United States Office of Air Quality EPA «50/5-82-ooa
Environmental Protection Planning and Standards OCTOBER 1992
Agency Research Triangle Park NC 27711
Air
HAZARDOUS AIR POLLUTANT
PRIORITIZATION SYSTEM
HAPPS
-------
ARGONNE NATIONAL LABORATORY
9700 South Cass Avenue,
Argonne, Illinois 60439
HAZARDOUS AIR POLLUTANT
PRIORITIZATION SYSTEM
(HAPPS)
by
A.E. Smith and D.J. Fingleton
Energy and Environmental Systems Division
October 1982
prepared for ~
Pollutant Assessment Branch
Standards and Air Strategies Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Under Interagency Agreement No. AD-89-F-1-344-0
Project Officer: Robert Schell
U.S. Envjrr.
Chicago, l_i
-------
DISCLAIMER
This report has been reviewed by the Office of Air Quality Planning
and Standards, U.S. Environmental Protection Agency, and approved for
publication as received from the Argonne National Laboratory. Approval
does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention
of trade names or commercial products constitute endorsement or
recommendation for use.
-------
CONTENTS
1 PURPOSE AND RATIONALE 1
1.1 SCOPE AND LIMITATIONS 1
1.2 RATIONALE FOR HAPPS FACTORS 5
1.3 RATIONALE FOR CRITERIA 10
1.4 FACTOR GROUPS AND WEIGHTS 42
1.4.1 General 42
1.4.2 Groups 45
1.5 Incergroup Weights 49
2 PRIORITIZATION METHODOLOGY 54
REFERENCES 55
APPENDIX A: Tables, Worksheets, and Abbreviations Used in RTECS 58
TABLES
1.1 Factors in HAPPS and the ORNL Procedure 6
1.2 Criteria for Oncogenicity 12
1.3 Criteria for Mutagenicity 18
1.4 Criteria for Reproduction and Developmental Toxicity 21
1.5 Criteria for Acute Lethality 24
1.6 Criteria for Effects Other than Acute Lethality 31
1.7 Criteria for Production Volume 35
1.8 Criteria for Vapor Pressure 37
1.9 Criteria for Bioaccumulation 39
1.10 Criteria for Existing Standards 40
1.11 Utility of Normalization 43
1.12 Groups of Factors 46
1.13 Intergroup Weights 50
1.14 Sensitivity Analysis 52
FIGURE
1 Scales for Equivalent Volumetric and Mass Concentration Units 26
111
-------
1 PURPOSE AND RATIONALE
This section discusses the purpose of the Hazardous Air Pollutant
Prioritizacion System (HAPPS) and the rationale used in developing the
system. Subsections 1.2, 1.3, 1.4, and 1.5 provide, respectively, the
rationale for choosing the particular set of eight factors, the rationale for
the specific criteria and weights within each factor, the rationale for
grouping the factors and assigning the intragroup weights, and the rationale
for the intergroup weights used in the final ranking. These rationales can be
properly understood only when the scope of HAPPS and the limitations imposed
by this scope are taken into account. Subsection 1.1 discusses these limita-
tions .
Section 2 provides instructions for using HAPPS. Appendix A contains a
set of worksheets and tables for documenting a prioritization. Both Section 2
and the Appendix have been placed after the explanatory material in Sec. 1 to
aid in copying the working material in the appendix. Section 1 assumes some
familiarity with HAPPS and readers may want to scan the instructions and
materials in Sec. 2 and Appendix A prior to reading Sec. 1,
1.1 SCOPE AND LIMITATIONS
The strategies and Air Standards Division (SASD) of U.S. SPA's Office
of Air Quality Planning and Standards periodically selects new substances for
assessment to determine whether regulatory development under the Clean Air Act
should begin. Ideally, a full range of toxicological and epideiniological
information coupled with detailed estimates of current emissions and human
exposure would be available to aid in such decisions. Hovever, such complete
information is seldom available and early assessment is often made on the
basis of incomplete and/or dated information. This is particularly true of
EPA's hazardous air pollutant assessment effort, since a large number of
organic and inorganic substances are potential candidates for study. However,
even if only a small number of potential candidates existed, the resources
involved in producing complete scientific information preclude the develop-
ment of such information for each substance until there is some certainty
that regulation is appropriate. Thus, a procedure for initially prioriti-
zing substances on the basis of limited, readily available information
-------
is needed so that resources for detailed studies might be allocated effi-
ciently, that is allocated first to detailed studies of substances anticipated
to be significant air pollution problems and later to substances anticipated
to be lesser problems. HAPPS provides a means of producing such a prioritiza-
tion.
It is recognized that a prioritization with limited data as opposed to
extensive, detailed data might produce substantially different results. HAPPS
is only intended to provide a reasonable prioritization to aid EPA in deciding
which substances to study first based on readily available information. As
such HAPPS is intended to be used as part of EPA's internal planning process.
It is important to recognize that even substances ranked very high by HAPPS
might never be regulated. Many subjective decisions must be made and detailed
objective studies done and evaluated between the time a substance is ranked
highly by HAPPS and a decision is made to regulate that substance as an air
pollutant. In other words, HAPPS must be viewed a's an initial, tentative step
within the context of the overall regulatory program. Even the initial
prioritizacion produced by HAPPS will be subject to additional screening by
experts to eliminate any obvious anomalies. In addition, the methodology will
be applied periodically to incorporate new information as it becomes avail-
abl;.. Such periodic reviews could result in changes in the relative rankings
of various substances, reflecting the new information. There may also be
progrsmaiatic reasons for overriding the indications -given by HAPPS. For
example, it might occur that some particular class of compounds like heavy
metals is receiving special attention throughout air programs, a considera-
tion which could lead to alterations to a. list produced by HAPPS. Regulatory
decisions will not be made on the basis of a substance's ranking by the HAPPS
procedure. Health assessments, exposure assessments, and other information
must all be evaluated prior to making the decision to regulate. It might be
found that the health effects associated with a substance ranked highly by
HAPPS were not serious enough, that control of the substance was technically
infeasible, or that the likelihood of exposure was not sufficiently great to
justify regulation.
To be useful, a procedure must be tailored to the expertise and exper-
ience of potential users. No macter how complete or precise a procedure is in
theory, it is useless if its application requires detailed or specialised
-------
knowledge not available Co Che user group. With Chese points in mind, discus-
sions with EPA personnel defined several guidelines which HAPPS would need to
follow Co be useful in the prioritization effort:
Generally, readily available summary documents or compu-
terized data bases will be used; searches of primary
sources are precluded by the intended use of the proce-
dure and the inability to justify allocating significant
resources to this preliminary prioritization step.
The methodology is purposely designed not to utilize
expert judgment in prioritizing chemicals. Such expert
judgment is more appropriate as part of subsequent regu-
latory decision making than to the preliminary prioriti-
zation stage where HAPPS is used.
The preliminary nature of the prioritization means that
the procedure should be simple enough to permit a single
user to prioritize several substances per day and prefer-
ably ten or more.
Personnel using the procedure should have only limited
expertise in toxicology or related subjects and only
limited familiarity with some of the sources of emissions
of the substances being ranked. Hence, the procedure
could not rely on decisions requiring expert judgment or
special knowledge related to these areas.
A particular set of substances should receive the same
ranked order when prioritized by two different persons.
Thus, insofar as possible, the procedure should be objec-
tive and the sources of data should be identical for all
users. Complete agreement between two different users of
HAPPS is unlikely because the goal of complete objectiv-
ity could not be attained; some factors still require the
use of informed, as distinct from expert, judgment in
choosing between criteria.
The system should be sufficiently flexible to permit up-
dates for a substance for which additional data becomes
available in the standard references.
The procedure is only intended to produce reasonable,
initial rankings. Detailed studies will take place after
the prioritization to develop sufficient information to
determine whether or not regulation is required. Ques-
tions of data interpretation, the validity of data and
similar technical items are left for experts to decide at
later stages in the assessment process.
These guidelines place considerable constraints upon the procedure by restric-
ting the source of data to be used and by limiting the effort in prioritizing
a single compound. In particular, searching the literature and contacting
other workers involved in prioritizing hazardous compounds led to an early
recognition that the most suitable summary reference appeared to be the
-------
Registry of Toxic Effects of Chemical Substances (RTECS).* RTECS is a
concise, easily used summary of toxic effects and is kept current by continual
updates in a computerized format and by quarterly updates in microfiche copy.
RTECS also contains data related to most of the criteria used for ranking sub-
stances in the scoring procedure developed by Oak Ridge National Laboratory.
As discussed below, the Oak Ridge procedure provided the basis for HAPPS thus
making RTECS an excellent match to the data requirements of HAPPS. That RTECS
was the most easily accessible reference had a significant impact on the
choice of factors and the structure of criteria for individual factors. In
addition, the need for consistency between users and the expected expertise
of users limit the types of decisions users should be expected to make and
make a straightforward approach with little chance for individual deviation
desirable.
Many ranking or scoring procedures for prioritizing chemicals exist
(see, for example, Refs. 2-25). Rather than develop an entirely new system,
consultation with EPA indicated the desirability of using a draft EPA multi-
media ranking procedure as the basis for HAPPS. That scoring procedure had
been developed by Oak Ridge National Laboratory (ORNL)21 for EPA's Office of
Pesticides and Toxic Substances (OPTS) and itself drew heavily on existing
scoring systems. Using an existing procedure as a basis for HAPPS was consi-
dered to be efficient in that a significant amount of duplicative work could
be avoided. Much thought and developmental work had already gone into the
ORNL procedure and its predecessors in choosing factors, criteria, and
weights. After an initial review, it was clear that much of what had already
been done was relevant to air programs and might need only minor revision or
expansion. It should be noted that the ORNL procedure was still in draft form
at the time HAPPS was developed. At the date of this report, the ORNL proce-
dure has not been completed and applied because objectives other than priori-
tization have become priority items in OPTS programs. Thus, the ORNL proce-
dure was used as the principal basis for HAPPS even though still in draft
form. In choosing the specific factors and criteria, additional existing
scoring systems, frequently those used in developing the ORNL procedure
itself, were consulted and used in developing HAPPS.
However, the ORNL procedure differed in scope and purpose from HAPPS.
The ORNL procedure was intended to consider multimedia exposures through
various routes (air, water, consumer usage, etc.) and was also intended to
-------
present data from the literature and data submitted in compliance with the
8(a)-Level-A rule of the Toxic Substances Control Act (TSCA) in a form suit-
able for review and final ranking by experts, both intentions making the ORNL
procedure inappropriate under the HAPPS guidelines. In developing HAPPS, the
ORNL procedure was used to select some of the factors, criteria, and weights
and then tailored to the specific concerns of air programs. Since the output
of HAPPS will provide program planning information for the Office of Air
Programs, it was not necessary to consider total human exposure; consideration
of exposure through the air route only is sufficient for such planning pur-
poses.
1.2 RATIONALE FOR HAPPS FACTORS
HAPPS prioritizes substances by scoring them in eight factors chosen to
reflect the concerns of air programs and issues deemed important by EPA.
Table 1.1 presents the eight factors used in HAPPS and the twenty-five factors
used in the ORNL procedure. The broader range of impacts on humans, animals,
plants, and the environment in the ORNL procedure shows clearly when the two
sets of factors are compared.
Among all the aspects of human health, the Office of Air Programs felt
that carcinogenesis should receive special attention in accordance with the
public's concern with carcinogens. The ORNL procedure already contained two
factors, oncogenicity and mutagenicity, (items 1 and 2 in Table 1.1) related
to carcinogenesis and these were retained in HAPPS. The oncogenicity factor
contains both malignant and benign tumors. Mutagenicity is related to car-
cinogenicity and evidence of mutagenic potential is frequently used as an
indicator in screening for carcinogens as in the Ames test. Although not a
current major concern for air programs, retention of the separate factor for
mutagenicity was reasonable for two reasons even though it is later grouped
together with oncogenicity under the carcinogenicity group. First, oncogeni-
city and mutagenicity are distinct effects; all mutagens are not carcinogens;
however, most carcinogens are mutagens. In addition, there is only a limited
amount of data available from carcinogenicity testing, making the use of the
more extensive surrogate data from mutagenicity testing desirable in order to
maximize the number of substances that could be scored using data reasonably
related to carcinogenic potential. Second, data are available separately for
each factor in RTECS, the principal data source, tasking scoring a compound
-------
Table 1.1 Factors in HAPPS and the ORNL Procedure
Procedure
Item
ORNL3
HAPPS
Component
Factor
Factor
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17-21
22
23
24
25
26
Chronic Toxicity
Acute Toxicity
Environmental Exposure
Occupational Exposure
Consumer Exposure
Oncogenicity
Mutagenicity.
Embryo-Fetotoxicity
Reproductive Effects
Terrestrial Animals
Aquatic Animals
Plants, Fungi, Bacteria
Terrestrial Animals
Aquatic Animals
Plants, Fungi, Bacteria
Production Volume
Environmental Release"
Transport & Transformation
Bioconcentration
Weighted Quantity
Processed13
Weighted Quantity
in Products^
(Five Separate Factors)0
Weighted Quantity
in Products'5
Number Exposed
Frequency of Exposure
Intensity of Exposure
Oncogenicity
Mutagenicity
Reproductive and Develop-
mental Toxicity
Effects Other than Acute
Lethality
Acute Lethality
Potential for Airborne
Release
Bioaccumulation
Existing Standards
aAdapted from Ref. 21.
"Requires information available from manufacturer compliance with 8(a)-Level-A rule of
TSCA.
-------
unambiguous if two factors are used while confusion might result with a
composite factor based on both oncogenicity and mutagenicity data.
Oncongenicity and mutagenicity were separated from other toxic effects
because the Office of Air Programs is primarily concerned with carcinogenicicy
rather than the toxic effects considered under the other toxicity-related
factors: reproductive and developmental effects (items 3 and 4), effects
other than acute lethality (item 5), and acute lethality (item 8). It is also
generally recognized chat oncogens and some mutagens have no thresholds
whereas the effects dealt with in the factors for items 5 and 8 normally
exhibit thresholds giving the separation a reasonable conceptual basis.
Assignment of appropriate weights to the carcinogenicity and toxicity groups,
as discussed later, was used to combine all toxic effects including oncogeni-
city and mutagenicity in the final prioritization. Finally, RTECS provides
separate data for oncogenicity, mutagenicity, reproductive and developmental
effects, and the two toxic effects factors so that the factors used in HAPPS
match the available data,.thereby reducing the likelihood of error.
The ORNL factors for items 3 and 4 were combined into a single factor
for reproductive and developmental toxicity in HAPPS. Increasing concern for
developmental effects has been shown in recent years as evidence accumulates
revealing the high sensitivity of human embryos, fetuses, and young to certain
substances. Reproductive effects could have long-term impacts on the popula-
tion and might be considered severe. However, data in RTECS does not dis-
tinguish between the two types of effects and hence the separate ORNL factors
were combined. As discussed above, these effects were separated from tine two
factors most directly related to carcinogenicity which is currently the
principal focus of air programs. Although it has been theorized that repro-
ductive and developmental effects exhibit thresholds they were also separated
from the two factors for toxic effects because this separation matches the
form of the data as presented in RTECS.
Since HAPPS emphasizes human health, the ORNL factors for aquatic
animals and for plants, fungi, and bacteria (items 6,7,9, and 10 in Table 1.1)
were eliminated as being poor indicators of human health effects. In addi-
tion, no readily available source of information was found for toxicity
effects in plants, fungi, and bacteria making the related factors inappropri-
ate for HAPPS even if they were relevant to the major concern with human
-------
health. Factors for toxic effects in terrestrial animals including humans
(items 5 and 8) but distinct from oncogenic, mutaganic, and reproductive and
developmental effects were retained. Effects in nonhuman terrestrial species
were included as being reasonable indicators of potential effects in humans.
Such inclusion, at least for higher mammals, is in accordance with standard
toxicological procedures that use animal studies to anticipate health effects
or toxicity in humans. Such procedures frequently involve tests on higher
mammals and results of such toxicity studies are used in scoring the two
toxicity factors in HAPPS.
The factor for production volume (item 11) was modified for HAPPS. In
assessing exposure via the ambient air, it is desirable to know how much of a
substance becomes airborne. Such detailed information is not available for
most substances at the prioritization phase of study so production volume was
considered as a surrogate following the ORNL procedure. Because airborne
release was the particular interest in HAPPS, the factor was modified to
include the consideration of vapor pressure in estimating the potential for
airborne release, since, other conditions being the same, a substance with a
high vapor pressure will become airborne more readily than a substance with a
low vapor pressure. (Vapor pressure was also considered in the ORNL procedure
in one of the factors related to occupational exposure.)
Of the five ORNL factors related to environmental exposure (items
12-16), four were eliminated and one was retained. The factors for environ-
mental release (item 12) and for the weighted quantities processed and in
products (items 15 and 16) require data that is to be submitted in compliance
with the 8(a)-Level-A rule of TSCA and may be of use in the future. However,
these factors are aimed at assessing overall environmental exposure and human
exposures through routes in addition to ambient air giving them a broader
scope than appropriate for air programs. Thus, these factors were dropped
from HAPPS. The factor for transport and transformation (item 13) was consi-
dered to be important in view of possible chemical transformations and resi-
dence time in the atmosphere. However, no summary sources of relevant data
were found and no sources seem likely to become available in the near future
thus eliminating this factor from consideration under the guidelines.
The factor for bioaccumulation (item 14) was retained in HAPPS. Air
pollutants can either be deposited mechanically or absorbed directly by food
-------
or be deposited on soil and absorbed later by plants. Such deposited sub-
stances can then either pass up the food chain through animals to humans with
possible biological magnification or be ingested directly. In both cases,
toxic levels can result directly or build up in the body over time. Data for
this factor was available in summary form and consequently was retained in
HAPPS.
Overall, then, three factors were used as surrogates for exposure:
production volume and vapor pressure combined to estimate the potential for
airborne release and bioaccumulation. Information for these three factors is
readily available but is not readily available for other indicators of ulti-
mate fate and exposure like residence time in the atmosphere and atmospheric
reactions. In addition, these latter two indicators were considered too
detailed for the preliminary nature of this prioritization.
The five ORNL factors relating to occupational exposure were deleted in
consultation with SASD, because occupational exposure does not fall within the
ambit of the Clean Air Act. In addition, only one of the five factors, level
of potential occupational exposure, does not depend upon data to be gathered
under the 8(a)-Level-A rule of TSCA. This ORNL actor scores compounds based
on exposure concentrations experienced by workers. While some of the criteria
used for the factor relate to the ease with which workers could contact the
airborne chemical, the information required would not generally be available
and the factor was dropped from HAPPS. However, vapor pressure could be used
for scoring liquids under this criterion in the ORNL procedure and was
retained as one of the indicators of the potential for airborne release.
