatory product is for the Agency to have an in-
house core of capable scientists who understand the regulatory
and standard setting requirements, who can perform the necessary
literature searches, can perform their own research and evaluation,
and can freely attend professional scientific meetings where
discussions and information exchanges occur.
777
-------
V. OBSERVATIONS OF CURRENT EPA RESEARCH AND DEVELOPMENT
A. Identification of Research Needs
ORD can be viewed as a large multifunction apparatus
capable of responding in a variety of! modes if appropriate
planning of the necessary dynamics anjl a complete "tune-up"
occur prior to "start-up." The initial step is to identify the
required outputs. ORD outputs should be responsive to regulatory
needs, in the short or long term. At present and, indeed, during
the entire history of EPA, short term* R&D needs have been
stressed. We do not see any conflict; between simultaneously
sustaining research programs with long (years) and short term
(months to years) goals, provided Program Office-ORD concurrence
is reached as to these goals. ;
Historically, Program Offices outlined needs according to
their perceptions of the problem. It* was a hierarchical
planning process which gave the scientists at the laboratory
little understanding of what was need:ed or why. Laboratory
scientists often communicated with loWer level Program Office
staff who did not fully understand the needs and priorities of
their program. I . .
. There seems to be no systematic identification of
information gaps (research needs) in the Agency. This
identification should take place as sioon as EPA receives
legislation on which it must act; it requires close cooperation
between the appropriate Program Off ice and" ORD scientists,
especially those in the laboratories.. These staff members
should carefully analyze the Act to a.ssess what the Agency must
do to gather the needed information and to fulfill the
requirements of the Act. Additional research needs come from the
process of drafting regulations and from writing the criteria
documents when perceived needs for information are recognized.
Better identification of needs takes place when there is a close
association between ORD and the Program Office, but this must be
directed throughout the Agency in a systematic way.
Long-term (anticipatory) research in subj.ect, areas central
to Agency responsibilities should be^planned as a natural
extension of the identification of gaps in the data baseT It
cannot be designed in a vacuum, as an activity to be initiated
or terminated at will. When effective cooperation occurs
between ORD laboratory and Program Office personnel and when
effort is expended to define common objectives, goals, and
plans, opportunities are likely to arise for defining relevant,
long-term research programs. ;
The perception of needs for longer term research arises
from the interaction of key regulatory people and creative
researchers who are in touch with the issues and the scientific
literature. People who do research, read scientific literature,
172
-------
attend meetings and work cooperatively with the Program Offices
are those with the best resources to define needs. The
Committee believes that the stress on identifying long-term
research needs must come from ORD and that more attention must
be devoted to identifying these needs and; pursuing the
associated research studies.
The pilot research committees have helped to identify gaps
deserving further research effort, to date only short term; but
even this has helped to gain better insight into Agency
priorities. Because of the large number oif people involved,
these pilot research committees are cumbersome, but they have
forced a meeting of minds among key people in the Program
Offices and ORD. In fact, the identification of research needs
by individuals with diverse backgrounds and responsibilities is
a very strong feature of the pilot research committee effort and
should be retained regardless of the ultimate fate of the
activities of these committees. This should be expanded to
include identification of long term needs.
Several efforts at identifying research gaps and
implementing research should be highlighted. The Drinking Water
Program has been an example of effective .cooperation in .
identifying and i mplementing. research ne.ed.s,, whe.reas the Human
Inhalation Exposure program, at''HERL,. RTR~,( Chapel. Hill) and the
Animal Exposure Program at HERL, Cincinnati are examples of very
poor coordination. In the area, of pollutant inhalation studies
on human subjects, the _sci;ent"i s.ts,, j)f •.Jhe_LChape f. Hi 11 f aci Tity
have attempted to implement longer range1studies to predict and
assess more accurately the potential adverse health effects of
selected chemical agents. In general, ORD administrators have
been sympathetic to funding short-term inhalation projects, but
have not been supportive of longer term inhalation research
programs. The Inhalation Toxicology (animal model) Program at
HERL, RTP, on the other hand, was enthusiastic about its
relationship with the Program Office. This group is well
supported, largely as a result of a sustained effort by the
section leader to keep close contact with ORD and Program Office
personnel in Washington. Development of new methodologies was
considered to be a major responsibi1ity"pf the group "working "on
animal inhalation toxicology; they expreissed .the desire to be
involved, in toxic substances support as well. - This group also
supervised contracts and grants. Management of both grants and
contracts in addition to the "in-house" respons.i bi 1 i ty was seen
as a .desi rable comporierit of the total jo:b done by'the Inhalation
Toxicology Section. A key element of this program seemed to be
the desire on the parts of the Program Office and the laboratory
to engage in cooperative planning and goial setting. The result
is a~very spirited and productive group :of researchers. -
173
-------
Scientists in the Diesel Exhaust Program at Center Hill
(Cincinnati) clearly foresaw the emerging importance of diesel
engines and attempted to start long-range research several years
ago. These projects were turned down by ORD staff members in
Washington, who have recently recognized the need for such
studies. Work is now frantically underway to obtain needed
results to meet the statutory deadline;for establishment of
diesel emissions criteria. ;
B . Plan nin g Rese a r c h Projects ;
1. Budget Formulation \
During the period of our Committed review, the Agency was
in the second year of zero based budgeting (ZBB), i.e., fiscal
years 1979 and 1980 budgets were in progress. Funds are
authorized and appropriated directly tp ORD in categories
related to enabling legislation or special projects.
Prior to the introduction of the ZBB process, senior ORD
personnel often established project allocations without
communicating with Program Office managers. The zero based
budgeting process has been an exasperating (but probably
desirable) experience for all concerned — Program Offices, ORD,
and laboratories alike. It has forced: a .certain amount-of
communication and has led to some good^ though torture-d,
outcomes, especially in the pilot research committees. However,
communications are/still occurring only, between ORD and Program
Office personnel of relative seniority. We perceive that many
bench scientists in ORD do not understand the relationship of
their work to overall ORD and Agency goals. If communication
involved the laboratory investigators doing the work, even more
effective decisions could be reached, while simultaneously
gaining the commitment of the researchers to the work.
An additional budgeting problem is the mismatching of
personnel ceilings and funding for specific programs and
laboratories. Numerous examples were found in which program
areas in specific laboratories had very few or no people
assigned and relatively large amounts of funds available.. In a
few instances, relatively large numbers of personnel were
assigned with limited funds available.' At the headquarters
level"," the view was frequently expressed that OMB had" mi nimized
management's latitude for shifting personnel between programs to
better match program needs and fund allocations. Laboratory
personnel expressed a feeling of hopelessness in dealing with
the problem and were, on occasion, forced into the unrealistic
posture of showing, for the record, programs with substantial
fun-ding managed with zero personnel ; obvi ously this does not
ha'ppen. The people who are assigned to manage the program
simply charge their time to some other program that has a more
adequate manpower ceiling. The result is manpower accounting fay
progam that is suspect, at best, and probably of limited value.
174
-------
Clearly, if laboratory directors are to be effective research
managers, they must be given the latitude: to utilize assigned
personnel without rigid program area constraints. A change in
approach should allow laboratory directors to place increased
emphasis on developing the appropriate mix of disciplinary
skills of their staffs to better serve current and future
program needs. i
Allocation of travel funds is another budget problem. When
travel funds are allocated to the laboratories, consideration
should.be given not only to the number of scientists in the
laboratories, the degree of participation in extra laboratory
Washington mandated activities, and the required extramural
program monitoring required, but also to;the geographic location
of the laboratories with respect to these activities and to the
location of national scientific meetings. Furthermore,
increased flexibility should be given to the laboratory
directors for control and- utilization of travel funds. For
example, the laboratory director at the ERL in Duluth should be
authorized to approve travel for his staff to go to Canada. One
of the major functions of this laboratory is scientific
cooperation with their counterparts in Canada. Yet this
collaboration is minimal because travel to the: .Canadian . .
-laboratory,in Thunder.Bay is considered foreign travel and must
be .approved each time, well in advance, by ORD headquarters in
Washi ngton. " i
'-'•'•'•:.---,'....-.--2. Resear.ch,..Progr'am--Eormul ation v . ;.;; _/.V.
The Committee senses that the major; contribution of the
pilot research committees in program formulation has been to
overcome previous inadequacies in planning and to initiate
discussions of research by the many individuals with an interest
in the outcome and utilization of the work. The previous "old
system" of hierarchical planning failed to establish
understanding and commitment by those who should have been
involved. The pilot research committee Approach to planning has
been warmly endorsed, by laboratory staffjs because they,
personally, provided inputs-and gained familiarity with and
perspective of the entire program and an awareness of their
projected contributions to the entire prbgram. This type of
"grass-roots" motivation must be retained, but the leadership
must_.alsp be _.i nvol ved in the process. Methods need to be
established to institutionalize the i nvol veme.nt. and commitment
of the .staff through proper involvement of laboratory directors,
as well. Pilot research committees are a useful means to an
end, but they are no substitute for accountable leadership,
whjch.must_beresponsible for the integrity and quality of the
final—product. " ~"' "' ' >' • •'-
When laboratory personnel did feel ;that they had an
influence in setting priorities, they became i nvolved with input
to the Program Offices, became i nvol ved ,1' n the objective
175
-------
setting, and became involved in the design of protocols to meet
objectives. The drinking water projects are outstanding
examples and illustrate many of the elements of success that
need to be emulated by others. The reputation of the people,
their professional standing, and thelhistory of performance
stemming from the Cincinnati laboratory and its predecessor, the
Taft Center, are influential factors which command the respect
and attention of the Program Office. A critical factor in
responsive and quality programs is the need to maintain a
continuum of qualified, knowledgeable; personnel. Also, it is
important to recognize that, in the drinking water program
office, there are counterparts to ORE) staff who understand the
scientific and technical issues. ;
3. Pre-project Evaluation :of Productivity and
Costs i
The laboratories in ORD are mostly media oriented, and
scientific program projects and resources are assigned
accordingly without assessment of the cost-effectiveness of
performing research in each specific-laboratory.
..ORD, or an outside agency, should perform a yearly
assessment of each 1 aboratory's pas.t .performance with respect
to"the quality of the research information produced, the
timeliness of delivery of research results, the cost-
effectiveness of.the laboratory, and ;other factors which deal
with a 1aboratory's .performance and prbductiyity.; Only after
such assessment has"been performed and defi ciences corrected
should the scientific work (decision units) and resources be
assigned to a specific laboratory.
4. Good and Poor Planning:
a. Some examples of good responses by ORD
-The drinking water program at Cincinnati
-The animal inhalation toxicology program at
;. __../; ' .. .... -.-. RT.P. ;. . -, .,.;..._ .._ _..„•_;._.:.•.;...
-The pesticide pilot research program
involving program and laboratory personnel
.. , ... -T.h.e Wenatchee Laboratory studies of field
exposure, of applicator to pesticides
, : ^'(relevant workigoes back in history and
":._.. ""should be better" uti 1 i zed)
These good responses all have a,very important common
element; namely, the participants work at good communication.
.-Objectives, goals and plans are understood by the affected *
parties. Solid scientific approacheis are being utilized and
researchers in the laboratory are involved with personnel in
..the Program Offices. _ -. .
176
-------
Figure
Diagrammatic Representation of Old and New Systems to Develop
Experimental Protocols at Bench Level of Investigation
Old System
New System
Assessment of Research
Needs by ORD Headqtrs
(from internal ORD and
external sources)
Lists of In-
dividual Pro-
gram Research
Priorities
priori t-
ization
ORD Lists of
Research Pro-
gram Priori-
ties
Development of ubjecfive
Statement by Hdqtrs ORD
(includes statements of
scientific objective plus
dollars and manpower)
Senior Managers
Intermedia Ranking of
Research Priorities
To Laboratories for
Development of Accom-
plishment Plans
Decision Units for Research
funding 'of
decision
unit ;
Laboratory Development
of Work Unit Document
for Each Accomplishment
Plan (retained at Lab-
oratory)
To Laboratories to Develop
Work Unit Documents
Lab Development Protocols
for Individual Project Con-
tracts or Grants
Lab Develop Protocols
Individual in-house
Projects, Contracts,
or Grants
777
-------
b. Some examples of poor responses by ORD
The Human Inhalation Facility at; Chapel Hill is an unusual
facility, engineered to deliver the desired exposure levels,
but the scientific program or plan toi utilize it is totally
inadequate-- a very conspicuous wastej.
The Diesel Exhaust Program at Center Hill was prevented
from doing adequate dosage response tests because of directives
from Washington. The Epidemiology Program associated with the
Diesel Emissions Program lacked adequate and mature direction.
C. Performance of Research ;
EPA's intramural health effects; research is conducted in
two major laboratories and in portions of three other
laboratories, which were established primarily for other
purposes. The major laboratories are Research Triangle Park,
North Carolina, and Cincinnati, Ohio. '• Small programs are in
effect at the environmental biology laboratories at Duluth, Gulf
Breeze, Narragansett, and the Environmental Monitoring and
Support Laboratory, Las Vegas. There are also health-related
field laboratories in Wenatchee.,—Washington and W* Kingston,
Rhode Is! and. ..!....
All of the laboratories have close relationships with
-neighboring universities; in some cases the laboratories are
1 ocated on university campuses (the main Ci ncirinati Laboratory,
the W. Kingston Laboratory, and the Human Inhalation Facility at
the University of North Carolina, Chapel Hill).
1. Adequacy of Facilities for Research
" ' ""™ i ~'
The facilities of the health effects laboratories are
generally excellent. The major exceptions are the RTP
laboratory and the W. Kingston facility, neither of which was
built for biomedical research purposes. Some laboratory
buildings, on the other hand, were constructed for biomedical
research- within the past five years (e.g., £l!icj_nnatj). in
spite of limitations of physic al p T a n t, s u c h" ~ as t h e If b s e n c e of
modern animal care facilities at Research Triangle Park, EPA
laboratory staff have improvised and created the physical
conditions necessary for good research. The laboratories are,
in general, notably well-equipped for physical and chemical
analysis and modern biologic research; they also appear to have
adequate library, data processing and statistical services on
the premises or conveniently accessible.
""- ^The Committee did riot conduct a formal audit"of good
laboratory practice at any laboratory visited. However, the
Committee did consider as part of their general review many of
the items that would be considered in'such an audit. It was the
178
-------
Committee's perception that additional attention is needed in
this area if EPA laboratories are to achieve the same standards
that EPA expects from research conducted outside the Agency and
submitted to the Agency.
Some of the specialized physical facilities are unique
in the cabability of their chambers to provide accurate
concentrations of gasses and aerosols at viery low
concentrations for human exposure. The inhalation facilities
at Cincinnati for experimental animal exposures and the
Inhalation Exposure Facility at Chapel Hill for controlled
human exposures are good examples, •
Housekeeping and safety programs wer;e generally quite
satisfactory. Animal facilities in only two laboratories were
examined (Cincinnati and RTP). The facilities at Cincinnati
have been approved by a national animal facility accreditation
committee, while no such accreditation hasj been attempted at RTP
due to its many deficiencies. Our Committee agrees with the
findings of the accreditation committee and suggests that EPA
devote the necessary resources to bring t^je RTP animal facility
into similar compliance. '
2. Staffing for Research •«-—.- -
The Committee recognizes the role of history in present
EPA staffing, not only the legacies of personnel from the
predecessor agencies and programs that were coalesced into EPA
in 1970 but also the effects of legislative actions, OMB
decisions, and civil service regulations. The Committee,
therefore, addressed only limited aspectS'Of the total problem,
including the effects of imbalance between funds available for
extramural research and professional staff available to monitor
the research, the availability of research staff to make
effective use of special facilities, and the utilization of
scientists from academic institutions to supplement EPA
research staff.
h
Over the past three years, there have been, several
increases in research appropriat ions,-without proportional
increases in personnel (Energy-Environment Act, TOSCA, CAA
amendments, etc.). One result is. an increase in the burden of
monitoring extramural grants ,,and contracts. We found great
variability from one research^ program to another in the
distribution and intensity of the .monitoring""! oa-d. There was
also much variability in attitudes toward1 an extramural "
program. Ideally an extramural project should complement and
enrich the intramural scientific endeavor. The individual
research worker= may or may not wish to exp.and his (her) own
research effort through an extramural grant or two.
179
-------
The Committee found that some EPA scientists were
attempting to monitor six or more extramural projects and had
no time for their own research. In one instance, every member
of a laboratory division was fully occupied monitoring grants
or contracts; there was no intramural ^research. This is an
unsatisfactory method for establishing and maintaining a
program of high quality; it is made even worse when
appropriations are increased without additional staff
increases, as frequently happens. •
EPA's special inhalation facilities were costly to build
and are expensive to maintain (over $1 million annually for one
facility). It is important that such facilities be competently
and fully staffed to be effectivly used. In fact, these
facilities are seriously underutilized!, due both to lack of
skilled personnel and to lack of funds for research projects.
At the same time multi-billion dol1ar decisions are being made
which would benefit greatly from the l<:ind of information these
laboratories could provide (for example, the standard setting
for ozone and NC^).
One practice which increases available manpower and
promotes intellectual quality is the exchange of staff between
universities, industry, and the Agency"(Ihteragency Personnel
Agreement-IPA). The exchange is largely from academic
institution to research laboratory, and we found universal
enthusiasm for this arrangement within the laboratories.
However, there seems, to -be 1 ittl e.-systiema.tip effort to recruit
IPAs; most of the arrangements develop out of personal
acquaintances. While these arrangements are mutually benefical
and should be encouraged, EPA has recently adopted a policy
which will make university recruitment much more difficult—an
academic institution must guarantee a position for a returning
IPA. This would severely limit opportunities for young
scientists in the early post doctorate period of their careers.
3. Accountability for Expenditures
The Committee did not discover any managerierl accounting
and auditing efforts within ORD -to--(a)-anal-yze the success or
failure of research projects after the,ir conclusion or (b)
apply account i ng methods... to individual projects to determine
dollar allocations to equipment, salaries, travel, and ,
."services. There is a remarkable and conspicuous lack of
.managerial aud-iting procedures in the ORD operation. After
initial formulation of the decision units and their overall
budgets, the laboratories are assigned the implementation of
projects. In general, it is at the laboratory level that work
-yjijt ..productivity .and cosis must be tracked on a continuing
basis and evaluated for effectiveness and adherence to or
departure from categorical costs of ORD operations. The
insensitivity to project evaluation after completion of effort
was reflected by attitudes of managers and bench scientists.
The unawareness of costs was also widespread.
180
-------
D. The Quality of Health Effects Research
i
The quality of research in EPA is important not only
because any worthwhile research should meet standards acceptable
to the scientific community but also for reasons derived from
the regulatory nature of the Agency. Presumably all research
supported by EPA should be related in the short or long term to
the development of a regulation or standard. In this context
scientific information is likely to be examined critically in an
adversary relationship. Any sloppiness in conduct or
interpretation of the work is likely to weaken or destroy EPA's
position. ;
Another characteristic of a regulatory agency is the
importance of the credibility of research supported by the
Agency. Just as research supported by industry is often
suspected of bias, whether justified or not, so research
supported by EPA is often alleged to be biased toward the
overzealous protection of public health., This question of
credibility is a difficult one and is never easily solved. For
EPA it implies a great need not only for|the highest standards
of quality in scientific work but also fpr active and constant
efforts o-f EPA scientists to participateMn and have the support
of the scientific community. "'
It was our experience in visiting "the health effects
research laboratories and Program Offices that EPA has many
scientists who would be welcome in the nation's universities and
private research institutions. Many, of .fehfi: scientists we talked
to" were clearly dedicated to the best traditions of public
service in carrying out the missions of EPA. The Committee
found areas of high morale and sense of accomplishment, but was
disturbed to find areas of low morale and frustration from
frequent changes of research direction or even the absence of a
sense of direction, often stemming from frequent changes in
leadership. i
In trying to assess quality, the Coimmittee used what it
could of the usual criteria for evaluation. The legal
counsel's interpretation of the Privacy Act did not permit the
Committee to request a curriculum vitae of any scientist, but
many offered them voluntarily.The following information was
usually obtained from each research unit: the number of staff
with research doctorates; the scientific publication record of
the unit, in peer reviewed journals and others; the statistical
and computational resources' of the unit; the procedures used for
peer review; and a sense of the intellectual climate of the
unit. !
-.r-:-.:.;.....T.he _Commi ttee al so .e.xami ned the .procedures used in conduct
of "extramural" research through grants and contracts.
Consultants were added as necessary to evaluate specific
programs and special facilities such as animal housing and care.
These and other aspects of quality assurance are described under
the headings that follow.
181
-------
1. Publication and Reporting of Research
Results '
!
Scientific investigators are part of a tradition
which places great importance on scientific peer review of
results prepared for publication in professional journals. As
with other characteristics, there was high variability of
attitudes and procedures among the different laboratories and
divisions of laboratories. Some resembled university
laboratories in their emphasis on scientific peer review of
research plans and peer review of manuscripts before submission
to high quality journals. In these cases publication was seen
as an incentive for promotion and professional advancement.
Publication in peer reviewed journals enhances the probability
that a product of research will "stand up in court." These
research units usually had strong interactions with local
universities and promoted attendance at scientific meetings,
development of symposia and workshops,, and participation by
IPAs.
At the other extreme were units that appeared to put no
emphasis on publication in the scientific literature and who
sensed that there was no incentive in EPA for such publication.
Others recognized the desirability ofi such publication but felt
so overwhelmed by other responsibilities that they; could not
find time to publish. Some felt that internal reports were all
that the Agency expected.