The four ORNL factors for consumer exposure (items 22-25) were dropped
from EAFPS. These factors are oriented toward exposures due to use of house-
hold and consumer products and would include many routes of exposure inappro-
priate to air programs. Although indoor air pollution problems due to use of
household and consumer products would logically come under the criteria, no
summary source of the data required by the criteria for these factors was
found. In addition, indoor exposure is not the focus of the air programs
office. Thus, these criteria were dropped and the factors for potential for
airborne release and bioaccumulation are the only factors related specifically
to the potential for human exposure via the ambient air.
-------
10
A factor for existing standards (item 26), not found in the ORNL
procedure, was added to HAPPS. The factor was based on the MITRE procedure^1-1
and is intended to be scored based on standards set by the Occupational Safety
and Health Administration (OSHA). Establishment of an OSHA standard requires
a finding of potential toxic effect. Such a finding was considered to be an
important indicator that a substance might need to be considered in more
detail even though the OSHA standards are intended to apply in the workplace
where concentrations are likely to exceed ambient levels. It was also felt
that existing standards would be useful for prioritization, especially if data
for the other factors was sparse or if two or more substances were scored
relatively close.
1.3 RATIONALE FOR CRITERIA
This section discusses the reasons for choosing the criteria and the
associated weights within the individual factors. As already noted, the draft
ORNL procedure^! provided the principal model for HAPPS. However, in develop-
ing the specific criteria, several other procedures were frequently consulted.
References 2, 9, 17, 18, 20, 22, and 23 were found to be particularly useful
and include an earlier version of the ORNL procedure itself (Ref. 20) as well
as several systems used in the development of that procedure. These referen-
ces were selected after review of Refs. 2-24 as being most nearly suitable for
HAPPS. Thus, the criteria finally used were chosen from among several sets
available in the literature with possible additions and modifications to
conform to the guidelines for HAPPS and to make them useable with the RTECS
data base. With regard to other systems, it should also be noted that the
Multimedia Environmental Goals (MEGS)^ was considered as a basis for HAPPS.
Review indicated that for ambient air all the parameters used by MEGS were
already represented in HAPPS in a form more appropriate for prioritizing
compounds. MEGS establishes "estimated permissible concentrations" (EPC's)
and "minimum acute toxicity levels" (MATE's) from threshold limit values
(TLV's), National Institute for Occupational Safety and Health (NIOSH) stan-
dards, and toxicity data available in RTECS and/or other standard references.
The SPC's and MATE's are oriented towards establishing ambient and emission
limits for sources rather than toward comparisons between different pollu-
tants. In establishing ambient and emission limits, MEGS provides a tool for
-------
11
detailed evaluation of the impacts of particular sources, not a means of
comparing different substances emitted by sources with many different emission
characteristics. Thus, the use of MEGS as a basis for HAPPS was rejected.
Oncogenicity. Table 1.2 compares the criteria used for oncogenicity in
HAPPS and the ORNL procedure. It is anticipated that most of the data used in
scoring a compound using these criteria will relate directly to carcinogeni-
city but the definition of the factor and the data in RTECS include neoplastic
and equivocal effects as well. It was considered reasonable to use all data
related to tumerogenic effects during the prioritization leaving distinctions
between types of effects and their relationship to cancer for later considera-
tion by experts.
The HAPPS criteria involve several modifications of the ORNL criteria
for this factor. First, since Air Program's principal concern is with human
health, criteria explicitly recognizing this concern were added and weighted
more heavily than criteria related to evidence based only on animals. Addi-
tional weight was also assigned to oncogenetic effects in humans if caused by
inhalation giving the highest weight to the route of exposure of interes.t to
air programs (items 1 and 2). The distinction based on inhalation was not
made for the criteria based on evidence in animals, because positive studies
by any route of administration in nonhuman species were considered as reason-
able indicators that additional study would be warranted. As in the ORNL
procedure, evidence in two or more animal species was considered a reason for
greater concern than evidence in a single species (items 3 and 4). A cri-
terion determined by a substance's status under the National Toxicology
Program's (NTP) Carcinogenesis Testing Program was introduced into HAPPS (item
8). Selection for testing under this program requires a determination that
concern over a substance's carcinogenic potential is justified and that the
degree of concern is greater than that associated with substances not selected
for testing. Both of these determinations were considered sufficient reason
for air programs to consider looking at a substance in more detail but were
not considered as important as actual data. The ORNL criterion based on
determining that a substance was a precursor to cancer was dropped because
summary data would not be found. Both of the criteria requiring professional
judgment were dropped, because such expertise could not be assumed under the
-------
12
X
4»J
«r4
U
*4
C
41
ac
o
o
c
o
w
0
CO
1-1
U
Hj
4J
p^
Vrf
U
CN
^4
0)
i *
_2
CO
H
CO
J
Z
o
&
CO
,p4
^4
41
4J
vl
lj
O
41
h
O
U
CO
co
Cu
0.
^
K
J3
to
t*4
kl
41
4J
ft
k.
o
X
w
co u
0 £.
C 00
0 *
U 0)
0) 3
to
X
CO £
E 90
U CU
ft. 7
e
41
*-<
^4
o
o
4J
u
o
CD
C
CO
i
5;
i^
e i
O co
*4 »-4 i»4
4J to to
to j: E
t i C -H
to '^ C
£ C to
C 0
* C 4)
M
xx o
O ^ S
co co >-
c c . o
CO !0 C
E E 0 0
^ Z AJ ^*
r
r** u^
* *
o o
in * co
i CM e*i
CO
0)
o
41
a
a
4)
f 1 4J
CO 3
S 0
«4 M
C
to x
c
4) tQ
0 X
E ^
01
3
0
>^
c
CO
x
.a
to
41
H
U
41
a.
to
u
-c *
1 4! U
tB C CO 4J
41 41 CO g
-t G 4) .O «
0 O 1- -< 1-
o. .
CO CO <
3 01 4J
CO U-4 CO
0 1 90 1
JJ W 1 01 1
3 a c
e c 41 41
CO O 41 4J CJ
« E
Z J3 .
4J
4>
^-4
a
u
w
o
iu
T3
41
e <
a.
c o.
o e
41 .^
e c
41 0)
CO >
41 **
1 U 00
CN a
4>
41 U
41 CO
ce to
tO '^
0 'C 4)
U 4) 4J
^4 W
O t-i O
41 U
-------
13
guidelines for HAPPS. The criterion based on tnutagenicity testing was dropped
for two reasons. The data in RTECS could not be used to identify those
particular tests for mutagenicity which, like the Ames test, are standard
protocols for assessing carcinogenic potential. More important, even if such
tests for mutagenicity could be identified in RTECS, it would have been
improper to use them to score a substance in both the oncogenicity and muta-
genicity factors because such a procedure would amount to a double counting.
The HAPPS procedure corresponds to the presentation of the data in RTECS,
avoiding both the possibility of error involved in using some mutagenicity
data to score the mutagenicity factor and other mutagenicity data to score the
oncogenicity factor and, more importantly, the error of double counting.
Mutagenicity tests carried out to screen for carcinogens are included in
scoring the mutagenicity factor and hence affect the overall scoring of a
substance in the carcinogenicity group which depends upon the two separate
factors for oncogenicity and mutagenicity. Thus, mutagenicity screening for
carcinogenicity is taken into account in the final ranking of a substance even
though such tests are not explicitly singled out in a specific criterion under
oncogenicity. The criterion for negative evidence was retained but with no
judgment as to the adequacy of the evidence being required of the user.
Finally, a criterion for "no data" was added to HAPPS.
Weights were assigned in HAPPS to be reasonably consistent with the
weights assigned in the ORNL procedure and thus to retain as much as possible
of the expert judgment as to the relative importance of the criteria that
went into developing the ORNL weights. Generally speaking, an attempt was
made to match each HAPPS criterion to a similar ORNL criterion allowing for
the differences between the two procedures. Where reasonable matches could be
made, the matched criteria were used as benchmarks and assigned weights equal
to the corresponding ORNL scores. The weights of unmatched HAPPS criteria
lying between two matched criteria were obtained by interpolation and the
weights of unmatched HAPPS criteria lying beyond the range of the ORNL cri-
teria were obtained by extrapolation. In some instances, this procedure
produced inconsistencies in the relative weights of various HAPPS criteria.
These instances are noted in the following discussions of the individual
factors. To guarantee complete consistency would have required changing the
relative scores (weights) assigned to the benchmark criteria; retention of
-------
14
these relative weights was considered more important than the achievement of
complete internal consistency.
The HAPPS and ORNL criteria for items 3, 4, and 9 in Table 1.2 were
felt to correspond reasonably well and these weights were used as benchmarks
in developing the other weights in HAPPS. Since the ratios of weights within
a factor and not the difference between them give relative measures of the
importance of the corresponding criteria, a constant scaling factor can be
applied to the individual weights without changing the relative importance of
the criteria. For oncogenicity, a factor of 1/3 was chosen. Scaling in this
way compresses the scale and thus maximizes the effect of additional secondary
weight throughout the range of primary weights. In general, the scaled
weights in HAPPS are related to the corresponding ORNL scores by
(Weight in HAPPS) - (Scaling Factor) x (ORNL Score).
Applying this equation with a scaling factor of 1/3, the weight for item 3 in
HAPPS is 3 (=» 9/3) and the weight for item 4 is 2 (» 6/3). (Note that the
ratio of the weights of the two criteria in HAPPS is the same as the ratio of
the scores of the corresponding criteria in the ORNL procedure (3/2 - the
ratio of the weights in HAPS = 1.5 9/6 = the ratio of the ORNL scores).
HAPPS contains two criteria (items 1 and 2) designed to emphasize the
primary concern of air programs with human health and exposure by inhalation.
Both of these criteria would correspond to a single criterion (item 3) in the
ORNL procedure so that assigning weights to them required extropolation beyond
the initial correspondences between the HAPPS and ORNL criteria established
above. In the ORNL procedure, evidence of oncogenicity in humans would
receive a score of 9 (item 3) and evidence of oncogenicity in one animal
species would receive a score of 6. Thus, the score (or weight) attributed
to evidence in humans is 502 greater than that attributed to evidence in one
animal species. It was felt reasonable to use half as great an increase (25%)
in HAPPS as a factor for determining the increase in weight to be associated
with evidence of oncogenicity in humans by a noninhalation route (item 2) in
comparison with evidence from two or more animal species (item 3). This
procedure gives additional weight to human data reflecting the concerns of air
programs but less additional weight than the ORNL procedure gives to evidence
in a second animal species, the lesser weight reflecting the fact that the
expert opinion embodied in the ORNL procedure considered human evidence or
-------
15
evidence in a second animal species equally important. The same factor was
also applied to extropolate from evidence in humans by noninhalation (item 2)
to evidence in humans by inhalation (item 1). Thus, the weight for item 2 in
HAPPS is
(Weight for item 3) x (Extropolation Factor)
- 3 x 1.25 - 3.75 * 4
and the weight for item 1 in HAPPS is
(Weight fo'r icetn 3) x (Extropolation Factor) x (Extropolation Factor)
- 3 x 1.25 x 1.25 * 4.68 *5.
The HAPPS criterion for item 8 (scheduled for testing) was considered
to correspond most closely to the ORNL criteria requiring expert judgment
(items 6 and 7), since it was felt that expert judgment is frequently used in
developing testing schedules. The weight for this criterion was chosen to be
the geometric mean of the ORNL scores for the two criteria requiring expert
judgment taking the scaling factor into account. Thus,
(HAPPS Weight for item 8) - (Geometric Mean of ORNL Scores for
items 6 and 7) x (Scaling Factor)
= V3 x 1 K (1/3) = 0.57 » 1.
"Finally, the criterion for no data (item 10) was assigned a weight of
zero, the same as negative evidence. The references do not reliably dis-
tinguish between no data and negative results so equal weights for these two
criteria were considered reasonable. Strictly speaking, this weight assign-
ment means that there is less need for concern about a substance for which no
data is available in the standard sources (weight m 0) than there is for
concern about a substance which has been scheduled for testing (weight * 1).
However, in practice, the results of a prioritization will be reviewed and
substances with a significant lack of data will be identified and handled
separately. In terms of the initial ranked list needed to start a program of
detailed evaluation, HAPPS emphasizes substances for which the most data is
available and, hence, presumably substances for which additional evaluation
will be easiest in the early stages of the evaluation program. Program-
matically, this would permit time for data to be gathered on data-deficient
substances identified exogenous to HAPPS as being of potential concern.
Furthermore, results of prioritization by HAPPS could still be of use in this
-------
16
effort by indicating the relative need for concern based on the limited
available data.
In addition to the primary weights just discussed, HAPPS assigns
secondary weights for the oncogenicity factor. Given the criteria, it is
possible for a substance to satisfy two or more criteria simultaneously. For
example, there could be evidence of oncogeoicity in humans by the noninhala-
tion route and evidence from a single animal study. These secondary weights
serve the purpose of giving a substance with data satisfying several criteria
a total higher score than a substance with data satisfying fewer criteria when
the highest weighted criterion satisfied by both substances is the same.
Furthermore, the secondary weights have been assigned so that a substance
could never receive a total score higher than a substance which satisfied a
single more heavily weighted primary criterion simply by satisfying multiple
secondary criteria. This relationship between the primary and secondary
weights is based on the view that while concommitant evidence should have some
positive influence on the final score for a substance, the additional weight
should not be sufficient to raise a substance to the score of the next higher
ranked criterion. For example, a substance with evidence of oncogenicity in
humans by a noninhalation route and evidence of oncogenicity in two animal
species should not receive a higher score than a substance with evidence of
oncogenicity in humans by the inhalation route. That the secondary weights
achieve this aim can be confirmed by noting that, except for a substance whose
primary weight is 5, the maximum sum of the secondary weights is 0.55 (« 0.5 +
0.05), because the HAPPS criteria for items 3 and 4 cannot be satisfied
simultaneously. This maximum sum for the secondary weights is less than the
minimum difference of one between any two primary weights. In making this
test it is not necessary to consider the maximum possible secondary weight for
a substance whose primary weight is the maximum possible, a substance whose
primary weight is 5 in the oncogenicity example, because there are no more
highly weighted criteria. The addition of secondary weight can never raise
the total weight of a substance with the maximum primary weight above the
weight associated with a substance satisfying a more heavily weighted single
criterion, because there is no such criterion.
As used in HAPPS, additional secondary weight is assigned only when
different criteria are satisfied; additional weight is not assigned when
-------
17
several data satisfy the same criterion. It was felt that some sense of the
quality of the data was necessary in assessing whether multiple data satisfy-
ing the same criterion were better than data based on a single study satisfy-
ing only one criterion. For example, a substance which has positive evidence
in two animal tests of poor quality should probably not be ranked higher than
another substance which has positive evidence in only one animal test of
superior quality. HAPPS would, however, rank the former compound higher than
the latter if additional secondary weight were assigned but since the quality
of the studies involved is not reported in the readily available literature,
such occurrences could not be checked for in the HAPPS procedure. Therefore,
it was decided, based on considerations of this type of situation and the
unavailability of data on study quality, not to assign additional weight based
solely on meeting the same criterion with data from multiple studies.
Mutagenicity. Table 1.3 compares the HAPPS and ORNL criteria for
mutagenicity. Many of the differences between the two sets of criteria are
similar to the differences already discussed for oncogenicity and will not be
fully discussed in this subsection. Although information specific to humans
is likely to be unavailable, evidence of mutagenicity in mammalian test
systems was given more weight in both HAPPS and the ORNL procedure than
evidence in nonmammalian systems. In HAPPS, evidence obtained from inhalation
studies in mammals (item 1) receives additional weight as being directly
related to the route of administration of interest in air programs. The ORNL
criteria (items 9 and 10) requiring expert judgment were dropped from HAPPS as
were the corresponging criteria for oncogenicity. As for the oncogenicity
factor, a criterion was added to HAPPS to reflect whether a substance has been
scheduled for or is undergoing mutagenicity testing under the National Toxi-
cology Program. Data in RTECS do not unambiguously identify tests showing
germinalcell DNA interactions as required by the ORNL criteria for items 7 and
8. Thus, this determination, although desirable, was dropped from HAPPS while
the remainder of the ORNL criteria specifying the type of test was retained.
Finally, the criterion for no data (item 13) was added.
With these changes, the HAPPS criteria were expanded according to the
following scheme. Evidence of mutagenicity from mammalian test systems was
considered a better indicator of potential effects in humans than evidence
-------
13
1
^
tf4
y
c
01
ac
9)
y
JC
b
0
IM
eg
.^
u
OI
u
*-(
u
^
ft
*
«
01
i i
.C
eg
H
eg
z
oi
O
>
r;
eg
I*
Oi
AJ
t»4
U
o
u
O
y
VJ
en
0.
a.
^
^
,£
eg
H
b
01
AJ
«-i
p
u
u
CO U
3 J=
S DC
O "*
y oi
oi 3
en
X
eg .=
E 90
H *f-4
14 0)
e. 3
S
C ^J 1
CO 00 W
H 0)0;
^* ij jj
CO C
^ 0 >H
S i-
eg AJ <
E -HZ
> Q
41
1-1 C -*
O « 1-4
£ 01
5 oi y
C 1
0) C r-4
u eg
o e E
Q eg **"4
V4 AJ C
O 99 01
AJ 4) 10
41 CO k-i
e o> o u
O AJ 2£ 0
CT* ^O
.
93
CO 93 CO
4) 01 U C
AJ AJ AJ C eg
eg -H
sec -* <
9j 91 as r* AJ eg
rf -H .H eg 93 E
§01 £
AJ eg
S E E 4) eg e
E E E AJ S e 2
3J CQ CQ £ CQ Q
E E ECO -< c
93 C -*
O O Q -H eg o
> > >- -- O E U
4 .»4 AJ E AJ
> > . -H E ' eg .^
S > S > E >
e c c «g e
* ' -H C E C O C
e u .* .-i c -H
Oi o 01 eg Q
C *H C ^ 4J W 4J O 4)
OAjQegbAJt-4 ^b
03 j= 0 -H o -HO
AJUCS>S >E
co eg co »»^
9]^3 egc« b C w* C w*
4IC 0)0 O-i-l O 03 -( O M
4 <*4 i 1 C AJ AJ
O 41 O CO 4) O 91
4-i^Aj>%3 c jyoj c j*cu
^^3
^c cy u &. AJ 4) cu
AJ a. y E
AJ C U-l 4) 60 4)
CO -»4 C O C. T3 >
cu O n 3 -^
O AJ < AJ 3 "I AJ
> Z -c C 91 1 eg |
-4 O a 0) 4) -J 1 90 1
AJ <-* eg 60 3 C C
c - ^ js -a ,00
rt > 4) 3 < 01
y AJ I eg co AJ
C C 1 U AJ 00 9)
O -H r-i 4) -J cfl 0) 3
J 9lD.eg9j-ovt4 cr
AJOICCOCAJ o 4)
O C *^ 3 O CO O l^ "C
cO (2 £ ^^ * ^ "0 2» Q" ^1
4) 41
f-i > eg
C 3 4 AJ
.* -o AJ eg
cy eg Tj
4) .CM
c y 4) O
O en z Z
V^i
-^ CM O O
* i i
o coo
p4
10
01
b
eg
eg
fn4
Vj
OI
AJ
*4
y
en
a.
a-
:S
V
ij
M-J
O
93
C
0
AJ
eg
u
H
U-l
f-4
y
OI
a.
eo
01
AJ
OI
c_
E
O
y
w
S
u
0
VA4
a
^
Oi
AJ
O
"oo
c
'
"O
4)
4 e
eN 4)
en
4)
4) 0.
eg
§ '£
V- 4) <
«-l U
13 U --4
v v ?
AJ C
C. AJ O)
eg co c.
^j C "«
< z <
eg .a
-------
19
from nonmammalian test systems. Thus, mammalian tests were given higher
weights than nonmammalian tests just as in the ORNL procedure. Within each of
these groups, in vivo tests carried on in the living body of an organism were
weighted higher than in vitro tests conducted outside living organisms. This
is an extension of the separation of in vivo and in vitro mammalian tests in
the ORNL procedure (items 2 and 7). Finally, HAPPS gives additional weight
when evidence of mutagenicity is available in more than one test system except
for in vivo mammalian tests for which a distinction based on inhalation or
noninhalation route of administration was thought to better represent the
concerns of air programs. Distinctions based on both the number of tests and
the route of administration within the in vivo mammalian category would have
been too detailed for this screening level of analysis. Assignment of addi-
tional weight for multiple tests is consistent with the treatment of in vitro
tests with no germinal cell interaction in the ORNL procedure (items 7 and 8).
As shown on Table 1.3, items 2, 7, 8, and 12 were used as benchmarks
in developing the weights in HAPPS. The weight for item 11 (scheduled for
testing) was assigned as the geometric mean of the two ORNL criteria requiring
expert judgment, just as was done for oncogenicity. Thus, the weight is
2 (Y3 x 2 = 2.4 **2). For the four criteria listed as items 3-6, interpolation
in equal multiplicative steps was used. For five steps between 6 and 9, the
factor is about 1.0844 (6 x 1.08445 « 9). Using this factor and rounding to
the nearest tenth, the weights given in Table 1.3 were obtained. For example,
the weight for the criterion for two or more in vivo nonmammalian tests
(item 5) is 6 x 1.0844^ » 7.055 * 7.1, because it lies two steps above the
benchmark criterion (item 7). The weight for item 1 was obtained by extrapo-
lation using the same factor as was used for oncogenicity: 9 x 1.25 «sll.
It should be noted that the same factor of 1.25 could be derived by using one-
half the increase of 50Z applied by ORNL to account for additional in vitro
tests or mammalian tests of mutagenicity (see items 2,7, and 8) thus indicat-
ing that a consistent approach was applied in developing the ORNL scores.
The type of inconsistency noted in the general discussion shows up for
mutagenicity criteria. Several examples, but not a complete list, follow.
HAPPS and ORNL increase weights by a factor of 1.5 as evidence becomes avail-
able in a second nonmammalian test system in vitro (items 7 and 8). It would
be desirable to have the same factor apply to the criteria for nonmammalian
-------
20
in vivo tests (items 5 and 6) and to mammalian in vitro tests (items 3 and 4)
but in these cases the factor involved is only about 1.1 (as7.1/6.5 and
«3.3/7.7) which is, of course, simply the interpolation factor used to
determine the weights. Similarly, it would be desirable for the ratios
between the weights for criteria which are the same except for the type of
species involved to be equal. Thus, the ratio of the weights of items 4 and 8
which correspond to evidence from one in vitro test in nonmannnals and mammals,
respectively, should ideally be equal to the ratio of the weights for items 3
and 7 which correspond to evidence from two or more in vitro tests in mammals
and nonmammals, respectively. However, the first ratio is about 1.9 0*7.7/4)
while the second is only about 1.4 (*8.3/6). The weights could have been
assigned to avoid this type of inconsistency but only if experts had been
available to provide a reasonable set of weights for the components to be
matched. In view of the unavailability of the needed experts, the procedure
used here of retaining the expert opinion embodied in the ORNL scores and
interpolating and extropolating to express the needs of air programs and to
more fully utilize the data in RTECS to spread out the final scores was felt
to provide a reasonable approach.
As is the case for oncogenicity- HAPPS assigns secondary weights for
mutagenicity when more than one criterLon is satisfied by a substance. The
previous discussion for oncongenicity is equally applicable to mutagenicity
and a similar confirmation can be made chat additional secondary weight can
never cause a substance with a particular primary weight to receive a greater
total score than a substance with the next higher primary weight. Additional
weight is not assigned under mutagenicity when a substance has multiple data
satisfying the same criterion.
Reproductive and Developmental Toxicity. As noted in Sec. 1.2, the
ORNL factor for embryotoxicity and fetotoxicity and the ORNL factor for
reproductive effects were combined into the single HAPPS factor for reproduc-
tive and developmental toxicity. Since the structure of the ORNL criteria for
both factors is very similar, Table 1.4 lists them together in the interests
of simplicity for comparison with the HAPPS criteria. For this factor, the
rationale for the changes between HAPPS and the ORNL procedure is the same as
that for the two previous factors. Additional weight is assigned to evidence
-------
21
^
4J
If4
U
x
o
H
i i
to
4J
C
01
s
a
c
01
>
0)
o
c
OJ
OI
.>
u
3
TT
O
kl
a
01
a:
ki
0
03
H
kl
01
u
H
kl
0
-3-
^
OI
l»H
X!
0)
H
03
Z
as
O
^
H
^j
(U
4J
*(
^j
o
0)
ki
o
u
CO
in
a.
a.
y
Xi
0]
kl
41
4-1
*4
kl
CJ
>^
^
Q XJ
c x;
C 00
O i
O 01
oi S
C£
>,
kl 4J
o! x:
S 00
1-1 **
k, 4)
a. 3
S
CJ
u
) 1
i
U
O-
o
0
5
^4
o
09
C
0}
E
X
^s
«
C 1
O "3
4 1*4 ^
U W Q
as x: S
-J C «
03 .« e
i § "
- C 01
kl
XX O
XI J3 E
CO 03 1-4
c c o
OJ fQ C
£ S O O
2 ~ w H
r*» in
O 0
U*^ +jf fO
~H c^ en
9
0)
H
u
^
3)
( 1
aj
s
c
4)
03
H
kl
a
01
4J
3
0
W
x
C
0]
>^
XI
01
01
4
U
0)
a.
01
1 -<
O «3
ki u C
CO O- 4J 03 O E ^ u "^
4) 4) 00 3 01
a. -o T3 y a «<-i
0 4) 1» 3 * O
> CO "-5 kl J-l k.
^ -H as 4-1 3 a
03 X X! -< « X>
E 3 03 E C u
H -HO 4J C 03 «3 O C
C U O O W 4J 4)
(8 C4) til 'H <8 oj-op
O U & CO On ^ 4) 00
oi u c 03 oo co T;
C COI SllW-i O«33
(5 MT5 W^-iO ZXl"-i
SO «^ CO "^
^
o
to
01
»4
U
OI
a.
03 1 1 1
1 1 1
oj OI
E u
.- 3
c; o
re u
0) X
e e
O 03
m
0
< 0)
0
*
oo
c
H
4J
CO CU
01 U
4J C
01
i- -o
o *
U-l >
0)
o
01 01
i 4 > 03
3 -< u
"O 4J 03
4) 03 *O
x: oo
u oi c
w z z
i/*i
o o o
o o o
-< O O
ao a* o
4
p
O
1
03
5-*\
u
^4
U
H
x
0
J-*
0
4J
0!
^4-J
a
e
03
>,
4J
F^
o
«H
X
0
o
^4
i
0)
kl
o
UJ
»^
fsl
U-l
4J
C
»<
03
p
O
4J
O
03
14^
O
4J
kl
<4«4
T3
4i
jj
a
93
w
<
CQ
03
H
4)
H
kl
U
u
H
01
x:
4^
kl
0
U^
01
U
^
3
4j
U
3
W
u
00
Vd
ct]
f-^
*f^
E
^*
en
cu
>
03
x:
x:
u
H
r-
3
14-1
O
Si
o
XI
CO
4J
Cj
4)
U-l
14-1
01
41
<^
4-1
U
3
"O
o
kl
c.
01
kl
w
ex
a.
^
=
4)
£.
^
^M
o
00
fi
o
i-4
4J
03
U
M
vu
w4
O
4)
a
03
0)
4J
OI
v 4
a.
£
O
u
M
g
O
vu
T3
01
C
01
4J
kl
0
CO
e
,r*
V~
OI
4J
c
01
CO
4)
t.4
e.
OS
»4
kl
0)
.^4
kl
u
4J
CO
O
s
XI
^
X
*4
o
c
a.
c.
<
c
c
01
^
H
00
41
kl
03
4
kl
01
u
H
ki
O
CO
4J
(J
0)
vw
VM
4)
01
iJ
4J
U
3
o
0
kl
0.
01
ki
kl
0
kl
o
u
cd
U-l
^J
Z
06
C
0)
x:
4J
C
^
p-^
e
o
91
kl
03
01
a.
c.
S3
e
o
U
o>
4J
.^,4
kl
U
co
£.
r-
o
-------
22
of effects in humans and to evidence of effects caused by inhalation in huaians
to emphasize the interests of air programs. Three ORNL criteria requiring
expert judgment (items 5, 6, and 7) were omitted. A criterion related to a
substance's testing status and a criterion for scoring substances with no data
were added.
The assignment of weights for this factor parallels that for oncogeni-
city. An overall scaling factor of 1/3 was applied and items 3, 4, and 9
served as the benchmarks. An increase of 25% was used to extropolate to the
criteria for humans (item 2) and for humans by inhalation (item 1). This
increase is one half the increase of 50% assigned by ORNL for positive results
in an additional species (compare ORNL scores of 9 and 6 for items 3 and 4,
respectively). The HAPPS weights for items 1 and 2 can be extropolated from
the HAPPS weight for item 3 by using an extrapolation factor for a 25%
increase just as was done for oncogenicity. Thus, in HAPPS
(Weight for item 2) » (Weight for item 3) x (Extropolation Factor)
- 3 x 1.25 - 3.75 w 4
and
(Weight for item 1) » (Weight for item 3) x (Extropolation Factor)
x (Extropolation Factor)
* 3 x 1.25 x 1.25 - 4.68 * 5.
As for oncogenicity, the criterion for compounds scheduled for testing (item
8) was weighted at the geometric mean of the ORNL criteria (items 5, 6, and 7)
requiring expert judgment and taking the scaling factor into account. Thus
(Weight for item 8 in HAPPS) " (Geometric Mean of ORNL Scores for
Items 5, 6, and 7) x (Scaling Factor)
- (4 x 3 x 1)1/3 x (1/3) * 0.76 * 1.
Secondary weights were also assigned in the same fashion and subject to the
same Limitation as was done for oncogenicity and mutagenicity. These latter
two effects, however, are generally believed to exhibit either no or very low
thresholds while it has been theorized that reproductive and developmental
effects exhibit thresholds. However, the thresholds may be very low or even
zero so the system of secondary weights was used rather than a system based on
lowest effective dose as might be done for effects exhibiting thresholds when
-------
23
dose-response data is available. In this regard, the criteria for oncogeni-
city, mutagenicity, and, with the above caveat in mind, reproductive and
developmental toxicicy should be compared to the criteria for effects which
exhibit thresholds. In these factors, no secondary weights are used; a
substance's rank depends only upon the lowest recorded dose producing the
effect under consideration.
Acute Lethality. Table 1.5 compares the ORNL and HAPPS criteria for
the acute lethality factor. Both systems base their rankings on data for
lethal dose and lethal concentration. Rather than retain the standard toxi-
cological term "acute toxicity" used in the ORNL procedure, the name of the
factor was changed to "acute lethality" in HAPPS to reflect this use of data
on lethal doses. This change emphasizes to the nonexpert that only data on
lethal doses should be used in scoring this factor and, in fact, HAPPS speci-
fies that only data so identified in RTECS be used. Data on chronic lethality
and nonlethal effects both chronic and acute are to be considered in scoring
under the factor for effects other than acute lethality. If only chronic
effects were considered in the factor for effects other than acute lethality,
then the two factors would correspond relatively well to the traditional acute
and chronic categories in toxicology. Since both factors are scored using
distinct sets of toxicological data as reported in RTECS, the approach taken
in HAPPS corresponds to the available data. An appropriate balance between
the two factors was obtained by the assignment of their relative weights in
developing an overall score for the toxicity group.
The major difference between the two procedures was the deletion
of criteria for the intravenous, subcutaneous, and intraperitoreal routes of
administration from HAPPS. While these routes of administration are important
in detailed toxicological evaluation, they were considered less closely
related to exposure through the ambient air than the inhalation, oral, and
dermal routes. In addition, these later three routes were the only three
considered by the Iteragency Testing Committee^ and in the Michigan Critical
Materials Register*7 which were used in developing the ORNL procedure. ORNL,
however, opted to be more detailed in their approach.
A second difference between the two procedures does not show up on
Table 1.5 and concerns the definition of the period of exposure which defines
-------
24
^
^^
03
"3
1*
jj
2
^
u
3
-3
"^
U
^
w
4
^
, ,
s~'
'J*"l
*
«*
w
*
^
"^
."*
?"*
a
z
z
O
^
^,
i.
^
1
'
1
41
y
«
.3
a
a
a
^
c
a
Is
a
41
U
3
a
a
a.
x
u
"3
u
A
U
3
^
o
c
^
V
u
2
a
^
^
a
a
X
'J*
1
1
3
5
u
y
a
*
a
w
O
e
a
u a
fl 9
"4 y
«» >-<
~
e
**
u
a
u
40
*
a
a
a
a
~-«
9
U
O
i
%
S *
3 C
w
4
^
41
jg
w
' «
,fl
1
4)
4
U
*<
VJ
>^
W w
2
~ ~
i. 3
=
^
-«
O O
a o o o
o a o a ><
1 1 NX 1 1 1 1 1 /\ 1
«i a a >
a a -I
u
3
o a
a a a «
a a a a a n 4
1 in m I i I 1 I in u
\y\^ u*imaayN>a
' ^ a - z
fl
M
5J
a z
a
a a a
a a a - a
__ a a a a a
u^ i ^-. i i i i i a
X. 'v- -3 a a a
'^ m a a /v
vrt a
m
a
a a a
a a a a a
u*^ i/> m in ^n a
X, I X- ' ' ' ' ' "*
m ^i a a /N
-x"l O
a a a a a
.5 .3 .3 .5
c c c c c
! ccaceacccce
sasaasa^aao
S|'-3S--S2S33
^j*;^;^^^^S
'-""-O^O^OOCOOO
sj " 1 ,-"» O *
*
*« *
a
x u
"* U*
3
C *
U O
a. a
a x
< 4)
c »
.* 4
«
e c
u o
> u
* QO ^
u
U CM
U ^
at u
u
4) -O
U C
^ D
U "
01
W3 3
SU O
£U U
< " C
=: o «:
4* U 3 i
« y
u *j «l3 *
3 M
u u a
a u '«.
a 41 a c
S a < 3 -<
a -s o« a
.- - ecu
u «* « 4) C
« a « >
v a a ^
« -u u
- u X
-a a = »j
41 C
a" g a ^
~- a u
u 10 u rj u
uS ^s 1)
^ oo a
« « . a u
C. a 41 J w
a ii 3 s
a a a u
u a - * TJ a
30 a s
3 3 ac
S u c- o =
^ a 4) 3a *^
O vu, 2 a
* j= a 3
> u U
as - x
il c. 3 a j
C 3. MJ 41
u i 50^ ** a
h j; ao u «
a e -v. g ao a u
S 3 30
"31 -0
U f*H » ^
. s * | 3
_ .-, « a * i
-*( ^ w ei 3 ?!,
^3 JS ^ W 5 5J ''
^ C ^ «) ^ ^
*< w >~ O a - 3 a
41 e -35
2C 41 a 4) Sfl
a .. 4)
- s i,
a a i. 41 i-
i. O < SI
< y a s - a
« -3 U -3 a *
-------
25
an acute exposure for the inhalation route. The ORNL procedure and its
predecessors follow the standard toxicological definition by considering
inhalation exposures as acute if they are under about 4 hours in duration but
do allow the use of professional judgment in evaluating studies with longer
exposure times. Expecting such judgments would, however, violate the HAPPS
guidelines and in HAPPS, inhalation exposure times up to 24-hours in duration
would still be scored under the factor for acute lethality. This value was
chosen because it covers the complete range of the current short-term ambient
air quality standards and because it is frequently used in discussing the
health effects of air pollution.
As shown in Table 1.5, only inhalation doses for gases given in ppmv
could be scored using the ORNL scale. RTECS gives exposure data for solids in
units of mg/nH which cannot be meaningfully converted to ppmv, a natural unit
for gases because equal volumes of different gases at the same conditions of
temperature and pressure contain equal numbers of moles. Reference 25 gives a
toxicity ranking for dusts and mists in mass concentration units and the
corresponding volumetric concentrations for particular gases. For example, a
particular mass concentration of a certain solid might be considered "highly
toxic" while a particular volumetric concentration of a certain gas would also
be considered "highly toxic." Only two pairs of corresponding concentrations
were available. In Fig. 1, these pairs have been plotted on log-log graph
paper and the straight line labeled "Acute Lethality" has been drawn through
them and extropolated. The line so generated was used to establish the mass
concentrations measured in mg/nr* coresponding to the ORNL volumetric concen-
trations in ppm's. In this way, the HAPPS scale for solids was constructed.
Log-log paper was chosen for the extropolation because the toxicity scales are
generally close to being logarithmic rather than linear as might be expected,
since the exposure ranges involved in toxicological experiments can cover
several orders of magnitude. Also shown in Fig. 1 is a line representing the
correspondence between volumetric concentrations of gases and mass concentra-
tions of solids for use in ranking existing standards. The development of the
points through which this line is drawn is described later in this section.
Although developed heuristically, three of the points lie on a straight line
that is exactly parallel to the line drawn for acute lethality. This paral-
lelism shows some degree of underlying consistency between the approach used
-------
26
e
o
H
4J
C
01
CJ
33
en
a
u
4J
IU
o
c
01
O
U-l
en
a)
^H
S3
y
C/3
eo
(Aiudd) UOIJBJ4U80UO
-------
27
for acute lethality and the approach used for existing standards and lends
support to the extropolation used for the "Acute Lethality" line. Another
approach to ranking both gases and vapors would have been to have converted
the volumetric concentrations for gases to mass concentration units. Since
such conversions involve the molecular weight of the gas, this approach could
lead to cases where two gases with equal lethal volume concentrations but with
different molecular weights would be given different weights on a mass concen-
tration scale. On the ORNL scale for volumetric concentrations they would
receive the same rank. In addition, Ref. 25 indicated that the mass concen-
tration scale should be used for dusts and mists. To avoid inconsistencies
with the established volumetric concentration scale for gases and in view of
the recommended use of the mass concentration scale, separate scales for gases
and solids were included in HAPPS.