The policy on review of manuscripts varied from in-house
review only to submission of the document to up to five
external reviewers. Some scientists hot only met the formal
requirements but also sent their manuscripts to one or two
personal acquaintances whose opi ni ons'. they particularly valued.
To ensure acceptability of research results, the studies
must be reviewed by one's scientific peers and published in a
reputable journal. Failure to so treat results of research
investigations involves the risk that review will occur at a
later date, with possible refutation of results" and""
embarrassment to the Agency. "Specif ic mechani sms= must be
established to require peer review of.. ORD results and to
encourage prompt publication in peer reviewed journals.
Attendance at professional scientific meetings to present
research results is not:consistentTy encouraged. "
It has been argued by some laboratory staff that peer
review and publication are not necessary for mission-oriented
i-researchj-the EPA focus.- The Committee rejects this viewpoint;
applied research, often with short-term goals, must be reviewed
and published as surely as that related to more fundamental
investigations. Applied research is the •••f-i rial-product of years
of basic research and should receive even greater review.
752
-------
2. Quality Assurance in Grants and Contracts
Examinations of this important component of the
health effects research program revealed .serious problems,
which affect in-house performance as well; as the quality and
relevance of extramural research. One aspect is wide
variability in funding from year to year iand the assignment of
funds without any addition of personnel (ithis happens with the
Energy-Environment "pass-through" appropriation, for example).
Another serious problem is the uneven distribution of
monitoring responsibility among scientists in a laboratory
unit; some are overloaded to the extent they cannot possibly do
a satisfactory job. j
Both the old and new planning systems give authority to
laboratory directors to obtain extramural: services through
award of contract or grant funds. Laboratory directors rely
upon their managers to allocate resources under their juris-
diction to complete work unit tasks. Thus there is local or
section management of contractors performing services for ORD.
In depth examination of several of the laboratory, sub-unit
extramural program procedures for contractor selection,
monitoring and evaluation revealed good examples of contractor
or grantee selection , based on submissions and competitive
selection. There were also examples of selection of weak or
incompetent applicants, failure of 1aboratory" staff, to monitor
performance, and almost a total absence of evaluation of the
final submission and its relevance to the ORD program and EPA
in general. - • ~ - .,-;.;•.--;•-- - •'••
Some scientists see grants and contracts as a desirable
extention of the scope of their personal efforts and en-
hancement of their contacts with the scientific community.
Indeed, a healthy balance between intramural and extramural
work can benefit both EPA and the universities. These kinds of
relationships do not currently appear to; be the norm.
- Three kinds of arrangements are used for support of the
extramural research program: contracts, grants, and
cooperative agreements. Increasingly^ contracts~have~ also been
used to provide operations and mafntenanc-e- services d-vrect-ly
supporting in-house efforts. The Committee did not
systematically examine the quality of contract . research and did
not look at all of the cooperative agreements, a recent
development which has been Tittle used so far.
EPA has more specific ' requirements'for the award of
contracts than for grants. The Committee was told repeatedly
that grants are being used increasingly,! because processing
~ them-is-easier- anxl takes -less time (three or four months •,-,"-- —
in.stead of six months to a year for a contract).
--Examination of selected files i indicated^ that -the review
procedures for grants were being abused :in at least one
183
-------
laboratory. There were examples of critical reviewers recom-
mending that the work not be funded of stating that the
proposed project was only marginally acceptable. Yet the
project officer proceeded to rationalize the reviewer's comments
and indicated alterations in the study protocol of the grant
applicants which would overcome the objections of the reviewers.
Because the proposed project review and the project officer's
revisions were performed near the end; of the Federal fiscal
year, the funds were awarded without either further submissions
or a modified submission by the applicant. In one example,
inquiry revealed that one year later the project monitor still
did not know if the grantee had modified the protocol, added
additional personnel, etc., as was recommended by reviewers and
as was rationalized by the project officer in justification of
awarding the grant. ;
In other examples the Committee found that external reviews
were not obtained before award of grants. {Some EPA staff
informed the Committee that soliciting external reviews of
contract proposals was illegal, except with permission of the
appl icant s . ) :
i
Scientists were encountered who had difficulty keeping
track of" the number of awards they were assigned to monitor;
they -were not familiar with the details of extramural contract
or grant work as it progressed.- The quality of investigatory
work external to EPA laboratories and: supported by ORD funds was
highly variable and of great concern, mainly, because ORD
oversight was usually lacking. It requires project monitoring
effort to ensure that contractors or grantees perform responsive
work on a timely basis. There is an efficient "mix" of one's
own research and that of others that can be effectively
monitored. Conversations with ORD laboratory staff suggested
that monitoring one or two contracts 6r grants totalling perhaps
$100-150,000 per year would be a stimulus to a senior ORD
scientist. More extensive monitoring responsibility is a burden
to the ORD scientist and, even more important, he/she cannot
efficiently discharge the monitoring responsibilities. Some
research units-are so heavily committed to monitoring grants and
contracts that no scientist i n "the unfit has -any t-i-me for his/her
own research. TFfe lesson is a clear one; Congress should not
increase R&D funding without concomitant ly increasing ORD
staffing or without identifying alternative approaches.
- A frequent complaint was that monitoring was handicapped by
the absence of travel funds for the project officer to visit the
institution where the research was being done.
..app.l i cat-i on,s, are of two •- types—sol i cited -and - ------
unsolicited. The latter presumably represents the spontaneous
interest of university scientists to do research on
-.environmental problems in which EPA mi gh.t.. be. .interested. .The
184
-------
common response to the Committee's inquiry was that unsolicited
grant proposals have almost no chance of .being funded, primarily
because they are judged "not relevant." Ijt seems clear that EPA
scientists are using grants in lieu of contracts, that they
monitor them like contracts, and that the're is little
opportunity for "investigator initiative.;"
i
The mechanisms for soliciting grant ^proposals vary from one
unit to another. We found little evidence that EPA has found
effective ways to interest university scientists in its problems
on a sustained basis. j
Another practice, employed to extend the time for longer-
term research but with the potential for [abuse, is the "front-
end loading" of a newly awarded grant, in this practice the
amount of the award may be as much as twice the amount of the
first year's budget. The investigator can then request an
extension for a second year without additional funds, an action
routinely granted without a critical review of research
progress. The Committee does recognize the need for assured
funding of projects that may require more than one year to
complete. However, if funds required for more than the first
year's operation must be obligated, the project must be
carefully monitored to assure that funds1for the second year are
required and appropriately usedi . / . , ,:-,-. -... —7••_:•• - :,- _....r~-
Another shortcoming of the present EPA system is the
absence of a routine operational, audit .of.-the qualjty of
extramural research. Individual scientists and laboratory
directors told us that a contractor or grantee who performed
poorly was not likely to obtain another jgrant or contract. This
informal and spottily used system is not adequate to assure the
high quality of extramural performance. \
ORD's entire program to make extramural awards of funds
under contracts, grants or cooperative agreements requires a
thorough overhauling. Extensive standard operating procedures
for awarding grants and contracts exist In the Agency; they are
voluminous, difficult to comprehend, and are avoided by-
laboratory staff. —It is necessary to establ i sh si mpl e,.-'e-xpl i ci t
procedures to be followed by laboratory directors and scientists
throughout the life of an extramural award. At present,
laboratory directors are expected to satisfactorily complete
work unit tasks; extramural projects are their choice and
responsibility. The Committee recognizes the need for ..... .-..
extramural assistance, particularly if the trend continues to
increase ORD dollars without increasing the number of positions
for investigators, but the procedures for extramural programs
mcrst be p.1aced on a more-defensible.--basis throughout 0-RD.
785
-------
3. Career Opportunities :
The civil service system was examined as an influence on
the quality of research programs and on career opportunities for
EPA scientists. There were several examples of negative effects
of the civil service system; for example, it does not permit the
flexibility to hire new people or to move people as program
orientation shifts. Consequently, there are cases in which
excellent scientists are placed on projects where their
expertise is not needed and where they have to be "re-tooled".
Although the Committee talked to people who had been
promoted because of the quality of their research, more
frequently promotion related to the assumption or increase of
administrative responsibility. Many times a good scientist
makes a poor administrator, but the scientist takes the
administrative position for the highef salary, not because he or
she has management skills. Talented researchers must be
encouraged to continue as investigators. Mechanisms must be
instituted to further their professional development and their
allegiance to the Agency. •
i
It appears that the policies ancl procedures for
advancement do not encourage the emergence of e.ither top,,.
scientific or managerial-performance.. The system does encourage
job-hopping by bright people, particularly those in Program
Offices. A promotion ladder based on;scientific achievement
rather than admrni strat i ve res'pans i bi 1 rty wbuld help to solve
this problem. Many industrial research laboratories use dual
ladders for advancement-- administrative and research. Senior
research personnel are rewarded with remuneration and privileges
comparable to those of a senior manager. ORD is experiencing
difficulty in retaining research physicians, epidemiologists,
and toxicologists, among others. At the time of this writing,
the Human Inhalation Laboratory in Chapel Hill, N.C., a unique
facility, is virtually without physicians to perform the
research vital to scheduled regulations in the air media.
_." . Admi ni strati ve mechanis"ms~s~h'ou1 d "be deyel oped to offer a
challenging career ladder to these pirpfes'sTohats"Tf"first rate
health effects research is to be performed in ORD. The
Committee recognizes that many of the reforms addressed
elsewhere in this report will improve conditions for these
professionals, but an explicit analysis of conditions and
^incentives related to a research career in ORD must be performed
"and improvements implemented where necessary.
4. Other Components of Quality Assurance
Performance evaluations of individuals and laboratories
are often perfunctory. Many individual scientists were unclear
about the criteria applied to their evaluations and advancement.
186
-------
Evaluation of laboratories is not being done in terms of good
laboratory practices, rewards and incentives, budget and
resource allocations, and accountability!
Personal scientific integrity is difficult or impossible to
determine in a study of this kind. To the extent that personal
conversations, attitudes expressed, and .measures taken to assure
the quality of research, design, and analysis can be used to
assess scientific integrity, the Committee was favorably
impressed. If there were subtle biases in the interpretation of
research results, they were not detected:in this study.
There are periodic "program reviews" in which head-
quarters' staff members visit the laboratories. These are
described by the laboratory scientists as superficial "show and
tell" sessions. There is limited scientific feedback from
headquarters' staff, and the only benefit to the laboratory is
the stimulus to prepare material for presentation.
By contrast, it was noted that when NIH is involved in a
jointly sponsored project, there is a vijsit by NIH staff
members, who conduct an intensive critical analysis of the
proposed research- project. EPA- staff who have thus been "nailed
to the wall" to defend their projects say they would welcome
this kind of evaluation of EPA projects.
There appears to be a general lack of understanding of the
Science Advisory Board and its constituent committees by
laboratory staff. In view of this, it was not surprising that
the Science Advisory Board was criticized for its lack of
scientific interaction, failures in communica-tion, and lack of
subsequent feedback.
i
5. Interagency Agreements :
The Interagency Regulatory Liaison Group (IRLG) is a
hew activity which seems to "be off to a'promising start. Si nee
it is a developing program, no attempt was made to evaluate it.
Other programs involving interagency agreements have had
mixed success, at. best. EPA has substantially supported the
National Center for Toxicologic Research since its inception,
with little evidence of any product benefiting EPA. Disap-
pointment was also expressed about interagency agreements with
Los Alamos and Oak Ridge National Laboratories and three of the
National Institutes of Health. >
A significant portion of EPA's health effects research is
supported by interagency .agreement for the special Energy-
Environment appropriation. No attempt, was made to examine this
program .in detail. '
187
-------
E. Other Relevant Topics i
1. Long Range, or Core, Program Research
There are subjects for research which are important
to several of the media programs. Examples are the properties
of particle dispersions, be they in air or water, because of
their relevance to collection of the disperse phase prior to
effluent discharge, to particle deposition in the human
respiratory tract and to particle retention or solubi1ization
in the human gastrointestinal tract; iepidemiol ogical
methodology because it is a major tool for relating exposures
to pollutants to potential effects in the exposed population;
and techniques of risk assessment and: presentation of the
implications prior to judging acceptability of risk. There
should be a long term ORD investment in researchers and
facilities to develop highly active and productive groups in
those areas of research which are central to large segments of
Agency regulatory activity. This investment is currently being
augmented by initiation of extramural university centers. It is
planned to shuttle ORD staff between :their resident
laboratories and the centers for "leaves of absence" during
which they can pursue studies in core- areas while upgrading
thei r capabi 1 i t i es on a university campus. We applaud this
plan, but also see the need for" small;, active core" research
groups in ORD la,bora.tories. Allocation" of a specific
percentage, at least 10%, of the ORD budget for relevant
research in core subject areas, but not on projects
specifically traceable ..t.o....immediate., program needs (6 months-2
years), is a reasonable assignment of funds.'""There is rib
obstacle to this programming of funds under the present
procedures for funds authorization. They are part of the funds
assigned to research for the specific statutes, because results
will be applicable to those statutes, as well as to others.
2. ORD/Congressional Stafif Information Transfer
The relationship and relevance of ORD projects to
regulatory needs is not always obviou's, particularly to non-
scientists. It is-essential— that members of- Congress and their
staffs understand the efforts of ORD.Sueh understanding does
not develop accidentally. ORD should develop a plan to
regularly inform interested members of Congress and their . .
staffs of the results of ORD efforts ;and the manner in which
they further the goals of statutes administered by the Agency.
ORD1 s ,i nvestment in what is .essenfrdalTy an educational program
for legislators should invo-lve Q-RD's most senior scientific
staff. It is critical * that this communication effort include
laboratory personnel who are directly involved in the conduct
of,research. .We note the 1978 and 1979 Research Outlook efforts
by ORD, but believe efforts must go far beyond this and must
incorporate personal communications, as well as transfer of
printed information. The conLcep.ts.: x>f .chroni c disease, multiple
etiologies of disease, host factors, and cumulative effects, to
name only a few, are complex and crucial to understanding the
underlying approaches to research in ORD.
188 i
-------
VI. UTILIZATION OF ORD RESULTS
Different Program Offices utilize ORD research results to
different extents. Senior program managers indicated that they
did not look to ORD for results; rather, ;they sought capable
laboratories and investigators related to their needs, be they
within or outside the Agency. A Radiation Program manager
indicated that ORD has little capability to assist them; ORD
has no capabilities in the area of biological effects of noise.
ORD appears to have little involvement with the Toxic Substances
Office. The Water Program draws heavily on ORD at the present
time, and recently ORD had a major involvement in the
formulation of criteria documents for 65 water pollutants.
The input of research to the screening test and risk
assessment process was clearly evident from the Drinking Water
Research Program in Cincinnati and the Pesticide Programs at the
Gulf Breeze and Wenatchee Laboratories. Their scientific
standing is recognized. The respective leadership has
maintained the kinds of communication necessary (with the help
of pilot research committees) to keep this personnel in
Washington knowledgeable and involved.
It is not surprising to find that the utilization of
results from ORD projects is not..carefully tracked when tlie
joint planning of research by ProgranrOffices and ORD is in its
infancy with the pilot research committee program. Program
managers elaborated on=,many..needs- not ke'tng met by ORD.; there
were few illustrations of ORD responsiveiness to programs and-
subsequent incorporation of results into regulatory programs. On
the other hand, ORD staff were often praised for their responses
to requests for preliminary review of regulatory documents,
consultation on imminent regulatory submissions to the courts
and, in general, what can be characterized as technical support
to the Program Offices. The Committee was not able to estimate
the average percentage of ORD professional staff time devoted to
technical support; it varied with individual research sections.
It was clear that in some instances it represented a significant
portion of some individuals' time-.—This-technical -support has -
on some occasions played a critical role in. the.Agency.1 s . ... ..
formulation and defense of regulations.
The ORD function in the Agency is defensible mainly on the
basis of program utilization of insights and results developed
intramurally or extramural ly under its auspices a.nd guidance. :
The Committee found that ORD did not fully recognize or accept -
this criterion for judging its efficacy, had not developed
mechanisms for efficient utilization of research results by
Program Offices, and did-not maintain records of results .which
had been incorporated into regulations.;
189
-------
VII. STATUS OF IMPLEMENTATION OF TWO SETS OF NATIONAL ACADEMY
OF SCIENCES (NAS) RECOMMENDATIONS TO EPA
The analytical study of Research and Development in the
Environmental Protection Agency conducted by the Environmental
Research Assessment Committee (Joh-n M. Neuhold, Chairman), of
the National Academy of Sciences, National Research Council,in
1974 and 1975 set forth a number of useful recommendations.*
Before that, a Review Committee on the Management of EPA's
Research and Development Activities (Robert W. Berliner,
Chairman) had developed recommendations submitted to the Agency
on August 27, 1974. Our Committee (HERRG), therefore, in its
collective judgment, has attempted to evaluate the extent to
whic'h former recommendations have or have not been implemented.
This final exercise was undertaken at the end of our study when
all visits had been completed. It was possible by this means to
add a different, but closely related, viewpoint against which to
compare our own observations of performance and changes during
the past four years.
Although there has been significant improvement in
selected aspects of EPA research planning and management, most
notably the development of pilot research committees with
representatives from across the Agency, the overall planning and
management system is still unsatisfactory. Many of the reasons
for inadequacies in the system in 1974 still exist today and
will be enumerated fn thevrf o-l Towi n g. - -^ 1 --•-. .-;1_;~' 1.1 .:.•_; 1.1;*.
A. Recommendations from the Environmental Research
Assessment Committee of 1975 **
(1) "EPA's research and development should
concentrate primarily on support of the Agency's
decision making and anticipation of future
problems."
There are improvements arising from better communications
between reseefrch workers in the'laboratories an"d~'the'"'Pr"ogram"
Offices. The pilot research committees have~helped establish
communications and understanding.
~"*Analytica1 Studies for the U.S. Environmental Protection
Agency, Volume IHj "Research and Development in the ~~
Ertvironmental Protection Agency," Environmental Research
Assessment Committee, Commission on Natural Resources, The
National Research Council, National Academy of Sciences,
Washington, D.C. 1977. • ..-..-.-.
**Ibid. page 2. ;
19O
-------
(2) "EPA should supplement its primary research
responsibilities with some fundamental research
to help advance understanding !i n environmental
sciences and technology."
Planning for fundamental or longer ,term research is still
inadequate. However, to achieve the rigJit kind of balance there
first needs to be a close and direct relationship between
researchers and program managers. Both 'must understand the
research process and information needs pf the regulatory
process. :
(3) "A new legislative mandate will be required if
EPA is to conduct effective anticipatory and
fundamental research." i
The HERRG Committee does not agree jthat additional
legislation is needed to fund and conduct "anticipatory and
fundamental research." I
(4) "We recommend that the Office of Science and
Technology Policy (OSTP) develop a federal
environmental research, development, and
demonstration strategy that -includes designation
of the appropriate roles of all participating
federal agencies and existing interagency
~ coord.i nat~i ng committees ,. and delineation of_the
relationships between federal ; and nonfederal -•.»•••-••
research and development. The OSTP should
coordinate the implementation;of the strategy
through its mandated consultations with the
Office of Management and Budget (OMB) about the
scientific programs of federal agencies."
This recommendation has not been followed, per se.
However, the Interagency Regulatory Lia;ison Group is seen as an
excellent initiative which has the potential of reducing
dupl icat ion and conf usi on among a gene i es .—• Better-e-oo-r-di-nati-on
of research efforts and better agre-ement -on the-met-hodo.log-i.es.
applicable to hazard assessment are encouraged by this
Committee. . . • i
i
• (5) "We recommend that the management of all research
--- and development in E:PA be central fzed in the
-••"-•• Office of Research and Development (OR&D)."
There seems to be progress in centralizing the management
of-R&D within ORDi but a number of Program Offices administer
R&D contracts and grants directly. The Committee urges th'at
this Academy recommendation be implemented to assure that proper
oversight and scientific peer review be applied whenever
research is conducted by the Agency. •
191
-------
(6) "EPA's research program needjs to be better
organized for balance and continuity, through
planning developed around a logical conceptual
framework of environmental protection..."
A number of areas within the present EPA research and
development program are still not aligned within a logical
conceptual framework of .environmental protection and thus are
not as effective as they could be. Th;e conceptual framework
proposed in the earlier NAS/NRC report (1977) still appears to
offer a sound framework for the assessment of research needs,
the planning and conduct of research, iand the utilization of
research results. The framework is sh'own below:
Framework for Environmental Protection
Natural,
Economic, and
Human Resources
Production Processes
(including control technology),
Transportation, and Use
Emissions
(residuals)
Other Environmental •
Impacts Including :
Unintentional Release
of Pollutants
Transport,
Transformation,
Accumulation
Ambient" Environmental
Quality (air, water,
soil, biota)
Decisions to Control
Effects
752
-------
(7) "A central function of scientific support to
decision making should be to provide integrated
assessments of available scientific, technical,
and economic data pertinent to pending decisions
in forms suitable for use by Agency decision
makers. We recommend that the importance of this
function be recognized by giving it formal status
and organization in OR&D." ',
The importance of integrated assessments continues to be
recognized, and the Agency is moving toward establishing the
formal organization required to make such assessments. When
such an organization is fully operational, it should be of major
assistance in providing information that; is useful to the
regulatory decision makers; but of equal; importance is inform-
ation that is crucial for the planning of a responsive research
program. Carefully conducted assessments can identify gaps in
research information or parameters that tiave the greatest
influence on the effects of emissions. In the absence of such
assessments there is a risk that research efforts may be
directed to developing information that may have limited value
in establishing or reassessing standards, or in guiding their
enforcement. •-'•,•
(8) "The research planning system now in use 'in OR&D,
characterized as "top-down" in structure, should
be retained f or .resear.ch i n .s.upp..qrtL, of decision
making. For anticipatory and fuTi"d"amenta4:—- :
research, however, we recommend a "bottom-up"
scheme that relies on the scientific community to
identify research needs." '
Except for the pilot research committees, the planning
process remains "top-down." Substantial improvements are needed
to achieve involvement of those generating and using the data.