A final difference concerns the relative weights assigned doses greater
than those corresponding to item 6 in Table 1.5. In HAPPS, these exposure
levels are all scored as having a weight of zero, the same weight assigned to
negative evidence. For the inhalation route, the dosages associated with
these items are all higher than 50 ppm, a value rarely exceeded in the ambient
air unless under accidental conditions. In the absence of other indicators
and since the effects being considered exhibit thresholds, these high dosages
were thus considered as poor indicators of a need for concern by air programs.
Little need for additional study seemed warrented if the threshold was
unlikely to be exceeded. Death from nonthreshold effects like cancer, fetal
death, and death from developmental disorders are considered in the factors
for oncogenicity, mutagenicity, and reproductive and developmental toxicity.
Criteria ac the corresponding dosage levels for the oral and dermal routes of
administration were also given weights of zero, because they are obviously of
less concern to air programs than the inhalation route.
The remaining differences between HAPPS and the ORNL procedure were
made for reasons similar to those discussed above in connection with other
factors. Given the concerns of air programs with human health and exposure
via the ambient air, extra weight was assigned to substances documented as
being lethal to humans by inhalation at the dosages deemed to be of interest
(items 1-6). The ORNL criterion for low or no biological activity was modi-
fied to require negative evidence in order to remove any need to make an
-------
28
expert judgment as to what constituted low biological activity. In HAPPS, low
biological activity is operationally defined as dosages above those corre-
sponding to item 6. As with the previous three factors, a criterion for no
data was added.
The criteria for animals at low and moderate doses (items 3 and 6) are
identical in HAPPS and the ORNL procedure and served as benchmarks. The
criteria listed under item 10 were also considered sufficiently related to
serve as a third benchmark. Because HAPPS combined the three ORNL criteria
for the highest dose ranges (items 7, 8, 9), retention of the ORNL score of 6
for the benchmark item 6 would have created a large gap at the bottom of the
HAPPS scale between the criterion for animals at moderate doses (item 6 with a
score of 6) and the criterion for high doses (item 7 with a score of 9). To
avoid distortion from such a large gap, a scale factor of 1/3 was used to
reduce the HAPPS weight for item 6 to 2.0. The weights for items 4 and 5 were
determined by interpolation between the benchmark items using a factor of
1.1447 (6 x 1.1447^ = 9) appropriate to three equal steps. Also, as with the
previous factors, the weights for the criteria for human noninhalation expo-
sures (item 2) and human inhalation exposures (item 1) were determined by
extropolation using a factor of 1.25.
As noted for the mutagenicity factor, the interpolation and extrapola-
tion from the benchmarks leaves a certain degree of inconsistency in the HAPPS
weights. For example, the ratio of the weights for human exposures by differ-
ent routes of administration should be the same regardless of the dose level.
However, in the low dose range the ratio is about 1.3 («4.7/3.7) for items 1
and 2 while in the moderate dose range the ratio is about 1.1 (=s2.6/2.3) for
items 4 and 5. Similarly, the additional importance attached to human data in
comparison to animal data should be independent of dose range for a given
route of administration. However, for inhalation exposures, the ratio of the
weights for humans and animals at low exposure levels is around 1.6 (**4.7/3.0)
(items 1 and 3) and around 1.3 (= 2.6/2.0) at moderate doses (items 4 and 6).
As previously noted, this type of inconsistency was deemed less important than
retention of the relative importance of the benchmarks; and despite this
inconsistency, the ratios are fairly close to being equal.
Secondary weights were not assigned for satisfying more than one
criterion under acute lethality. Since these lethal effects generally exhibit
-------
29
thresholds, the best indication of the need for further assessment of a.
substance was considered to be the lowest documented dose. In other words, a
substance causing death at 1 ppm and 1000 ppm was not considered a better
candidate for further assessment than a substance causing death at 1 ppm.
Both substances should be and, under HAPPS are, given equal weights, assuming
that both sets of data corresponding to 1 ppm are for the same species and
route. This situation is different than for the nonthreshold effects, onco-
genesis and mutagenesis, for which all the criteria satisfied by the data for
a substance are considered in developing an overall score for a particular
factor.
Effects Other than Acute Lethality. Unfortunately, little data is
available for scoring this factor. Ideally, this catchall factor could be
split into several categories such as acute (nonlethal), subacute (lethal and
nonlethal), and chronic (lethal and nonlethal) with, perhaps, additional
breakdowns into toxic and irritant effects. It is precisely some of these
chronic and irritant effects which correspond to the types of exposures and
effects frequently of concern in dealing with populations repeatedly exposed
to low concentrations over long times. In dealing with toxic pollutants,
however, air programs is primarily interested in serious health effects. Very
little data on chronic exposures is available due to the time and expense of
conducting controlled chronic experiments. Since the data necessary to make
distinctions between various types of effects are not readily available, all
effects other than acute lethality are scored under a single factor in HAPPS.
HAPPS is also an extension of the ORNL procedure which specifically requires
chronic exposures with repeated doses over a period of time like weeks or
months. In fact, the QRNL factor is titled to indicate that chronic toxicity
is being scored. The extension in HAPPS to include acute nonlethal effects
makes fuller utilization of the data available in RTECS. It should be noted
that the terminology used in HAPPS is employed to match the material presented
in RTECS which itself does not always employ the terminology of standard
toxicological protocols and that the studies reported in RTECS were not always
done under standard protocol. For example, lethality should not occur during
standard chronic tests so that the term chronic lethalitv should not be used
-------
30
for standard tests. However, since it is used to describe some tests reported
in RTECS, it was considered appropriate to use it in HAPPS.
Another important difference between the two procedures involves the
absence of the severity subfactor used in the ORNL procedure. A substance
is scored by the dose at which an effect is observed and separately according
to the severity of that effect, for example, severe incapacitation or mild
incapacitation. Then the overall ORNL score for chronic toxicity is obtained
by multiplication of the dose score and the severity score. HAPPS uses only a
single score which reflects dose but not severity. No standard list of
effects categorized by severity was found although typical effects in similar
categories of severity are given in Ref. 17. Assignment of effects to the
severity categories would require expert judgment and would probably vary with
the operator. Thus, assignment of severity rankings was not acceptable for
inclusion in HAPPS. Even if a standard list of the relative severity of
different effects were available, the primary literature would need to be
consulted to determine the nature of the effect in a majority of the cases
reported. RTECS distinguishes between toxic effects and irritant effects and
ranks the latter according to the degree of irritant severity. However, only
dose data with no indication of the type of effect, let alone the severity of
the effect, is given for toxic data. Such data could only be found by refer-
encing the primary literature, a procedure precluded by the guidelines for
HAPPS as being too detailed at this preliminary stage. In summary, all
compounds have the potential to elicit deliterious health effects at some
level; however, this methodology emphasizes effects related to carcinogenesis
and lethal effects as being most important to air programs at this time and
does not address the character or severity of other effects. This procedure
was considered reasonable given the preliminary nature of ranking by HAPPS,
the judgemental nature of ranking these other effects, and the improbability
that such effects would occur at ambient concentrations. Dropping the ranking
based on severity also makes the schemes used in HAPPS for the factors for
acute toxicity and for other effects consistent in structure and application.
Having noted these differences, the HAPPS criteria for effects other
than acute lethality and the ORNL criteria for the dose component of chronic
toxicity in terrestrial animals are compared in Table 1.6 The ORNL procedure
did not provide a scale for scoring inhalation data given in ppm. Since such
-------
31
Table 1.6 Criteria for Effects Other than Acute Lethality
Item
1
2
3
4
5
6
7
8
9
Primary
Weight
7
6
5
4
3
2
1
0
0
Species
Human
Human
Human
Human
Animal
Animal
Human/
Animal
Human/
Animal
Human/
Animal
HAPPS3
Exposure Route and Dose°
Inhalat ion
Gas Solid6 Oral Dermal
,<1 %<10
a a
1-10 10-100
1-10 1-10
10 MOO >10 >10
Negative or insignificant
evidence.
No data.
ORNLa'c
,d
Score Dose
-
-
-
-
3
-
-
-
-
<1
2 1-10
1
Low or no
biologica
activity.
MO
1
aMost criteria given in shortened form; complete specifications of
the HAPPS criteria are given in Appendix A.
''Dosage units: Inhalation - ppmv for gases, mg/m^ for solids
Oral - mg/kg
Dermal - mg/kg
CORNL procedure considers only chronic toxicity.
dAdapted from Ref. 21.
eDerived from gas (ppm) scale by using acute lethality line in Fig. 1.
-------
32
data is Che primary concern of air programs, an inhalation scale was developed
for HAPPS. Reference to the ORNL scales for chronic toxicity (see Table 1.5)
shows that at low and moderate dose ranges, the numerical values of the
concentrations defining the ranges for ranking inhalation and oral exposures
are identical. For inhalation and dermal exposureSj the defining concentra-
tions are identical or comparable. The assumption was made that a similar
equivalence was reasonable for both chronic and acute nonlethal exposures.
The HAPPS scale for inhalation data for gases in volumetric units (ppm) thus
assigned defining concentrations in ppm's numerically equal to the ORNL doses
(see items 5, 6, and 7). The corresponding mass concentrations for solids
were taken from the acute lethality line in Fig. 1 again assuming that the
correspondence developed for acute data should be reasonably valid for chronic
data as we 11.
HAPPS distinguishes between effects in humans and those in animals
except at high doses above 10 ppm for gases and 100 rag/m^ for solids for in-
halation exposures and above 10 mg/kg for oral and dermal exposures (item 7).
Extra weight is assigned to human data to reflect the air programs' charge to
protect human health. The distinction was not made at the high dose levels,
because it was felt that human exposures at these high levels would probably
be experienced only accidentally and even then probably only in occupational
settings. Both of these types of exposures were felt not to be indicative of
the situation that would be regulated by air programs. Thus, human data at
high doses was not considered a better indicator of increased need for addi-
tional study than animal data at high doses.
At lower dose levels (items 1-6), HAPPS always weights human data more
heavily than animal data whereas the ORNL procedure weights both types of data
equally. This additional weight reflects air programs' primary concern with
human health but, as just discussed, was not assigned to human data at high
dose levels. It was recognized that information on human toxicity is rarely
available for this factor. However, where such information is available or
where it becomes available in the future it was felt that the significance of
human toxicity data should be emphasized by ranking it above animal toxicity
data at dose levels near those expected in the ambient air. Similarly, in the
HAPPS system, human data at moderate doses (items 3 and 4) is ranked above
animal data in the lowest dose range (item 5). This differs from the ORNL
-------
33
procedure and from the HAPPS assignment of weights for acute lethality in both
of which the range in which the dose lies is the primary determinant of the
weight assigned. Except where acute lethality is concerned, it was considered
more appropriate to weight evidence from huraans more heavily than evidence
from animals at the low and moderate dose levels. Such a weighting seemed
reasonable in view of the many different types of effects and levels of
severity of effects which were subsumed under this factor. For example, if
dose were the primary determinant of weight, mildly incapaciating effects at
low doses would be weighted more heavily than severly incapacitiating but
nonlethal effects at moderate doses. Modification of such an assignment would
require consideration of the severity of the effects. Since no way to incor-
porate relative severity was available to HAPPS, it was decided to give more
weight to effects in humans regardless of the dose involved as being more
valid indicators of the need for additional study in a program concerned
primarily with human health. Exactly as for acute lethality, additional
weight was assigned for inhalation effects in humans as being most closely
related to the goals of air programs.
Since the ORNL procedure did not distinguish between data based on
human exposures and data based on animal exposures, the choice of benchmarks
was somewhat more arbitrary for this factor than for those discussed previ-
ously. By using items 5, 6, and 7 as benchmarks and extrapolating, the ORNL
weights were consistently applied to animal data at high, moderate, and low
exposure levels. With these benchmarks, the HAPPS weights are easily obtained
by extrapolation using the 1.25 factor and rounding the results to the nearest
integer for convenience. For example, the weight for item 3, two steps above
benchmark item 5 is 5 (3 x 1.252 « 4.68 a 5).
Additional weight was not assigned when a substance satisfied several
criteria simultaneously. The reasons given for acute lethality also apply to
this factor for other effects. Finally, & criterion for no data has again
been included in HAPPS.
Potential for Airborne Release. As noted in Sec. 1.1, this factor in
HAPPS contains two subfactors, production volume and vapor pressure, each of
which either is or is associated with a separate factor in the ORNL procedure.
Both of these subfactors are being used as rough indicators of potential
-------
34
exposure and do not reflect other factors like the fraction of production
emitted to the atmosphere and the number of people potentially exposed.
However, all attempts in the literature reviewed to improve upon the use of
these particular subfactors appeared to require data which was generally
unavailable or was available under the 8(a)-Level-A rule of TSCA and thus not
specific to exposure via the air. When actually implemented, the systems
reviewed frequently relied on default values for many of the substances
scored. Other proposals have been made such as using the labor intensiveness
or price per weight as indicators of how valuable a substance is and thus how
well its release might be controlled. However, the indications were that the
data required for such efforts are not readily available and that such
approaches suffer from as much, if not more, uncertainty than is associated
with figures for production volume. In addition, measures of market economics
such as downward trends in production volume were not considered in the HAPPS
decision process because such data are not readily available and were consi-
dered too detailed for this effort. Thus, production volume seems to be the
only simple, easily accessible surrogate for exposure. In HAPPS, production
volume information was supplemented by scoring a substance by its physical
state: solid, liquid, or gas and within liquids by vapor pressure as an
additional measure of the potential for a substance to be released into the
atmosphere and thus as an additional rough indicator of the quantity poten-
tially released.
Table 1.7 compares the two sets of criteria for scoring production
volume. Little need be said to compare the sets because the criteria are
essentially identical except for slight differences in rounding and the
addition of the criterion for no data (item 8) to HAPPS. The relative weights
between various criteria are the same in the two systems and the additional
criterion for no data has again been weighted at the same level as the lowest
weighted criterion used when data is available.
Both sets of criteria are structured so that substances with high
production volumes are considered more likely candidates for additional study
than substances with low production volumes. Of course, some substances like
sucrose with very high production volumes but with no adverse effects could be
ranked higher than a substance with little production but with very high
toxicity. However, the reverse situation could occur when the factors related
-------
35
03
0)
c
5
^
0
c
o
"jj
u
3
T3
O
ti
(X
w
o
U-i
Q
^
w
0)
u
*
w
u
r^»
*
«*
-s
^*-»
jS
»«4
v^
^
^^
U
>^
**x.
&c
^
vC
O
*
ao
gj
3
E
01
4J
i-"
vC
O
O m
ON ^D
/\ ON
^«
O
en
O cs
m i
sr o
/\ in
O 00
_4
0
O i
O vn
C 1
o
/N. O
o
.^
o
t^! f^)
in CM
* 1
/\ O
l/N
^
O
-i CO
1 CM
ON ON
ON \O ON ON ON
ON O C ON
1 in 1 O 1
VO 1 NO 1
O ON ON V
m ON f^ CT*
m
in in
in co sj
H <*4
ac
c
01 0)
> i^
^4 CC
ac
m
0) .M
u u
a o>
^j
<8
*4 \*
1)
u M
-i a.
1- 0.
0 <
^i
tn oi
O J=
-2
^M4
CM
1)
3£
£
u
vu
o
41
jj
a.
«
*c
<^
y
-------
36
co coxicity were scored and even if a substance like sucrose were prioritized
at a relatively high level, review of the list prior to additional study
should eliminate most such anomalies.
The physical state of a substance may affect the potential of that
substance for being released to the atmosphere and thus the quantity of the
substance actually released. The subfactor for vapor pressure really looks
at the physical state and, for liquids, the vapor pressure of a substance.
Criteria for vapor pressure are presented in the ORNL procedure as guidelines
for use in scoring the level of potential occupational exposure when actual
exposure concentration data is unavailable. As modified for use in HAPPS, the
scale for scoring vapor pressure is identical to one used in the MITRE scoring
procedure (Ref. 9). During testing of HAPPS, it was found that vapor pressure
data was unavailable for some substances for which boiling point data was
available. A toiling point scale equivalent to the MITRE boiling point scale
was added to HAPPS for use when vapor pressure data was unavailable. Table
1.8 compares the two sets of criteria. The weights in HAPPS reflect the fact
that gases are generally more difficult to contain than liquids and solids and
hence, other things being equal, will be released in greater quantities. Two
liquids with equal production volumes will be emitted in proportion to their
vapor pressures if all other factors are equal. In the ORNL procedure for
scoring occupational exposure, the type of process is categorized by the
degree of containment: open, controlled release, or enclosed. Such con-
siderations would also be relevant to determining the quantity released to the
ambient air but the required data would not normally be available and such
determinations would probably require expert knowledge of the specific proces-
ses involved. Such considerations would be unsuitable for HAPPS and are too
detailed for a preliminary prioritization. Gases were ranked above all other
forms of substances in HAPPS as being the most difficult to contain and thus
the most likely to be released in large quantities. Solids are weighted equal
in importance to highly volatile liquids, not at the lowest level of impor-
tance where their very low vapor pressures would place them. However, the
factor of true concern is the quantity of material released and vapor pressure
is not the only consideration. The ranking for solids in HAPPS was based upon
consideration of the importance of solid particulate matter air pollution and
EPA's assessment of how important they considered solids compared to various
-------
37
CO
11
u
3
co
u
ex
^4
0
o.
CO
"*
M
o
U-l
CO
4
i.
4>
4J
*4
u
c^.5
00
H
u
2
CO
f4
O
.a"
j
*y
ex
0
/ s
oo
I
a.
^
A
a
3
cy
'J
O
M
to
a.
*
4
c
" H
^
1
CN
^.
CN
1
O
0
"~*
o
o
4
/\
-o
»H
o
to
CO
CO
u
01
0
o
to
.J
2
H
0
U N
u
50«
C ^w'
c
H
mt
^
0
Q
U
kl
3
a
go /-N
c
\
U £0
4 """
IH E
^ 'SM'
C
-3
>
jj
A
5C
-w
4«
II XI
II XI
1 1 X 1
1 1 X 1
1 1 1
1 1 1
1 1 O CO vo i
e
O
o «
« -. 0
o v o
CO Ol -H 1
v « /\
a, v a,
CQ , &
o
CO
CO
"O J-J
M -! CO
CO ' "O
0 0
CO O
z
0
0 0
o i sr
-1 V CN
XN ex v
a. > a.
> V >
-3-
CN
*^ f*") O CN * 1 .
2
!
S
£ 1
^
. JN t* ^ u-i vo r»
c
H
C
Ol
>
f^
ao
Ol
m
CO
*4
t^
01
iJ Ol
H JJ
M 3
to o u
OH -H
Cu tO 4J
< (X Ul
=: o- J2
< 3
C S W
H
a; cij
1) J2
03 OJ CQ
CQ CQ
OJ U-l
-* O "3
0) U
L* CO CO
C 3
01 O
C -! CO
O co
-2 O u
k- -! C
.«H I4_t >^
CO "^ O
o a.
U Ol
o a. so
"*-l <0 C
1-4
_4 4) 1
CO iJ -H
* 4J O
ij -H .0
e a
US Ol
tj O »C
O U -u
O.
. * M
? t 2
HO .0
u y-i co
O
C CU CO
-< .u e
P 3
T3 O
01
O * M >"
Ci. 60 3 U
S ec e£
C «J eo
cj i Oi E
u . k c
o a < a. w
3 w >n
4J -H X 1-
W * C "3
t. O T) C. H
O C ^3 w
u U > 6.
ti a a. a
e C e. "-I -s
c -z. < <
to £i y TJ
-------
38
liquids in terms of the quantity released, other things being equal. With
this ordering of the criteria, the actual weights were assigned. Once again,
a scaling factor of 1/3 was used in HAPPS. The identical criteria corre-
sponding to items 3 and 4 were used as benchmarks to the HAPPS weight for item
3 was 3 (10/3 - 3.3 « 3) and the weight for item 4 was 2 (8/3 = 2.7 «2). For
item 4, the weight was truncated to 2 rather than rounded to 3 to keep three
of the criteria (items 2, 3, and 4) from receiving equal weights. The weight
for solids (item 2) also received a weight of 3 equal to the weight for highly
volatile liquids as discussed in the previous paragraph. The weight for gases
was obtained by increasing the weight for highly volatile liquids by the same
factor as ORNL used to go between moderate and high volatility liquids. This
factor can be found from
(ORNL Score for High Volatility Liquids) = (ORNL Factor) x (ORNL Score
for Moderate Volatility
Liquids)
or 10 = (ORNL Factor) x 8 and
(ORNL Factor) = 10/8.
The HAPPS weight can be calculated taking the scaling factor into account:
(HAPPS Weight for Gases) = (ORNL Score for High Volatility Liquids)
x (ORNL Factor) x (Scaling Factor)
- 10 x (10/8) x (1/3) - 4.16 * 4.
As had been done for several other factors, the weight for low volatility
liquids which corresponded to two separate ORNL criteria (items 5 and 6) was
taken as the geometric mean of the corresponding ORNL criteria. Thus, item 5
received a HAPPS weight of 1:
(Weight for item 5 in HAPPS) = (Geometric Mean of ORNL Scores for
items 5 and 6) x (Scaling Factor)
= VTTl x (1/3) * 0.87 "* 1.
As was the case for previous factors, the criterion for no data was weighted
at the lowest level of importance.
The principal difference between HAPPS and the ORNL procedures is that
the scores for production volume and vapor pressure are multiplied together in
HAPPS to obtain a score for the potential for airborne release rather than
being used separately as individual scores for separate factors. Combining
-------
39
the subfactors provided a convenient way of placing the source-related surro-
gates for exposure into a single factor. As discussed in Sec. 1.4, this
source-related factor is combined with the receptor-related bioaccumulation
factor in scoring the exposure group. The individual scores were multiplied
rather than added for several reasons.16 First, at the factor level, it was
deemed desirable to avoid the problem of weighting the individual subfactors.
Subjective weighting decisions are always required when scores are added in
procedures like HAPPS. (The interfactor weightings used in HAPPS are discus-
sed in Sec. 1.4.) Second, use of the multiplicative method normally provides
a wider range of scores than does addition. The wider range, even when
normalized, tends to avoid equally weighted substances.
Bioaccumulation. The criteria for bioaccumulation are compared in
Table 1.9. No summary of bioconcentration factors seem to be available so the
criteria based on this parameter were not included in HAPPS. Instead, the
criteria were based on the octanol/water partition coefficient which is
related to the tendency of a substance to accumulate in fat rather than water
and hence to accumulate in .-tniirais. The criteria were based upon the fact
that higher values of the cctanol/water partition coefficient generally
correspond to a substance with a greater tendency to dissolve in and accumu-
late in fat. There are so1 e exceptions to this, particularly for values of
Table 1.9 Criteria for Bioaccumulation
HAPPSa'b
Item
1
2
3
4
5
Weight
10
8
6
1
0
LoglOp
>6
6-4
4-2
s<2
No data
Score
10
8
6
1
-
ORNLa > c
Bioconcentracion
Factor L
>4000
4000-1000
1000-300
<300
-
°£io?
>6
6-4
4-2
<2
-
aP is the octanol/water partition coefficient.
^Criteria given in shortened form, complete specifications
of the HAPPS criteria are given in Appendix A.
cAdapted from Ref. 21.
-------
40
greater than 6 but the data to correct for such exceptions is not
readily available. HAPPS uses the same criteria and weights as the ORNL
procedure uses for items 1-4 and, as with previously discussed factors and for
similar reasons, adds a criterion for no data at the lowest importance level.
Existing Standards. Table 1.10 compares the criteria used in HAPPS
with those used in the MITRE system^ upon which this HAPPS factor was based;
the factor for existing standards was not used in the ORNL procedure. The
criteria for gases are the same except at the low end (item 1) where the
criterion for ranking carcinogens was eliminated from HAPPS. Since HAPPS
assigns weight for carcinogenic activity under the factor for oncogenicity, to
include it again in the factor for existing standards could have resulted in
double counting.
A new sec of criteria for solids was introduced in HAPPS because both
solids and gases are air contaminants. None of the systems reviewed had a
model set of criteria for scoring standards for solids. The HAPPS criteria
for solids were developed by examination of the current OSHA standards. As an
initial attempt, the range of OSHA standards for solids was divided in a
manner proportional to the division of the OSHA standards for gases by the
MITRE criteria eliminating one substance with an extremely small time-weighted
Table 1.10 Criteria for Existing Standards
Item
1
2
3
4
5
6
7
Weight
6
5
4
3
2
1
0
HAPPSa
Gas
(ppm)
^5
5-10
10-25
25-100
100-200
>200
Solid
(mg/np)
$.25
.25-. 5
.5-1
1-5
5-10
>10
No standard
Score
5
4
3
2
1
0
-
MITRE3. b
Gas
(ppm)
<5 or carcinogen
5-10
10-25
25-100
100-200
>200
-
aMost criteria given in shortened fora; complete specifications
of the HAPPS criteria are given in Appendix A.
bAdapted from Ref. 10.
-------
41
average (TWA) from the process. The criteria so construted ran from <.01
mg/m^ for item 1 to >.4 mg/m^ for item 6. However, they were found to
deemphasize the importance of solids with respect to gases. Very few solids
had standards in the low end of the scale so constructed; only four solid
substances would have received weights of four or more given the current
standards while many gases would receive such weights. In order not to
deemphasize solids with respect to gases, these initial criteria for solids
were redefined in a sequence of steps. In the first step, the range of
current standards less than 1 mg/m^ was divided into four ranges corre-
sponding to items 1-4 in the table. The division was done in such a way that
approximately equal numbers of standards fell into each of the four ranges.
For standards above 5 mg/aH, a different approach was used. In this first
step, the second concentration class (corresponding to item 2 in Table 1.10)
covered a concentration range whose width from its lower concentration limit
of 0.1 mg/m^ co £ts Upper concentration limit of 0.5 mg/ta^ corresponded to
a factor of 5. This initial division point between the first and second
concentration classes is shown as the open circle near the line for existing
standards in Fig. 1. Similarly, the width factors corresponding to items 3
and 4 were 2 and 5, respectively. These width factors suggested a repeating
sequence of 5,2,5 as one moved from item 2 to item 3 to item 4. It should be
noted that this 5,2,5 sequence of width factors is destroyed by the final
adjustment of the boundary between the first and second concentration classes
from 0.1 mg/m^ upward to 0.25 mg/m-^ as described below. A tentative width
factor for item 5 was chosen to be 2, the next factor suggested by this
series. Then,
(Upper Bound for item 5) * (Lower Bound for item 5) x (Width Factor)
" (Upper Bound for item 4) x (Width Factor)
* 5 x 2 * 10
and item 6 would then correspond to any standards exceeding the upper bound of
10 mg/m-' established for item 5. The corresponding values for gas concentra-
tions in ppm's and for solid concentrations in mg/m^ were then plotted on
log-log graph paper (see the line for existing standards in Fig. 1). As noted
in the discussion of the factor for acute lethality, three of these points
were found to lie on a straight line parallel to the line already established
for acute lethality. As a last step, this standards line was used to adjust
-------
42
Che boundary between items 1 and 2 upward from 0.1 to 0.25 mg/m^ so that the
adjusted point would lie on the line and correspond to 5 ppmv on the MITRE and
HAPPS gas scales. Similarly, the boundary between items 3 and 4 (corre-
sponding to 10 ppmv) could also have been adjusted upward from 1.0 to 1.2
mg/ar. Such a small change was not considered worth making considering the
preliminary nature of the use of HAPPS and the heuristic method used to
develop the scale for solids.
The weights used in the MITRE system were modified slightly for HAPPS.
Even at the highest concentration levels (item 6), the existence of a standard
was felt to indicate a positive finding of adverse human health impact. It
was desired that the existence of no standard should receive even less weight.
The criterion for no standard was given zero weight and the weights for all
the other criteria were incremented by one over the corresponding MITRE scores
to avoid having any two HAPPS criteria having identical weights.
1.4 FACTOR GROUPS AND WEIGHTS
1.4.1 General
After scoring a substance in each of the factors, HAPPS departs sub-
stantially from the ORNL procedure although there are still some points of
similarity such as the grouping together of related factors. These differ-
ences arise mostly out of the differences in the scopes and purposes of the
two systems. In view of these differences, HAPPS will not be compared to the
ORNL procedure in discussing the groups of factors nor in discussing the final
ranking procedure.
HAPPS ranks substances by proceeding through three levels, beginning
with the most detailed level and aggregating at successive levels to provide a
final single rank for each substance. The first, most detailed level, scores
substances in each of the eight factors chosen as described above in Sec. 1.2.
These scores are chosen by comparing the available data against the criteria
described in Sec. 1.3. At the second stage, the scores in certain groups of
closely related factors are combined to give group scores. Section 1.4.2
describes the groups and the relative weights of the factors in them.
-------
43
c
0
u
N
»»«
CO
£
u
o
z
U-4
0
^,
^J
"^
5
*>H
*
rf
11
o
4J da O
U 3 W
0) O O
IJ 1.
wH '-^ «~1 H
U"\ W-l
. in O
< -H CN «N
ii a ii u
H -^ ^^ «<
X X X X
O >O
O 0
< O CN -H
+ * +
1
60
ri
CU
X
o;
0
CJ
C/5
41
r^ 14 «4 «H
X X X X
u-i O
u-i O
O t -H
OS i '<-> 14
(5 I , 1 1 M 1 1
C^J
! 1 "
T3
1
UM
! O 03
^J
3 (V
"C W
i, *»
N
!!!_'- 3
a. = S
1 .x u a u
!r- v; = o
^ C
g
cs
^1
0
JJ
u
flj
u.
^4
o
U-l
o
CN
1
0
o
c
09
0
o
«J
u.
^
0
o
^H
1
0
U-l
0
ffl
01
u
o
o
CD
IUI
o
01
eo
c
eg
1.
a
01
c
3
X
en
^
01
nH
3
c-
O!
)»4
C
0
ij
u
01
u
^
c
u
c
4J
X
01
c
H
c
o
4
to
CO
3
O
to
1-d
o
01
01
CO
*
CN
M
O
u
u
0
-------
44
Finally, the group scores are combined Co give the overall rank of the sub-
stance as described in Sec. 1.5, the overall ranks of different substances
giving them a numerical prioritization.
At each stage or level, HAPPS normalizes the score or rank to the
maximum value that could be obtained. Thus, the maximum score or rank
at any level will be one. This normalization procedure was adopted primarily
to aid in the assignment of the interfactor and intergroup weights required in
the second and third levels of the procedure. The relative score or rank of
different substances remains the same before and after normalization so the
important information is unaltered. However, the assignment of relative
weights becomes much easier when the factors being combined have all been
normalized to one (or to some other single value). For example, suppose that
two closely related factors were being grouped and that it was desired to give
each factor equal weight in prioritizing substances, that is, both factors
were considered as equally important indicators of the need for additional
study in relation to the group. Table 1.11 illustrates the utility of normal-
ization in this situation. A substance satisfying the midrange criterion
for the first factor and the highest priority criterion for the second factor
(Case I) should be ranked the same as a substance satisfying the highest
priority criterion for the first factor and the midrange criterion for the
second factor (Case II). The table illustrates the problem avoided by nor-
malization. This illustration assumes that equal weights (=1) have been
assigned to each factor. When normalized factor scores are used, a score of
1.0 corresponds to the highest criterion and a score of about 0.5 corresponds
to the midrange criterion for both factors. In the two cases considered, both
substances would receive the same group score (= 1.5) and hence the same
priority based on this group alone just as they should. However, if the
scores for the two factors range over different sets of values, say 0-10 for
the first factor and 0-20 for the second factor, then the unnormalized scores
corresponding to the highest and the midrange criteria depend upon the factor
being considered with the result that the two substances are no longer given
equal group scores. This situation could be corrected in the example by
assigning a weight of 0.5 to Factor 2 but then unequal weights would corre-
spond to equal rankings of importance, obscuring for someone interpreting the
results of a priorization the relative importance assumed for factors and
-------
45
groups. Normalization thus aids in assigning the relative interfactor and
intergroup weights; elements considered equally important could be assigned
equal weights when normalized scores were used without bothering to adjust the
weights for differences in the scales of individual elements. Of course, both
normalized and unnormalized weights would provide the same ranked list of
substances if the unnormalized group scores were corrected appropriately, but
normalization was used in HAPPS to aid the clarity of presentation and to make
the assignment of weights as simple as possible.
1.4.2 Groups
Certain factors are either closely related or are surrogates for the
same effect of real interest. For example, oncogenicity and mutagenicity are
closely related and potential for airborne release and bioaccumulation are
surrogates for human exposure. Scores for each factors are combined together
into group scores prior to final scoring of a substance. This procedure is a
matter of convenience only. The same rankings could be obtained by applying
an appropriately chosen set of weights to the eight factors individually
without going through the intermediate step of scoring within groups. How-
ever, use of groups makes it easier to assign relative weights both within a
particular group and between different groups.
Within a single group, only a few factors need to be considered at one
time. The number of decisions as to the relative weights to be assigned to
various factors is thus reduced to a level where the process becomes more
managable. In addition, the judgments required are between related factors
like oncogenicity and mutagenicity, a situation in which, for example, the
decision to weight them both equally or to give one factor five or ten times
the weight of the other is relatively easy compared to a situation in which
relative weights must be assigned to disparate elements like oncogenicity and
production volume. Of course, the use of groups only postpones the assignment
of weight to disparate elements until the intergroup weights must be assigned,
but because of the grouping of similar factors into groups, the number of
decisions to be made has been reduced and the comparisons to be made are
between the dissimilar groups. Problems encountered when considering similar
and dissimilar elements at the same time are avoided because the similar items
have been combined in the groups.
-------
46
In proceeding in this fashion, HAPPS assumes that linear expressions
are acceptable for prioritizating substances, that is, there are no inter-
actions between factors and/or groups. For example, HAPPS assumes that the
oncogenicity scores for two substances satisfying the same criterion for
oncogenicity should be the same even if one substance is toxic at low levels
and produced in high volumes while the other has substantial evidence of no
toxic potential and is produced in very small volumes. In fact, interactions
between factors may be very important in making decisions and are neglected in
HAPPS. Linear systems like HAPPS cannot account for interactions between
factors or groups; the weights simply tell how important an element is com-
pared to other elements for a specific set of scores for these elements.
Thus, HAPPS is an approximation to an ideal approach but an approximation
which is reasonable given the purpose of the system and the available data.
The groups of factors actually used in HAPPS are shown in Table 1.12
along with the relative weight of each factor within a group. Two of the
groups (items 2 and 5 in the table) contain only a single factor and need not
be discussed further. The remaining three groups are discussed below.
Carcinogenicity Group. As discussed in connection with the oncogeni-
city and mutagenicity factors in Sec. 1.2, carcinogenic potential provides the
Table 1.12 Groups of Factors
Item
1
Group
Carcinogenicity
Factor
Name
Oncogenicity
Mutagenicity
Weight
1
1/4.40
Reproductive and Reproductive and -
Developmental Developmental
Toxicity Toxicity
Toxicity Acute Lethality 1
Effects Other than 1
Acute Lethality
Exposure Potential for Air- 10
borne Release
Bioaccumulation 1
Standards Existing Standards
-------
47
basis for EPA's concern with oncogenicity and mutagenicity. Higher scores for
the two factors separately are intended to reflect increased concern that the
substance being scored is a human carcinogen acting through inhalation. It
was thus reasonable to combine these two closely related factors into a single
group. Since not all mutagens are carcinogens, somewhat less weight was
attached to evidence based on mutagenicity than to evidence based on oncogeni-
city even though some evidence for oncogenicity could relate to noncancerous
tumors. Put another way, the strongest evidence for mutagenicity was consi-
dered to be a less reliable indicator of a substance's carcinogenic potential
than the strongest evidence for oncogenicity. For both factors, the strongest
evidence corresponds to a normalized score of 1.0. For onccgenicity, for
example, a normalized score of 1.0 would require evidence of oncogenicity in
humans by both inhalation and noninhalation routes, evidence in two or more
animal species, and scheduling for carcinogenesis testing under the NTP. The
weighting factor of 1/4.40 ( = 0.23) was chosen so that the strongest evidence
from mutagenicity (normalized score = 1.0) would receive less weight in the
group than evidence of noncogenicity in one animal species (normalized score
» 2/6.25 * 0.32) but more weight than if the only evidence of oncogenicity was
scheduling for testing under the NTP (normalized score = 1/6.25 » 0.16). (The
criteria and scores can be checked by reference to the tables in Sec 1.3;
normalization factors can be checked by reference to the coring sheets in
Appendix A.) Any factor between 0.16 and 0.32 could have been chosen; the one
selected was 0.23 (* »0.16 x 0.32), the geometric mean of the two scores of
interest. The geometric mean was used to keep the relative ratios of the
weights the same, since it is the ratio of weights, not the difference between
them that is the measure of their relative importance. (Values quoted in the
text ha.ve been rounded for presentation and a check of the ratios using the
text values will show a slight inequality.)