(9) "We recommend that block funding of extramural
-grants, contracts, and-i nteragency agreements^ be ,
considered as a mechanism_tQ;.._e_stabl i sh centers of
excellence, federally funded contract research
and development laboratories, and umbrella
interagency agreements to supplement the intra-
— --mural research and development program.,"
To date, block funding mechanisms have.not".been extensively
used by ORD, although legislation has provided the opportunity
for use of cooperative agreements that may very well match ORD
needs-* -ORD. has made preliminary plans for using such agreements
and should proceed expeditiously to implement their use. Such
agreements offer an opportunity for a complementary approach to
the present system of grants and cpjrtracts for_ extramural
performance. ;
193
-------
i
(10) "All proposals and completed research should be
subjected to review on their technical merits by
scientific and technical peers."
Peer review of proposals and completed research was
inconsistent and, in many cases, inadequate.
(11) "We recommend the use of a parallel grade
advancement system, based .on performance of
research, that does not require researchers to
assume administrative or managerial tasks to
attain promotions." ;
There was little evidence of implementation of a parallel
grade system. In some cases, individuals have accepted
administrative or managerial assignments based on the perception
that such assignments are critical to obtaining promotions.
B. Recommendations of the Review Committee on Management
of EPA's Research and Development Activities *
The Review Committee report noted that the present (1974)
"Office of Research and Development planning and management
system fails to meet the. needs of the. Agency" and proceeded to
identify two main categories .of.failure: (1) the nature of the
system itself and (2) external constraints as perceived by the
Office of Research and ^Development an;d communicated to the
Review Committee^! /::7 " ";ivv\ \:":"rT;:IL?,i :.,-.„.. - '
1. The nature of the system itself.
a. "Planning is separated from
responsibility for executioin, leading
to severe resentment among performing
researchers. The assignment of
responsibility for specific actions and
decisions is difficult."
There fs""stili an inadequate linkage between planning and
responsibility for execution that is apparent, in"varying
degrees, at all levels of the organization below the Assistant
Administrator for Research and Development. An individual
researcher charged with.responsibility for performing a task may
:have no input to theL-planni ng of that; task.
b. "Priorities do not reflect the needs of
regulatory offices and regional offices
because of the 'vacuum cleaner1
' " "approach to soliciting ideas, and the -
system-induced barriers to using common
sense in the selection process."
bia.page 96.
194
-------
There has been improvement in the establishment of
priorities in selected areas, most notably those for which pilot
research committees have been established,, to yield a research
program potentially more responsive to the needs of the Program
Offices. In other areas, the research program is less clearly
defined and priorities have not been established. Faced with
necessarily limited resources, the responsible individuals have
frequently elected to continue work in all areas at a reduced
level of effort rather than electing to eliminate or defer the
lowest priority projects. The result is a reduced potential for
success in the highest priority areas because of lack of funds.
c. "Inadequate attention has been 'paid to
the possibility for trade-offs,; or
modifications in budgeted costs!, among
various projects. This has aided in
the development of a situation -where
there is only a series of discrete
projects and no Agency program. This
situation is further aggravated by the
absence of long-term (3-5 year);
pianni ng."
Long-range planning within the Agency remains, inadequate.
The large portion of the planning within |ORD-is necessarily
dependent upon the needs identified by the Program Offices.
These long-term needs have oft.en . been -/i nadequatelystated, if
at all, thereby handicapping the development of a responsive
long-term research plan. It was originally anticipated that
the pilot research committees would develop a strategic plan
for their areas of responsibility. However, this was not done,
in part because of the timing and pressure of the ZBB process
which forced the pilot research committees to take a shorter-
term outlook. An additional factor which should also be
recognized is the reluctance of some individuals to engage in
defining a strategic plan until they are certain that managers
are serious about the effort. ,
d. " The complexity of the system mak.es ,it -
counter-productive. The «large amount
of paperwork and excessive bureaucratic
review is a wasteful consumption of
time and energy. The needs of the
Agency are complex; however," this does -.._•----
not change — but rather heightens—the
need for a simple and understandable
planning and reporting system clearly
directed by the Assistant Administrator
and in which field personnel have a "
real participatory input.-" '
195
-------
The planning and management system is still extremely
complex, involves a large amount of paperwork, and is often a
waste of valuable time.and energy. An inadequate amount of
authority has been delegated downward;to the laboratory
directors and lower echelons of the Agency. In those cases
where authority has been delegated, there appear t.o be excessive
requirements for keeping all upper levels of the Agency
informed. One example is the use of the highly structured
quarterly "Project Status Reports," which include detail at the
task level (tasks ranging in expenditures of less than $10,000
to over $200,000 per year); the volume of material developed at
the laboratory scientist's level is passed successively to the
Division Director, the Accomplishment'P1 an Manager, and the
Office of Health and Ecological Effects and its various staff
units.
e. "Accountability is made impossible by
the parallel but separate management
systems — some for housekeeping and the
others for program content — and by the
hopelessly complex Program Area
Manager-Program Element Director-
Program Assessment Group-Strategies
system which obfuscates management
responsibility." ... : . ; ; • - - -.--
The chain of accountability is extremely difficult to trace
from the laboratory, scient/ist^ (ei±herJ.i n:-.hquse or engaged as a
contractor or grantee) to the-'Assistant-•-Ad-mi nist rat or "f or
Research and Development. The "chain 'of command" is excessive
with numerous intermediate steps that serve only to delay or, in
some cases, reprocess information without serving any clear
management functions to enhance research productivity,
efficiency, or responsiveness. Indeed, in many areas the number
of information reprocessors and/or relayers makes it difficult
to identify the laboratory scientist.
f. "Excessive requirements for detail at
all planning levels lead to ,an ---—
oversized headquarters staff and to the
stifling of innovation in the
laboratory." ,
The 1 eve! of .de.tai 1. required at all levels and the transfer
of materials with limited,informational -or management value
continues t.o, contribute to the maintenance of an overly large
Washington staff. In what appears to be a contradiction, the
Washington staff is understaffed in relation to the amount of
material beijig transferjie.d an.d processed. Unfortunately, much
of this effort is misdirected. Because of the attempts to
maintain detailed accountability of even extremely small
projects, the innovative responsibilities of ..the laboratory
scientists continue to be unfulfilled.
196
-------
g. "The existing management structure does
not allow for the corrective feedback
and flexibility which are essential to
any successful research and development
program." j
Because the "chain of command" is so long and the
communication pathways are jammed with trivia, corrective feed-
back does not occur at the level required for effective manage-
ment. The rigid system of accountability to the laboratory
directors diminishes the flexibility needed for operation of a
responsive and innovative research program.
h. "A long-term program designed,to meet
stated goals is missing and this is
vital for any scientific venture."
The ORD program has few clearly stated long-term
strategies, specific to each Program Office, with easily
identifiable objectives and goals. In the absence of long-term
objectives and goals, the Agency's research and development
resources seem excessively preoccupied with meeting short-term
goals, some of which are restatements of goals not previously
attai ned. ......
i. "A false sense of control is generated
by the highly structured mechanism for
planning.," - ; -_.; :-...._..,..,...... _:.._.._... ,.-,_. _;.. : ...
The highly structured planning and control system, which
generates considerable activity, has promoted the feeling
that something is happening that is of a positive nature. The
widespread lack of clearly stated and agreed upon long-term
objectives and goals, however, makes it difficult to determine
whether the movement is positive, negative, or random in nature.
j. "Relationships between the headquarters
and field are strained at best; a state
of frustration in the field staff is
apparent." -•••-
Considerable frustration is apparent in many of
the organizational units below the Assistant Administrator's
office. In many cases, the individuals have resigned themselves
to tolerating a work environment that is-constantly changing,
but rarely for^the better.
2. External constraints as perceived by the Office of
Research and Development. ; _..- ..
a. "Enabling legislation is noncoherent
and mandates a .set of unbalanced and
uncoordinated research objectives and
timetables."
197
-------
The enabling legislation for the Agency has been
and continues to be viewed as noncoherent, mandating a set of
unbalanced and uncoordinated research objectives and timetables.
Since the enabling legislation has not and may not be changed in
the near future, ORD has no real choice but to accept the
situation that exists and strive to adjust its planning and
operations accordingly,
b. "The lack of an integrated approach to
environmental pollution control in the
Agency as a whole makes an integrated
research and development program very
difficult to form."
Although some individuals view the Agency as not
having an integrated approach to environmental pollution
control, some progress has been made, and the use of approaches
such as the pilot research committees offers the opportunity for
developing an integrated research program with long-range
objectives and goals as recommended ini!974.
c. "Civil Service rules, parochial
political pressures, and human nature
combine as barriers to the ,. ; -
simplification, assembly into 'critical
masses,' and logical organization of
the research units which were inherited -.
by EPA when it was' created. "
i
Civil service rules, parochial political pressures, and
human nature continue to be barriers to simplification, assembly
into "critical masses," and logical organization of the research
units. Of perhaps equal importance has been the failure to
recognize that in the absence of a clearly recognizable research
and development strategies specific for the Program Offices, the
constraints of civil service rules, the influence of political
pressures, and human nature will have substantial adverse
impacts on the , re,s,earch program. An identifiable strategy with
well thought out-objectives and goals will..go a long way toward
minimizing the impact of factors that can push a reaction-
oriented program, with ill-defined objectives and goals, off
course. As addressed elsewhere in this report, civil service
rules do adversely impact the research program, and suggestions
for change are offered. However, In t-he absence of changes in
the ruleSjjthe situation must be accepted and plans developed
within the constraints of the rules. Parochial political
pressures have been, and probably will continue to be, brought
to bear. However, it should*be recognized that the Agency has
strong poTit'-ical supporters, who can counter parochialism if
they know that the Agency has^a "research program that is
scientifically and managerially sound and programatically
responsible with a plan for the future. Without question human
nature may at times offer constraints, ;but, if properly
directed, can also provide forward momentum.
198
-------
d. "A level budget (except for th:e energy
'roller coaster1 of FY 74,75,7J6)
prevents transitions which would be
possible in a steady growth situation.
An internal 'roller coaster' budget
appears to be particularly disruptive
to individual projects."
The level budgets of fiscal years 1:974, 1975, and 1976 were
given as the reason for the failure of the ORD planning and
management progression. The level budget was said to prevent
transitions that would be possible in a steady growth situation.
Recent budgets have shown an increase; however, transitions do
not appear to have occurred any more smoothly. A concern raised
even more frequently than the shortage of funds is the
.restriction on the number of full-time employees. Although the
impacts of the restriction are real, little has been or is
likely to be accompl is.hed by merely accepting the OMB mandated
personnel ceilings until they can be changed. Until changes are
made, it would seem prudent to exercise greater care in the use
of available personnel and to have a strategic plan for addition
of personnel when vacancies do occur. Such a strategy for the
management of personnel resources is an essential part of the
total Agency research and development plan and is the only way
the personnel resources (as to number of individuals with
specific types and levels of disciplinary training) can be
matched to the long-term needs . of the Agency-. .... .........
The 1974 letter report of the NAS/NRC Review Committee listed
four major recommendations.* The recommendations have been
implemented to varying degrees and, even where not fully
implemented, still, seem appropriate. Because they are still
germane, each is reviewed below. :
1. "The Environmental Research Objective Statement-
Research Objective Achievement PIari-Program Area
Manager-Program Element Direction-Program
Assessment Group-system should be abolished.
Responsibility for carrying "out a program.
designed to meet the goals of the Office of
Research and Development should be delegated
directly to, the National Environmental Research
Center directors. Resources of manpower and
money should be allocated-directly to each-
Na-ttonal Environmental Research Center."
*Ibid. • *e 98,
199
-------
The planning and management system referred to has
been largely abolished. It has not been replaced by a system
that is understandable to all parties involved; thus vestiges of
the old system remain. The five Pilot Research Committees cover
a portion of the ORD program and partially meet the planning
function requirement. The National Environmental Research
Centers and related field stations in existence in 1974 have
since been separated into 15 individual! laboratories, which
report through four deputy assistant administrators to the
Assistant Administrator for Research and Development. Although
allocations of resources are made directly to the individual
laboratories, there appear to be numerous strings attached which
severely restrict the authority of the;laboratory directors.
2. "The line reporting within the,Office of Research
and Development should be from the National
Environmental Research Center directors to the
Assistant Administrator. The Assistant
Administrator should have a small staff to
perform only staff functions and not to serve as
a filter or layer through which the National
Environmental Research Center directors report.
This should develop into a simple pyramidal
management system through which aTT~rfirection,
supervision, and evaluation is accomplished.
This would, in effect, eliminate all layers or
parallel management plans and result in a clear
chain of .authority from the individual
researchers to the Assistant Administrator for
Research and Development. The pyramid should
decentralize quickly from Washington Headquarters
to major field units. The Headquarter*s staff
should be trimmed appropriately and those
necessary for "Washington liaison" activities
clearly labeled. We did not h.ave sufficient time
to evaluate the role and position of the
Washington Environmental Research Center. Such an
evaluation should be made.
"Because of the recent formation of--the
Agency by coalescence of disparate portions
of other agencies, a particular need for
intra-agency communication exists. To this
ends a planned continuing rotation of field
pers'onnel into .andjpack fronr a small
jHeaciquarters staff unit and between other
units should be carried out. Short term,
non-government talent should also be worked
into this rotation system."
200
-------
--The Washington staff of ORD is still quite large with
a relatively large number of individuals serving in special
staff roles and on numerous ad hoc committees. Clear chains of
authority do not exist between individual researchers and the
Assistant Administrator for Research and Development; rather
there are numerous filters through which^information exchange
must take place. Despite the largeness of the Washington staff,
many appear overwhelmed by their work load, while others
apparently fill slots for which there are no longer meaningful
work assignments. Approximately 90% of the work load seems to
be carried by one-half of the staff.
Communication between Program Offices and the Office
of Research and Development has been virtually non-existent in
some areas. The five recently organized pilot research
committees appear to have helped improve; intra-agency
communication and offer considerable promise for further
improvement.
Rotation of field personnel into and back from head-
quarters has occurred to a limited extent, but more exchanges
are needed. A limited number of short-term, nongovernment
individuals have rotated through the system, however more
exchanges of this type are also needed. ,
3. "The function of the Assistant Administrator
for Research and Development should be to
assemble, analyze, and clearlyr defi/ne Agency
research and development needs and objectives
with the participation of the other Assistant
Administrators and the National Environmental
Research Center directors as the mechanism to
develop goals, programs, and priorities. He
should allocate objectives and the resources for
their accomplishment to the National
Environmental Research Centers. Once allocation
is decided upon, the performer of the research
or devel opment..shoul d be linked directly to the
user- of the projected output f or. i nf ormati on
exchange.
"A performance evaluation should be set up to
" include continued inputs from users, and outside
-visiting committees reporting at a high level
-.should be regularly employed. The system of
visiting committees employed by the National
Bureau of Standards should be studied for
jippl icabil ity.
"A plan for a 3-5 year period to be revised at
least annually should be deviel oped."
20?
-------
--The Assistant Administrator for Research and
Development has not systematically assembled, analyzed or
clearly defined research and development needs and objectives.
"The Research Outlook: 1978-1983",, which has been published, and
"The Research Outlook: 1979-1984",, currently nearing completion,
are perhaps the most definitive statements of research and
development needs and objectives. Ho.wever, neither document is
an adequate statement of near-term, mid-term or long-term plans
and objectives. Participatory discussions have apparently
occurred with^laboratories. Until initiation of the pilot
research committees, most planning activities were carried out
in headquarters with only limited and late stage input from the
laboratories. With the advent of the^ pilot research committees,
laboratory and Program Office input to near-term research
planning has occurred in those research areas for which
committees have been developed. This, has had a positive impact
on planning; however, in most cases where the laboratory
director was not involved in the committee's _acti v_itie,s_» .it. .has. ..
m-in-frrrized the role "O'f th'e" Ta~boratory" director in the planning
process. For a majority of the research programs, the
laboratory directors and staff have been involved primarily in
near-term planning and then most frequently at late stages of
the budget cycle. In many cases the input has been, fragmentary
and spurious, i.e., "What would you and your people like to do
n.ext year?" - * - • • ~ ; " '
Resource allocations (personnel and finances) are in a
continuous state of flux. -As •'-expected in relation to the
Federal budget system, chaffges aTe made up to the beginning of
the current fiscal year, but frequently continue on throughout
the year. The major certainty appears to be that change will
take place. The laboratory directors apparently are given
little authority for shifting resources within program areas and
even less authority for shifting resources between program
areas. This lack of flexibility, with continuous management
from headquarters, appears to have had a negative impact on the
productivity of the programs. EPA scientists, in many cases,
are confronted with changes in program direction and level of
effort with very short; notice. Extramural projects have, in many
cases, been"treated -as the most flexible portion.of the system.
Contracts that have been expanded or shifted in direction on
very short notice have served to alienate substantial portions
of the research community. Precipitous actions, discontinuation
of programs, or shifting of program direction raises legitimate
questions "concerning the adequacy1 "of'"Agency research and
development planning. Precipitous increases of funds, although
having associated moments of elation, are usually followed by a
recognition that the time and personnel resources available do
not allow careful selection of new contractors, resulting in
projects that are less successful than they should be.
202
-------
4. "Not only the changing nature of
environmental problems but also the
exigencies of the economy, suggest that it
would be inadvisable to build up a large
permanent staff. Rathera maintaining the
necessary competence to monitor grant and
contract work as needed would appear to be
a prudent course.
"A careful review of the contract and grant
procedures should be undertaken."
--The Agency has not given adequate attention to
developing a strategy for the implementation of its research
program, i.e., balance among intramural research, contracts,
grants and interagency agreements. Although the mandated
ceiling on numbers of personnel is recognized, the Agency has
not -m-ade- a-de-q-ua-te pl-a-ns fo-r -1-i.v_in.g. -w~ithin. Lhajt. celling. ..To..
circumvent the personnel ceiling, contracted personnel are used
on site at many laboratories to perform maintenance operations,
thereby extending the work force. There are numerous
individuals who are faced with a multitude of competing
responsibilities: performing hands-on research; supervising .
technicians who directly assist them; preparing orders and
monitoring the efforts of on-site contract personnel; soliciting
and reviewing research grants and proposals; monitoring research
being perf=q.r;med_ by contractors and prantees, either by personal
visit or review of innumerable reportsi expected of the
contractors and grantees; and participating in the preparation
and review of criteria documents and related material. In some
instances, there are experienced scientists and managers
available who do an excellent job of balancing and meeting these
competing demands. In a few instances, individuals, who have been
unwilling to accept the demands placed on them, have retreated
into their corners to do "their thing," ;i.e., perform specific
research in line with their interests, and are content to let
the system go on its own merry way. Although this has solved
th.ei r -i mmedi ate.....p.r.obj em, it has increased the workload and
demand placed on their col leagues. In many cases, the demands
are excessive in relation to the experience an"d training of the
staff member, and one or more of the aspects of the job are
performed poorly. ,
The impact on botK intramura,] .and extramural research is
apparent., jvh.e,Impact oh th.e intramural" program is discernible
by the fact that many EPA scientists do not publish because they
have performed relatively little research. A review of how
selected grants and contracts were initiated and monitored
suggests that, in some cases, the individuals involved did not
have adequate experience or time to perform their assignments.
A related and contributing factor has been the development of an
"unwritten" set of procedures for promoting the use of grants
rather than contracts because of the more cumbersome nature of
the contract award process.
2O3
-------
In summary, a careful review of contract and grant
procedures is as much needed now as i£ was at the time of the
NAS/NRC report. A key aspect of such a review should be the
development of a strategy dealing with how much research can be
appropriately performed in the Agency:and how extramural work
can best be performed.
2O4
-------
VIII. COMMUNITY HEALTH AND ENVIRONMENTAL SURVEILLANCE SYSTEM
(CHESS):. AN INVESTIGATIVE REPORT j
A. Background of the CHESS Program
The Community Health and Environmental Surveillance System
(CHESS) was initiated about 1970 and involved collection of data
during the period 1970 to, 1975. This research and surveillance
program was designed to investigate the relationship, if any,
between air pollution and health in human populations (up to a
few thousand persons), studied at single Contacts or followed
for short periods of time (up to two years), for
characterization of health status. These observations were
coordinated with observations on air pollution in the
environments of the study populations. The populations and
areas included for study were selected to; represent pairs or
larger sets of contrasting exposures, for example, a "clean" and
a "dirty" town or a series of several communities with a known
or suspected substantial range of air pollution conditions. Most
populations consisted of persons not previously known to have
any special health problems, although some studies within CHESS
were directed at groups defined by disease conditions, for
example, known ast-hma patients. . T..
The program operated from 1970 to 1975 and resulted in a
major publication in May 1974 (Health Consequences of Sulfur
Oxide: A Report from CHESS, 1970-1971).. That publication
included analysis and interpretation of the first two data
collection years. Other smaller papers and presentations
involved these and some later years1 data. The major review in
1974 implicated sulfates, sulfuric acid, and sulfur dioxide as
causing health effects, chiefly respiratory tract disease or
disturbance of pulmonary function, at or near levels of these
pollutants common.ly considered "safe." That report was
extensively reviewed by a number of individuals and groups and
received both praise and criticism. In part because of some of
the criticism, CHESS, in its original form, was discontinued.