Toxicity Group. Two of the factors rank substances by their toxic
effects: acute lethality and effects other than acute lethality (item 3 in
Table 1.12). As discussed previously, both of these factors deal with tradi-
tional toxicological data except for the effects of special interest dealt.
with under oncogenicity, mutagenicity, and reproductive and developmental
toxicity. Generally speaking, the acute lethality factor will score data on
-------
48
acute exposures while Che factor for other effects will score data on chronic
exposures although nonlethal acute effects would also be scored under the
latter factor. Even with this potential mixing of acute and chronic effects
in the factor for other effects, it is reasonable to group both factors
together into a toxicity group which summarize the need for concern based on
data from standard toxicological tests. Were some of the data from other
types of tests, say epidemiological studies of human populations, the grouping
is still sensible, as both factors measure the degree of concern based on
evidence of toxic effects in humans or other species.
Equal weights were assigned to both acute lethality and to other
effects in deriving an overall score for the toxicity group. Since most of
the concentrations used in the experiments which provide the data used in
scoring the two constituent factors exceed the concentrations likely to be
encountered in the ambient air, the weights were chosen based on consideration
of the lowest concentration ranges scored. Furthermore, since the principal
interest of Air Programs is in effects on human health caused by air contami-
nants, the considerations were restricted to inhalation effects in humans.
From the viewpoint of the need for additional assessment, it was felt that a
substance producing acute, lethal effects in humans at very low doses should
be of as much concern as a substance producing other effects (probably chro-
nic) in humans at very low doses, that is, that the two factors should be
equally weighted. Use of weights of one accomplishes this goal. Inspection
of Table 1.5 shows that the normalized score for acute lethality in humans in
the very low dose range is 1.0 (= 4.7/4.7) for inhalation. Likewise, Table
1.6 shows that the normalized score for other effects in humans in the very
low dose range is also 1.0 (= 7.3/7.3) for inhalation. Thus, the choice of
equal weighting factors (=1) does provide the intended equal relative weights
to the two factors for the species, concentration range, and exposure route of
greatest interest in air programs.
Exposure Group. The factors measuring potential for airborne release
and bioaccumulation are surrogates for human exposure. As already pointed
out, they leave much to be desired in terms of being reliable indicators but
within the constraints placed on HAPPs, they appeared to be the only reason-
ably available indicators; detailed exposure analyses must await further study
-------
49
in cases where Che need for concern indicated in prioritizing by HAPPS is
substantiated by additional preliminary information and expert judgment.
In weighting the two factors in the exposure group, the potential for
airborne release was considered to be more indicative of the need for addi-
tional study than the potential to bioaccumulate. Airborne release is more
directly related to the charge of air programs; a substance which accumulated
in humans but whose exposure medium was drinking water would not come within
the purview of air programs. Specifically, a reasonable weighting scheme was
considered to be one which would rank a substance with a high ability to
bioaccumulate approximately equivalent, in terms of public exposure potential,
to a liquid substance with a moderate production volume and a low vapor
pressure. It was also felt reasonable to consider as approximately equivalent
a substance with a high ability to bioaccumulate and a gas with a relatively
low production volume. In both cases, a high potential to bioaccumulate would
correspond to a normalized score of one (see item 1 in Table 1.9). In the
first case, the normalized score for potential for airborne release would be
0.10 (» 4 x 1/40) where moderate production volume has been chosen as item 4
in Table 1.7 and a liquid with low vapor pressure corresponds to item 5 in
Table 1.8. To make this normalized score weigh equally with the normalized
score of 1.0 for bioaccumulation would require multiplication by a weighting
factor of 10.0 (10.0 x 0.10 = 1). Similarly, the normalized score for air-
borne release in the second case would be 0.10 (= 1 x 4/40) where item 7 in
Table 1.7 corresponds to low production volume and item 1 in Table 1.8 corre-
sponds, to gases. This score would also need to be multiplied by 10.0 to
receive equal weight with the score of the high potential bioaccumulator.
Although the use of two independent, reasonable ways of determining a weight-
ing factor would not ordinarily result in equal estimates of that factor, such
was the case here and 10.0 was taken as the weighting factor for the potential
for airborne release factor in the exposure group.
1.5 Intergroup Weights
The final step in the HAPPS procedure combines the normalized scores
for the five groups into an overall score or rank. This combining requires
that the groups be weighted to indicate their relative importance. As pointed
out previously, this task was difficult because it required that decisions be
made as to the relative importance of dissimilar elements. Different indivi-
duals could reasonably be expected to disagree on the relative importance cr
-------
50
weight to be assigned to a. particular group. Procedures like decision analy-
sis capable of assisting in such tasks and of ensuring internal consistency do
exist but could not be applied to HAPPS within available resources. Rather,
Air Programs developed a set of weights believed to approximate fairly well
the importance given to various groups in the past when decisions were
required as to whether or not further assessment of a substance was warranted.
A sensitivity analysis on the weights indicated that shifts in the priority
levels of substances were small for practical purposes. The remainder of this
section discusses this process in greater depth.
Table 1.13 lists the relative weights of the five groups. Three of the
groups (items 1, 2, and 3) deal directly with data related to health. Of
these three, toxicity was considered to be the least important because most of
the concentrations used in developing the data for toxicity would exceed
ambient levels. Although the same is probably true of the concentrations used
in developing the data for the carcinogenicity and the reproductive and the
developmental toxicity groups, the effects considered in these latter two
groups probably exhibit no thresholds; for carcinogens this lack of a
threshold is almost certainly true. It was still felt, however, that tradi-
tional toxicity data by itself should provide sufficient justification for a
closer look at a compound. Making carcinogenicity twice as important as
toxicity was felt to reasonably balance these two considerations. Even though
carcinogens are currently a major concern within EPA, reproductive and devel-
opmental toxicity was weighted as being of the same importance as carcino-
genicity, because both groups deal with severe health effects that may well
occur at ambient levels of concentration.
Table 1.13 Intergroup Weights
Item
1
2
3
4
5
Group
Carcinogenicity
Reproductive and Develop-
mental Toxicity
Toxicity
Exposure
Standards
Weight
2
2
1
5
0.5
-------
51
The exposure group (item 4) was assigned a weight of 5 to make it equal
in importance to the three primary health effects groups together. EPA
considers both exposure and health effects in making regulatory decisions.
For example, it is unlikely that even a potent carcinogen would be regulated
unless there were significant exposure via the ambient air. Conversely, it is
unlikely that a widely distributed substance exposing many people would be
regulated in the absence of severe health effects. A substance ranked at the
top of all three health-related groups (the normalized score for each group
* 1) would receive an accumulated unnormalized score of 5 for these three
groups with the weights shown in Table 1.13 (2x1 +2x1 + 1x1* 5). A
substance ranked at the top of the exposure group would receive an unnormal-
ized score of 1 for group before the assignment of the relative weight.
Choosing a weight of 5 would make the score for the exposure group equal in
importance to that of the three health-related groups together. Although the
choice of weights made in HAPPS clearly cannot reflect all the nuances invol-
ved in considering health effects and exposure for regulatory purposes, the
equal weights given to the exposure group and the three health effects groups
together w2re considered to provide a reasonable approximation to the type of
thinking cfone in the past, particularly in terms of determining the order in
which sub. tances should receive additional study.
The assignment of tha weight to the standards group was perhaps the
most arbitrary assignment of weights. This group was considered important
because it i-dicated a past concern with human health. However, the concen-
trations invcJved and the exposure conditions assumed in setting these stan-
dards are significantly different from those experienced in exposure via the
ambient air. It was fell that a similar degree of concern might be appropri-
ate for tvc substances one of which had a top score of one in either the acute
lethality or other effects factors and no data (score = 0) in the other and
the second of which had a top score of one in the existing standards factor.
In this situation, the toxicity group would have a normalized score of 0.5 (=
[1 xl + 1 x 0]/2) and the standards group would have a normalized score of 1.
To make the standards group equal in importance to the toxicity group in the
final ranking thus required the assignment of a weight of 0.5 to the standards
group. Again, it should be emphasized that both the toxicity group and the
standards group are given low weights because the concentrations needed to
-------
52
elicit the responses dealt with under these groups are generally higher
than those encountered under ambient conditions.
Since there was subjectivity in the assignment of the intergroup
weights, a sensitivity analysis was conducted using different values for
the intergroup weights. The analysis indicates that the assignment of
intergroup weights is not all that critical in determining the rank of a
substance within reasonable bounds. Thus, given Air Programs' own
uncertainty about what weights are best, the overall rank of a substance
is sufficiently accurate for their purposes.
-------
53
2 PRIORITIZATION METHODOLOGY
The first task that must be completed in a HAPPS analysis is daca
collection. All the information necessary to identify the criteria satisfied
for each of eight factors must be assembled. Seven basic reference sources
are needed to complete a HAPPS analysis. Toxicological information, stand-
ards, and some physical and structural properties are obtained from RTECS.^6
The status of a compound within the National Toxicological Program (NTP) is
available from the NTP Carcinogenesis Testing Program list of Chemicals on
Standard Protocol*? or NTPs Annual Plan.28 Production volume data is obtained
from the SRI Chemical Economies Handbook.^9 The state of matter (solid,
liquid, or gas) for the chemical is obtained from The Merek Index^O while
vapor pressure information is from the Handbook of Chemistry and Physics,31
and octonal/water partition coefficients are from Let et al.^2
The procedure for completion of 'a HAPPS analysis is presented in a
worksheet format that leads the analyst step-by-step from data collection
through prioritization of any substance. Worksheets 1-9 are used for data
collection. Each worksheet explains what information is needed and where to
find it. Worksheets 10-17 are used in conjunction with Tables 1-9 and Work-
sheets 1-9 to assign a normalized weight for each of the eight factors.
Worksheet 18 helps the analyst combine the eight factors into five groups and
calculate normalized group weights. The final prioritization is accomplished
using Worksheet 19.
To facilitate their use apart from this document, Tables and Worksheets
used in the HAPPS analysis are presented in the Appendix and are numbered from
1-9 and 1-19, respectively. Abbreviations used in RTECS also are presented in
the Appendix in Tables A-l to A-3.
-------
54
REFERENCES
1. Lewis, R.J., and R.L. Tatken, eds., 1979 Ed., Registry of Toxic Effects of
Chemical Substances, VoLs. I and II, U.S. Dept. of Health and Human
Services, National Institute for Occupational Safety and Health,
Cincinnati, Oh. (Sept. 1980).
2. As till, B.D., at al., Sequential Testing for Chemical Risk Assessment,
Health, Safety, and Human Factors Laboratory, Eastman Kodak Company.
3. Babcock, L.R., Jr., and N.L. Nagda, Popes - Ranking Air Pollution
Sources by Population Exposure, EPA-600/2-76-063 (1976).
4. Cleland, J.G., and G.L. Kingsbury, Multimedia Environmental Coals for
Environmental Assessment, Vol. I, U.S. Environmental Protection Agency
Report No. EPA-600/7-77-136a, Research Triangle Park, N.C. (Nov. 1977).
5. Council on Environmental Quality. TSCA Interagency Testing Committee.
FR 42 No. 197 (Oct. 12, 1977).
6. Carroll, J.W., Formulation and Assessment of Air Pollutant Abatement
Strateaies and Priorities, Task 1: Air Pollutant Prioritization
Methodology, GCA-TR-73-14-G (1974).
7. Carroll, J.W., and N.F. Suprenant, Implementation of the CCA Prioritiza-
tion Methodology for Selected Chemicals, final report, GCA-TR-76-10-G
(1976).
8. Environmental Protection Agency, Pesticide Chemical Active Ingredients;
Proposed Registration Standards Ranking Scheme, FR 45 No. 222 (Nov. 14,
1980).
9. Fuller, B., et al., Preliminary Scoring of Organic Air Pollutants,
PB-264442, MITRE Corp., McLean, Va., METREK Div. (1976).
10. Fuller, B., et al., Scoring of Organic Air Pollutants , MTR-7248, Rev. 1
(1976).
11. Final Report of NSF Workshop Panel to Select Organic Compounds Hazardous
to the Environment (Oct. 1975).
12.' Fong, C.V., and R.J. Clerann, Hazard Evaluation of Hew Chemicals,
Approaches to Level I Test Selection, MTR-79W00347, MITRE Corp., McLean,
Va. (1979).
13. General Procedures for Scoring Air and Water Pollutants, draft report,
Clement Associates, Inc., Washington, B.C.
14. Gevertz, J.N., and E. Bild, Chemical Selection Methods: An Annotated
Bibliography, EPA 560/TIIS-80-001 (1980).
-------
55
15. Griesemer, R.A., and C. Cueto, Jr., Towards a Classification Scheme for
Degrees of Experimental Evidence for the Cardnogenicity of Chemicals
for Animals, reprinted from Molecular and Cellular Aspects of Carcinogen
Screening Tests.
16. Margler, L.W., M.B. Rogozen, R.A. Ziskind, and R. Reynolds, Rapid
Screening and Identification of Airborn Carcinogens of Greatest Concern
in California, JAPCA, 23(11):1153-1157 (Nov. 1979).
17. Michigan Critical Materials Register 1980, Michigan Dept. of Natural
Resources, Lansing, Mich., Publication No. 4833-5324.
18. Michigan Air Priority Chemicals List 1980, Michigan Dept. of Natural
Resources, Lansing, Mich., Publication No. 4833-5324.
19. Preliminary List of Chemical Substances for Further Evaluation, TSCA,
Interagency Testing Committee (1977).
20. Ross, R.H., and P. Lu, Chemical Scoring System Development, draft report,
Oak Ridge National Laboratory, Oak Ridge, Tenn. (Dec. 1980).
21. Ross, R.H., and P. Lu, Chemical Scoring System Development, draft report,
Oak Ridge National Laboratory, Oak Ridge, Tenn. (June 1980).
22. Ross, R.H., and J. Welch, Proceedings for the EPA Workshop on the
Environmental Scoring of Chemicals, ORNL/EIS-158, EPA-560/11-80-010
(1980).
23. Scoring Chemicals for Health and Ecological Effects Testing} TSCA-ITC
Workshop, unnumbered Enviro-Control, Inc., report (no date).
24. Welch, J.L. and R.H. Ross, An Approach to Scoring of Toxic Chemicals for
Environmental Effects, paper presented at the First Annual Meeting of
the Society of Environmental Toxicology and Chemistry, Arlington, Va.
(Nov. 1980).
25. Kohan, A.M., A Summary of Hazardous Substance Classification Systems,
Office of Solid Waste Management Programs, report No. SW-171, Washing-
ton, D.C. (1975).
26. Lewis, R.J., and R.L. Tatken, eds., Registry of Toxic Effects of Chemical
Substances, October 1981, microfiche Ed. No. 210-81-8101, U.S. Dept. of
Health and Human Services, Cincinnati, Oh. (1981).
27. National Toxicology Program, Chemicals on Standard Protocol, Carcinogen-
esis Testing Program, National Toxicology Program, Sethesda, Md. (1982).
28. National Toxicology Program, Annual Plan, National Toxicology Program,
Bethesda, Md. (1982).
-------
56
29. Stanford Research Institute, Chemical Economics Handbook, Stanford
Research Institute, Menlo Park, Calif., (1982).
30. Windholz, M., S. Budavari, L.Y. Stroumtsos, and M.N. Fertig, eds., The
Merck Index, Merck and Co., Inc., Rahway, N.J. (1976).
31. Weast, R.C., ed., Handbook of Chemistry and Physics, The Chemical Rubber
Co., Cleveland, Oh. (1971).
32. Leo, A., C. Hanscle, and D. Elkins, Partition Coefficients and Their
Uses,' Chemical Reviews, 71:525-616 (1971).
-------
57
APPENDIX A
TABLES, WORKSHEETS, AND ABBREVIATIONS USED IN RTECS
-------
Table 1 Criteria and Associated Weights for Oncogenicitya
Primary Secondary
Index Criteria13 Weight Weight
1 Evidence of oncogenicity in humans by 5.0
inhalation route.
2 Evidence of oncogenicity in humans by 4.0 0.7
noninhalat ion route.
3 Evidence of one
ogenicity in two or more 3.0 0.5
animal species by any route of adminis-
tration.0
4
5
6
7
Evidence of oncogenicity in one animal
species by any route of administration.0
Compound scheduled for or currently
undergoing oncogenicity testing.
Negative or equivocal results from
oncogenicity testing.
No data.
2.0
1.0
0.0
0.0
0.3
0.05
0.0
0.0
aMost available data will relate to carcinogenicity .
Table A-l in the appendix for a complete list of abbreviations
used in RTECS , Table A-2 for species abbreviations, and Table A-3
for route of administration abbreviations.
clf the data satisfy the criteria for Index 3 then Index 4 should
not be considered.
-------
Table 2 Criteria and Associated Weights for Mutagenicity
Primary Secondary
Index Criteria3 Weight Weight
1 Evidence of mutagenicity (in vivo) in at least 11.0
one mammalian test species by the inhalation
route. b
2 Evidence of mutagenicity (in vivo) in at least " 9.0 0.7
one mammalian test species by the non-inhala-
tion route.
3 Evidence of mutagenicity (in vitro) in two or 8,3 0.5
more mammalian test species. ci^
4 Evidence of mutagenicity (in vitro) in one mam- 7.7 0.4
malian test species. ^
5 Evidence of mutagenicity (in vivo) in two or 7.1 0.25
more nonmammalian test species by any route of
administration.6
6 Evidence of mutagenicity (in vivo) in one non- 6.5 0.2
mammalian test species by any route of admini-
stration.6
7 Evidence of mutagenicity (in vitro) in two or 6.0 0.15
more nonmammalian test species. ^
8 Evidence of mutagenicity (in vitro) in one 4.0 0.1
nonmammalian test species. *
9 Compound scheduled for or currently undergoing 2.0 0.05
mutagenicity testing.
10 Negative or equivocal results from mutagenicity 0.0 0.0
testing.
11 No data. 0.0 0.0
aSee Table A-l in the appendix for a complete list of abbreviations used
in RTECS, Table A-2 for species abbreviations, and Table A-3 for route of
administration abbreviations.
°If the route of administration is specified the test is in vivo.
clf the route of administration is not specified the test is in vitro.
dlf the data satisfy the criteria for Index 3 then Index 4 should not be
considered.
elf the data satisfy the criteria for Index 5 then Index 6 should not be
considered.
the data satisfy the criteria in Index 7 then Index 8 should not be
considered.
-------
Table 3 Criteria and Associated Weights for Reproductive
and Developmental Toxicity
Index
1
2
3
4
5
6
7
Criteria3
Evidence for reproductive or developmental
effects in humans by inhalation route.
Evidence for reproductive or developmental
effects in humans by noninhalat ion route.
Evidence for reproductive or developmental
effects in two or more animal species by
any route of entry. ^
Evidence for reproductive or developmental
effects in one animal species by any route
of entry. k
Compound scheduled for or currently under-
going testing for reproductive and develop-
mental effects.
Negative or equivocal results from testing
for reproductive or developmental effects.
No data.
Primary
Weight
5.0
4.0
3.0
2.0
1.0
0.0
0.0
Secondary
Weight
0.7
0.5
0.3
0.05
0.0
0.0
aSee Table A-l in the appendix for a complete list of abbreviations used
in RTECS, Table A-2 for species abbreviations, and Table A-3 for route
of administration abbreviations.
^If the data satisfy the criteria for Index 3 then Index 4 should not be
cons idered.
-------
Table 4 Criteria and Associated Weights for Acute Lethality
Route of Exposure and Criteriaa
Index
1
2
3
4
5
6
7
8
9
Species
Human
Human
Animal
Human
Human
Animal
Human/
An ima 1
Human/
Animal
Human/
Animal
Inhalation
Gas(ppm) Solid(mg/mJ)
X<5
-
X<5
550
Negative or
or animals.
No data.
X<50
-
X<50
50500
insignificant
Oral
(mg/kg)
-
X<5
X<5
-
550
results
Dermal Primary
(tng/kg) Weight
-
X<5
X<5
-
5200
in humans
4.7
3.7
3.0
2.6
2.3
2.0
0.0
0.0
0.0
aSee Table A-l in the appendix for a. complete list of abbreviation-
used in RTECS, Table A-2 for species abbreviations, and Table A-3
for route of administration abbreviations.
-------
Table 5 Criteria and Associated Weights for
Effects Other than Acute Lethality
Route of Exposure and Criteria3
Index
1
2
3
4
5
6
7
8
9
Species
Human
Human
Human
Human
Animal
Animal
Human/
Animal
Human/
Animal
Human/
Animal
Inhalation Oral Dermal
Gas(ppm) Solid(mg/m3) (mg/kg) (mg/kg)
xa
-
i10
Negative or
or animals.
No data.
xao
xa xa
io100 X>10 X>10
insignificant results in humans
Primary
Weight
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
0.0
aSee Table A-l in the appendix for a complete list of abbreviations
used in RTECS, Table A-2 for species abbreviations, and Table A-3
for route of administration abbreviations.
-------
Table 6 Criteria and Associated Weights
for Production Volume (PV)
Criteria
Index
1
2
3
4
5
6
7
8
106 kg/yr
PV>450
2301000
5io
-------
Table 7 Criteria and Associated Weights
for Vapor Pressure (VP)
Index
1
2
3
4
5
6
Criteria3'13
VT(mmHg) bp(*C)
Gas
Solidc
VPHOO bp<80
24100
No data.
Primary
Weight
4.0
3.0
3.0
2.0
1.0
1.0
aVapor pressure data should be reported at 25°C
and 760 mm Hg.
°If vapor pressure data is unavailable use the
substance boiling point (bp), at 760 ram Hg, as
a substitute.
CA substance should be considered a solid if its
melting point is greater than 25*C.
-------
Table 8 Criteria and Associated
Weights for Bioaccurau-
lation
Index
Criteria
Primary
Weight
1 Log P>6.0 10.0
2 4.0
-------
Table 9 Criteria and Associated Weights
for Existing Standards
Index
1
2
3
4
5
6
7
Criteria
Gas ( ppm)
X<5
5200
No
Solid
(mg/m^)
X<0 . 25
0.2510.0
standard.
Primary
Weight
6.0
5.0
4.0
3.0
2.0
1.0
0.0
aBased on OSHA time-weighted-average (TWA)
standards or threshold limit values (TLV)
when TWAs are not available.
-------
WORKSHEET 1. CENTRAL INFORMATION*
Chemical Name:
RTECS Number:5
CAS NUMBER:0
MOLECULAR WEIGHT:d g/g-mol
MOLECULAR FORMULA:6
MELTING POINT:f : *C BOILING POINT:f "C
SUBSTANCE PHYSICAL STATE:8 Gas Liquid Solid Unknown
RTECS Edition:h
MERCK INDEX Edition:*
HANDBOOK OF CHEMISTRY AND PHYSICS Ed it ion:J
aAll data taken from the Registry of Toxic Effects of Chemical Substances
(RTECS). The latest microfiche edition of RTECS should be used for the
analysis and can be obtained from the Superintendent of Documents, U.S.
Government Printing Office, Washington, D.C.
bUsing the RTECS data field RTECS Accession Number, enter the RTECS number.
cUsing the RTECS data field CAS NUMBER enter the Chemical Abstract Service
Registry Number.
dUsing the RTECS data field MOLECULAR WEIGHT, enter the molecular weight.
eUsing the RTECS data field MOLECULAR FORMULA, enter the molecular formula.
%sing the most recent edition of the Merck Index, enter the melting/boiling
points of the substance under analysis. If the substances is not in the Merck
Index use the most recent edition of the Handbook of Chemistry and Physics
(Tables of Physical Constants of Organic/Inorganic Compounds) for the melting/
boiling points.
SUsing the information on physical characteristics from the Merck Index and
the melting/boiling point information determine the physical state of the sub-
stance under analysis.
hEnter the edition of RTECS used for this analysis.
i-Enter the edition of the Merck Index used for this analysis.
JEncer the edition of the Handbook of Chemistry and Physics used for this
analvsis.
-------
WORKSHEET 2. ONCOGENICITY
Using Che RTECS data field TUMORIGENIC DATA, record the route of administra-
tion and test species for all citations that show the Toxic Effect (TFX) as
TFX:CAR, TFX:NEO, or TFX:ETA.
ROUTE OF TEST ROUTE OF TEST
ADMINISTRATION SPECIES ADMINISTRATION SPECIES
Using the latest edition of the Carcinogenis Testing Program: Chemicals on
Standard Protocol3 from the National Toxicology Program, record the status
of the substance in the Carcinogenesis Testing Program.
Q Not on list (should not be scored on Worksheet 10)
Q Scheduled for or currently undergoing testing
Q Negative or equivocal results
Edition Usedb:
aA copy can be obtained from: Technical Information Section, Carcinogenesis
Testing Program, National Toxicology Program, Landow Building, Room A306,
Bethesda, MD 20205.
bEnter the edition of the Carcinogenis Testing Program:Chemicals on Standard
Protocol used for this analysis.
-------
WORKSHEET 3. MUTAGENICITY
Using the RTECS data field MUTATION DATA, record the test species and route of
administration (if applicable) for all citations. Below the heading MUTATION
DATA, each mutation data line includes, in sequence, the mutation test system
utilized, the species of the tested organism (and where applicable, the route
of administration or cell type), and the exposure concentration or dose.
TEST ROUTE OF TEST ROUTE OF
SPECIES ADMINISTRATION SPECIES ADMINISTRATION
Using the latest edition of the National Toxicology Program Annual Plan,3
record the status of the substance in the Mutagenicity Testing Program.
Tables in the section on cellular and genetic toxicology should be used for
the following analysis (See for example, pages 35-73, Tables 2-8, of Che NTP
Aur.ual
n Not on list (should not be scored on Worksheet 11)
Q Scheduled for or currently undergoing testing
Q Negative or equivocal results
Edition
aA copy can be obtained from: Technical Information Section, National
Toxicology Program, Landow Building, Room A306, Bethesda, MD 20205.
''inter the edition of the National Toxicology Program Annual Plan used for
this analysis.
-------
WORKSHEET 4. REPRODUCTIVE AND DEVELOPMENTAL TOXICITY
Using the RTECS data field REPRODUCTIVE EFFECTS DATA, record the rouce of
adminiscration and test species for all citations.
ROUTE OF TEST ROUTE OF TEST
ADMINISTRATION SPECIES ADMINISTRATION SPECIES
Using the National Toxicology Program Annual Plan3, record the status of
Che substance in the Teratogenic Testing Program. Tables in the section on
reproductive and developmental toxicology should be used for the following
analysis (sse for example, pages 125-134, Tables 17-20, of the NTP Annual
Plan28).
Q Not on list (should not be scored on Worksheet 12)
Q Scheduled for or currently undergoing testing
Q Negative or equivocal results
Edition Usedb:
aA copy can be obtained from: Technical Information Section, National
Toxicology Program, Landow Building, Room A306, Bethesda, MD 20205.
nEnter :ne edition of the National Toxicology Program Annual Plan used for
th is anaivs is.
-------
WORKSHEET 5. ACUTE LETHALITY
Using the RTECS data field TOXICITY AND DATA REFERENCES, record che route of
administration, test species, dose, and length of exposure for all cita-
tions (other than TFX:CAR, TFX:NEO, or TFX:ETA) that are of the form LDLo,
LD^Q, LCj^Q, and LC5Q that give the route of administration as ihl, orl, or skn
and length of exposure less than or equal to 24 hr.a
ROUTE OF TEST LENGTH OF
ADMINISTRATION SPECIES DOSEb>c EXPOSURE
aAny data for LDLo> LD5Qj LC^o' or ^5Q with exposure times greater than 24 hr
should be included in Worksheet 6 (Effects Other Than Acute Lethality).
^Concentrations for solid substances should be recorded as mg/m3 and cannot
be converted to ppm. If the physical state of the substance is unknown,
record the units given in RTECS and do not attempt to convert to ppm.
Occasionally, the dose for solids will be given in ppia's (usually when the
melting point is near 25*C), the dose then should be recorded in ppm's and
not converted to mg/m^.
cFor gases and vapors, concentrations in mg/m^ should be converted to ppm by:
ppm = rj£ x concentration (mg/m^) @ 25*C and 760 mo Hg
where MW = molecular weight (Worksheet 1).
-------
WORKSHEET 6. EFFECTS OTHER THAN ACUTE LETHALITY
Using Che RTECS data field TOXICITY AND DATA REFERENCES, record the route
of administration, test species, and dose for all citations (other than
TFX:CAfi, TFX:NEO, or TFX:ETA) that are of the form TDLo> or TCLo that give the
route of administration as ihl, orl, or skn. Citations of the form LD^QJ
LDjQ, LCLo> or LG50 cl*ac 8^-ve che route °f administration as ihl, orl, or skn
and length of exposure greater than 24 hr should be included here.
ROUTE OF TEST
ADMINISTRATION SPECIES DOSEa>b
Concentrations for solid substances should be recorded as mg/ai^ ancj cannot be
converted to ppm. If the physical state of the substance is unknown, record
the units given in RTECS and do not attempt to convert to ppm. Occasionally,
the dose for solids will be given in ppm's (usually when the melting point is
near 25*C), the dose Chen should be recorded in ppm's and- not converted Co
'DFor gases and vapors, concentrations in mg/m^ should be converted to ppm by:
ppm = - "*y?, x concentration (mg/nr) @ 25*C and 760 mm Hg
where MW * molecular weight (Worksheet 1).
-------
WORKSHEET la. PRODUCTION VOLUME (PV)
Using the SRI Chemical Economics Handbook find the total U.S. production for
the substance under analysis.
PV - x 106 kg/yr or PV = x 106 Ib/yr
Edition Used3:
aEnter the edition of the SRI chemical Economics Handbook used for this
analysis.
-------
WORKSHEET 7b. VAPOR PRESSURE (VP)
Using Che information from Worksheet 1, record the physical state (at 25*C and
760 mm Hg^ and boiling point (BP) of the substance under analysis.
D Gas
D Solid
Q Liquid BP - _ °C
Unknown
If the physical state is given as gaseous, solid, or unknown, continue on to
Worksheet 8.
If the substance is a liquid use the data from the Handbook of Chemistry and
Physics-^ to calculate the vapor pressure (VP), at 25*C and 760 mm Hg.a Enter
constants A and B, and the temperature range for which the equation and
constants are valid.
Temperature Range: "C
VP » mmHg @ 25 "C and 760 mmgHg
If 25*C does not fall within the temperature range above continue on to
Worksheet 8.
aVP (mm Hg) - antilog10 [(-7.3285 x 10~4 x A)+B]
where A = molar heat of vaporization
B * constant.
Constants A and B are obtained from the Handbook of Chemistry and Physics^*
pages D-151 to D-170 for organic compounds and pages D-171 to D-177 for
inorganic compounds.
-------
WORKSHEET 8. BIOACCUMULATION
Using Table XVII from Leo et al.32, find the MOLECULAR FORMULA (Worksheet l)
of the substance under analysis in the column headed EMPIRICAL FORMULA. Find,
in the column headed SOLVENT, the line corresponding to octanol. Match the
chemical name of the substnace under analysis with the chemical name in the
column headed NAME. Enter, in the space provided above, the value from the
column headed LOGP OCT. If more than one value of the LOGP OCT is given for
the octonal solvent system use the average of the LOGP OCT values. If octonal
is not one of the solvents listed for the given MOLECULAR FORMULA and NAME
enter the average LOGP OCT estimated from the other solvent systems.
-------
WORKSHEET 9. EXISTING STANDARDS
Using Che RTECS data field STANDARDS AND REGULATIONS, record che entry
for the OSHA time weighted average (OSHA STANDARD-air:TWA) in the space
provided below. If OSHA Standard-air TWA is not given use the RTECS data
field REVIEW, record the entry for the threshold limit value (THRESHOLD LIMIT
VALUE-air:) in the space provided below.
STANDARDS AND REGULATIONS: OSHA STANDARD-air:TWA « ppm (gas)a.b
mg/m3 (solid)3.b
REVIEW: THRESHOLD LIMIT VALUE-air: ppm (gas)a>b
mg/m3 (solid)aib
aStandards for solid substances should be recorded as mg/m3 and cannot be
converted to ppm. If the physical state of the substance is unknown, record
the standard in the units given in RTECS and do not attempt to convert to
ppm. Occasionally, the standard for solids will be given in ppm's (usually
when the melting point is near 25*C), the standard then should be recorded in
ppm's and not converted to mg/m3.
"For gases and vapors, standards in mg/m3 should be converted to ppm by:
ppm - 2.4v^.5 x standard (ng/m3) $ 25'C and 760 mm Hg
where MW - molecular weight (Worksheet 1).
-------
WORKSHEET 10. ONCOGENICITY FACTOR SCORE
The completion of Worksheet 10 require* the uae of Worksheet 2.
1. Circle Che index number*, in the t«ble below, that corretpond co Che
criteria that »re satisfied by Che data from Worksheet 2.
2. Starting at Index 1, read down Che Index column until the fint circled
index number i« encountered. Record the Primary Weight anociated vith
that index number in the corre*pondiog Criteria Weight column.
3. Continue reading down the Index column to each successive circled index
number and record the Secondary Weight aitociated with each index number
in the corresponding Criteria Weight column.
A. SUB all value* recorded in the Criteria Weight column and record in the
pace labeled £(Criceria Weight) below.
5. Divide Che reiulcs from Seep 4 by 6.25 Co obtain the Normalized Factor
Score for Oncogenicicy (ONCOnorm).
Criteria and Aisociated Weight* for Oncogenicicy4
Index
1
2
3
4
5
6
7
Criteria*1
Evidence of oncogenicity in huaan* by
inhalation route.
Evidence of oncogenicity in human* by
aoninhalat ion route.
Evidence of oncogenicity in £vo or more
animal ipecie* by any route of adminis-
tration. c
Evidence of oncogenicity in one animal
ipecie* by any route of adminiccrac ion.c
Compound icheduled for or currencly
undergoing oncogenicity testing.
Negative or equivocal result* from
oncogenicity testing.
Mo data.
?rimary
Weight
5.0
4.0
3.0
2.0
1.0
0.0
0.0
Secondary
Weight
0.7
0.5
0.3
0.05
0.0
0.0
Criteria
Weight
_ . w.*^
2,(Criteri* Weight) ONCOnorm
'Host available data vill relate to carcinogenicity.
bSee Table A-l in the appendix for a cooplete Hit of abbreviation* used in
RTECS, Table A-2 for specie* abbreviation*, and Table*A-3 for rouce of admini-
ttration abbreviation*.
clf the data iati*fy the criteria for Index 3 then Index 4 *hould not be consi-
dered.
-------
WORKSHEET 11. MUTAGENICITY FACTOR SCORE
The completion of Worksheet 11 requires Che use of Worksheet 3.
1. Circle Che index numbers, in Che Cable belov, chat correspond to the
criteria that are satisfied by the data from Worksheet 3.
2. Starting at Index 1, read dovn the Index column until Che first circled
index number is encountered. Record the Primary Weight associated with
Chat index number in the corresponding Criteria Weight column.
3. Continue reading dovn the Index colunn to each successive circled index
number and record the Secondary Weight associated with each index number
in the corresponding Criteria Weight column.
4. Sum all values recorded in the Criteria Weight column and record in the
space labeled £( Criteria Weight) below.
5. Divide the results from Step 4 by 12.65 to obtain the Normalized Factor
Score for Mutagenicity (MUTnorm).
Criteria and Associated Weights for Mutagenieicy
Primary Secondary Criteria
Index Criteria' Weight Weight Weight
1 Evidence of mutagenicity (in vivo) in at least 11.0
one mammalian test species by the inhalation
route.b
2 Evidence of mutagenicity (in vivo) in at least 9.0 0.7
one mammalian test species by the non-inhala-
tion route.
3 Evidence of nutagenicity (in vitro) in two or 8.3 0.5
more mammalian test species.c>d
4 Evidence of nucagenicity (in vitro) in one mam- 7.7 0.4
ma Inn cesc species.^
5 Evidence of nutagenicity (in vivo) in tvo or 7.1 0.25
more nonaaomalian test species by any route
of administration.'
6
7
8
9
10
11
Evidence of mutagenicity (in vivo) in one non-
maomalian test species by any route of admini-
stration.'
Evidence of mutagenicity (in vitro) in tvo or
ore nonmamalian test species, f
Evidence of mutagenicity (in vitro) in one
nonmaomalian test species. f
Compound scheduled for or currently undergoing
outagenicity testing.
Negative or equivocal results from mucagenicity
testing.
(to data.
6.5
6.0
4.0
2.0
0.0
0.0
0.2
0.15
0.1
0.05
0.0
0.0
_________________ i t? *< <
Weight) ' '-"-1 MOTnoro
See Table A-l in the appendix for a complete list of abbreviations used in RTECS,
Table A-2 for species abbreviations, and Table A-3 for route of administration
abbreviations.
blf the route of administration is specified the test is in vivo.
clf the route of administration is aot specified :he test is in vicro.
d!f the data satisfy the criteria for Index 3 then Index 4 should aot be consi-
dered.
If the data satisfy the criteria for Index 5 then Index 6 should aot b« consi-
dered.
'If the data satisfy the criteria ia Index 7 then Index 8 should not be consi-
dered.
-------
WORKSHEET 12. REPRODUCTIVE AND DEVELOPMENTAL TOXICITY FACTOR SCORE
The completion of Worksheet 12 require* the uie Worksheet 4.
1. Circle Che index nuabers, in the table below, that correapond to the
criteria that (re aaciafied by the data from Worksheet 4.
2. Starting »t Index 1, read down the Index column until the fir*t circled
index nusber ia encountered. iecord the Primary Weight aaaociated vith
that index amber in the corresponding Criteria Weight column.
3. Continue reading dovn the Index column to each *uccei*ive circled index
number and record the Secondary Weight aiaociated vith each index number
in the corresponding Criteria Weight column.