It was recommended, however that add.itiohal substantial efforts
be made to.optimally use the collected data beyond those uses
reported in 1974. Special features to be considered in further
work were to include: (1) analysis of extensive data collected
from 1973 to 1975 and not included in thb 1974 report; (2)
improvements of statistical data and analytic techniques; (3)
assessment of validity of coded data and- of extent of coding
- errors or other correctable problems in the data set; (4)
increased objectivity i'n interpretation :of findings; and (5)
assessment of confidence range of estimates of pollution.
2O5
-------
B. Findings of the Subgroup
During the site visit in September 1978, the status of the
CHESS program was reviewed and a summary follows. The mechanism
for continuing work on CHESS is a contract from the Environ-
mental Protection Agency to the University of North Carolina,
Chapel Hill, principal investigator Dr. Carl Shy. This contract
work is closely followed by members of the epidemiology division
and the statistics unit of the Health Effects Research
Laboratory, Environmental Protection Agency. Dr. Shy was
formerly extensively involved with the CHESS project as a member
of the epidemiology unit; he is now a member of the faculty,
University of North Carolina. The plan is to review all of the
CHESS data collected for 1970 to 1975. The contract to the -
University was let in September 1977.
To date there has been a major effort to validate the CHESS
data sets. This was projected to require two years but is now
expected to be completed about eight months ahead of schedule
because special priority was given to the validation project.
This has been accomplished in spite of a budget deletion of the
funds planned for this purpose, thereby making it necessary to
discontinue other work to meet this mandated task. The
validation project is designed to identify discordances between
manually recorded original data and tape recordings on exposure
(pollution), outcome (health measures);, and control demographic
and confounding variables. It is being done very effectively
under the direction of Mr. Gerald Nehls, Director of the Data
Management Unit in the Health Effects Research Laboratory. It
must be noted that any validation of these old data is now
limited to validation of the previous coding and automating and
not to any review of the correctness of initial observations of
symptoms and other health effects.
!
A standing committee has been created, reporting to Dr. Shy
and supported under the research contract, to review all planned
publications of the CHESS data. The committee presently
consists of Dr. Warren Winkelstein (University of California),
Dr. James Grizzle (University of North Carolina), and Dr.
Michael Lebpwitz (University of Arizona). This committee has
just been funded, and"its effectiveness cannot yet be judged.
The membership seems appropriate, and the plan for a standing
procedure for outside review is a useful move in response to
criticism regarding objectivity of reporting.
•---"•• A report of a current analysis of a portion of the CHESS
data from the Southeast region (Charlotte, North Carolina and
Birmingham, Alabama) was presented to the site group by Ms. Shi-
Ping Lan. The analysis and presentation indicated a high degree
of statistical competence and good collaboration among Dr. Shy,
Ms. Lan, and Dr. Hasselblad of the Health Effects Research
Laboratory. The material presented will presumably be in a form
for publication soon. A principal feature of the new analysis
is more adequate use of the symptoms data from the health
survey, employing a 5-level symptom scale rather than the
dichotomy used in earlier analyses.
205 !
-------
The information that can optimally be obtained from this
Southeastern study is limited, however, because any possible
effect of air pollution on the measured health indices is lower
by factors of 10 to 100 than effects of smoking or job
exposure. Even though a pollution (intercity) association is
found, it remains possible that this association is not causal
but is due to a variable related to the stronger effects of
smoking or job exposure or to other confounding variables for
which no observations are available.
While the acronym CHESS is understood to apply to the 1970
to 1975 group of studies, certain new work in progress
follows the general outline of that program. The study most
clearly conforming to that design is in four Utah communities,
in which 1976 observations are being compared with former 1970
CHESS observations of chronic respiratory disease and of acute
lower respiratory tract disease, as related to increasing S02
pollution in the region.
A substantial change in the operation of CHESS and related
studies has been made in the past three years with a change in
emphasis from in-house research to research grants and
contracts. This appears to be a result, in part, of the
extensive criticism of the previous CHESS program and is
reflected in the entire activity of the Epidemiology pi vision.
Only four professional researchers from a previous epidemiology
staf f of ,15 , rem.ai n In that division. Three new, young junior
investigators have recently joined the division. The reduced
staff is essentially completely occupied with their duties as
project officers on contracts and grants. The result of this
change from intramural to extramural with regard to CHESS
appears not to be obstructive and may offer certain advantages.
C. Steps Taken by EPA to Meet Brown Committee
Recommendations
Public Law 95-155, passed by the 95th Congress, mandated a
review of and a report on "the findings and recommendations of
the report to the House Committee on.Science and Technology
entitled 'The Environmental Protection Agency's Research
Program with Primary Emphasis on the Community Health and
Environmental Surveillance System (CHESS): An Investigative
Report.1" It was further specified that; special attention be
focused on "procedural safeguards required to preserve
scientific integrity of such research and to insure the ,
reporting and use of such research in subsequent recom-
mendations."
207
-------
Although Chairman Brown emphasized the desirability of a
positive.attitude in the letter of transmittal of the Committee
Report, the document impressed some members of the subgroup as
often being hypercritical and demanding an approximation to
perfection that is not obtainable in studies of human
populations. The EPA has published a response to the recom-
mendations of the Investigative Committee in the EPA Research
Outlook of March 1978. The report of this subgroup will
addre,ss only those recommendations that deal with on-going
activities related to CHESS or other epidemic!ogical and bio-
statistical work at HERL/RTP. Recommendations will be ident-
ified by the numbers used in the Investigative Report and in
the Agency's response.
3(a): EPA should publish an announcement regarding the
limitations of the CHESS Monograph.
3(c): EPA should publish an adderidum to the CHESS
Monograph including most of the Investigative Report.
Subgroup findings; It is believed that the EPA response
covers these recommendations satisfactorily, although it is
difficult to see how the response can be delivered to all
holders of the CHESS Monograph. .Most scientists, however, will
be aware .of the limitations of the data in this Monograph.
4(a): Legislation should be reexamined regarding
unrealistic procedures and schedules, i
Subgroup findings: The legislative mandate for a study of
air pollution and its effects on the Gulf Coast (Houston) area
appears to require an unreasonably rapid approach to a very
complex problem. The epidemiology group expressed an interest
in investigating this situation in a systematic, planned
fashion. They doubted that the mandated crash approach would
be maximally productive but stated their intent to obtain as
much valid data as possible. It is not known to what extent
this legislative mandate was reexamined. No evidence was found
at this Jeyel to indicate that reexamjRation was effective in
producing any important changes. Current procedures referred
to in the Agency's response in the EPA Research Outlook do not
appear to be adequate to solve problems caused by unrealistic
legislative mandates.
tions
~4(d): EPA should advise. Congress if budgetary restric-
wijl impact completion of major projects/
"*"""' *~ 'F ' "" ' "" " " ' , •«-
-------
5: OMB should be asked to develop procedures for prompt
review of questionnaires. , '
Subgroup, findings: The Population Studies Division has
found OMB responsive to their need for quick approval of
questionnaires. The subgroup supports the EPA position that
its questionnaires for volunteers in research projects should
not require submission to OMB.
6(a): CHESS data analyses should be carried out only on
data with high validity potential. .
Subgroup findings: Dr. Shy's group at the University of
North Carolina and the epidemiologists and statisticians at
HERL have reviewed the CHESS data and have decided which data
sets warrant analysis for publication.
6(b): EPA should publish research in refereed journals in
a timely fashion.
6(c): EPA should not publish large projects solely in
monograph form.
6(d): EPA should not initiate projects for policy
consideration unless they can be completed in a realisti'c time
frame. -
Subgroup findings: Staff indicated their desire to see
results published in scientific peer reviewed journals but
emphasized their lack of time to do or report their research or
the findings of contractors. It is reasonable to assume,
however, that most grant recipients and contractors will
publish their findings in appropriate journals. It should be
noted, however, that a document entitled "CHESS Bibliography,
December 1, 1977" lists, for the period 1/75 to 12/77, only one
journal article, seven government publications, and ten EPA in-
house publications, plus three more in-house publications that
are undated but whose authors or titles suggest that they
belong in this time period. For 1977, the bibliography lists
..only one ..government publication, which must have been-planned
well in advance of -the Brown Committee report. ... . . .
.It seems unlikely that the EPA responses to this
recommendation can be properly assessed until the epidemiologic
staff.... is increased to a size more commensurate with its duties.
-7(a):- 'EPA should strengthen the CHAMP aerometric and
quality control programs.
7(b): EPA should shorten the time between data acquisi-
tion and quality assurance analysis of data.
7(c): EPA should stop employing development stage
instruments before qualification testing,
7(d): EPA should not use laboratory models of instruments
in"the field until they have been field checked and operating
personnel trained.
205
-------
7(e): EPA should reevaluate the opening of the CHAMP
operations contract to competition.
Subgroup findings: CHAMP is no ;longer at HERL. We were
informed that it no longer exists as ian identifiable unit
separate from other monitoring activities.
7(f): EPA research and monitoring personnel should
closely coordinate regarding chemical species.
!
Subgroup findings: Coordination of CHAMP with health
effects personnel is now potentially more difficult because of
the transfer of the responsibil ities ;0f CHAMP to another
laboratory. It is still too early to tell whether the transfer
will help by strengthening this type of monitoring activity or
will hinder the accomplishment of the Agency's mission by
impeding coordination.
10(a): An interdisciplinary task force should draw up an
integrated air epidemiology exposure assessment program plan
for EPA.
Subgroup findings: There.is a_desire for an advisory
group not only to meet this recommendation for assessing health
effects of air pollution but also to provide consultation for
other epidemiologic studies, both intra- and extramural.
10(cj: EPA should have epidemi©logical questionnaires and
panel selection criteria approved by peer groups.
Subgroup findings: Aside from a comparison of self-
administered versus interviewer-administered questionnaires,
the work related to this recommendation is limited to the
information that can be gathered from the extensive analyses of
CHESS data being carried out by Dr. Shy. The panel data are not
scheduled for analysis. :
Planning for a second round of CHESS or for investigation
of aJr .pol1ution,"episodes" was not, ment.lon.ed,. It is difficult
to see how very much can be done" along this line with the
limited staff. It seems reasonable to delay planning for a
second round of CHESS until the current analyses are completed.
210
-------
> APPENDIX A :
I UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON. D.C. 20460
JUL 12 1978 !
: THE ADMINISTRATOR
TO: Dr. Emil M. Mrak
Chairman '
Executive Committee, Science Advisory Board
THRU: Dr. Richard M. Dowd
SUBJECT: Charge to the Science Advisory Board's Health Effects
Research Review Group i
\
The Authorization Act of 1978 for Research and Development, PL 95-155,
requires that a special evaluation report on th^ Agency's health effectis
research efforts be prepared by the Science Advisory Board (SAB). The Act
specifically outlines what is expected to be included in the report regarding
your assessment of our health effects research programs, and the procedures
fo.r the conduct, review, reporting and use of such research.
To -delineate the Congress's charge more sharply; I urge the Study Group
to define health effects research to include all planned activities, col-
lection and analyses of data done within the Agency for the purpose of
adding to the scientific basis for understanding the effects of environmental
factors on human health. This definition would, include those activities
within the Agency which may be used to assess human risk, and which support
standard setting and regulatory decisions, and any activity which gathers
new knowledge about human health, or improves our understanding of human
health either directly or which can be used to jextrapolate to human health
impacts. I am happy to hear that Dr. James Whittenberger and Dr. Roger
McClellan-will chair and co-chair this review group.
I can assure you that your assessment of the Agency's activities*?
within the scope of this definition will be appreciated and that^you will
have our full cooperation in this endeavor.
Doug
211
-------
-1- ,
PUBLIC LAW 95-155—NOV. 8, 19*7 91 STAT. 1257
Public Law 95-155 ;
95th Congress
An Act
To authorize appropriation*) tor activities of the Environmental Protection Nov. 8. 1977
Agency, and for other purposes. [H.R. 5101]
lie it enacted by the Senate and House of Ke-prcxcntatives of the
United States of Aincnaa in (Jonyrasx axxrmbli'd, That this Act may Environment*!
bo cited as the "Environmentul Ke.suarch, Development, and Demon- Research,
Stratkm Authorization Act of 1978". Development,
SKC. 2. (n) There, arc authorized to bo appropriated to the Environ-
mental Protection Agency for environmental research, development.
11 A i* tt
-------
91 STAT. 1258
PUBLIC LAW 95-155— NO.V. 8, 1977
Ante, p. 687.
Appropriation
authorize tion.
Transfer of funds,
restriction.
Budget
, projections.
42 USC 4361*.
42 USC 4361.
(b) In addition tb any other sums authorized by this section or by
other provisions of law — ,
(1) there are authorized to bo, appropriated to the Adminis-
trator of the Environmental 1'rotection Agency for fiscal year
1978, $10,000,000 for long-term research nnd development in
accordance with section C of this Act;
(2) there are authorized to be appropriated to tin1. Adminis-
trator, for fiscal year 1'JTH, $2,0(X),000 for training of health sci-
entists needed for environmental research and development in
fields where there arc national shortages of trained personnel;
and
(3) there arc authorized to be appropriated to flic Administra-
tor, for fiscal year 1078, $:}, 000,000 to implement the study author-
ized in section 103(d) of the Clean Air Aet Amendments of 1977
(Public Law 05-95).
(c) There is authorized to be appropriated to the Administrator
$10,000,000 for fiscal year 1978 for program management and support
related to environmental research and development.
(d) Xo funds may be. transferred from. any particular category
listed in suljsection (a) or (b) to any other category or categories
listed in either such subsection if the. total of the funds so transferred
from that, particular category would exceed 10 per centum thereof,
and no funds may be transferred to any particular category listed in
subsection (a) or 00 from any other category or categories listed in
cither such subsection if the total of the ;funds so transferred to that
"particular category "would exceed 10 per centum thereof, unless —
(1) a period of thirty legislative days has passed after the
Administrator of the. Environmental Protection .Agency or Ins
designee. lias transmitted to the Speaker of the House of Kepre-
sentativcs and to the President of the Senate a written report
containing a full and complete statement concerning the nature
.:.-_ of the transfer aiid the reason 'there f of. or '- -••--,
(2) each committee, of the House of Representatives and the
Senate, having jurisdiction over the subject matter involved,
before the expiration of such period, has transmitted to the Admin-
istrator written notice ro the effect -that such committee has no
objection to the proposed action.
SKC. 3. Appropriations made pursuant to the authority provided
in section 2 of this Act shall remain available for obligation for
expenditure, or for obligation and expenditure, for such period or
periods as may be specilied in the Acts making sneh appropriations.
SKC. 4. The Administrator of the Environmental Protection Agency,
Public sector
agencies, grants.
42 USC 300j~3a.
sion shall include a detailed explanation of the relationship of each
budget projection to the existing laws which authorize the Adminis-
tration's environmental research, development, and demonstration
programs. - - -
SEC. 5. (a) The Administrator of the Environmental Protection
Agency shall offer grants to public sector agencies for the purposes
of— • .
(1) assisting in the development and demonstration (includ-
ing^ const ruction) of any project which will demonstrate a new
or improved method, approach, or technology for providing a
dependably safe supply of drinking water to the public; and
273
-------
-3-
PUBLIC LAW 95-155-^NOV. 8, 197? 91 STAT. 1259
(2) assisting in -the development ami demonstration (includ-
ing construction) of any project which will investigate and dem-
onstrate health and conservation implications involved in the
reclamation, recycling, and reuse of waste-waters for drink-
ing and the, processes and methods for the preparation of safe
and acceptable drinking water.
(b) Grants made by ilhe Administrator under this section shall be Grants,
subject to the following limitations: < limitation*.
(1) Grants under this section shall not exceed 60% per centum
of the total cost of construction of any facility;aud 75 per centum
of any other costs, as determined by the. Administrator.
(2) Grants under this section shall not be made for any proj-
ect involving the construction or modification of any facilities
for any public witter system in a State unless such project has
been approved by the State agency rharged with the responsi-
bility for wifety of drinking water (or if there is no such agency
in a State, by the State health authority).
(3) Grants under this section shall not be made for any proj-
ect unless the Administrator determines, after consultation, that
such project will serve a useful purpose relating to the develop-
ment- and demonstration of new or improved techniques, methods,
or_technologies for the provision of safe water to the public for
drinking.
(e) There are authorized to be appropriated for the purposes of this
section $25,000,000 for fiscal year 1978...... .;...,.... ...
SKC. 6. (a) The Administrator of the Environmental Protection
Agency shall establish a separately identified program to conduct con-
tinuing and long-term environmental research and development.
Unless otherwise specified by law, at least 15 per centum of any funds
appropriated to the. Administrator for environmental research and
development under section 2(a) of this Act.-or under any other Act
shall be allocated for long-term environmental research and devel-
opment- under this section. ;
(b) The Administrator, after consultation with the Science Advisory
Board, shall submit to the President and the* Congress a report con-
cerning the desirability and feasibility of establishing a national
environmental laboratory, or a system of such laboratories, to assume
or supplement, the long-term environmental research functions cre-
ated bv subsection (a) of this section. Such report shall be submitted
on or before March 31, 1078, and shall include findings and recom-
mendations concerning—
(1) specific typos of research to be carried out by such labora-
tory or laboratories; :
(2) the coordination and integration of research to be con-
ducted by such laboratory,or hibpratorjes_wjt.h research conducted
&y existing Federal or other research facilities: ,
(3) methods for assuring continuing long-range funding for
such laboratory or lulKmitories; and
(4) other a'clm'mtstrative or legislative, actions necessary to
facilitate the establishment of such laboratory]or laboratories.
SKC. 7. (a) The Administrator of the Environmental Protection 42USC4364.
Agency shall assure that the expenditure of .any funds appropriated
pursuant to this Act or nny other provision of liiw^for environiMpn*i»l
research and development related to regulatory jirosiram nc*5v!'ips
bo coordinated with and reflect the research needs and priorities
Research and
development
program.
42 USC 4363.
Report to
President and
Congress.
Contents.
214
-------
-4-
91 STAT. 1260
PUBLIC LAW 95-155—
8, 1977
Program offices.
Report to
President and
Congress.
Science Advisory
Board.
Establishment.
42 USC 4365.
Membership.
42 USC 4361.
Report to
Administrator,
President, and
Congress.
of the program offices, as well as tho overall research needs and priori-
ties of the Agency, including those defined in the five-year research
plan. ;
(b) For purposes of subsection (a)j the appropriate program
offices are— :
(1) the Office of Air and Waste Management, for air quality
activities;
(2) tho Office of Water and Hazardous Materials, for water
quality activities and water supply activities;
(3) the Ofltce of Pesticides, for environmental effects of
pesticides; !
(4) the Office of Solid Waste, for solid waste- activities;
f5) the Office of Toxic Substances, for toxic substance activities;
(6) the Office of Radiation Programs, for radiation activities;
and ;
(1) the Office of Noise Abatement and Control, for noise
activities.
(c) Tho Administrator shall submit to the President and the Con-
gress a report concerning the most appropriate means of assuring, on
a continuing basis, that the research efforts of the Agency reflect the
needs and priorities of the regulatory program offices, while main-
taining a high level of scientific quality., Such report shall be submitted
on or before March 31,1978.
SEC. 8. (a) Tho Administrator of the Environmental Protection
Agency shall establish a Science. Advisory Board which shall provide
such scientific advice as the .Administrator requests.
(b) Such Board shall lx», composed of at least nine members, one of
whom shall be designated Chairman, and;shall meet at such times and
places as may be designated by the Chairman of the Board in consul-
tation with the AdministratoV. Kaeh .member of the Board shall bo
•qualified-by-education, training, and experience, to evaluate scientific
and^technical information on matters referred to the Board under this
section. ;
(c) In addition to providing scientific advice when requested by tho
Administrator under subsection (a), the Board shall review and
comment on the Administration's five-year plan for environmental
research, development, and demonstration provided for by section 5
of Public Law 94—t75 and on each annual revision thereof. Such
review and comment shall be transmitted to the Congress by the
Administrator, together with his comments thereon, at the time of the
transmission to the Congress of tho annual revision involved.
(d) The Board shall conduct a review:of and submit a report to the
Administrator, the President, and the Congress, not later than
October 1,1978, concerning—
(1) "the hcalth-elFccts research authorized by this Aet-and other
laws;
(2) the procedures generally used in the conduct of such
research;
(3) the internal and external reporting of the results of such
research; •
(4) the review procedures for such research and results;
(5) the procedures by which such results are used in internal
and external recommendations on policy, regulations, and legisla-
tion; and
(0) the findings and recommendations of the report to the
House Committee on Science and Technology entitled "The
215
-------
-5-
PUBLIC LAW 95-155—NOV. 8, 19J7
Environmental Protection Agency's lie-search Program with
primary emphasis on the Community Health and Environmental
Surveillance System (CHK.SS) : An Jtivwitiga'tive Kcport".
The review sliall foc.iis special uttc.ntion on the procedural safeguards
requirtul to preserve the .scientific integrity of sucli research and to
insure reporting iincl use. of the results of such research in suhseq.ue.nt
recommendations. The report sliall include specific-recommendations
on the results of the review to ensure scientific integrity throughout
the Agency's health effects research, review, reporting, and recom-
mendation process. j
(e) (1) The Administrator, at the time any proposed criteria docu-
ment, standard, limitation, or regulation under the dean Air Act, the
Federal Water Pollution Control Act, the Kesource, Conservation and
Kceovery Act of )!)7c chaired by a member of the Board.
(h)(l) Upon the recommendation of the Board, the Administrator
shall appoint a secretary, and such other employer-sUis deemed neces-
sary to exercise and fulfill iho Board's powers and responsibilities.
The compensation of all employees appointed under this paragraph
shall be fixed in accordance with chapter f>l and siibchapter III of
chapter flSof tit le it of the United States Code. I
(2) Members of the Board may be compensated atn rate to be fixed
by the President but. not in excess of the maximum rate of pay for
grade GS-18, as provided in the General Schedule under section 53S2
of title 5 of the United Stales Code. '
•• (i)-In carrying out the/functions assigned by this section^ the Board
shall consult and "coordinate,Its activities with the Scientific Advisory
Panel' established by the Administrator pursuant Ito section 25(d)
of the. Federal Insecticide, Fungicide, and Kodenticidc Act, as
amended. :
SKC. 9. (a) The. Administrator of the Environmental Protection
Agency, in consultation and cooperation with the heads of other Fed-
eral agencies, shall.take such actions on :i continuing basis as may be
necessary or appropriate—
(1) to identify environmental research, development, and
demonstration activities, within and outside the.;Federal Govcrn-
91 STAT. 1261
42 USC 1857
note.
33 USC 1251
note.
42 USC 6901
note.
42 USC 4901
note.
15 USC 2601
note.
42 USC 300f
note.
Member
committees and
investigative
panels.
Secretary,
appointment.
5 USC 5101,
5331.
5 USC 5332 note
7 USC 136w.
42 USC 4366.
216
-------
-6-
91 STAT. 1262
Report.
42 USC 4361,
Report to
i'rciwient *nd
Congress.
Legislative
recommenda-
tion*.
Presidential
report to
Congreis.
42 USC 4361b.
Personnel
PUBLIC LAW 95-155—NOV. 8,1977
mcntj which may need to be. more, effectively coordinated in order
lo minimize unnecessary duplication of programs, projects, and
research facilities;
(2) to determine the: steps which might be taken under existing
law, by him und by the. bends of such other agencies, to accomplish
or promote such coordination, and to provide for or encourage the
taking of such steps; and ;
(8) to determine the. additional legislative, actions which would
be needed to assure such coordination to the maximum extent
possible.
The Administrator shall include in each annual revision of the five-
year plan provided for by section 5 of Public Law 94-475 a full and
complete report on the actions taken ami determinations made during
the, preceding year under this subsection1, and may submit interim
reports on sucli actions and determinations at such other times as ho
deeiMS appropriate. _ ;
(b) The Administrator of the Environmental Protection Agency
shall coordinate- environmental research,: development, and demon-
stration- programs of such Agency with the heads of other Federal
agencies in order to .minimizeunnecessary duplication olprograms,
projects, and research facilities.
(o)(l) In order to promote the coordination of environmental
research and development activities, aiuljo assure that the action taken
and methods used (under subsection (a) and otherwise) to bring about
s"ch ^ordination will be asell'eetuv as possible for that purpose, the
Council on Environmental Quality in consultation with the Oflice of
Science and Technology Policy shall promptly undertake and carry
out- a joint study of all aspects of the coordination of environmental
research and development. The Chairman, of the Council shall pre-
pare a report on the results of such study, together with such recom-
mendations (including legislative recommendations) as he- deems
appropriate, and shall submit- such report to the President and the
Congress not later than May HI, 1978.
(2) Not later than September 30, 1078, the President shall report
to the Congress on steps he has taken to implement the recommenda-
tions included in the report under paragraph (1), including any rec-
ommendations he may have for legislation.
SEC. 10. The Administrator of the Environmental Protect ion Agency
shall implement the 'recommendations of-the report 'prepared for the-
House Committee on Science, and Technology entitled ''The Environ-
mental Protection Agency Research Program with primary emphasis
on the Community Health anil Environmental Surveillance System
(CHESS): An Investigative Report", unless for any specific rec-
_ ommendation ho determines (.1) that such recommendation has been
implemented, (2) that implementation of 'such recommendation would
not enhance the quality of the research, or (3) that implementation of
such recommendation will require funding which is not available.
Where such funding is not available, thiV Administrator shall request
the required authorization or appropriation for such implementation.
The Administrator shall report the status of such implementation
in each annual revision of the five-year plan transmitted to the Con-
groi-s under section 5 of Public Law 1)4-475.
SKC. 11. The Administ rat or ot the Environmental Protection Agency
shall increase, the number of personnel positions in the Health and
Ecological Effects program to 862 positions for fiscal year 1978.
217
-------
-7-
PUBLIC LAW 95-155—NOV. 8, 1977
91 STAT. 1263
SEC. 12, (a) Each officer or employee of the Environmental Protec-
tion Agency who—
(1) performs any function or duly under this Act; and
(2) has any known iinancial interest in any person who applies
for or receives grants, contracts, or other forms of iinancial assist-
ance under this Act,
shall, beginning on February 1, 1978, annually file with the Admin-
istrator a written statement concerning all such interests held by such
officer or employee during the preceding calendar year. Such statement
shall be available to the public.
(b) The Administrator shall— ;
(1) act within ninety days after the date of enactment of this
Act—
(A) to define the term "known financial interest" for pur-
poses of subsection (a) of this section; and
(B) to establish the methods by which;the requirement to
file written statements specified in subsection (a) of this sec-
tion will be monitored and enforced, including appropriate
provision for the filing by such officers and employees of such
statements and the review by the Administrator of sxich state-
ments; and - .. • , ••-•;. •. _• .,
(2) report to the Congress on June 1 of each1 calendar year with
respect to such disclosures and the actions, taken in regard thereto
during the preceding calendar year. •
(c) In the rules prescribed under subsection (b) of this; section, the
Administrator may identify specific positions of a nohpblicymaking
nature within the Administration and provide that officers or'employ-
ees occupying such positions shall be exempt from the requirements of
this section. >
(d) Any officer or employee who is subject to, and knowingly vio-
lates, this section, shall be fined not more than $2,500 or imprisoned
not more than one year, or both. '<
SEC. 13. It is the national policy that to the maxhnum extent possible
the procedures utilized for implementation of this Act shall encourage
the drastic minimization of paperwork.
Approved November 8, 1977.
Annual
statement, filing,
42 USC 4367.
Report to
Congress.
Violation,
penalty.
Paperwork
minimization,
encouragement.
LEGISLATIVE HISTORY: ;
HOUSE REPORTS: No. 95-157 (Comm. on Science and Technology) and No. 95-722
(Cornm. of Conference).
SENATE REPORT No. 95-188 accompanying S. 1417 (Comm. on Environment and
Public Works).
CONGRESSIONAL RECORD, Vol. 123 (1977):
Apr. 19, considered and passed House. ,
May 27, considered and passed Senate, amended, in lieu of S. 1417.
Oct. 20, Senate agreed to conference report. •
Oct. 25, House agreed to conference report. i
275
-------
Appendix 6
COMMITTEE MEMBERS AND CONSULTANTS
»
1. Subcommittee Core Members
Chairman: Dr. James L. Whittenberger
Professor of Physiology
School of Public Health
Harvard University
Co-chairman: Dr. Roger 0. McClellan
Director of Inhalation Toxicology
Research Institute
Lovelace Foundation ;
Members: Dr. Peter Bloomfield
Associate Professor ;
Department of Statistics
Princeton University
Dr. George W. Comstock ......
Professor of Epidemiology ~
Johns Hopkins Training Center
Dr. Morton Corn
. Professor of Industrial Health and
Ai r Engineering \
Graduate School of Public Health
University of Pittsburgh
Dr. Julius E. Johnson
Consultant
Dow Chemical Company
Dr. Wendell Kilgore
Professor of Toxicology
Department of Environmental
Toxicology
University of California at Davis
Dr. Robert A. Neal
Director, Center in Toxicology
Department of Biochemistry
Vanderbilt Medical School
219
-------
Dr. Gerard A. Rohlich
Professor of Environmental
Engineering, Department of Civil
Engineering, University of Texas
SAB Staff Officer: Dr. Frpde Ulyedal
Supervisory lexicologist
Office of Research and Development
Environmental Protection Agency
2, Consultants
Dr. Edwin Lennette, Biomedical Labs, Cali-
fornia State Department of Health
expertise: microbiology, virology
1
Dr. Jeanne Manson, Ketterihg Laboratory
University of Cincinnati
expertise: reproduction,teratology
Dr. Sol M. Michaelson, Professor of Radiation
Bi ol-ogy. a.nd Bi ophysics , Uni versity
.pf Rochester ;
expertise: non?-ionizing radiation
i
Dr. Steven M. Horvath, Director, Institute of
; , ~ ..En vi ronmental Stress ,.Uni versity
of Californi a ,
expertise: pulmonary physiology,
inhalation toxicology
Dr. George Hutchinson, Professor of Epidemi-
ology, Harvard School of Public
Health
expertise: epidemiology,
microbiology
Dr. James
G* Fox, Director^ Laboratory of
Animal Medicine, Massachusetts
Institute of Technology
expertise: laboratory animal care
and facilities
!
Or. Jennifer L. Kelsey, Associate Professor
of Epidemiology, Department of
Epidemiology and Public Health,
Yale University School of Medicine
expertise: epiidemiology of chronic
disease
Dr. Ralph C. Buncher, University of Cincinnati
Medical Center
expertise: epidemiology
220
-------
DATE
APPENDIX C '
MEETING AND TRAVEL SCHEDULE FOR HERRG
LOCATION . : PARTICIPANTS
21 June 78
13-14 July 78
20-21 July 78
23 Aug. 78
25 Aug 78
25-27 Sept. 78
28 Sept. 78
5-6 Oct. 78
16-18 Oct. 78
19 Oct. 1978
Preliminary meeting, ; Dr. Whittenberger
with Dr. Hueter, ! Dr. Ulvedal
HERL/RTP :
Public meeting,
Washington, D.C.
Envi ronmental
Research Lab
Duluth, Minn.
Office of Water &
Waste Management
Washington, D.C.
Office of Toxic
Substance's
Washington, D.C.
Health Effects
Research Lab
Research Triangle
Park,, N.C.
Preliminary Mtg.
-with Dr*—Garner
HERL/Cincinnati
Envi ronmental
Research Lab
Gulf Breeze, Fla,
Health Effects
Research Lab
Ci nci nnati, Ohio
.Health Effects
Research Lab.
Fie!d Station
Wenatchee, Wash
HERRG
Dr. McClellan
Dr. Kilgore
Dr. Ulvedal
Dr. Rohlich
Dr. Neal
Dr. Johnson
Dr. Ulvedal
Dr. Neal
Dr. Ktlgore
Dr. Johnson
Dr. Ulvedal
HERRG and
Dr. Manson
Dr. Michaelson
Dr. Horvath
Dr. Hutchinson
Dr. Fox
Dr. Kelsey
Dr. Ulvedal
Dr. McClellan
~Dr..Ulvedal,
Dr. Whittenberger
Dr. Kilgore
Dr. Ulvedal
HERRG and
Dr. Lennette
Dr. Hutchinson
Dr.- Fox
Dr. Buncher
Dr. McClellan
Dr. Johnson
Dr. Kilgore
Dr. Ulvedal
227
-------
-2-
DATE
LOCATION
PARTICIPANTS
24 Oct. 78
26 Oct. 78
27 Oct. 78
30 Oct. 78
8 Nov. 78
9 Nov. 78
13-14 Nov. 78
13 Nov. 78
13 Nov. 78
Office of Ai r,
Noise, & Radiation
Environmental
Research Lab.
Narragansett, R.I.
Health Effects
Research Lab
Field Station
W. Kingston, R.I.
Office of Planning
and Management
Washington, D.C. ,
Region I
Boston, Mass.
Environmental Monr
it oring & Support
Laboratory,
Las Vegas, Nev.
Public Meeting
Washington, D.C.
Office of Planning
and Management
Washington, D.C.
Office of Research
and Development
Washington, D.C.
Dr. Whittenberger
Dr. Corn
Dr. Bloomfield
Dr. Ulvedal
Dr. Whittenberger
Dr. Lennette
Dr. Ulvedal
Dr. Whittenberger
Dr. Lennette
Dr. Ill vedal
Dr. McClellan
Dr. Ulvedal
Dr. Whittenberger
Dr. Ulvedal
Dr. McClellan
Dr. Ulvedal
HERRG
Dr. Corn
Dr. McClellan
Dr. Johnson
Dr. Bloomfield
Dr. Whittenberger
Dr. Kilgore
Dr. Neal
222
-------
APPENDIX D
PRINCIPAL EPA PERSONNEL PROVIDING INFORMATION TO HERRG
!
* Interviewed
+ Provided written Information
!
i
Office of the Administrator
Douglas M. Costle*+
Administrator \
Dr. Ri chard Dowd*
Science Policy Advisor to the1 Administrator
Staff Director, Science Advisory Board
Dr. Toby Clark*f
Special Assistant to the Administrator
!
Regional Offices
William R. Adams, Jr.*
Regional Administrator, Region I
Dr* Richard Keppler*
Director, ORD, Region I
Office of General Counsel
i
James C. Nelson*-*
Attorney Advisor
John W. Lyon*
Attorney
Edward Gray*
Deputy Associate General Counsel for Program Support
Office of Legislation ;" "
Marianne Thatcher*
Congressional Liaison Specialist
Alice White*
Legislative Reference Specialist
Office of International Activities
Jack E. Thompson* -~ .
Director, International Organizations and Western Hemisphere
Division
223
-------
-2- |
Office of International Activities (Continued)
Thomas Lepine+
Chief, Scientific Activities Overseas Branch
Office of Planning and Management
Roy N. Gamse*
Deputy Assistant Administrator for Planning and Evaluation
Frans J. Kok*
Director, Economic Analysis Division
Marian Ml ay*;
Director, Program Evaluation Division
Associate Deputy Assistant Administrator for Resource Management
Raymond A Pugh*+
Director, Budget Operations Division
Donald Hambri c +
Chief, Cost Review and Policy Branch
Contract Management Division (CMD)
: Vi ncerit-.d.ay,+.. ,
Chief, Interagency Agreements Branch, CMD
Carlene Foushee*
Grants Specialist, Grants Division
Office of Water and Haste Management
Thomas C. Jorlijng*
Assistant Administrator for Water and Waste Management
Allen Cywi n**
Senior Science Advisor
Swep T. Davis*
Deputy Assistant Administrator for Water Planning and Standards
Albert""J™. Erickson*
Associate Deputy Assistant Administrator for Water Planning
and Standards
- John T. Rhett*
Deputy Assistant Administrator for Water Program Operations
Henry L. Longest*
Associate Deputy Administrator for Water Program Operations
224
-------
-3-
Offlce of Water and Waste Management (Continued)
Kenneth Mackenthuh*
Director, Criteria and Standards Division
i
Gary N. Dietrich*
Direcotr, Office of Program and Management Operations
Victor J. Kimm*;
Deputy Assistant Administrator for Drinking Water
John P. Lehman*
Director, Hazardous Waste Management Division
Joseph Cotruvo*t
Director, Criteria and Standards Division
Office of11 Drink-trrg—W-at-er
Shelly Williamson1**
Epidemiologist:
Office of Air. Noise and:Radiation
David 'GT Hawkins*
Assistant Administrator for Air, Noise and Radiation
Rudo 1 -pfr-M-.- -Mar razz o*
Science Assistant to the Deputy Assistant Administrator for
Noise Abatement and Control
William A. Mills*
Director, Radiation Criteria and Standards Division
Acting Deputy Assistant Administrator for Radiation Programs
Walter C. Barber, Jr.*
Deputy Assistant Administrator for Air Quality Planning and
Standards
John O'Connor**
Strategies and Air Standards Division
Joseph Padgett*
Director, Strategies and Standards Division
Michael P. Walsh*
Deputy Assistant Administrator for Mobile Source Air Pollution
Control
Stan Blacker*
Special Assistant to DAA for Mobile Source Air Pollution Control
225
-------
Office of Toxic Substances
Steven D. Jel11 nek*
Assistant Administrator for Toxic Substances
Warren R. Mulr*
Deputy Assistant Administrator for Testing and Evaluation
John DeKane^*
Deputy Assistant Administrator for Chemical Control
Edwin L. Johnson*
Deputy Assistant Administrator for Pesticide Programs
James M. Conlon*
Associate Deputy Assistant Administrator for Pesticide Programs
William S. Murray*+
Director, Technical Services Division
Jack Griffith*
Chief, .Human Effects Monitoring Branch, Technical Services
Di vi si on
Don Barnesf
Special Assistant to the Assistant Administrator for Toxic
. Substances
Norbert Page*
Director, Health Review Division
James R. Bea'll*
Toxicologist, Health Review Division
i
David Gould*
Toxicologist, Health Review Division
-—David Anderson*
Biochemist, Health Reyiew Division
Carl Morri s*
Pharmacologist, Health Review Division
Office of Research and Development
Stephen J. Gage*+
Assistant Administrator for Research and Development
Sam Rondberg*+
Director, Office of Planning and Review
Dennis Tirpak+
Special Assistant to AA for Research and Development
226
-------
-5- • :
Office of Research and Development (Cont.)
Randal 1 W. Shobe*
Director, Technical Information Division
Robert W. Lane*+
Special Assistant to AA for Research and Development
Delbert Barth* !
Deputy Assistant Administrator for Health & Ecological Effects
William S. Murray*
Associate Deputy Assistant Administrator for Health & Ecological
Effects \
Roger Cortes 1*
Director, Criteria Development and Spetial Studies Division
Davi d Flemer* ',
Director, Ecological Effects Division
George Armstrong**
~ Director, Health Effects D'i vision
Alphonse Forziati+
Director, Stratospheric Modification Research Staff
William A-; Cawley*
Director, Technical Support Division
Office of Monitoring and Technical Support
Michael Mastracci*
Director, Regional Service Staff
Office of Monitoring and Technical Support
Gerald J. Rausa+;
Program Officer9 Energy Related Health Effects
Office of Energy, Minerals and Industry
William A. Rosenkranz*
Director, Waste Management Division
Office of Air, Land and Water Use
,Wi Vson Tal 1 ey* i .
Former Assistant Administrator for-Research -and- Development
Mel Myers*
Technical Assistant to AA for Research and Development
Richard E. Marland-f
Special Assistant to AA for Research and Development
227
-------
-6- ;
Office of Research and Development (Cont.)
George Simon+
Supervisory Health Scientist Administrator
Bernie McMahon+
Chief, Administrative Management Staff
i
i
Robert Edgar*
Chief, Planning iStaff
';
Robert l.ee+
Management Analyst
Denise Zwink:+
Health Scientist
Oeanie Loving*
Health Scientist
Robert E. McGaughy+
Senior lexicologist, Cancer Assessment Group
Health Effects Research Laboratory, Cincinnati, Ohio
Dr. R. John Garner*-*-
• ,. Di rector
Dr. James B. Lucas*
Deputy Director
Dr. Elmer V. Akin*
Chief, Viral Disease Group
Dr. Peter J. Bercz*
Chief, Chemical and Genetic Effects Group
Dr^- David A. Brashear*
Mi crobi ologist
Dr. Richard J. Bui 1* +
Chief, Toxicological Assessment Branch
..._;;. Mr...J.,K..:.8ur.k3r.d*,
Chief, Mechanical Group
Dr. Kirby I. Campbel 1*
Acting Chief,'Functional Pathology Branch
Dr. Kenneth P. Cantor*
Epidemiologist
228
-------
-7-
HERL, Cincinnati (Continued)
Dr. Norman A. Clark*
Director, Laboratory Studies Division
Mr. Emile W. Coleman*
Research Chemist;
Mr« Gunther F,. Craun*
Chiefs Epidemiology Branch
Dr. B.F. Daniel*
Genetic lexicologist
Mr. R.M. Danner*
Acting Chief, Biochemistry Group
Mr. T.H. Erickson*
Microbiologist
Mr. D.G. Greathouse*+
Chief, Chronic Diseases and Biostatistics Group
Dr. W.E. Grube*+ " ;:
-Acting Director, Program Operations Staff
Mr. A.E. Hammonds*
Computer Specialist
Mr,, W. Paul Heffernan*
Chief, Developmental Toxicology Group
Mr. R.G. Hinnsers*
Chief, Exposure Systems Branch
Mr. Walter Jakuborvyski*
Chief, Bacterial and Parasitic Disease Group
Dr. F.C. Kopfler*
Chief, Exposure Evaluation Branch
Dr. Norman Kowal*
Research Medical Officer
Mr. D.A. Laurie*
Physiologist-- -
Dr. R.D. Lingg*
Research Chemist
Mr. Edwin Lippy*
Chief, Outbreak Investigation Group
229
-------
-8-
HERL.CINCINNATI (Continued)
Mr. Myron Malanchuck*
Chief, Experimental Aero;metry Group
Mr. Lei and J. McCabe*
Director, Field Studies Division
Dr. R.G. Milton*
Chief, Organics Identification Group
Dr. Robert Mi day*
Medical Officer
Mr. G.E. Michaeil*
Environmental Health Scientist
Mr. R.G. Miller*
Chief, Tissue Analysis Group
Mr. James Milletjte*
Chief, Particulate Analysis Group
Dr. John G. O.rthoefer*+
Chief, Pathology Group
Mr. Herbert L. Pahren*
.Physical Science Advisor
Dr. W-.E. Pepelkp*
Chief, Physiology feroup
Dr. Michael Pereira*
Research Pharmacologist
Mr. Merrel Robinson*
Biologist
Dr. Frank W; Schaefer*
' Microbiologist
Ms. Cynthia Sonich*
Environmental Health Scientist
Dr. Robert W. Tuthill*
- Epidemiologist
Ms. Nancy S. Ulmer*
Research; Chemis't
Dr. Jean M. Wiest,er*
Research Physiologist
230
-------
-9-
HERL, Cincinnati^ Cont . )
Mr. P.P. William^*
Microbiologist:
Health Effects Research Laboratory ,; Marine Field Station
West Kingston, Rhode Island
Dr. Victor J. Cabelli*-*-
Director, Field Station
Dr. Morris Levinje*
Research Microbiologist
Dr. Alfred Dufouir*
Research Microbiologist
Dr. Paul Cohen*
Chairman, Microbiology Dept., University of Rhode Island
Health Effects Research Laboratory, Research Triangle Park, NC
Dr. F. Gordon Huet£r*+
Di rector .; .
Dr. Robert E. Lee*
Deputy Director
Dr. R.J.M. Horto'n*
Senior Research Adi/isor
Mr. Or in W. Stopin^ski*
Physical Scientist
Mr. James R. Smith*
Physical Scientist
Dr. ..QonaJd. K^,,,Hin.Kl,e*
Veterinari an ;
Dr., Thomas M. Wagner**
Acting Director, Program Operations Office
Ms. Ann H. Akland*
-Superv-isory,, R to gr!,am. Analyst
Ms. Margaret C. Mickelson*
Administrative Officer
Dr. William C. Neison*
Acting Chief, Statistics and Data Management Office
231
-------
-10-
HERL,RTP (Cent;.)
Dr. Victor Hasselblad*
Supervisory Mathematical Statistican
Dr. John P. Creason*
Supervisory Mathematical JStatistician
Dr. Daniel F. Cahill*
Director, Experimental Biology Division
Dr. Neil Chernoff*
Research Biologist
Dr. Lawrence Retter*
Research Pharmacologist
Dr. John W. Las key*
Supervisory Research Biologist
\
Dr. Joe Elder*
Chief, Neurobiology Branch
s
Dr. Carl G. Hayes*
Chief, Air Pollutants Branch
Dr. D.G. Gillette*
Economist ,
Dr. Willson B, Ri;ggan*
Research Health Scientist; (Statistics)
Dr. Dorothy Calaf;iore*
Epidemiologist
Dr. Robert S. Chapman*
Medical Officer (Research)
Dr. G.S. Wilkinson*
Epidemiologist
Dr. Gregg Pran:g*
Epidemiologist
Dr. Mi'chael D. Waters*
Chief, Biochemistry' Branch
Dr. Joel!en L. Huisingh*
Supervisory Research Chemist
Mr. Larry Claxton*
Biologist
232
-------
-11- ;
HERL.RTP (Cont)
Ms. Martha Brown*
Biologist i
Dr. Stephen Nesnow*
Supervisory Research Chemist
! '
Dr. William F. Durham*
Director, Environmental Toxicology Division
Dr. Ronald L. Baroin*
Physical Science Administrator
Mr. August Curley*
Chief, Toxic Effects Branch
i
.Or.. T.M. Scottl*
Medical Officer, Pathology
Dr. C.Y. Kawanishi*
Research Microbiologist
Dr. Jeffrey Charles*
Research Phafmacplogist/Toxicologist
Dr. Joseph Roy croft*
Pharmacologist
Dr. John H. Knelsbn*
Director, Clinical Studies Division
Dr. Ralph W. Stacy*
Research Health Scientist
Dr» Donald E. Gardner*
Chief, Biomedical Research Branch
Dr. "John" 0' Neil*
Re sea re h Physi ologi st
Mr. Jerome M. Kirtz*
Engineer
,Dr. Edward .
Microbiologi^st '
Dr. Mary Jane K. Sel grade*
:M i c r 6 b i 6 1 o g i s t ~ "
233
-------
-12-
HERL.RTP (Cont.)
Dr. George M. Goldstein*
Chief, Clinical Pathology Branch
Dr. Mirzda Peterson*
Research Microbiologist
Dr. E.D. Haak, Jr.*
Chief, Physiology 'Branch
i
Mr. Matthew Petrovick*
Research Biomedical Engineer
Dr. Vernon A. Benignus*
Research Psychologist
Dr. David A. Otto*
Research Psychologist
Dr. Brock T. Ketcham*
Medical Officer
Dr. Milan Hazucha*.
Medical Officer
Mr.-. Walter L. Crider*
Chief, Research Services Branch
Health Effects Research Laboratory, Field Station
Wenatchee, Washington"
Mr. Homer R. Wolf*+
Director and Research Entomologist
Dr. James E. Davis*
Deputy Chief and Biochemist
Dr. Donald C. _Staiff*
Research Chemist
Dr. Larry Butler*
Research Chemist
Environmental Research Laboratory, Narragansett, R.I.
Dr. Eric D. Schneider**
..._._ J3.j.rjELCtQ.r
Dr. Richard W. Latimer*
Director, Laboratory and Program, Operations Division
234
-------
-13-
ERL, Narragansett (Cont)
Dr. J. Prager*
Ecolegist
P. Yevich*
Research Biologist and Pathologist
Dr. P. Rogerson*
Chief, Analytical Chemistry Branch
Dr. G. Hoffman* .
Research Chemist
Dr. G. Zaroogian*
Research Chemist
Dr. G. Gardner*
Aquatic Biologist,
Dr. A.R. Malcolm*+
Research Chemist :
Dr. E. Jackim*
Re sea re h Ch e m i s t
Dr. G. Persch* '
Aquatic Biologist
Environmental Research Laboratory, Duluth, Minn.
Dr. J. David Yount*+
Deputy Di rector
Dr. Wil1iam A. Brungs*
Director, Office of Technical Assistance
Dr. Kenneth E. Biesinger*
Director, Office of Extramural, and.Interagency Programs
Ms. Evelyn P. Hunt*+
Chief, Research,Support :Section
Dr. Gary E. Glass*
Research Chemist
Dr. James M. McKim*
Chief, Physiological Effects of Poillutants Section
Mr. James H. Tucker*
Aquatic Biologist
235
-------
-14- |
ERL, Duluth (Corjt.)
Dr. Gil man D. Verth*
Research Chemist
Dr. William A. Spoor*
Aquatic Biologist
Mr. Charles. E. Stephan*
Environmental Scientist
Dr. Bernard R. Jones*
Director, Duluth Research Branch
Mr. Armond E. Lemke*
Ecologist .
Dr. Glenn M. Christiansen*
Research Chemist
Mr. Frank H. Pulglisi*
Chemist
Mr.. Douglas W. Kuehl*
Research Chemist
Mr..Richard E. Siefert*
Chief, Physical Pollutants and Methods Section
Dr. Philip M. Cook+
Research Chemist
Dr. Richard L. Anderson*
Research Entomologist
Mr. Anthony R. Carlson*
Aquatic Biologist
M"rV J'ohn H. McCormick*
Aquatic Biologist
i
Mr. John" I. Teasley*
Research Chemist
Mr. John G. Eaton*
Chief, Chemical Pollutants Section
Mr. Robert W. -Andrew*
Research Chemist
Mr. Leonard H. Mueller*
Research Chemist
236
-------
-15- |
ERL, Duluth (Cont.)
Mr,, Robert A. Drummond*
Aquatic Biologist
Dr. John E. Poldoski*
Research Chemist
Environmental Research Laboratory, Gulf Breeze, Fla.
Dr. T.W. Duke**
Di rector
Dr. T.T. Davis*i+
Deputy Director
Dr. N.L. Richards*
Associate Director for Extramural Activities
Dr. J.A. Couch*
Coordinator, Experimental Biology Team
Dr. W.P. Schoor*
Aquatics-Biologist '
Dr. J.I. Lowe*
Chief, Experimental; Environments Branch
Dr. D.R. Nimmo*
Research Ecologist
Dr. G.E. Walsh*
Research Ecologist
Mr. D.J. Hansen*
Aquatic Biologist
Mr. S.C. Shimmel*
Aquatic Biologist
Dr. N.R. Cooley*
Research Microbiologist
Dr. Richard Garner*
Research Chemist
Environmental Monitoring and Support Laboratory, Las Vegas
Dr. G.B. Morgan*
Director
Dr. R.E. Stanley*
Deputy Director
237
-------
-16-
EMSL, Las Vegas (Cont.)
Mr. W.E. Petrie+
Director, Office of Program Management and Support
Dr. J.A. Santolucito*+
Director, Monitoring Systems Research and Development Division
!
Dr. Pong Lern*
Researach Chemist
Dr. J.V. Beh'ar*
Director, Monitoring Systems Design and Analysis Staff
Dr. Robert. Papcher*
Medical Officer
Dr. E. Meier*
Methods Development and Analytical Support
Mr. A. Jarvis*
Chief, Quality Assurance Branch
Dr. G. Wiersma*
Chief, Pollutant Pathway Branch , - _•-.
Dr. G. Potter*
Chief, Exposure/Dose Assessment ..Branch
Dr. D. Smith*
Chief, Farm and Animal Investigation Branch
235
-------
THURSDAY, NOVEMBER 30, 1978
PART II
UNITED STATES
ENVIRONMENTAL
PROTECTION
AGENCY
s>EPA
REGULATORY AGENDA
239
-------
56158
NOTICES
[6560-01-M]
ENVIRONMENTAL PROTECTION
AGENCY
[FRL 983-5]
AGENDA OF REGULATIONS
AGENCY: Environmental Protection
Agency.
ACTION: Agenda of Regulations.
SUMMARY: Four times a year the
Agency publishes a summary of the
significant regulatory actions under
development to help assure that inter-
ested parties have an early opportuni-
ty to participate in shaping our regula-
tions. We call the summary our
Agenda of Regulations.
FOR FURTHER INFORMATION
CONTACT: For information about
any particular item on the Agenda
contact the individual identified as the
contact person for that item. For gen-
eral information about public partici-
pation in the regulatory process con-
tact:
Chris Kirtz, (PM-223), Standards
and Regulations Evaluation Divi-
sion, Environmental _ Protection
Agency, 401 M Street, SW Washing-
ton, D.C.20460.
SUPPLEMENTARY INFORMATION:
On March 23, 1978, President Carter
signed Executive Order 12044, Improv-
ing Government Regulations, which
directed all executive agencies to
adopt procedures to improve existing
and future regulations. One procedure
which the Order required all agencies
to adopt was the publication twice a
year of a list of significant regulations
which are under development or
review. The Order also directed that
the Agenda provide the following in-
formation about the potential regula-
tions:
• A brief description
• A citation of its statutory authori-
ty
• Its status
• The name and phone number of a
knowledgeable official
• Whether we will prepare a regula-
tory analysis due to the regulation's
potentially major economic conse-
quences
• Whether the listed item is an ex-
isting regulation which we are reeva-
luating
The Order also directed that the
Agenda provide the status of all items
listed on the previous Agenda.
EPA's previous Regulatory Agenda
was published April 6, 1978
COVERAGE
We have tried to list all significant
actions which are going through the
Agency's formal regulation develop-
ment process, but we may have inad-
vertently omitted a few. Appearance
or nonappearance in the Agenda car-
ries with it no legal significance.
Executive Ord< r 12044 gave general
guidelines on determining what regu-
lations were significant and which,
therefore, should be included on the
Agenda. It directed each agency to de-
vt lop specific criteria for identifying
significant regulations. We will de-
scribe cur criteria for determining sig-
nificant regulations in our final report
responding to the Executive Order. I
will be signing this report soon, and
you will be able to obtain copies of it
from Philip Schwartz (PM-223),
Washington, D.C., 20460.
The Agency's formal process of regu-
lation development starts when an As-
sistance Administrator sends a notice
form to the Administrator and other
senior management. This form notifies
all EPA offices that a regulation is
about to be prepared and allows these
offices to plan their participation.
Different events might trigger the
start of the Agency's formal regula-
tion development process. The most
common event is the passage of new
legislation. Other common triggers in-
clude new scientific studies; advances
in technology; petitions for rulemak-
ing sent in from outside EPA; judicial
documents such as court orders and
consent agreements; and simply, oper-
ating experience with a particular reg-
ulation which may suggest ways that
we can improve it.
EXPLANATION OF INFORMATION IN THE
AGENDA
The Agenda lists prospective regula-
tory actions authorized by the follow-
ing laws:
• the Clean Air Act (CAA)
• the Motor Vehicle Information
and Cost Savings Act (MVICSA)
• the Safe Drinking Water Act
(SDWA)
• the Noise Control Act (NCA)
• the Federal Insecticide, Fungicide,
and Rodenticide Act (FIFRA)
• the Atomic Energy Act (AEA)
• the Public Health Service Act
(PHSA)
• the Resource Conservation and
Recovery Act (RCRA)
• the Toxic Substances Control Act
(TSCA)
• the Federal Water Pollution Con-
trol Act as amended by the Clean
Water Act (CWA)
The first column of the Agenda pro-
vides the following information about
each regulation:
• A citation from the Code of Feder-
al Regulations
• A short title
• A citation of statutory authority
• A description, including whether
the item is an existing regulation
which we are reevaluating
If the regulation may have economic
consequences large enough to require
a regulatory analysis, an asterisk (*)
appears at the beginning of the entry.
The second column lists the date we
proposed a regulation in the FEDERAL
REGISTER or the month in which we
expect to propose it.
The third column lists the date we
published a final regulation or the
month in which we expect to publish
the final regulation.
The fourth column provides the
name, address, and phone number of
whom to contact for each regulation.
DOUGLAS M. COSTLE,
Administrator.
NOVEMBER 20, 1978.
MAJOR EPA REGULATIONS UNDER CONSIDERATION
Name and description of regulation
Proposal date in FEDERAL REGISTER Final date in FEDERAL REGISTER
Contact person and address
THE CLEAN AIR ACT
We are developing the following seven items under the au' hority of sees. 108 and 109 of the CAA which direct the Administrator to establish national Ambi-
ent Air Quality Standards (NAAQS). To write a NAAQS for an, pollutant, we first prepare a criteria document which contains the latest scientific knowledge on
the kind and extent of public health and welfare problems cau-.cd by the presence of the pollutant in the air. If we revise the criteria document, we may find it
necessary to also change the NAAQS.
A National Primary Ambient Air Quality Standard define;, the Maximum amount of an air pollutant which the Administrator of EPA determines is compati-
ble with an adequate margin of safetyto protect the public h< Ut.h. A National Secondary Ambient Air Quality Standard defines levels of air quality which the
Administrator Judges necessary to protect the public welfare from any known or anticipated adverse effects of a pollutant.
40 CFR 50 'Review of NAAQS for Photochemical Junt 22, H'78.
Oxidants. CAA 108. The proposed regulation
would change the existing primary, health-based
standard to 0.10 ppm for a 1-hour average from
the existing 0.08 ppm standard. The secondary,
welfare-based standard would remain at 0.08 ppm
for 1-hour average. The pollutant we control
would be changed from photochemical oxidants
to ozone, which is the principal measurable in-
gredient in photochemical oxidants.
December 1978..
Joe Padgett (MD-12).
Environmental Protection Agency,
Research Triangle Park, N.C.
27711, 919 541 5204, FTS 8-629-
5204.
FEDERAL REGISTER, VOL. 43, NO. 231—THURSDAY, NOVEMBER 30, 1978
240
-------
NOTICES
MAJOR EPA REGULATIONS UNDER CONSIDERATION—Continued
56159
Name and description of regulation
Proposal date In FEDERAL REGISTER Final date in FEDERAL REGISTER
Contact person and address
THE CLEAN AIR ACT
February 1980.
December 1980.
June 1979.
40 CFR 50 -NAAQS for Lead. CAA 108. EPA pro- Dec. 14. 1977 Oct. 5, 1978
posed an ambient lead standard of 1.5 micro-
grams per cubic meter averaged over 30 days.
Public reaction has been mixed. Federal agencies
and public interest groups support the proposal.
Industry argues that: (1) the health data ;ind
analyses do not support the standard, <2> large
parts of the secondary lead and foundry indus-
tries are Technically unable to comply, and <3)
plant closures for economic and technical rea-
sons will result from enforcement of the stand-
ard.
40 CFR 50 'Review 6{ NAAQS for Carbon Monux- September 1979
ide. CAA 108. The health basis for control of tins
pollutant will we reviewed. This requires prepa-
ration of an updated criteria document and anal-
ysis of whether or not NAAQS should be revise d.
40 CFR 50 'Rei'ieir of NAAQS lor Sulfur Oxniea. May 1980
CAA 108. A review of the health basis for control
of this pollutant will require ['reparation of an
updated criteria document and analysis of
whether or not NAAQS should be revised
40 CFR 50 *Rci-ieu- oj Long Term NAAQS for \i- January 1979...
trogcn Dioxide CAA 108. The NAAQS for nitro-
gen dioxide is undergoing review. ORD will com-
plete a revised criteria document by January
1979. Under the CAA amendments, the criteria
and the decision to revise the standard must ad-
dress both the long-term effects of NO2. and ef-
fects associated with other nitrogen species in
the air, particularly nitrates, and nitric acid aero-
sol.
40 CFR 50 'Rei'iew of NAAQS for Particulates. May 1980 December 1980
CAA 108. A review of the health basis for control
of this pollutant will require preparation of an
updated criteria document and analysis of
whether or not NAAQS should be revised.
40 CFR 50 * Development of Short Term NAAQS January 1979 June 1979
for Nitrogen Dioxide. CAA 109. The Clean Air
Act Amendments of 1977 require proposal and
promulgation of a 1-3 hour standard for NO2
unless EPA finds that such a standard is not nec-
essary to protect the public health.
We are developing performance standards to control emi.ssions from the following industries under sec. lll(b) of
Administrator develop New Source Performance Standards (NSPS) for stationary sources which significantly contribute
the best system of continuous emission reduction which has been adequately demonstrated. The standards would apply
which are modified after approval of the regulation.
40 CFR 60 'NSPS—Fossil Fuel Steam Generators Sept. 19, 1!*78 March 1979
(.Revision). CAA 111. Revised standards are being
proposed for utility boilers for control of KO2,
NOx and participates. The revised NSPS will
apply to any fossil-fueled utility boiler with a
heat Input of 25© million Btu/hour or greater.
The NSPS will require a percent removal of
sulfur dioxide and will include an emission ceil-
ing and an emission flow.
40 CFR 60 NSPS—Petroleum Liquid Storage Ves- May 18, 1978 do
teU. CAA 111. This is a revision of 1974 NSPS.
The revised standard will propose the use of
double seals rather than single seals on floating
roofs. The standard, as currently being devel-
oped, will essentially eliminate one of two types
of seals currently in use.
40 CFR 60 NSPS—Glass Manufacturing. CAA February 1979....
111. This regulation will address the problem of
emissions from new glass manufacturing fur-
naces. The Governor of New Jersey requested
that EPA develop national standards.
40 CFR 60 NSPS—Internal Combustion Engines.
CAA 111. These regulations will require the ap-
plication of best demonstrated control technol-
ogy to control emissions from stationary internal
combustion engines. It will also require States to
act under sec. lll(d) to regulate these com-
pounds from existing sources.
40 CFR 60 NSPS—Sulfur Recovery in Natural July 1979. May 1980
Gas Fields. CAA 111. This regulation will control
emissions of total reduced sulfur compounds.
40 CFR 60 NSPS—Non-MetaUic Minerals. CAA January 1979.._ December 1979.
111. Particulate emissions from quarrying oper-
ations and related facilities will tie controlled.
40 CFR 60 NSPS—Organic Solvent Metal Clean-
ing. CAA 111. This rule will control evaporative
emissions from metal cleaning and degreasing
operations.
Do.
Do.
the CAA. This section requires that the
to air pollution. The NSPS are based on
to both new sources and existing sources
Don Goodwin (MD 13).
Environmental Protection Agency.
Research Triangle Park, N.C
27711, 919-541-5271, FTS 8 629
5271.
December 1979
December 1978 .. do
March 1979 January 1980.
Do.
Do.
Do.
Do.
Do.
Do.
FEDERAL REGISTER, VOL 43, NO. 231—THURSDAY, NOVEMBER 30, 1978
241
-------
56160
NOTICES
MAJOR EPA REGULATIONS UNDER CONSIDERATION—Continued
Name and description of regulation
Proposal date in FEDERAL REGISTER Final date4n FEDERAL REGISTER
Contact person and address
THE CLEAN AIR ACT
Do.
40 CFR 60 NSPS—Surface Coating Operations for February 1979 cio
Auto Assembly Plants. CAA 111. Evaporative
emissions from coating operations in the auto
and light trurk industry will be controlled.
40 CFR 60 NSPS-Synthetic Organic Chemical March 1979 do
manufacturing. CAA 111. Selection of a degree
of control of emission from manufacture of over
100 major organic chemicals is to be made. A
series of standards will be proposed.
40 CF'R 60 NSPS-Can Coating. CAA 111. This November 1979 September 1980.
regulation will establish emission standards for
volatile organic emissions from can coating oper
ations.
40 CFR 60 NSPS—Pressure Sensitive Tapes and Januarv 1980 November 1980 ..
Labels Coating. CAA 111. This regulation will en
tablish emission standards for volatile organic
emissions from pressure sensitive tapes and label
operations.
40 CFR 60 NSPS—Metal Furniture Surface Coat- December 1978 December 1979..
ing. CAA 111. This regulation will establish emis-
sion standards for volatile organic emissions
from metal furniture operations.
40 CFR 60 NSPS-Lead Battery Manufacturing. April 1979 February 1980...,
CAA 111. This regulation will establish emission
standards for lead and sulfuric acid mist emis-
sions from lead battery manufacturing facilities.
The action on H2S04 will key the requirement
that States regulate existing sources under sec.
lll(d).
40 CFR 60 NSPS—Gas Turbines. CAA 111 This Oct. 3, 1977
regulation will establish limitations on oxide of
nitrogen emissions from stationary gas turbines.
40 CFR 60 NSPS-lndustnal Boilers. CAA 111. October 1980
This regulation will control the emissions of par-
ticulates, NOx and SO2.
40 CFR 60 NSPS-Phosphate Rock. CAA 111. This May 1979 March 1980.
regulation will control the emission of particu-
lates.
40 CFR 60 Aluminum Plant Flouride Control— Januarv 1979
Existing Plants. CAA lll(d). These are guide-
lines for State control of flouride emissions from
existing aluminum plants.
40 CFR 60 Guidelines for Existing Krajt Pulp Feb. 23 1978
Mills. CAA lll(d). These are guidelines to con-
trol sulfur (odors) from existing Kraft pulp mills
will allow States flexibility in establishing con-
trols.
40 CFR 60 List of New Source Performance Aug. 31.1978 May 1979.
Standards. CAA lll(f). The 1977 Clean Air Act
requires the Administrator to list the categories
of major stationary sources that are not already
controlled by NSPS. He must then issue stand-
ards for these categories within 4 years.
We are developing emission standards for hazardous air pollutants under sec. 112 of the CAA. This section requires that the Administrator develop National
Emission Standards for Hazardous Air Pollutants iNESHAPS) for emissions which cause or contribute to air pollution which results in an increase in mortality, or
an increase in serious or incapacitating illness. The standards would apply to both new sources and existing sources.
February 1979.
August 1981
November 1979 .
January 1979.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
September 1979 July 1980.
40 CFR 61 NESHAPS: Asbestos-Iron Ore Benefi-
ciation. CAA 112. This regulation would estab-
lish limits on asbestos emissions from iron ore
beneficiation facilities.
40 CFR 61 NESHAPS: Vinyl Chloride Amend- June 7,1977 Indeterminate...
ments. CAA 112. The proposed regulations have
called for increased-control of existing sources,
stringent control of new sources, and a zero emis-
sion goal.
40 CFR 61 NESHAPS: Handling and Storage. August 1979 June 1980
CAA 112. This regulation would control the han-
dling and storage of benzene and benzene-rich
liquids.
40 CFR 61 NESHAPS: Gasoline Distribution Sys- Indeterminate Indeterminate...
terns. CAA 112. This regulation would control
benzene emissions from major marketing sources
such as bulk terminals, bulk plants, and service
stations.
40CFR 61 NESHAPS—Refinery Sources. CAA 112. September 1979 November 1980 .
This regulation would control the emission of
benzene from point sources as well as from fugi-
tive sources (pumps, valves, etc.) and waste dis-
posal.
40 CFR 61 NESHAPS-Maleic Anhydride. CAA 111!. January 1979 November 1979 .
This regulation would control the emission of
benzene in the manufacture of maleic anhydride.
40 CFR 61 NESHAPS—Ethyl Benzene. CAA 112. March 1979 January 1980
This regulation would control the emission of
benzene in the manufacture of ethyl benzene.
Don Goodwin (MD-13).
Environmental Protection Agency,
Research Triangle Park, N.C.
27711, 919-541-5271, FTS 8-629-
5271.
Do.
Do.
Do.
Do.
Do.
FEDERAL REGISTER, VOL 43, NO. 231—THURSDAY, NOVEMBER 30, 1978
242
-------
NOTICES
MAJOR EPA REGULATIONS UNDER CONSIDERATION—Continued
56161
Name and description of regulation
Proposal date In FEDERAL REGISTER Final date In FEDERAL REGISTER
Contact person and address
THECLXAX AIR ACT
June 1979
May 1980
April 1980
March 1981
40 CFR 61 NESHAPS-Styrene. CAA 112. This
regulation would control the emission of benzene
In the manufacture of styrene.
40 CFR 61 NESHAPS: Asbestos Released from
Crushed Stone. CAA 112. Use of crushed serpen-
tine rock for roadway surfacing may release sig-
nificant quantities of. asbestos, A monitoring pro-
gram is under way and results indicate standards
will be proposed.
40 CFR 61 NESHAPS Coke Oven Emission- December 1978 September 1979.
Charging Operations. CAA 112. The regulation
would define coke oven emissions as a hazardous
air pollutant. Charging operations would be reg-
ulated first. Regulations on top side leaks would
follow.
40 CFR 61 NESHAPS. Arsenic. CAA 112. A health December 1979
risk assessment is being conducted. If i is deter-
mined that arsenic emissions (primarily from
copper smelters) are a hazardous air pollutant,
then emission standards would be proposed.
40 CFR 57 Primary Nonferrous Smelter Orders.
CAA 119. These regulations will establish the
substantive requirements of initial primary non-
ferrous smelter orders (NSO's) and the proce-
dures to be used in issuing them. NSO's will
allow certain copper, lead, and zinc smelters to
delay compliance with the requirements for con-
stant control of sulfur dioxide emissions and let
them use tall stacks and supplementary control
systems to meet ambient standards.
40 CFR 56 Npncompliance Penalties. CAA 120,
EPA is required to establish a penalty program
to start collecting money from polluters after
mid-1979 in an amount equal to the money the
polluter saves by failing to obey the law.
40 CFR 51 Tall Stack Regulation. CAA 123. The November 1978 April 1979 .
regulations will specify what height stacks may
be given credit for dispersion under State imple-
mentation plans.
December 1978.
April 1979 .
,.do Undetermined
40 CFR 51.240 Regulations Providing for State/ May 18, 1978
Local Consultation. CAA 121. The regulations
will ask the States to provide a satisfactory proc-
ess of consultation with local governments, elect-
ed officials, and Federal land managers. The reg-
ulations will also require the States to choose a
lead planning organization to coordinate the
State Implementation Plan revisions for oxidants
(smog) and carbon monoxide.
^979 Listing of Radioactive Pollutants. CAA August 1980 Undetermined.
122. Determine whether radioactive pollutants
shall be classified as 108, 111. or 112 pollutants or
none of these categories.
40 CFR 51 Emission Offset Policy Regulations. Dec. 21, 1976 „ November 1978
CAA 129. These regulations address the issue of
whether and to what extent the national ambi-
ent air quality standards established under CAA
restrict or prohibit growth of major new or ex-
panded air pollution sources. These proposed re-
visions reflect the public comments (including
four public hearings on the December 21 ruling
and the changes required by CAA Amendments
of 1977).
40 CFR 51 and 52 Prevention of Significant Dele- December 1979 October 1980...
rioratton (PSD). Set II. CAA 166. These regula-
tions will insure that areas which are in compli-
ance with hydrocarbon, carbon monoxide, photo-
chemical oxidant, and nitrogen oxide standards
will remain in compliance.
Visibility Protection. CAA 167(a). EPA is re- October 1979 August 1980
quired to prepare a report to Congress and guide-
lines which require SIP's to address visibility
problems.
40 CFR 85 Requirement* to Build Demonstration
Cars Meeting 0.4 Ornm/Mile NOx Standard. CAA
202. All manufacturers with a least a 0.5 pet
share of the U.S. passenger car market will have
to build research vehicles which meet the 0.4
grams nitrogen dioxide per mile research objec-
tive. This regulation will be published in interim-
final form.
40 CFR 86 Light-Duty Diesel Particulate Stand-
ards. CAA 202. EPA is required to set paniculate
standards for mobile sources starting In 1981.
The regulation will contain 1981 standards and
more stringent standards for 1983 and later
model years.
December 1978 July 1979.
..do _ July 1979.
Do.
Do.
Do.
Joe Padgett (MD 12).
Environmental Protection Agency,
Research Triangle Park. N.C.
27711 919-541-5204, FTS 8-629
5204.
Judith Larsen (EN 341).
Environmental Protection Agency,
Washington, D.C. 20460, 202-755-
2583.
Bob Homiak (EN 341).
Environmental Protection Agency,
Washington. D.C. 20460, 202-755-
2542.
Dick Rhoads(MD 15).
Environmental Protection Agency,
Research Triangle Park, N.C.
27711, 919-541 5251. FTS 8-629-
5251.
John Hidinger (AW-445).
Environmental Protection Agency,
Washington. D.C. 20460, 202-755-
0481.
William A. Mills (AW-460).
Environmental Protection Agency,
Washington. D.C. 20460. 703-557-
0704.
Kent Berry (MD-11).
Environmental Protection Agency,
Research Triangle Park. N.C.
27711, 919-541 5343, FTS 8-629-
5343.
Dick Rhoads (MD-15).
Environmental Protection Agency,
Research Triangle Park. N.C.
27711, 919-541-5261, FTS 8-629-
5251.
Joe Padgett.
Environmental Protection Agency.
Research Triangle Park, N.C.
27711. 919-541-5204, FTS 8-629-
5204.
Karl Hellman.
Emission Control Technology Divi-
sion, Environmental Protection
Agency. 2566 Plymouth Rd.. Ann
Arbor, Mich. 48105, 313-668-4246.
Merrill Korth.
Emission Control Technology Divi-
sion, Environmental Protection
Agency, 2565 Plymouth Rd., Ann
Arbor,.Mich. 48105. 313-668-4299.
FEDERAL REGISTER, VOL 43, NO. 231—THURSDAY, NOVEMBER 30, 1978
243
-------
56162
NOTICES
MAJOR EPA REGULATIONS UNDER CONSIDERATION—Continued
Proposal date In FEDERAL REGISTER Final date in FEDERAL REGISTER
Contact penon and address
THE CLEAR AIR ACT
40 CFR 86 Heavy Duty Diesel Particulate Stand- December 1980 August 1080
arris CAA 202. Although required by CAA for
1981 models, there is no test procedure available
I hat can be used a.s the basis for a standard. A
1983 model year is targeted.
40 CFR 86 Text Procedures for Measuring Heavy December 1978 August 1980
Duly Evaporative Emissions. CAA 202(a> The
Clean Air Act requires that a test procedure be
promulgated which will require measurement of
evaporative emission from the vehicles as a
whole. EPA will promulgate test procedure and
standards
40 CKR 8ti Heavy Duty Evaporative Emission January 1979 August, 1980 ,
Standards. CAA 202(3>. The CAA requires EPA to establish
emission standards for engines for heavy-duty ve-
hicles over 8,500 pounds. Standards for HC and
CO are a 90 percent reduction from baseline
emissions for 1983 model year. EPA is in the
process of developing a new test procedure for
measuring exhaust emissions and measurements
of baseline emissions.
40 CFR 86 NOx Emission Standard for Heavy December 1919 September 1980
Ditty Vehicles (Over 8,500 Pound* >. CAA
202ia)(3). The CAA requires EPA to establish
emission standards for heavy-duty vehicles (over
6.000 IDS. GVWR). A 75 percent reduction for
NOx beginning with 1985 model year. EPA is in
the process of developing a new te«t procedure
for measuring exhaust emissions and must then
measure baseline emissions.
Fill Pipe Standards. CAA 202(a)<5>. At such September 1979 June 1980
time as phase II vapor recovery regulations are
promulgated. EPA is required to set standards
for vehicle refueling orifices and associated parts
of the fuel system to provide effective connection
between the fill pipe and vapor recovery refuel-
ing nozzles. The effective model is to be deter-
mined on the basis of lead time required for
design and production of the required systems.
The type of fill pipe needed depends of whether
phase II or on-board HC control is selected by
EPA.
On-Board Hydrocarbon Technology. CAA September 1979 June 1980
202<6). Under this section EPA is required to
determine whether onboard HC controls are fea-
sible and more desirable than Phase II: Vapor
Recovery, taking into consideration such factors
as fuel economy, costs, adminlnstrative burdens,
equitable distribution of costs and safety. If
found feasible and desirable, onboard HC control
standards are to be set by EPA, with such lead
time as is needed for implementation. In issuing
such regulations, EPA is required to consult with
the Department of Transportation regarding the
safety of the controls.
40 CFR 86 Interim High Altitude Requirements. December 1978 August 1979
CAA 202. The regulations will
identify the components and specifications that
are a required part of motor vehicle certification;
the parameters of allowable deviation of parts;
and the specifications for the certification tests.
Do.
Mike Lelferman.
Environmental Protection Agency,
Ann Arbor, Mich. 48105. 313-668-
4271.
Do.
William Houtmann.
Environmental Protection Agency,
Ann Arbor, Mich. 48105, 313-668-
4272.
Chet France.
Environmental Protection Agency,
Ann Arbor, Mich. 48106, 313-668-
4338.
Do.
Ernie Rosenberg (AW-455).
Environmental Protection Agency,
Washington, D.C. 20460. 202-756-
0596.
Paul Stolpman (AW-443).
Environmental Protection Agency.
Washington. D.C. 20460. 202-426-
2484.
William Houtmann.
Environmental Protection Agency,
Ann Arbor, Mich. 48105. 313-668-
4272.
Tom Preston (EN-340).
Environmental Protection Agency,
Washington, D.C. 2O460, 202-755-
0944.
Do.
FEDERAL REGISTER, VOL. 43, NO. 231-THURSDAY, NOV-MBER 30, 1978
244
-------
NOTICES
MAJOR EPA REGULATIONS UNDER CONSIDERATION—Continued
56163
Name and description of regulation
Proposal date in FEDERAL RBOISTKK Final date in FKDOIAL Rxcisrn
Contact person and address
THE CLJEAN AIR ACT
40 CFR 86 Selective Enforcement Auditing of Mo- Holding
torcycles. CAA 206(b). The regulation will estab-
lish a program for testing motorcycles at the as-
sembly line to assure compliance with emission
standards.
40 CFR 86 Selective Enforcement Auditing of December 1978 February 1979....
Heavy Duty Engines and Vehicles. CAA 206(b>.
The regulation will establish a program for test-
ing heavy duty engines and vehicles at the as-
sembly line to assure compliance with emission
standards.
40 CFR 86 Engine Parameter Adjustment Regula- Oct. 21. 1977 November 1978..
tions. CAA 206
-------
NOTICES
MAJOR EPA REGULATIONS UNDER CONSIDERATION—Continued
Name MM! (4). These
regulations will authorize States to establish reg-
ulatory programs for the discharge of dredge and
fill material to supplement State 404 permit pro-
grams.
40 CFR 233 Modification of Secondary Treatment Apr. 25, 1978.
Requirements for Marine Dischargers. CWA
301(h). The 1977 amendments of the Clean water
Act allow EPA to modify the treatment require-
ments for existing ocean dischargers from Pub-
licly Own Treatment Works (POTW's) in regard
to the required degree of removal of Biological
Oxygen Demand (BOD), Total Suspended Solids
(TSS), and pH. Applicants are required to meet
eight specific 301(h) criteria in addition to any
other applicable criteria of the Act. The receipt
of modification would not relieve a POTW from
compliance with performance standards which
EPA will later publish to reflect Best Practicable
Wastewater Treatment Technology (BPWTT).
This rule establishes the criteria which EPA will
apply and the procedures it will follow in its eval-
uation of application for a modification.
40 CFR 124 Extension of Pollution Control Dead-
lines for Publicly Owned Treatment Works and
Other Point Sources Planning to Discharge to
Those Publicly Own Treatment Works. CWA
30KJ). This regulation establishes criteria which
EPA and NPDES States will use in reviewing re-
quests for 301(i) extensions from the July 1.
1977, treatment requirements.
40 CFR 125 Requirements for Application for 301 Sept. 13 1978, interim final
(c) and (g) Variances. CWA 30KJM1XB). These
regulations require discharges desiring 301 (c)
and (g) variances to file Initial applications by
Sept. 25, 1978, or 270 days after promulgation of
BAT limitations whichever is later .
Effluent guidelines representing best available treatment technology, new source performance standards, and pretreatment standards are being developed
for the following industries to comply with the Act and a court order mandating control of certain toxic substances in industrial effluents. CWA 301 304 306 and
307.
May 16, 1978, interim final.
Will be incorporated into NPDES
program regulations 40 CFR 122
to 125.
January 1979, will be Incorporated
into NPDES program
regulations 40 CFR 122 to 125.
Ed Kramer (EN-336).
Environmental Protection Agency,
Washington. D.C. 20460, 203-765-
0750.
Scott Sleslnger (EN-336).
Environmental Protection Agency.
Washington, D.C. 20460, 202-755
0750.
40 CFR 420 Iron and Steel Manufacturing.
November 1979 .
May 1980
40 CFR 435 Petroleum Refining March 1979.
Ernst Hall (WH-552).
Environmental Protection Agency.
Washington, D.C. 30460. 203-426
2576.
October 1979 Robert Dellinger (WH-552).
Environmental Protection Agency,
Washington. D.C. 20460. 202-426-
3497
40 CFR 429 Timber Products Processing May 1979 December 1979 John Riley (WH-652).
Environmental Potection Agency.
Washington. D.C. 30460. 202-426-
6664
40 CFR 423 'Steam Electric Power Plants do do John Lum (WH-552).
Environmental Protection Agency,
Washington. D.C. 20460 202-426-
40 CFR 436 Leather Tanning and Finishing January 1979 August 1979 William Sonnett (WH-552).
Environmental Protection Agency.
Washington, D.C. 20460 202-426-
3440.
40 CFR 421 Nonjerrous Metals Manufacturing August 1979.... March 1980 Patricia Williams (WH-662).
Environmental Protection Agency,
Washington, D.C. 20440, 302-424-
2684.
PEDEftAL UeiSTER, VOL 43, NO. 231—THUtSDAY, NOVEMBBI 30, 1978
246
-------
NOTICES 56165
MAJOR EPA REGULATIONS UNDER CONSIDERATION—Continued
Name and description of regulation Proposal date In J-'EDERAL REGISTER Final date In FEDERAL REGISTER Contact person and address
THE CLEAN WATEH ACT
40 CFR 46 Paint and Ink Formulation September 1979 April 1980 Richard Gigger iWH 55'2V
Environmental Protection Agency,
Washington, DC, 20460. 202 426
2583.
40 CFR 448 Printing and Publishing Services November 1979 June 1980 Do.
40 CFR 440 Ore Mining and Dressing do July 1980 Gail Coiui (WH 586).
Environmental Protection Agency,
Washington, D.C" 20460, 202 426
2503.
40 CFR 434 Coal Mining Dec. 'mb.'i 1979 JKmelPHO William Tellinni ' V.'H 586 i.
Em iroiimer.lnl Protection Agency,
Washington. DC 20460. 202 426
2726.
40CFR414 Organic Chenm-alii M(inufiu'h.,-i AuguM 19riO Paul Kan nUioi.i < VA H :>f>2>.
Environmental Pi »;. <•! ion Aj.:encv.
Washington. !.)(' 'JuUiu. 2'.J2 42ti
2497
40CFK415 Inorganic rhciniciili, Manujictun-!g.. Septcmt. 1"79 A;>',M!»HU Wult IT H ur.l . WH f>;,2>
Environmi-iil.U 1 , .>teet.".>n AM-IK \ .
W;u lnrii'.'on I) C LV-i'iO 20. 426
2724
40 CFR 410 Textile Mills V:i\ l'.»7<> ... D.-cii-ihet l'.'7<> J:un< - i in! ii;;. i XV i! 552 >
t.in ircMirn. iil-il P; ol ••<•',.>n Agency,
vV;Lshit!i '•.;;. !U .:0460. 202-426-
2.r).:i4
40CKR416 Plastics and SvnU;i-ti<-Material January 11)80.... August 1980 Paul KiUin-i.liu.iKi • WH 552).
EIH iroiimentiU Protection Agency,
Washington, D.C'. 20460. 202-426-
2497.
40 CFR 430 Pulp and Paper February 1980 do Bob De'.llnger (WH 552).
Environmental Protection Agency.
Washington, D.C 20460. 202 426
2554.
40 CFR 428 Rubber Processing June 1979 January 1980 Do.
40CFR417 Soap and Detergents Manufacturing.. July 1980 July 1981 Sammy Ng (WH 586).
Environmental Protection Agency.
Washington. D.C. 20460, 202 426
2503.
40 CFR 444 Auto and Other Laundries Decemhei 1979 July 1980 Richard GigRer'WH 552).
Environmental Protection Agency.
Washington, D.C. 20460. 202 426-
2583.
40 CFR 456 Miscellaneous Chemicals—Adhesivcs February 1980 August 1980 F.I wood Forsht (WH 552).
and Sealants. Environmental Protection Agency,
Washington. D.C. 20460. 202 426
2707.
40 CFR 457 Miscellaneous Chemicals--Explosives December 1979 July 1980 Elwood Martin (WH 552).
Manufacturing. Environmental Protection Agency.
Washington. DC. 20460, 202 426
2440.
40 CFR 454 Miscellaneous Chemicals—Gum and August 1979 March 1980 Richard Williams (WH 552).
Wood. Environmental Protection Agency,
Washington, DC. 20460. 202 426
2555.
40 CFR 455 Miscellaneous Chemicals—Pesticides.. March 1980 October 1.980 George Jett (WH 552).
Environmental Protection Agency.
Washington, D.C. 20460. 202 426
2497.
40 CPR 439 Miscellaneous Chemicals—Pharma- Decembci 1979 July 1980 Joe Vitalis (WH 552).
ceuticalt. Environmental Protection Agency.
Washington. D.C. 20460. 202 426
2497.
40 CFR 413 Electroplating March 1980 October 1980 Maurice Owens (WH 586)
Environmental Protection Agency.
Washington. DC. 20460. 202 755
1331.
40 CFR 459 Machinery and Mechanical Prod- February 1980 August 1980 Ernst Hall (WH 552).
ucts—Photographic Equipment and Supplies. Environmental Protection Agency,
Washington. DC. 20460. 202 426
2576.
40 CFR 433 Machinery and Mechanical Prod- August 1980 March 1981 Do.
ucts—Mechanical Products.
40 CPR 469 Machinery and Mechanical Prod- March 1980 October 1980 Do.
ucts—Electrical and Electronic Components.
40 CFR 464 Machinery and Mechanical Prod- October 1979 May 1980 Do.
uctt—Foundry Off ration*.
40 CPR 468 Machinery and Mechanical Prod- April 1980 November 1980 Do.
ucts—Copper and Copper Alloy Products.
40 CFR 461 Machinery and Mechanical Prod- March 1980 October 1980 Do.
uctt—Battery Manufacturing.
40 CFR 465 Machinery and Mechanical Prod- August 1979 March 1980 Do.
ucts—Coil Coating.
40 CPR 443 Machinery and Mechanical Prod- October 1980 May 1981 Do.
*ct*—PUutics Proceuing.
40 CFR 466 Machinery and Mechanical Prod- October 1979 May 1980 Ernst Hall (WH-552).
ucts—Porcelain Enamel. Environmental Protection Agency,
Washington. DC. 20460. 202-426-
2576.
40 CFR 467 Machinery and Mechanical Prod- March 1980 October 1980 Do.
ucts—Aluminum Forming.
FEDERAL ROISTER, VOL tt, MO, HI—THURSDAY, NOVEMKt 30, 1978
247
-------
56166
NOTICES
MAJOR EPA REGULATIONS UNDER CONSIDERATION—Continued
Name and description of regulation
Proposal date in FEDERAL REGISTER Final date in FEDERAL REGISTER
Contact person and address
THI CLEAN WATER ACT
Jan. 6, 1978 May 23, 1978.
January 1979 Will be incorporated into NPDES
program regulations 40 CFR 122
to 125.
March 1979 March 1980.
(29 pollutants) March 1979 September 1979 .
(36 pollutants) July 1979 December 1979...
Aug. 23, 1978 April 1979.
Sept. 1, 1978.
40 CFR 124 and 125 Veto Modification. CWA
301(b)(a), 304(1), 307(a), 402(b), 501(a). This reg-
ulation revises existing regulations to conform to
the requirements in the NRDC versus Train Con-
sent Decree June 8, 1976 and to clarify the proce-
dures under which EPA will exercise its power to
object to (veto) State issued NPDES permits.
40 CFR 125 Substantive Criteria for 30KO and
(«r) Variances from BAT Requirements. CWA 301
(c) and (g). This criteria will establish informa-
tion necessary for assessment of economic and
environmental variance requests.
40 CFR 130.17 'Revision of Water Quality Stand-
ards Regulation (Part 130.17). CWA 303. This
regulation will amend the existing regulation
covering State Water Quality Standards to estab-
lish requirements regarding States adopting
standards for toxic pollutants when EPA has
issued national ambient water quality criteria for
those pollutants. One effect of this amendment
will be that dischargers (both municipal and in-
dustrial) may have to install treatment technol-
ogy beyond that required by Best Available
Wastewater Treatment Technology (BPWTT) or
Best Available Technology (BAT) guidelines.
40 CFR Quality Criteria for Water: Volume 11.
CWA 304(a). Ambient water quality criteria will
be established for 65 pollutants.
40 CFR 400 to 469 Secondary Industry Review
CWA 304(b). This regulation will provide for pro-
mulgated of Best Practicable Conventional Pol
lutant Control Technology (BTC) for certain
subcategories of the "secondary industries" in-
dustries not covered by the NRDC Settlement
Agreement. For other subcategories, Best Availa-
ble Technology (BAT) limits will be suspended.
The methodology that will be used for BCT for
secondary industries will also be applied to BCT
for primary industries at the time that BAT reg-
ulations are established.
40 CFR 125 Criteria and Standards for Imposing
Best Management Practices for Ancillary Indus-
trial Activities. CWA 304(e). This regulation will
indicate how "best management practices" for
on-site industrial activities may be imposed in
NPDES permits to prevent release of toxic and
hazardous pollutants to surface waters.
General Pretreatment Regulations for Existing and
New Sources of Pollution. CWA 307
-------
NOTICES
MAJOR EPA REGULATIONS UNDER CONSIDERATION—Continued
56167
•nd dcau'lptton of regulation
Proposal date in FEDBRAL Rcoismt Fmal date In PKDKXAL Rmsra
Contact person and address
Tm CUBAN WATTS ACT
(Federal Water Pollution Control Art as amended by the dean Water Act Amendment* of 1*77)
4* CPR 231 Ocean Discharge Criteria. CWA
403(c). These guidelines pertain to discharges to
the ocean. They are based on prevention of envi-
ronmental degradation of waters of the territori-
al seas, the contiguous zone, and the oceans.
Both Industrial and municipal dischargers would
have to meet these criteria
40 CFR 230 Guidelines to Protect the Aquatic En-
vironment, Including Wetlands, From the Dis-
charge of Dredged or Fill Material CWA
404(4xB) program. Failure to
comply with these guidelines justifies denial of
permit applications and return of State permit
programs to the Corps of Engineers. Sept. 5,
1975, interim-final guidelines are being revised
and expanded by this effort
40 CFR 123 Procedural Regulations Concerning
State Qualifications lor Assuming the Section
404 Permit Program. CWA 404(g), Certain re-
quirements that musi be met for States to
assume permitting authority under sec. 404
-------
56168
Name and description of regulation
Proposal date In FEDERAL REGISTER Final date in FEBEKAL Rmsm
Contact person and address
THE NOISE CONTROL ACT
40 CFR 205 Light Duty Motor Vehicles. NCA 5. Work plan unuer development William Roper (AW-490).
This action will result in a decision regarding Environmental Protection Agency,
whether or not light duty vehicles are or are not Washington, D.C. 20480. 703-567-
a major noise source. If they are found to be, 7747.
then resulting noise emission and/or noise label-
ing standards will be prepared.
40 CFR 205 Buses. NCA 5/6. This regulation will Sept. 12, 1977 June 1979 Do.
set noise emission standards for new inter-State,
inner-city, and schoolbuses.
40 CFR 204 Truck-mounted Solid Waste Compac- Aug. 26, 1977 June 1979 Kenneth Felth (AW-4«0).
tor. NCA 5/6. The regulations sets noise emission Environmental Protection Agency,
standards for solid waste compactors. Washington, D.C. 20480, 703-557-
2710.
GFR 206, 207 Lawnmowers. NCA 5/6. The regulia- October 1979 October 1980 Henry Thomas (AW-490).
tion sets noise emission standards for new lawn- • Environmental Protection Agency,
mowers Washington, D.C. 20460, 703-557-
7743.
40 CFR 204 Pavement Breakers and Rock Drills. June 1979 June 1980 Kenneth Feith (AW—490)
NCA 5/6. The regulation sets noise emission Environmental Protection Agency,
standards for new pavement breakers and rock Washington, D.C. 20460, 703-557-
drills. 2710.
40 CFR 204 Truck Transported Refrigeration Developmental work halted ., Do.
Units. NCA 5/6. The regulation sets noise emis- pending analysis of regulatory
sion standards for new truck transport refrigera- alternatives.
tion units.
40 CFR 204 Wheel and Crawler Tractors. NCA 5/ July 11, 1977 June 1979 Henry Thomas (AW-490).
6. The regulation sets a noise emission standard Environmental Protection Agency.
for new wheel and crawler tractors. Washington, D.C. 20460, 703-557-
7743.
40 CFR 205 Motorcycles. NCA 5/6. This regula- Feb. 15, 1978 October 1979 William Roper (AW-490).
tion sets noise emission standards for motorcy- Environmental Protection Agency,
cles and replacement exhaust systems. Washington, D.C. 20460, 703-557-
7747.
40 CFR 211 Labeling: Hearing Protectors. NCA 8. June 22, 1977 January 1979 Henry Thomas (AW-490).
The regulation requires the labeling of hearing Environmental Protection Agency,
protectors. Washington, D.C. 20460, 703-557-
7743.
40 CFR 211 Labeling: General. NCA 8. The regu- do do Do.
lation establishes general labeling provisions.
40 CFR 210 Administrative Hearing Procedures. Aug. 3, 1978 December 1978 Jim Kerr (EN-387).
NCA 11. These procedures will apply to hearings Environmental Protection Agency,
for the issuance of remedial orders under sec. Washington, D.C. 20480 703-557-
ll(d) of the Act. As mandated, these are adjudi- 7410.
calory hearings under the Administrative Proce-
dure Act, 5 U.S.C. 554.
40 CFR 203 Low Noise Emission Products. NCA May 27, 1977 May 1979 Henry Thomas (AW-490)
15. This regulation allows a determination of Environmental Protection Agency,
when a product is a low noise emission product Washington, D.C. 20460 703-557-
and whether it is suitable for special considera- 7743.
tion in Federal purchasing.
40 CFR 205 Interstate Rail Carriers. NCA 17. December 1978 February 1979 William Roper (AW-490)
This regulation sets noise emission standards for Environmental Protection Agency
railroad ••facilities." EPA has prepared this regu- Washington, D.C. 20460 703-557-
lation as a result of a successful lawsuit brought 7747.
by the Association of American Railroads which
said EPA's regulations setting noise emission
standards for locomotives and cars failed to ad-
dress the related problem of noise from facilities
such as railroad yards. The Court ordered EPA
to adopt final regulations controlling railroad
facilities—everything in addition to the cars and
locomotives.
40 CFR 201 Special Local Conditions. NCA Nov 29, 1976 Henry Thomas (AW-490)
17(02/18(02. The-regulation establishes proce- Environmental Protection Agency
dures permitting adoption by a State or other- Washington D C 20480 703-557-
wise preempted State and local rail and motor 7743.
carrier noise regulations when necessitated by
special local conditions..
40 CFR 202 Interstate Motor Carrier. NCA 18. Work plan under development William Roper (AW-490)
This action will update the noise emiss10n sta.nd- Environmental Protection Agency
ards for interstate motor carriers to reflect in- Washington. D.C. 20460, 703-557-
creased knowledge about available noise abate- 7747
ment technology.
FEDERAL REGISTER, VOL 43, NO. 231— THURSDAY, NOVEMBER 30, 1978
250
-------
NOTICES
MAJOR EPA REGULATIONS UNDER CONSIDERATION—Continued
56169
Name and description of regulation
Proposal date In FEDERAL RECISTEB Final date in FEDERAL REGISTER
Contact person and address
THE FEDERAL INSECTICIDE, FUNGICIDE, AND RODENTICIDE ACT
40 CFR 162 'Pesticide Registration Guidelines:
Introduction. FIFRA 3. This subpart B (will July 10, 1978 April 1979
become A) includes the general purposes of all of
the guidelines, degree o.f flexibility in require-
ments and in interim data usage, definition of
terms used throughout the guidelines, and re-
quirements for retention of data and test sam-
ples at laboratories.
Experimental Use Permits. FIFRA 3. This sub-
part A (will become subpart C) specifies the data
that must be submitted in support of an applica-
tion for an experimental use permit.
Chemistry Requirements. FIFRA 3. This sub- July 10, 1978 April 1979
part D covers data submission requirements re-
lating to chemistry of pesticide products' active
ingredients and their formulation components
and manufacturing impurities. (Chemical study
requirements dealing with environmental fate of
pesticides may be included here or be moved to a
new subpart.).
Hazard Evaluation: Wildlife and Aquatic Or- do May 1979
ganisms. FIFRA 3. This subpart E outlines the
data submission requirements for studies of pes-
ticide effects on birds, wild mammals, fish, and
other aquatic animals.
Hazard Evaluation: Humans and Domestic Aug. 22. 1978 June 1979
Animals. FIFRA 3. This subpart F delineates the
data submission requirements for studies of pes-
ticide effects in laboratory animals involving
oral, dermal, and inhalation uptake routes,
acute, subchronlc, anbd chronic exposures, and
including local or systemic injury and maladies
such as oncogenic. teratogenic, mutagenic, and
neurotoxic effects.
Product Performance. FIFRA 3. This subpart December 1978 August 1979
O specifies the data submission requirements
that registrants must submit to demonstrate that
the prospective pesticide product will control the
pests or control undesired growth or behavior as
specified in label claims.
Label Development. FIFRA 3. This subpart H March 1979 October 1979
describes all essential parts of a pesticide product
label, how labeling and label statements must
comply with the Act, and how claims and direc-
tions must correspond to evidence presented or
on hand in data on efficacy and safety.
40 CFR 162 Pesticide Use Restrictions. FIFRA 3. December 1978 January 1979
This regulation will classify pesticide uses for re-
stricted use..
Bill Preston (TS-769).
Environmental Protection Agency,
Washington, D.C. 20460, 703-557-
7351.
Do.
Do.
Do.
Conditional Registration Regulation. FIFRA February 1979
3(c)(7) (A) and (B). This interim/final regulation
would establish procedures for conditional regis-
tration of pesticide products which are identical
or substantially similar to those currently regis-
tered or new uses of existing pesticide products.
Conditional Registration Regulation. FIFRA July 1979
3(c)(7)(C). This regulation provides for the condi-
tional registration of new chemicals when certain
data are missing.
40 CFR 162.9, 173 Registration Data Compenaa- June 21. 1977 February 1979
tion. FIFRA 3(C)(1)(D). These rules provide for
compensation when one pesticide registrant
relies on test data generated by another regis-
trant.
40 CFR 172 State Experimental Use Permits. Sept. 30. 1975, intei im final
FIFRA (5)f. The regulation defines the scope of
State Jurisdiction to allow experimental uses of
pesticides.
40 CFR 165 Storage and Disposal Practices (Pro- Oct. 15, 1974 Will not be issued.
hibition). FIFRA 19. These rules will prohibit
dangerous or environmentally unsound pesticide
storage practices.
40 CFR 162 State Registration to Meet Special Sept. 3, 1975 March 1979
Local Needs. FIFRA 24(c). This part defines the
scope of State jurisdiction over the registration
of pesticides.
40 CFR 162.16 Pesticide Special Packaging Reau- Feb. 16, 1977 December 1978
lations. FIFRA 25. The rule prescribes when and
what form of child-proof packaging is required
40 CFR 162 Exemption of New Human Drugs. Oct. 13, 1978 do
FIFRA 25(c)(2). This part would exempt from
FIFRA pesticides that are also new drugs regu-
lated by FDA.
Walt Waldrop (TS-770).
Environmental Protection Agenry.
Washington. D.C. 0460, 202-755-
7014.
Bob Rose (TS-767).
Environmental Protection Agency,
Washington. D.C. 20460. 202-426-
2510.
Ed Gray (A-132).
Environmental Protection Agency.
Washington. D.C. 20460. 202-755-
0846.
Phil Gray (TS-770).
Environmental Protection Agency,
Washington. D.C. 20460. 202 755-
7014.
John Lehman (WH-565).
Environmental Protection Agency,
Washington. D.C. 20460. 202-755-
9185.
Phil Gray (TS-770).
Environmental Protection Agency.
Washington, D.C. 20460. 202-755-
7014.
Maureen Grimmer (TS-766).
Environmental Protection Agency.
Washington, D.C. 20460. 202-755-
8030.
Dave Brandewein (TS-766).
Environmental Protection Agency,
Washington, D.C. 20460, 202-755-
8037.
FEDERAL REGISTER, VOL 43, NO. 231—THURSDAY, NOVEMBER 30, 1978
251
-------
56170
NOTICES
MAJOR EPA REOmJlTIONB UNDER CONSIDERATION—OORtfaraed
Name and description of regulation
Proposal date tn FEDERAL RTOISTCT Final date in FEDERAL REGISTER
Contact person and address
Tire ATOMIC EKEROY ACT
Protective Action Guidelines for Nuclear Emerges- September 1979 .
cies. AEA 274
-------
NOTICES
MAJOH EPA HEGUI.ATIONS UNDER CONSIDERATION—Continued
56171
Name and description of regulation
date In FEDERAL Rec.ibTE*
Final date In FEDERAL KroiSTZR
Contact person and address
THE RtsouRcr CONSERVATION AND RWOVCKY ACT
June 1979
40 CFR 250 GuHtrlinex for Stale Hazardous Feb 1. 197H January 1979
Waste Pru'jram'i. RCRA 3006 These guidelines
aro to assist Si.at'-s in the development of their
own ha/.ardous waste regulatory programs. The
guideline;; also specify minimum re.ivnreincnUs
States must me''t in order to be authon/.ed by
EPA to implement their hazardous waste pro-
grams.
40 CFR 250 Notification System for Hazardous July 11. 1978 August 1979..
Waste Generators, Transporters. Storeri. Treat
ers. ar.d Disposers. RCRA 30!0 The regulation
describes thP one-time notification requirement
for generators, transporters, trt aters. siorcrs.
and di.,poseis of hazardous wa_s'.e. which will
bring them to the attention of the persons ad-
ministering RCRA's hazardous wa..te program.
40 CFR 256 Guidelines for State S'.lid Waste Pro- Aug. 28. 1978
grains. RCRA 4C02(b). These guidelines are to
assist States in the development and implemen-
tation of solid wa.ste management programs.
40 CFR 257 Criteria for Clas^irication of Solid Feb. 6. 1978..
Waste and Dispose! Facilities. RCRA 400-lci).
These criteria provide a basis against which solid
waste land disposal facilities can be evaluated in
order to determine probability of adverse effects
on health or the environment.
Guidelines for Federal Procurement Practices.
RCRA 6002(e). These guidelines will assist Fed-
eral agencies to comply with the RCRA's re-
quirement that procured materials be composed
of the highest percentage of recovered materials
practicable:
Utilization of Fly Ash and Slag
Use of Recycled Paper in Paper Products....
Use of Waste in Construction Products
July 1979.
Dan Derkus .c)
9190
Timothy Fields (WU 565)
Environmental Protection Agency.
Washington. DC 20400. 202 755
91!06
George Garland (WH-555).
Environmental Protection Agenrj
202 753-9125
Kenneth Shuster (WH-564).
Environmental Protection Act-no;,
Wa.shington.DC 20460.202-755
9116
Stephen Lingle (WH-563).
Environmental Protection Agency.
Washington, D.C. 20450. 202 755
9140
April 1979
June 1979
July 1979
July 1979
September 1979 .
October 1979
THE Toxic SUBSTANCE CONTROL ACT
40 CFR 740 to— 'Testing of Chemical Substances December 1978 Mar. 1979. 749..
and mixtures. TSCA 4. These regulations require
testing of chemical substances that may present
an unreasonable risk to human henlcii or the en-
vironment, or are produced in substantial ouanti-
ties but are not supported by adequate test data.
EPA is prepanun two testing regulations: on co-
Remcity testing and environmental fate testing.
40 CFR 720 Premanufaclure Notification. TSCA December 1978 April 1979
5. This regulation will establish the procedure
whereby a company will notify EPA of its intent
to manufacture a new chemical. The regulation
will prescribe the required premanufacture noti-
fication form, describe the procedure for EPA
review, and contain testing guidelines.
40 CFK 761 PCB's Manufacture and Distribution. June 7. 1978 January 1979
TSCA 6. This regulation bans the manufacturing
and distribution of PCBs and products contain-
ing PCBs.
Control of Polybrominatcd Biphenvls TSCA 6. January 1979 July 1979
The regulation would control the use of polybro-
minated biphenyls.
C7iloro/7uorocar6on Emissions. TSCA 6. This To be determined
regulation would apply to nonaerosol uses of
chlorofluorocarbons.
May 1979 December 1979..
40 CFR 730 Reporting on Substances Recom-
mended for Testing TSCA »"" The regulation
requires reporting of existing health and safety
studies for chemical categories as recommended
for testing.
40 CFR 720 Records of Adverse Reaction. TSCA March 1979 October 1979
8
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/1-80-021
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Addendum to "The Health Consequence; of Sulfur Oxides
A Report from CHESS,. 1970-17971," May 1974
5. REPORT DATE
April 1980 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/00
15. SUPPLEMENTARY NOTES
Original report, EPA-650/1-74-004; available from NTIS PB 234-920
16. ABSTRACT ' ~~ ~~~"~
A controversy about the scientific credibility of results from the Community Health
and Environmental Surveillance System (CHESS) study prompted a series of Congressional
hearings in 1976 ("The Brown Report") with subsequent legislation (Public Law 95-155)
to enact the Environmental Research, Development and Demonstration Authorization Act
of 1978. This addendum has been compiled to satisfy Recommendation 3(c) of The Brown
Committee Report, entitled "The Environmental Protection Agency's Research Program
with Primary Emphasis on the Community Health and Environmental Surveillance System
(CHESS): An Investigative Report." It contains the following materials which concern
the 1974 CHESS Monograph and various CHESS studies, in addition to EPA's research and
development program in general: as follows The Brown Committee Report; P.L. 95-155;
Appendices from EPA's Research Outlook for 1978 and 1979; and the Science Advisory'
Board's Health Effects Research Review Group Report to Congress in February 1979.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
258
20. SECURITY CLASS /This page I
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PCJEVIOJS EDITION >i OBSOLETE
------- |