4. Sun til valuei recorded in the Criteria Weight column and record in the
pace labeled ^(Criteria Weight) below.
5. Divide the results froa Step 4 by 6.25 to obtain the Normalized Factor
Score for Reproductive and Developmental Toxicity (RDTnorm).
Criteria and Aaaociated Weight! for Reproductive and Developmental Toxicity
Index
Criteria*
Primary Secondary Criteria
Weight Weight Weight
1
2
3
Evidence for reproductive or developmental
effecta in humana by inhalation route.
Evidence for reproductive or developmental
effecta in humans by noninhalat ion route.
Evidence for reproductive or developmental
5.0
4.0
3.0
0.7
0.5
effecti in two or more animal species by
any route of entry.^
Evidence for reproductive or developmental 2.0 0.3
effecti in one animal species by any route
of entry.b
Compound scheduled for or currently under- 1.0 0.05
going testing for reproductive and develop-
mental effects.
Negative or equivocal results from testing 0.0 0.0
for reproductive or developmental effects.
No data. 0.0 0.0
2,(Criteria Weight)
T 6.25
SDTnorm
'See Table A-l in the appendix for a complete liat of abbreviations vised in RTECS,
Table A-2 for apecies abbreviations, and Table A-3 for route of administration
abbreviations.
'If the data satisfy the criteria for ladex 3 then Index 4 should not be considered.
-------
WORKSHEET 13. ACUTE LETHALITY FACTOR SCORE
The completion of Worksheet 13 require* the use of Worksheet 5.
1. Circle the index numbers, ia the table belov, chjt correspond Co the
criteria thie are tatisfied by the data from Worksheet 5.
2. Starting at Index 1, read down the Index column until the first circled
index number if encountered. Record the Primary Weight associated vith
that index number in Che (pace labeled Criteria Weight below.
3. Divide the value obtained in Step 2 by A.7 to obtain the Normalized
Factor Score for Acute Lethality (AlZTHnora).
Criteria and Asiociated Weight* for Acute Lethality
Index
1
2
3
4
5
6
7
8
9
Specie*
Huaan
Human
Animal
Huaan
Human
Animal
Human/
Animal
Human/
Animal
Human/
Animal
Route
of Exposure
Inhalation
Ca«(ppm) Solid(mg/m3)
X<5
-
X<5
5OC<50
-
550
negative
human* or
No data.
X<50
-
X<50
50500
and Criteria*
Oral Dermal
(mg/kg) (mg/kg)
-
X<5 X<5
X<5 X<5
-
550 X>200
or in«ignif icant re*ult* in
animal*.
Primary
Weight
4.7
3.7
3.0
2.6
2.3
2.0
0.0
0.0
0.0
Criteria Weight
ALETHnorm
'See Table A-l in the appendix for a complete lilt of abbreviations
used in STICS, Table A-2 for specie* abbreviation*, and Table A-3 for
route of adminiitration abbreviation*.
-------
WORKSHEET 14. EFFECTS OTHER THAN ACUTE LETHALITY FACTOR SCORE
The couplet ion of Worksheet 14 require! Che use of and Worksheet 6.
1. Circle Che index nunbers, in Che table below, that correspond Co Che
criteria that are satisfied by the data froa Worksheet 6.
2. Starting at Index 1, read down the Index column until Che firsc circled
index number is encountered. Record the Primary Weight associated with
that index number in the space labeled Criteria Weight below.
3. Divide the value obtained in Step 2 by 7.0 to obtain the Normalized
Factor Score for Effect* Other Chan Acute Lethality (NLZTHnonn).
Criteria and Associated Weights for Effects Other than Acute Lethality
Route of Exposure and Criteria*
Index
1
2
3
4
5
6
7
8
9
Species
Human
Hunan
Hunan
Human
Animal
Animal
Human/
Animal
Human/
Animal
Human/
Animal
Inhalation Oral Dermal
Gaa(ppm) Solid(mg/m3) (mg/kg) (og/kg)
X<1
-
110
Kegacive
in humans
Mo data.
X<10
X<1 X<1
10100 X>10 X>10
or insignificant results
or animals.
Primary
Weight
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
0.0
Criteria Weight
T 7.0 «
HLETHnorm
See Table A-l in th« appendix for a complete lisc of abbreviations
used in RTECS, Table A-2 for species abbreviations, and Table A-3 for
route of administration abbreviations.
-------
WORKSHEET 15. POTENTIAL FOR AIRBORNE RESEASE FACTOR SCORE
The completion of Vorkiheet 15 require* the use of Worksheet* 7i and
7b.
Production Volume (PV)
1. Circle the index number, in the table below, that correspond* to the
?V interval that the PV data from Work»heec 7a fall* into.
2. Stad the Primary Weight correiponding to the index number circled in
Step 1 and record it in the space labeled PV Weight below.
Criteria and Associated Weight! for Produc-
tion Volume (PV)
Criteria
Index
1
2
3
4
5
6
7
8
10° kg/yr
PV>450
230100
So data
Primary
We ight
4.0
3.0
3.0
2.0
1.0
1.0
5. Divide the Product cf PV Weight and VP Weight by 40.0 to obtain the
normalized Factor Score for Potential for Airborne Release (AIRJOnorm).
T 40.0
PV Weight VP Weight ' v' AIRBOnona
'Vapor prexure data thould be reported at 25*C
and 760 OB Bg.
"If vapor pre**ure data i* unavailable use the
substance boiling point (bp), at 760 mm Hg, a*
a (ubttitute.
CA subcance ihould be con*idered a (olid if it*
citing point is greater thin 25'C.
-------
WORKSHEET 16. BIOACCUMULATION FACTOR SCORE
The completion of Worksheet 16 requires the use of Worksheet 8.
1, Circle the index numbers, in the table below, that correspond to the
Log,0 P interval that the Log10 P data from Worksheet 8 falls into.
2. Read the Primary Weight corresponding to the index number circled in
Step 1 and record it. in the space labeled Bioaccumulation Weight below.
3. Divide Bioaccumulation Weight by 10.0 to obtain the Normalized Factor
Score for Bioaccumulation (BIOAnorm).
Table 8 Criteria and Associated
Weights for Bioaccumu-
lation
Index
1
2
3
4
5
Criteria
Log10P>6.0
4.0
-------
WORKSHEET 17. EXISTING STANDARD FACTOR SCORE
The completion of Worksheet 17 requires the use of Worksheet 9.
1. Circle the index numbers, in the table below, that correspond to the
TWA (or TLV, if TWA unavailable) interval that the TWA (or TLV) data from
Worksheet 9 falls into.
2. Read the Primary Weight corresponding to the index number circled in
Step 1 and record it in the space labeled Standard Weight below.
3. Divide Standard Weight by 6.0 to obtain the Normalized Factor Score
for Existing Standard (ESTDnorm).
Criteria and Associated Weights for Exist-
ing Standards
Index
1
2
3
4
5
6
7
Criteria
Gas (ppm)
X<5
5200
No standard.
Solid
(mg/m3)
X<0 . 25
0.2510.0
Primary
Weight
6.0
5.0
4.0
3.0
2.0
1.0
0.0
- 6 n T-
Standard Weight
ESTDnorm
aBased on OSHA time-weighted-average (TWA)
standards or threshold limit values (TLV)
when TWAs are not available.
-------
WORKSHEET 18. CALCULATE NORMALIZED GROUP SCORE
CAXClSOCEHIClTf CROC? (CAR)
Record the Normalized Factor Scorei (HFS) for Oncogenicity (ONCOnorm from
Workfheet 10) «nd Mutagenicity (MUTnorn from Worksheet 11) and complete
the requefted mathematical procedures to calculate the normalized group
core for CAR (CAKnorm).
7 4.40 r 1.23 -
ONCOnorm I MUTnorm ' M ' CAtaorm
REPRODUCTIVE AMD DEVELOPMENTAL TOX1CITY GROUP (REPRO)
i
1. Record the NFS for Reproductive and Developmental Toxicity (RDTnorm from
Worksheet 12). Because there ii only one factor in the REPRO group,
RDTnorm it equal to the nornalized group score for REPRO (REPROnorm).
RDTnorm
REPROnorm
TOXICin GROUP (TOX)
Record the NTS for Acute Lethality (ALETHnorm froa Worksheet 13) and
Effects Other than Acute Lethality (NLETHnorm froo Worksheet 14 and
complete the requested mathematical procedures to calculate the normalized
group score for TOX (TOXnona),
T 2.0
ALETHnona NLETHnorm ' ' TOXnorm
EXPOSURE GROUP (EXPO)
Record the NFS for Potential for Airborne Release (AIRBOnorm from Work-
sheet 15) and Bioaccuaulation (BIOAnonn from Worksheet 16) and complete
the requested mathematical procedures to calculate the normalized group
score for EXPO (LXPOnora).
I ^ AIRBOnorn * 10-°j * BIOAnorm I ?
11.0
EXPOnorm
STANDARDS CROUP (STD)
5. Record the NFS for Existing Standards (ESTDnorm from Worksheet 17). Be-
cause there is only one factor in the STD group, ESTDnorm is equal to the
normalised group score for STD (STDnorm).
H -. « STDnorm
ESTDnorm
-------
WORKSHEET 19. SUBSTANCE PRIORITIZATION
The completion of Worksheet 19 requires the use of Worksheet 18.
1. Record the values for CARnorm, REPROnorm, TOXnorm, EXPOnonn, and STDnorm
from Worksheet 18 in the appropriate space below.
2. Calculate the Normalized Substance Rank.
2 x
CARnorm
j * V2 X REPROnorray I TOXnorm ' (5 X EXPOnonn I"1"
0.5 x
STDnorra /
T 10.5
Normalized Substance Ranka
aValues for Nortna'i^ac Substance Rank will range from 0.014 to 1.0.
-------
Table A-l. Comprehensive List of All Abbreviations
Used in RTECSa>b
ALR - allergenic effects
AQTX- Aquatic Toxicity
asn - Aspergillus nidulans (mold)
ast - Ascites tumor
BCM - blood clotting mechanism effects
bcs - Bacillus subtilis (bacteria)
bfa - body fluid assay
BLD - blood effects
bmr - bone marrow
BPR - blood pressure effects
brd - bird (domestic or lab)
bvd - wild bird species
C - continuous
CAR - carcinogenic effects
cat - cat
cc - cubic centimeter
end - child
ckn - chicken
CL - ceiling concentration
clr - Chlamydomonas reinhardi (protoza)
CNS - central nervous system effects
COR - corrosive effects
GRIT DOC - NIOSH criteria document
ctl - cattle
CUM - cumulative effects
CVS - cardiovascular effects
cyt - cytogenetic analysis
D - day
dck - duck
DDP - drug dependence effects
DEF - definition
dlt - dominant lethal test
dog - Drosophila melanogaster (insect)
-------
Table A-l. (Cont'd)
dnd - DNA damage
dni - DNA inhibition
dnr - DNA repair
dns - unscheduled DNA synthesis
dog - dog
dom - domestic
DOT - Department of Transportation
dpo - Drosophila pseudo-obscura (insect)
emb - emb ryo
EPA - Environmental Protection Agency
esc - Escherichia coli (bacteria)
ETA - equivocal tumorigenic agent
eug - Eugiena gracilis (protoza)
eye - administration into eye (irritant)
EYE - eye effects (systemic)
fbr - fibroblast
frg - frog
GIT - gastrointestinal tract effects
GLN - glandular effects
gm - gram
gpg - guinea pig
grb - gerbil
grh - grasshopper
H - hour
ham - hamster
hla - HeLa cell
hma - host mediated assay
hmi - Haemophilus influenzae (bacteria)
htnn - human
hor - horse
I - intermittent
ial - intraaural
IARC - International Agency for Research on Cancer
iat - intraarterial
-------
Table A-l. (Cont'd)
ice - intracerebral
lev - intracervical
idr - intradermal
idu - intraduodenal
ihl - inhalation
imm - immersion
imp - implant
ims - intramuscular
inf - infant
ipc - intraplacental
ipl - intrapleural
ipr - intraperitoneal
IRDS- primary irritation dose
irn - intrarenal
IRR - irritant effects (systemic)
isp - intraspinal
itr - intratracheal
ivg - intravaginal
ivn - intravenous
kdy - kidney
kg - kilogram
kip - Klebsiella pneumoniae (bacteria)
L - liter
LC5Q- lethal concentration 50 percent kill
L^Lo" lowest published lethal concentration
- lethal dose 50 percent kill
- lowest published lethal dose
leu - keukocyte
Ing - lung
Ivr - liver
lym - lymphocyte
M - minute
nH - cubic meter
mam - mamal (species unspecified)
-------
Table A-l. (Cont'd)
man - man
mg - milligram
MGN - multigeneration
mky - monkey
ml - milliliter
MLD - mild irritation effects
mma - microsomal mutagenicity assay
MMI - mucous membrane effects
mmo - mutation in microorganisms
mmol- millimole
mmr - mammary gland
mnt - micronucleus test
MOD - moderate irritation effects
mol - mole
mppcf - million particles per cubic foot
mrc - gene conversion and mitotic recombination
msc - mutation in somatic mamallian cells
MSK - musculo-skeletal effects
MTDS- mutation dose
MTH - mouth effects
mul - multiple routes
mus mouse
NEO - neoplastic effects
ng - nanogram
nmol- nanomole
nsc - Neurospora crassa (mold)
nse - non-standard exposure
DBS.- obsolete (trade name)
ocu - ocular
omi - other microorganisms
oin - other insects
open- open irritation test
orl - oral
OEM - Other Regulated Material (DOT)
-------
Table A-l. (Cont'd)
OSHA - Occupational Safety and Health Administration
oth - other cell types
otr - oncogenic transformation
ovr - ovary
par - parenteral
pg - picogram
pgn - pigeon
pic - phage inhibition capacity
pig ~ pig
Pk - peak concentration
pmol - picomole
PNS - peripheral nervous system effects
post - after birth
ppb - parts per billion
pph - parts per hundred
ppm - parts per million
ppt - parts per trillion
pre - prior to copulation
preg - pregnant
PSY - psychotropic effects
PUL - pulmonary system effects
qal - quail
rat - rat
RBC - red blood cell effects
rbt - rabbit
rec - rectal
REGS - standards and regulations
rns - rinsed with water
RPDS - reproductive effects dose
RTECS - Registry of Toxic Effects of Chemical Substances
S - second
sal - salmon
sat - Salmonella typhimurium (bacteria)
see - sister chromatid exchange
-------
Table A-l. (Cont'd)
WBC - white blood cell effects
wmn - woman
Y - year
% - percent
aFrom RTECS microfiche edition, October 1981 (Lewis and
Tatken 1981).
recent editions of RTECS may use abbreviations not
included on this list. Any abbreviation can be iden-
ified by using the Key to Abbreviations in the Appendix
of the RTECS edition being used.
-------
Table A-2. Species Abbreviations Used in RTECSa>b
Mammalian
cat - cat, adult
ctl - cattle
chd - child
dog - dog, adult
dom - domestic animal (goat, sheep)
grb - gerbil
gpg - guinea pig, adult
ham - hamster
hor - horse, donkey
hmn - human
inf - infant
mam - mammal (species unspecified in reference)
man - man
mky - monkey
mus - mouse
Pig - pig
rbt - rabbit, adult
rat - rat
sql - squirrel
wmn - woman
Nonmammalian
asn - Aspergillus nidulans (mold)
bcs - Bacillus subtilis (bacteria)
brd - bird (any domestic or laboratory bird reported but
not otherwise identified)
bwd - bird (wild bird species)
ckn - chicken
clr - Chlamydomonas reinhardi (protoza)
dck - duck
dmg - Drosophila melanogaster (insect)
dpo - Drosophila pseudo-obscura (insect)
esc - Escherichia coli (bacteria)
-------
Table A-2. (Cont'd)
eug - Eugiena gracilis (protoza)
frg - frog
grh - grasshopper
hmi - Haemophilus influenzae (bacteria)
Kip - Klebsiella pneumoniae (bacteria)
nsc - Neurospora crassa (mold)
pgn - pigeon
qal - quail (laboratory)
sal - salmon
sat - Salmonella typhimurium (bacteria)
slw - silkworm
smc - Saccharmyces cerevisiae (yeast)
srm - Serratia raarcescens (bacteria)
ssp - Schizosaccharomyces pombe (yeast)
tod - toad
trk - turkey
aFrom RTECS microfiche edition, October 1981 (Lewis and
Tatken 1981).
recent editions of RTECS may use abbreviations not
included on this list. Any abbreviation can be identi-
fied by using the Key to Abbreviations in the Appendix
of the RTECS edition being used.
-------
Table A-3. Route of Administration Abbre-
viations Used in RTECSa>b
eye - eyes
ial - intraaural
iat - intraarterial
ice - intracerebral
icv - intracervical
idr - intradermal
idu - intraduodenal
ihl - inhalation
imp - implant
ims - intramuscular
ipc - intraplacental
ipl - intrapleural
ipr - intraperitoneal
irn - intrarenal
isp - intraspinal
itr - intratracheal
ivg - intravaginal
ivn - intravenous
mul - multiple
ocu - ocular
orl - oral
par - parenteral
rec - rectal
skn - skin
scu - subcutaneous
unk - unreported
aFrom RTECS microfiche edition, October
1981 (Lewis and Tatken 1981).
''More recent editions of RTZCS may use ab-
breviations not included on this list.
Any abbreviation can be identified by us-
ing the Key to Abbreviations in the Ap-
pendix of the RTECS edition being used.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing]
1. REPORT NO.
EPA 450/5-82-008
3. RECIPIENT'S ACCESSIOf*NO.
4. TITLE AND SUBTITLE
Hazardous Air Pollutant Prioritization System (HAPPS)
5. REPORT DATE
October 1982
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
A. E. Smith and D. J. Fingleton
8. PERFORMING ORGANIZATION REPORT NO.
-------
TECHNICAL REPORT DATA
(Please read Inunctions on the reverse before completing)
1, REPORT NO.
EPA 450/5-82-008
3. RECIPIENT'S ACCESSION>NO.
4. TITLE ANDSUBTITLE
Hazardous Air Pollutant Prioritization System (HAPPS)
5. REPORT DATE
October 1982
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
A. E. Smith and D. J. Fingleton
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Argonne National Laboratory
9700 South Cass Avenue
Argonne, Illinois 60439
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
Interagency Agreement No.
AD-89-F-1-344-0 '
12. SPONSORING AGENCY NAME AND ADDRESS
Pollutant Assessment Branch
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report presents a preliminary screening technique by which a large
number of potentially hazardous compounds can be numerically ranked using readily
available information on health effects and release to the ambient air. Factors
considered are oncogenicity, mutagenicity, reproduction and developmental toxicity,
acute lethality, effects other than acute lethality, production volume, vapor pressure,
bioaccumulation and existing standards.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
Prioritization System
Hazardous Air Pollutants
8. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
91
20. SECURITY CLASS (Thispage}
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |