EPA600/AP-93/001d
July 1993
Urban Soil Lead Abatement
Demonstration Project
Volume IV:
Cincinnati Report
Environmental Criteria and Assessment Office
Office of Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Printed on Recycled Paper
-------
DISCLAIMER
This document is an internal draft for review purposes only and does not constitute
U.S. Environmental Protection Agency policy. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
11
-------
EPA600/AP-S3/001d
July 1993
Urban Soil Lead Abatement
Demonstration Project
Volume IV:
Cincinnati Report
Environmental Criteria and Assessment Office
Office of Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Printed on Recycled Paper
-------
DISCLAIMER
This document is an internal draft for review purposes only and does not constitute
U.S. Environmental Protection Agency policy. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
u
-------
EPA 6007 AP-a3/001d
July 1993
Urban Soil Lead Abatement
Demonstration Project
Volume IV:
Cincinnati Report
Environmental Criteria and Assessment Office
Office of Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Printed on Recycled Paper
-------
DISCLAIMER
This document is an internal draft for review purposes only and does not constitute
U.S. Environmental Protection Agency policy. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
ii
-------
TABLE OF CONTENTS
LIST OF TABLES vii
LIST OF FIGURES xi
LIST OF AUTHORS AND CONTRIBUTORS xv
LIST OF ACRONYMS AND ABBREVIATIONS , xvii
1. EXECUTIVE SUMMARY 1-1
1.1 INTRODUCTION . . . 1-1
1.2 METHODS .... 1-2
1.3 RESULTS 1-3
, 1.3.1 Soil Lead 1-4
1.3.2 Exterior Dust Lead 1-4
1.3.3 Interior Dust Lead 1-4
1.3.4 Blood Lead 1-5
1.3.5 Hand Lead 1-5
1.3.6 Intercorrelation Among Environmental Lead Measures
and Blood Lead „ 1-5
1.3.7 Environmental Dust Lead Levels 1-6
1.3.8 Abatement Costs 1-6
1.4 CONCLUSIONS .-. 1-6
2. BACKGROUND 2-1
2.1 HYPOTHESES 2-1
2.2 PREVIOUS RELATED WORK 2-2
3. STUDY DESIGN AND METHODS 3-1
3.1 OVERALL DESIGN 3-1
3.2 NEIGHBORHOOD AND SUBJECT SELECTION,
RECRUITMENT AND RETENTION 3-1
3.2.1 Enrollment Criteria 3-1
3.2.2 Initial Soil Survey 3-4
3.2.3 Identification of Study Areas 3-6
3.2.4 Participant Enrollment Procedures . . 3-6
3.2.5 Community Relations 3-10
3.2.6 Subject Retention Plan 3-11
3.2.7 Human Subject Research Review Process 3-12
3.3 ABATEMENT PROCEDURES 3-12
3.3.1 Interior Dust Abatement Methods Development 3-14
3.3.1.1 Development of Abatement Methods for Wood
and Vinyl Floors 3-15
3.3.1.2 Development of Abatement Methods for
Cleaning Carpeting -. . 3-16
3.3.1.3 Interior Dust Abatement Procedure Summary .. 3-18
-------
TABLE OF CONTENTS (cont'd)
3.3.2 Exterior Dust Methods Development 3-19
3.3.2.1 Types of Pavement Cleaners Available 3-19
3.3.2.2 Exterior Dust Testing 3-20
3.3.2.3 Exterior Dust Abatement Summary 3-22
3.3.3 Soil Abatement Methods ; 3-22
3.3.4 Sequence of Abatement 3-25
3.3.5 Monitoring and Supervision 3-27
3.3.6 Contract and Specification Development 3-27
3.3.7 Bidding and Contract Letting 3-28
3.4 BLOOD COLLECTION AND ANALYSIS 3-29
. 3.4.1 Blood Collection 3-29
3.4.2 Analytical Procedures 3-30
3.4.3 Quality Assurance/Quality Control . 3-31
3.5 ENVIRONMENTAL SAMPLES 3-33
3.5.1 Environmental Sample Collection 3-33
3.5.1.1 Soil Sample Collection 3-33
3.5.1.2 Exterior Dust Collection 3-34
3.5.1.3 Interior Dust Collection 3-36
3.5.1.4 Handwipe Sample Collection 3-37
3.5.1.5 Water Sample Collection 3-38
3.5.1.6 Paint Collection by Portable X-ray
Fluorescence 3-38
3.5.2 Environmental Sample Analysis 3-39
3.5.2.1 Soil 3-39
3.5.2.2 Exterior Dust 3-40
3.5.2.3 Interior Dust 3-41
3.5.2.4 Interior Dustfall 3-42
3.5.2.5 HandDust 3-43
3.5.2.6 Water 3-44
3.5.2.7 Paint 3-45
3.6 HEALTH AND SAFETY PLAN . 3-46
3.6.1 Training 3-47
3.6.2 Medical Surveillance Program 3-48
3.6.3 Workplace Audits 3-51
3.6.3.1 Laboratory Audits 3-52
3.6.3.2 Interior and Exterior Field Audits 3-52
3.7 DATA MANAGEMENT 3-52
3.7.1 Data Management Objectives 3-52
3.7.2 Data Management System Development 3-52
3.7.3 Users' Needs Survey [ 3-53
3.7.4 System Analysis 3-53
3.7.5 System Design 3-53
3.7.6 Design of the Components of Data Management
Engine 3-54
iv
-------
TABLE OF CONTENTS
LIST OF TABLES vii
LIST OF FIGURES . . . xi
LIST OF AUTHORS AND CONTRIBUTORS xv
LIST OF ACRONYMS AND ABBREVIATIONS xvii
1. EXECUTIVE SUMMARY 1-1
1.1 INTRODUCTION l-l
1.2 METHODS 1-2
1.3 RESULTS 1-3
. 1.3.1 SoilLead 1-4
1.3.2 Exterior Dust Lead 1-4
1.3.3 Interior Dust Lead 1-4
1.3.4 Blood Lead 1-5
1.3.5 Hand Lead . 1-5
1.3.6 Intercorrelation Among Environmental Lead Measures
and Blood Lead 1-5
1.3.7 Environmental Dust Lead Levels 1-6
1.3.8 Abatement Costs 1-6
1.4 CONCLUSIONS 1-6
2. BACKGROUND , • 2-1
2.1 HYPOTHESES . 2-1
2.2 PREVIOUS RELATED WORK 2-2
3. STUDY DESIGN AND METHODS 3-1
3.1 OVERALL DESIGN 3-1
3.2 NEIGHBORHOOD AND SUBJECT SELECTION,
RECRUnMENT AND RETENTION 3-1
3.2.1 Enrollment Criteria 3-1
3.2.2 Initial Soil Survey 3.4
3.2.3 Identification of Study Areas 3-6
3.2.4 Participant Enrollment Procedures 3-6
3.2.5 Community Relations 3-10
3.2.6 Subject Retention Plan 3-11
3.2.7 Human Subject Research Review Process 3-12
3.3 ABATEMENT PROCEDURES . 3-12
3.3.1 Interior Dust Abatement Methods Development 3-14
3.3.1.1 Development of Abatement Methods for Wood
and Vinyl Floors 3-15
3.3.1.2 Development of Abatement Methods for
Cleaning Carpeting 3-16
3.3.1.3 Interior Dust Abatement Procedure Summary .. 3-18
-------
TABLE OF CONTENTS (cont'd)
3.3.2 Exterior Dust Methods Development 3-19
3.3.2.1 Types of Pavement Cleaners Available 3-19
3.3.2.2 Exterior Dust Testing . 3-20
3.3.2.3 Exterior Dust Abatement Summary 3-22
3.3.3 Soil Abatement Methods . . 3-22
3.3.4 Sequence of Abatement 3-25
3.3.5 Monitoring and Supervision ! 3-27
3.3.6 Contract and Specification Development 3-27
3.3.7 Bidding and Contract Letting 3-28
3.4 BLOOD COLLECTION AND ANALYSIS 3-29
. 3.4.1 Blood Collection 3-29
3.4.2 Analytical Procedures 3-30
3.4.3 Quality Assurance/Quality Control '. - 3-31
3.5 ENVIRONMENTAL SAMPLES 3-33
3.5.1 Environmental Sample Collection , 3-33
3.5.1.1 Soil Sample Collection 3-33
3.5.1.2 Exterior Dust Collection 3-34
3.5.1.3 Interior Dust Collection 3-36
3.5.1.4 Handwipe Sample Collection ' 3-37
3.5.1.5 Water Sample Collection 3-38
3.5.1.6 Paint Collection by Portable X-ray
Fluorescence 3-38
3.5.2 Environmental Sample Analysis 3-39
3.5.2.1 Soil 3-39
3.5.2.2 Exterior Dust 3-40
3.5.2.3 Interior Dust 3-41
3.5.2.4 Interior Dustfall 3-42
3.5.2.5 Hand Dust 3-43
3.5.2.6 Water 3-44
3.5.2.7 Paint 3-45
3.6 HEALTH AND SAFETY PLAN . 3-46
3.6.1 Training 3-47
3.6.2 Medical Surveillance Program 3-48
3.6.3 Workplace Audits 3-51
3.6.3.1 Laboratory Audits 3-52
3.6.3.2 Interior and Exterior Field Audits . . . . 3-52
3.7 DATA MANAGEMENT . . . 3-52
3.7.1 Data Management Objectives 3-52
3.7.2 Data Management System Development . . . 3-52
3.7.3 Users' Needs Survey 3-53
3.7.4 System Analysis 3-53
3.7.5 System Design - - - 3-53
3.7.6 Design of the Components of Data Management ;
Engine . . 3-54
iv
-------
TABLE OF CONTENTS (cont'd)
3.7.7 System Programming .. 3-55
3.7.8 System Implementation, Installation, and Training .... 3-56
3.7.9 Data Base Management Operations and Maintenance . . 3-56
3.7.10 A Summary of Project Data Base 3-57
3.8 GEOGRAPHIC INFORMATION SYSTEM 3-57
3.8.1 What is a Geographic Liformation System? 3-57
3.8.2 What Did We Plan To Achieve with Our Geographic
Information System? 3-57
3.8.3 Application of a Geographic Information System in
This Study 3-57
. 3.8.4 The Contribution of a Geographic Information System
to This Project 3-58
3.9 DATA ANALYSIS . 3-58
3.9.1 Data Analysis Objectives 3-58
3.9.2 Data Analysis Plan 3-59
3.9.3 Data Conversion 3-59
3.9.4 Clean-Up Data and Data-Coding Update 3-59
3.9.5 Data List Printing . . . 3-60
3.9.6 Preliminary Data Analysis 3-60
3.9.7 Correlational Data Analysis 3-61
3.9.8 Confirmatory Data Analysis 3-61
3.9.9 Prospective, Retrospective, and Sequences of
Cross-Sectional Data Analyses 3-61
3.9.10 Modeling 3-62
3.9.11 Statistical Methodology Development 3-62
3.10 PROJECT ADMINISTRATION 3-64
3.10.1 Financial Management 3-64
4. RESULTS 4-1
4.1 STUDY POPULATION 4-1
4.2 EXCLUSION FROM DATA ANALYSIS . . . ; 4-3
4.3 SAMPLE COLLECTION AND ABATEMENT SCHEDULE ... 4-4
4.4 BLOOD AND ENVIRONMENTAL SAMPLE QUALITY
ASSURANCE/QUALITY CONTROL RESULTS 4-5
4.4.1 Quality Control Evaluation for Blood 4-5
4.4.2 Quality Control Evaluations for Soil X-ray
Fluorescence 4-11
4.4.3 Quality Control Evaluation for Exterior Dust 4-17
4.4.4 Quality Control Evaluations for Interior Dust 4-18
4.4.5 Quality Control Evaluations for Interior Dustfall 4-22
4.4.6 Quality Control Evaluations for Hand Lead 4-23
4.4.7 Quality Control Evaluations for Water 4-25
4.4.8 Quality Control Evaluations for Paint 4-27
-------
TABLE OF CONTENTS (cont'd)
4.5 SOIL LEAD CONCENTRATIONS 4-27
4.5.1 Grass Cover 4-34
4.5.2 Component Neighborhoods Within Study Areas 4-34
4.6 EXTERIOR DUST 4-37
4.7 INTERIOR DUST LEAD 4-43
4.7.1 Net Change in Lead Loading 4-56
4.8 BLOOD LEAD 4-59
4.9 HAND LEAD • • • 4~63
4.10 INTERCORRELATIONS 4-72
4.11 MODELING 4'81
4.11.1 Modeling the Difference of Phase 1 and Phase 5 Blood
Lead for the Initially Recruited Families Who lived in
the Rehabilitated Housing Units 4-81
4.11.1.1 Correlation Analysis 4-94
4.11.1.2 Regression Modeling 4-97
4.11.1.3 Multiple Regression Modeling 4-98
4.11.1.4 Structural Equation Modeling 4-104
4.11.1.5 Comparison of Treatment-Effects Among
the Three Statistical Approaches . 4-106
4.11.1.6 Statistical Modeling Conclusions j 4-107
4 11.2 Cross-Sectional Structural Equation Models for Loading
Data 4~108
4.11.2.1 Summary and Conclusions from the
Cross-Sectional Structural Equation
Models 4-109
4.12 HEALTH AND SAFETY 4-112
4.12.1 Workplace Audits • • 4-112
4.12.2 Special Workplace Evaluations 4-124
4.12.2.1 Price Hill Facility 4-124
4.12.2.2 Noise Levels 4-124
4.12.3 Manuscript on Safety and Health Plan 4-125
4.13 ABATEMENT COSTS 4-125
5. CONCLUSIONS • • •. 5~l
5.1 RESPONSE TO QUESTIONS 5-1
5.2 GENERAL CONCLUSIONS 5-3
6. REFERENCES
6-1
VI
-------
LIST OF TABLES
Number Page
2-1 Environmental Lead Measures by Housing Type 2-6
3-1 Comparison of Street Sweeping Machines 3-17
3-2 Blood Collection Sampling Phases 3-29
4-1 Cincinnati Soil Project Enrollment of Subjects 4-2
4-2 Number of Study Participants by Age Range . 4-2
4-3 . Quality Control Evaluations for Each Sample Type 4-6
4-4 Absolute Difference Between Blind Duplicate Analyses of the
Same Human Blood Samples 4-8
4-5 Blood Lead Quality Control Samples Prepared by Centers for
Disease Control for the Cincinnati Soil Lead Project 4-9
4-6 Laboratory Performance in the Analysis of Isotope Dilution
Mass-Spectroscopy Reference Samples ., 4-10
4-7 University of Cincinnati Perfbrmace in the Analysis of
Hematology Controls 4-15
4-8 Serum Iron and Total Iron Binding Capacity Serum Pools 4-15
4-9 Soil Field Lab Blanks 4-16
4-10 University of Cincinnati Lab Performance in Analyzing
Environmental Monitoring Systems Laboratory Quality
Control Soil Samples by X-ray Spectrometer 4-17
4-11 Exterior Dust Field Lab Blanks . 4-18
4-12 Environmental Monitoring Systems Quality Control Soil
Samples Analyzed with Exterior Dust Samples 4-18
4-13 Performance on Analyses of Environmental Monitoring
Systems Laboratory Quality Control Dust Samples 4-19
4-14 Average Weights of Interior Dust Samples 4-20
Vll
-------
LIST OF TABLES (cont'd)
Number
4-15
4-16
4-17
4-18
4-19
4-20
4-21
4-22
4-23
4-24
4-25
4-26
4-27
4-28
4-29
4-30
4-31
4-32
Soil Project Interior Dust Quality Control Standards
Interior Dustfall Field Duplicates . . •
Quality Control Standards Analyzed with Dustfall Samples
Environmental Monitoring Systems Laboratory Quality Control
Hand Lead Samples •
Comparison of Duplicate Water Collections
Quality Control Water Samples Analyzed Blind
Soil Lead Concentrations :
Comparison of Soil Lead Concentrations According to Whether
or Not Rubble at Depth was Observed, Phase 00
Comparison of Soil Lead Concentrations by Sampling Pattern,
Phase 00
Soil Lead Concentrations in Surface Scrapings by Extent of
Grass Cover
Soil Lead Concentrations by Area Component Neighborhoods . . .
Soil Lead Concentrations (Bottom 2-cm Samples) in
Neighborhoods by Presence or Absence of Rubble
Observed at Depth
Soil Lead Concentration in Surface Scrapings by Neighborhood
and by Extent of Grass Cover
Exterior Dust Lead Concentrations
Exterior Dust Lead Concentrations
Exterior Dust Lead Loading . . .
Exterior Dust Lead Loading
Exterior Dust Loadings . .
Page
4-22
4-22
4-23
4-24
4-26
4-26
4-28
4-32
4-33
4-35
4-37
4-38
4-39
4-40
4-42
4-44
4-47
4-48
VUl
-------
LIST OF TABLES (cont'd)
Number page
4-33 Interior Entry Dust Lead: Concentrations and Loadings 4-49
4-34 Interior Entry Dust Lead (Paired Data) 4-55
4-35 Interior Floor Dust Lead (Paired Data) 4-60
4-36 Interior Window Dust Lead (Paired Data) . 4-64
4-37 Percent Reduction Between Phases in Geometric Mean Interior
Dust Lead Loadings (Paired Data) 4-65
4-38 Changes in Interior Dust Lead Loadings Between Sample
Collection Times 4-66
4-39 Interior Dust Loadings 4-67
4-40 Interior Dust Loadings 4-67
4-41 Blood Lead Concentrations by Phase and Study Area 4-68
4-42 Blood Lead by Area and Phase (Paired Data) 4-69
4-43 Ratio of Blood Lead Levels from Phase to Phase and Differences
in Blood Lead Levels Between Phases 4-70
4-44 Summary of Blood Lead by Age and House Type, Phase 01 .... 4-71
4-45 Ratio of Blood Lead Levels from Phase to Phase by Age and
Area „ 4-73
4-46 Hand Lead 4.75
4-47 Comparison of Hand Lead Values Between Pairs of Phases 4-76
4-48 Ihtercorrelations Between Blood Lead and Hand Lead and
Environmental Lead Measures and Age for Overall Study
Population and by Area 4-78
4-49 Intercorrelations Between Blood Lead and Hand Lead and
Environmental Lead Measures and Age for Overall Study
Population and by Area 4-80
IX
-------
LIST OF TABLES (cont'd)
Number
4-50
4-51
4-52
4-53
4-54
4-55
4-56
4-57
4-58
4-59
4-60
4-61
4-62
4-63
4-64
Interconelations Between Blood Lead and Hand Lead and
Environmental Lead Measures and Age for Overall Study
Population and by Area
Ihtercorrelations Between Blood Lead and Hand Lead and ,
Environmental Lead Measures and Age for Overall Study •
Population and by Area
Intercorrelations Between Blood Lead and Hand Lead and
Environmental Lead Measures and Age
Summary of Blood Lead and Hand Lead Correlations
latercorrelations Among Environmental and Blood Lead Data ...
Summary Statistics
Summary Statistics of DPbB, DPbH, DPbDIFL, and
DPbDELD by Area
Correlations Between Blood Lead and Other Important
Variables
Correlations Between Hand Lead and Other Important
Variables
Correlations Between DPbDIFL and Other Important Variables . .
Correlations Between Exterior Dust Lead and Other
Important Variables
Results of Structural Equations Modeling Using Phase 1
Lead Loading Data
Results of Structural Equations Modeling Using Phase 3 ;
Lead Concentration Data
Results of Structural Equations Modeling Using Phase 5
Lead Loading Data
Abatement Costs •
Page
4-82
4-84
4-86
4-92
4-93
4-97
4-98
4-99
4-100
4-101
4-102
4-110
4-113
4-116
4-126
X
-------
LIST OF FIGURES
Number Page
2-1 Reduced structural model for pathway from environmental lead to
blood lead for 18 mo old 2-3
2-2 Effect of housing type and condition on early childhood blood lead
concentrations 2-5
3-1 Abatement and monitoring time table 3_2
3-2 Blood and environmental sample collection 3.3
3-3 Schematic diagrams of the Pendleton (Area A) and Findlay
(Area B) neighborhoods 3.7
3-4 Schematic diagrams of the Back and Dandridge neighborhoods,
both in Area B 3_g
3-5 Schematic diagrams of the Glencoe-and Mohawk neighborhoods,
both in Area C 3.9
3-6 Decision criteria for soil abatement 3_23
3-7 Cincinnati soil project—budget tracking system flow diagram .... 3-65
4-1 Performance of the Cincinnati project in the Centers for Disease
Control proficiency program for blood lead analysis during the
period June 1989 to September 1991 4_12
4-2 Performance of the Cincinnati project in the College of American
Pathologists proficiency program for blood lead analysis during
the period June 1989 to September 1991 4-13
4-3 Performance of the Cincinnati project in the Centers for Disease
Control proficiency program for free erythroporphyrin
analysis during the period June 1989 to September 1991 4-14
4-4 Surface scraping soil lead concentrations by area during the
preabatement (Phase 0) and early postabatement (Phases 2
and 3) periods 4_29
4-5 Soil lead concentrations in the top 2 cm by area 4-30
4-6 Soil lead concentrations in the bottom 2 cm of a 15 cm core,
by area ' 4.3^
XI
-------
IIST OF FIGURES (cont'd)
Number
4-7
4-8
4-9
4-10
4-11
4-12
4-13
4-14
4-15
4-16
4-17
4-18
4-19
4-20
4-21
4-22
4-23
4-24
Distribution of lead concentrations by ground cover type
and area, prior to abatement
Exterior dust lead concentrations in streets, sidewalks,
and alleys by area
Exterior dust lead concentrations at targeted locations by area . . .
Exterior dust lead loading in streets, sidewalks, and alleys by
area
Exterior dust lead loading at targeted locations by area
Interior dust lead concentrations at the entry way
Interior dust lead loading at entryway by area .
Interior dust lead concentration on the floor
Interior dust lead loading floor by area . .-
Comparison of interior entryway dust lead concentrations between
Phases 1 and 2
Comparison of interior entryway dust lead loading between
Phases 1 and 2
Comparison of interior entryway dust lead concentrations between
Phases 1 and 5
Comparison of interior entryway dust lead loading between
Phases 1 and 5
Pre- and postabatement interior floor dust lead concentrations in
Area A
Pre- and postabatement interior floor dust lead loading in
Area A
Pre- and postabatement floor dustjead concentrations, all areas . .
Pre- and postabatement floor dust lead loading, all areas
Effect of child age and house type on blood lead
xii
'. Page
4-36
4-41
4-43
4-45
4-46
4-52
4-52
4-53
! 4-53
4-56
4-57
4-58
4-59
4-61
4-62
; 4-62
4-63
4-72
-------
Number
4-25
4-26
4-27
4-28
4-29
4-30
4-31
4-32
LIST OF FIGURES (cont'd)
Impact of abatement on children living in rehabilitated housing,
Phase 5 versus Phase 1
Impact of abatement on children living in rehabilitated housing,
Phase 9 versus Phase 1 , .
Comparison of hand lead loading between Phase 1 and Phase 3
Comparison of hand lead loading between Phase 1 and Phase 2
Structural equation analysis: relationship between blood lead
and environmental lead 4
Phase 1 structural equation model for loading data
Phase 3 structural equation model for loading data
Phase 5 structural equation model for loading data
4-74
4-74
4-77
4-77
4-105
4-121
4-122
4-123
Xlll
-------
-------
LIST OF AUTHORS AND CONTRIBUTORS
Co-Principal Investigators
Scott Clark, Ph.D., P.E., C.I.H.
Robert Bornschein, Ph.D.
Project Managers
Linda Conway-Mundew
Joanne Grote
William Menrath
Winkey Pan, Ph.D.
Sandy Roda
Administration
Subject Matters
Abatement and Environmental Monitoring
Data Analysis
Analytical Chemistry
Other Contributors
Shane Que Hee, Ph.D.
Douglas Linz, M.D.
Paul Succop, Ph.D.
Barbara Poppe
Carol Thompson
Penny Schmitgen
Donna Adams
Denis Boudreau
William Hansan
Lois Betka .
Steve O'Strodka
Pat VanLeeuwen, Ph.D.
University
University
University
University
University
University
Cincinnati
Cincinnati
Cincinnati
U.S. EPA
U.S. EPA
U.S. EPA
of Cincinnati
of Cincinnati
of Cincinnati
of Cincinnati
of Cincinnati
of Cincinnati
Health Department
Health Department
Health Department
Region V
Region V
Region V
xv
-------
-------
LIST OF ABBREVIATIONS AND ACRONYMS
AAS
ACGIH
ANCOVA
ANOVA
ARG/INFO
ASTM
ASV
BMDP
CDC
CSP
CUFS
dBASE
dL
DMS
Dust Loading
Dustfall
DVM
ECAO/RTP
EHRT
EMSL/LV
Atomic absorption spectroscopy
American Conference of Governmental Industrial
Hygenists
Analysis of covaiiance
Analysis of variance
Software package supporting geographic information
systems
American Society for Testing Materials
Anodic stripping voltammetry
Statistical software package
Centers for Disease Control
Cincinnati Soil Project
College and University Financial System
Database management software
Deciliter, used here as a measure of blood lead in
micrograms per deciliter
Data management system
Mass of dust per unit area
Dust collected on an upfacing surface
Dust vacuum method
Environmental Criteria and Assessment Office/ Research
Triangle Park, NC
Environmental Health Testing, Inc.
U.S. Environmental Protection Agency, Environmental
Monitoring Systems Laboratory at Las Vegas, NV
xvii
-------
IIST OF ABBREVIATIONS AND ACRONYMS
FEP
GIS
GM
GSD
Hematocrit
HEPA
BDMS
HUB
LCL
NBS
NIOSH
NIST
Nonparametric statistics
OSHA
Pb
Pb Concentration
Pb Loading
PbB
QA/QC
SARA
Free erythrocute protoporphyrin
Geographic Information Systems
Geometric mean
Geometric standard deviation
Volume traction of blood representing the cellular and
other particulate matter ;
High efficiency particle accumulator
Isotope dilution mass spectrometry
Internal Review Board
Lower confidence limit
National Bureau of Standards (now National Institute for
Science and Technology)
National Institute for Occupational Safety arid Health
National Institute for Science and Technology
Statistical methods for noncontinuous parameters
Occupational Safety and Health Administration
Lead
Mass of lead per mass of medium (soil, dust, water)
Mass of lead per unit area
Blood lead, measured in micrograms per deciliter
Parts per million, genearHy equivalent to micrograms per
gram
Qualtiy assurance/quality control
Superfund Amendments and Reauthorization Act
xvui
-------
SAS
SD
SES
SFe
SHO
SHP
SPSS
Structural Equation Modeling
SYSTAT
TEBC
T-test
UCL
U.S. EPA
XRF
ZPP
LIST OF ABBREVIATIONS AND ACRONYMS
Statistical software
Standard deviation
Socioeconomic status
Serum iron
Safety and health officer
Safety and health plan
Statistical software package
Model based on a system of simultaneous linear equations
Statistical software package
Total iron binding capacity
Statistical test for normal distribution
Upper confidence limit
U.S. Environmental Protection Agency
X-ray fluorescence
Zinc protoporphyrin
xrx
-------
-------
1. EXECUTIVE SUMMARY
1.1 EMTRODUCTION
In late 1987 Cincinnati was selected to be the location of one of three urban Soil Lead-
Abatement Demonstration Projects to implement Section lll(b)(6) of the Superfund
Amendment and Reauthorization Act (SARA) of 1986. This project was conducted by
investigators in the Department of Environmental Health at the University of Cincinnati in
collaboration with the City of Cincinnati Department of Health. The Cincinnati project made
use of extensive experience in the Department of Environmental Health in pediatric lead
exposure studies. The Cincinnati project was conducted in neighborhoods where most of the
housing had lead-based paint removed about two decades earlier as a result of a complete
rehabilitation carried out under U.S. Housing and Urban Development-supported programs.
The rehabilitation involved a "gutting" of the buildings and the complete replacement of
plumbing, wiring and heating systems and the installation of new walls, flooring, windows
and doors. Exterior brick areas were either sandblasted or chemically cleaned and sometimes
were re-painted. Most of the rehabilitated buildings were 3 to 4 story multi-family structures
and all were rental units. Soil in these neighborhoods was located primarily in small parks,
recreational areas and vacant lots and not specifically part of the same property containing
the rehabilitated housing. Therefore, in order to include all of the soil areas in the abatement
project the decision was made early in the project design to abate entire neighborhoods.
Thus, all soil areas in the neighborhoods were included, whether or not there were housing
units on the property. This neighborhood-wide design, rather than the scattered house
approach used elsewhere, was extended to include exterior surface dust on paved areas.
Therefore, a neighborhood-wide paved surface cleaning was also performed.
The two central hypotheses of this project were that:
(1) A reduction of lead in residential soil accessible to children would result in a
decrease hi their blood-lead levels, and
1-1
-------
(2) Interior dust abatement, when carried out in conjunction with exterior dust and
soil abatement, would result in a greater reduction in blood lead than would be
obtained with interior dust abatement alone, or exterior dust and soil abatement
alone.
A secondary hypotheses was that:
(3) A reduction of lead in residential soil accessible to children will result in a
decrease in their hand lead levels.
1.2 METHODS
Three areas, designated A, B and C, were selected for the study. A door-to-door
census was performed to determine the availability of an adequate number of children less
than six years of age. A preliminary, soil lead survey was also undertaken to provide
baseline soil lead values and to document the presence of lead-contaminated soil. The choice
of which neighborhoods to include was made on the basis of the census and preliminary soil
lead survey. During the 1989 abatement period Area A received interior and exterior dust
abatement and soil abatement, while Area B received interior dust abatement only. Exterior
dust and soil abatement were performed in Area B during the summer of 1990 and all three
abatements were performed in Area C after the last sampling phase in 1991.
The soil abatement plan involved removal of the top six inches of soil when either the
average or the top 2 cm lead concentrations was ^500 ppm, replacement with low lead soil
(<20 ppm) and fesodding. In areas where the grass cover was poor and the surface
concentration was ^300 ppm, removal, replacement and resodding also was performed.
Available vacuum-based street cleaning equipment was tested to determine which equipment
was capable of removing greater than 90% of the dust from rough paved surfaces. Extensive
testing of high efficiency particle accumulator (HEPA) vacuum cleaning of lead-dust
containing carpets was conducted. Based on the results of these tests, and tests of new
carpets embedded with dust in the laboratory, it was concluded that carpets could not be
satisfactorily cleaned. Substantial amounts of lead remained in the carpets after repetitive
cleaning and, in a number of cases, the lead dust loading on the surface was increased during
t
the vacuum cleaning. Therefore, carpet replacement was selected as the method of choice
for dealing with existing carpets. Based on the carpet cleaning experience and a limited
1-2
-------
amount of testing of cleaning methods for upholstered furniture, it was concluded that
contaminated upholstered furniture should also be replaced to the extent possible.
Environmental monitoring was performed during ten sampling phases before and after
the various abatement procedures were conducted. The types of environmental samples
collected and the number of times such samples were collected during a three-year period are
as follows: soil (seven times), exterior dust (seven tunes), interior dust (seven times), and
paint and water (once). Blood samples were collected from participants five times and hand
wipes seven times. Interior surface dust samples were collected with a vacuum procedure
that we had previously developed using a portable pump designed for collecting personal air
samples. These samples were collected at the ulterior apartment entry, on a door mat that
was supplied to the participants and on a composite of interior floor areas and on a composite
of window sill areas. The exterior surface dust samples were collected with a battery-
operated "auto-vac" at the exterior entry to the housing of study participants and on paved
areas (streets, alleys, sidewalks, parking lots, etc.) throughout the study neighborhoods.
Buffer areas, extending for about one block in all directions outside the recruitment areas,
were also included in the exterior dust abatement. This was done in an effort to delay the
recontamination of paved areas from adjacent unabated paved areas.
1.3 RESULTS
A total of 307 children were involved in the study including 291 recruited during the
two recruitment efforts and 16 children who were born to study families during the study.
The focus of the study was on 173 children less than six years of age who lived in
rehabilitated housing and who were in the initial recruitment (Spring, 1989).
The number of blood and hand wipe samples collected was 1,367 and 8,127,
respectively, 2,407 exterior dust, 3,332 ulterior dust, 8,127 soil and 324 water samples were
collected and analyzed. In addition, 580 X-ray fluorescence (XRF) in situ paint lead
determinations were also performed.
1-3
-------
1.3.1 Soil Lead
The geometric mean soil lead in top 2 cm cores in Area A decreased from 200 ppm
preabatement to 54 ppm postabatement. The 95 percentile value decreased by over
2,200 ppm (from 2,659 to 422 ppm). In Area B the geometric mean decreased from 161 to
60 ppm and the 95% value decreased from 1,509 to 249 ppm, a drop of 1,260 ppm. In all
areas soil lead concentrations in lots adjacent to buildings were much higher than in other
lots. For example, in preabatement samples in Area A, the geometric mean and
95 percentile concentration were 201 and 2,856 ppm, respectively, near buildings while they
were 93 and 579 ppm, respectively, in other lots.
1.3.2 Exterior Dust Lead
In Area A the exterior dust lead loading (mg Pb/m2) was 260 ppm before abatement
and 257 ppm after abatement in the buffer areas. In the nonbuffer areas (center portion of
study area) the loading was reduced slightly from 419 to 347. In Area B the first
postabatement dust sample was collected within about 24 h of the abatement, a [somewhat
shorter time period than for Area A the previous year, hi an attempt to get a clearer
indication of the abatement impact. For one of the neighborhoods in Area B the lead loading
decreased in the buffer area by 48 % and in the non-buffer area by 63 %. In the other Area B
neighborhood, the decreases were 59 and 72%, respectively. Exterior dust lead loadings in
the exterior housing entry area decreased in Area A from about 230 mg/m2 preabatement to
about 100 mg/m2 postabatement but increased to preabatement levels about 4 mo later.
1.3.3 Interior Dust Lead
f\ ,
Interior entry dust lead loadings in Area A decreased from 387 /*g Pb/m tpreabatement
-------
about 60%. They remained at that same level in samples collected about four months
postabatement and were 35% below preabatement levels in samples collected 10 mo
postabatement. In Area B, geometric mean dust lead loadings in composite floor samples
were 138 ftg/m2 preabatement and were 81 % lower in the first postabatement sample and
62% lower in samples collected four months postabatement. However, by ten months
postabatement the mean value had increased to well above the preabatement level.
1.3.4 Blood Lead
Following interior and exterior dust and soil lead abatement, blood lead concentrations
decreased in Area A from 8.9 to 7.0 (21 %) but increased to 8.7 10 mo postabatement. In
Area B (interior dust abatement only) blood lead concentrations decreased from 10.6 to 9.2
(13%) 4 mo postabatement and were 18% below preabatement levels 10 mo postabatement.
However, blood lead levels in Area C (no abatement) also decreased by 29% and 6% during
these same time periods. Other comparisons also revealed no effects of the soil or dust
abatement.
1.3.5 Hand Lead
A comparison of the geometric mean hand lead in the first postabatement sample with
the preabatement sample revealed a decrease of 0.92 /tg Pb in Area A, an increase of
0.14 jttg Pb in Area B and an increase of 0.41 jig Pb in the control group C. Four months
postabatement the geometric mean hand lead was 1.49 fig below preabatement in Area A,
0.74 below in Area B and 0.68 below in control Area C. In the samples coEected 10 mo
postabatement, geometric mean hand lead levels were much higher than preabatement levels
in all areas. Other comparisons before and after the exterior dust and soil abatement in
Area B in 1990 also did not reveal any potential effects of the abatement.
1.3.6 Intercorrelation Among Environmental Lead Measures and
Blood Lead
Correlations between blood lead and hand lead, or interior dust lead (ppm and j^g/m2)
or exterior dust (ppm) were observed as they were in an earlier study of children hi the same
neighborhoods but with a wider range of housing types. Similarly, correlations were
1-5
-------
observed between hand lead and interior and exterior dust lead and between interior and
exterior dust lead. Correlations involving paint lead were not significant probably because of
the very low levels and narrow range of paint lead in the abatement project.
1.3.7 Environmental Dust Lead Levels
Geometric mean concentrations of lead in exterior dust ranged from 416 ppm (Area C)
to 2,119 ppm (Area A) in samples collected at the entry areas of the housing of study
2 2
participants. Mean exterior dust lead loadings ranged from 88 mg/m to 225 mg/m . In
street, sidewalk and alley samples, mean levels ranged from 829 to 2,216 ppm and 231 to
534 mg/m2. Mean dust lead in interior entry samples ranged from 261 to 559 ppm and
150 to 387 /ig/m2.
1.3.8 Abatement Costs
Average soil abatement costs (including removal, disposal, replacement and resodding)
were $35/m2 in 1989, $27/m2 in 1990 and $32/m2 in 1991. Interior dust abatement costs,
including carpet replacement and the limited replacement of upholstered furniture, were
$16/m2 in 1989, $18/m2 in 1990 and $13/m2 in 1991. The interior dust abatement costs may
also be expressed on a per housing unit basis, $1,212 in 1989, $1,477 in 1990 and $1,124 in
1991. Exterior dust abatement costs were $0.97/m2 in 1989, $0.89 in 1990 and $0.92 in
1991. On a per study participant basis, average abatement costs, as a percent of the total for
all three abatements, were 56% for soil, 26% for interior dust and 23% for exterior dust.
1.4 CONCLUSIONS
Soil lead abatement was achieved and there was no evidence for soil reconteunination
over the period of measurement (up to about two years).
While some evidence of exterior dust abatement was found, its impact was often not
evident at the next sampling phase (one to 3 mo later). Recontamination of paved areas often
occurred immediately after exterior dust abatement.
1-6
-------
2. BACKGROUND
In late 1987 Cincinnati was selected to be the location of one of three urban Soil Lead-
Demonstration Projects to implement Section lll(b)(6) of the Superfund Amendment and
Reauthorization Act (SARA) of 1986. This project was conducted by investigators in the
Department of Environmental Health at the University of Cincinnati in collaboration with the
City of Cincinnati Department of Health. The Cincinnati project was conducted in
neighborhoods where most of the housing had previously been lead-based paint-abated about
two decades earlier as a result of a complete rehabilitation carried out under U.S. Housing
and Urban Development-supported programs. The rehabilitation involved a "gutting" of the
buildings and the complete replacement of plumbing, wiring and heating systems and the
installation of new walls, flooring, windows and doors. Exterior brick areas were either
sandblasted or chemically cleaned and sometimes were re-painted. Most of the rehabilitated
buildings were 3-4 story multi-family and all were rental units. Soil in these neighborhoods
was generally located primarily in small parks, recreational areas and vacant lots and not
specifically part of the same property containing the rehabilitated housing. Therefore, in
order to include all of the soil areas in the abatement project the decision to include entire
neighborhood areas was made early in the project design. Thus all soil areas in the study
areas were included, whether or not there were housing units on the property. This
neighborhood-wide design, rather than the scattered house approach used elsewhere, also
allowed us to perform a neighborhood-wide paved surface cleaning.
2.1 HYPOTHESES
The two central hypotheses of this project were that:
(1) A reduction of lead in residential soil accessible to children would result in a
decrease in then: blood-lead levels, and
(2) Interior dust abatement, when carried out in conjunction with exterior dust and soil
abatement, would result in a greater reduction in blood lead than would be obtained
with interior dust abatement alone, or exterior dust and soil abatement alone.
2-1
-------
Secondary hypotheses were that:
(3) A reduction of lead in residential soil accessible to children will result in a decrease
in their hand lead levels, and
(4) Interior dust abatement, when carried out in conjunction with exterior dust and soil
abatement, would result in a greater reduction in hand lead than would be obtained
with interior dust abatement alone, or exterior dust and soil abatement alone.
The specific questions to be answered in the Cincinnati Soil-Lead Abatement
Demonstration Project were as follows: ;
i
• Does interior dust abatement in rehabilitated (including lead-based paint abated)
housing in conjunction with soil-lead and exterior dust abatement, or interior dust
abatement alone, result in the reduction of blood lead (PbB) in children relative to
children living in rehabilitated housing in the control area where no abatement
occurs?
• Does soil or dust abatement result in a covariate adjusted reduction hi PbB relative
to each child's pre-abatement PbB?
i
• Does exterior abatement (exterior dust and soil) result in a significant reduction in
PbB relative to that in children in housing where no abatement occurs?
• Is the reduction in PbB and environmental lead transient or long term?
Secondary objectives were to:
• determine the effectiveness of the abatement procedures in reducing the quantity of
lead-contaminated dust in residences; ;
* determine the rate of reaccumulation of household dust-lead;- ;
• determine the factors associated with household dust-lead reaccumulation; and to
• estimate the rate of exterior and interior recontamination.
2.2 PREVIOUS RELATED WORK
An ongoing study of childhood lead poisoning (Bornschein et al., 1985 and Clark et al.,
1991) led to the development of a causal model (Figure 2-1) that demonstrated the existence
of a lead exposure pathway leading from lead in external soil/dust adjacent to housing units
2-2
-------
CD
J3
Q_
CO
U)
O
00
o
X
£
Q
.0
QL
CO
CO
00 CM CM
CO CM 10
0 0 0
CO
T—
O
5
CM
O O O
O + + +
:= 8
fi 5
°- «
*-• ra
O
'o
al
o
eo
2-3
-------
(PbSS) and paint lead (XRF) to interior house dust (PbD), then to hand lead (PbH) and
ultimately to blood lead (PbB). This pathway suggested that an intervention strategy
involving control of exterior dust/soil, paint, and interior dust could potentially lead to
statistically significant reductions in blood lead. Contaminated soil serves as one of the lead
reservoirs that contributes to external dust on the street and other areas in or near housing
units. The reduction in soil lead in areas where the house paint lead has already been
thoroughly abated would, therefore, lead to a reduction in exterior dust lead and
consequently interior dust lead, hand lead and ultimately blood lead. However, because of
the existing accumulations of exterior and interior dust lead in the housing environment, the
impact on blood lead reduction of soil lead abatement alone would be delayed and reduced if
dust lead abatement did not also occur. The Cincinnati Soil-Lead Abatement Demonstration
Project was therefore designed to evaluate three abatement components suggested by the
causal model to be significantly linked to elevated blood lead levels:
Soil Lead abatement j
Exterior dust abatement
Interior dust abatement
The impact of existing paint lead was not expected to interfere with the evaluation of
the effectiveness of soil and dust lead abatement if the focus was on lead-paint free
rehabilitated housing units. Cincinnati fortunately has a large stock of about
6,000 residential housing units, available to low and moderate income families, that have
been completely rehabilitated and deleaded under various U.S. housing programs, mainly
during the late 1960's and early 1970's.
Blood lead profiles of children up to 42 mo of age living in rehabilitated housing units
(low paint lead), public housing (also containing only very low levels of paint lead), newer
post World War IE housing (low paint lead), and 19th century private housing units (multiple
sources of lead), that are in deteriorated/dilapidated or satisfactory condition (by exterior
evaluation) are shown in Figure 2-2. The children with the lowest PbB reside in the post
World War and public housing, while those in the deteriorated housing, not unexpectedly,
have the highest PbB values. What was surprising from the data in Figure 2-2, however,
was that children residing in the rehabilitated housing had PbB values much higher than those
in public housing, higher by about 10 /tg/dL at 18 mo of age, even though paint lead levels
2-4
-------
CD
jQ
Q.
25
20
15
I 10
Dot
v Sat
- Reh.
(54)
(4) Post WWII
Pub.
_L
_L
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42
Child's Age (In Months)
Figure 2-2. Effect of housing type and condition on early childhood blood lead
concentrations.
Source: Clark et al. (1988).
were equally low in both housing categories (Table 2-1). Blood lead levels of children living
in the rehabilitated housing were similar to those living in satisfactory housing that had not
been rehabilitated. The higher than expected PbB of children in rehabilitated housing
appeared to be due to the fact that most of the rehabilitated housing was intermixed in
neighborhoods with non-rehabilitated housing, frequently in a poor state of repair, which
resulted in much higher interior and exterior dust lead levels in rehabilitated housing than hi
public housing (Table 2-1). Being in close proximity to housing with lead paint, the
rehabilitated housing became contaminated by lead-contaminated dusts which migrated to the
areas immediately outside and within the rehabilitated units. This project offered the
opportunity to test the hypothesis that the blood lead levels of children living in rehabilitated
and satisfactory housing could be reduced by reducing lead levels hi soil and dust in the
areas in which housing is located.
2-5
-------
ft
i «3
Pi
AD
d
«>
S
52,
r-< O C? ~
o °5.
JQ
-------
i-Q
IS
"II
•
$
Post-2nd W
Private
Satisfactory
Condition
en . t-
VO «-l N
«o -<
o\
\o
.
5?
Og
-
s
.
<£ S
T3 -^
«H
-------
fclo O
111
g
i
?? a 5s
lll
3
I
o o
•
-
•Cl os
3 s
8
1
SIS
in
en
S 6?
.4_> CO O\
8 ^ dv
151 -5
1 ^
p
«o
;S
"
1
•g
S
is I
sis
w H .=}
11S
080
"SIS.
.3'
'&
i'S
s|* I
w "53 to
tf J4I
•8 S 3
O TO r)
}|J
•all
til
111
I
CD
«t-l
1
11 e 1
, t .a" §
O C4-, O
«-i O »""
_<^ -R
5 o
"^ fcrf
3
£ O
2-8
-------
3. STUDY DESIGN AND METHODS
3.1 OVERALL DESIGN
The Cincinnati Project involved the selection of three areas (A, B, and C) with a
predominance of lead-based paint abated (rehabilitated) housing and with an adequate number
of children up to five years of age.
The schedule of environmental and biological monitoring and for abatement for each of
three study areas is outlined in Figures 3-1 and 3-2. Area A received all three abatement
treatments (lead, exterior dust and interior dust) in 1989. Area B received interior dust
abatement in 1989 and exterior abatement (soil and dust) in 1990 and the control Area C
received all three abatements in 1991. Environmental and biological monitoring was
conducted at a total of 9 times before and after each abatement cycle except that no
monitoring was performed after the 1991 abatement in control Area C.
3.2 NEIGHBORHOOD AND SUBJECT SELECTION, RECRUITMENT
AND RETENTION
As mentioned earlier, the design of the Cincinnati project involved the selection of
multi-block areas where the predominant type of housing was that which had previously
undergone extensive "gut" rehabilitation with the removal or encapsulation of most of the
lead-based paint areas. Therefore, a neighborhood area selection process was used which
required a complete census of all property owners and residents and an initial soil sampling
survey. This was needed to determine if there was an adequate number of children less than
six years of age (about 50 per study area) and accessible soil with elevated lead concentration
(greater than or equal to 500 ppm). The procedure for calculating the sample size required
is presented in Appendix A.
3.2.1 Enrollment Criteria
Selection of several study areas with similar characteristics was important in order to
determine the impact of the exterior and interior lead abatement procedures. This was
3-1
-------
Figure 3-1.
ABATEMENT AND MONITORING TIME TABLE
Year 1 - 1988 Soil Pb Surveys, Methods Development, Pilot Testing of Abatement
September - December (Phase 0) Methods, Training, and Negotiation with Property
Owners \
Year 2 - 1989
January - May
June/My
August/
September
September/
October
November
Year 3 - 1990
February
June
August
September
November
Year 4 - 1991
February
June
June/July
July-
December
Year 5 - 1992
Continuation of Above
Area A
Ma
abate
soil
exterior dust
interior dust
EMb
M
P/WM°
M
no abatement
EM
M
P/WM
M
no abatement
Sample and
AreaB
M
abate
interior dust
EM
M
P/WM
M
abate soil and
exterior dust
EM
M
P/WM
M
no abatement
data analysis
Complete sample and data
AreaC
M
no abatement
EM
M
•
P/WM
M
no abatement
EM
M
P/WM
M
abate soil,
exterior dust,
interior dust
analysis and report
[
(Phase 1)
(Phase 2)
(Phase 3)
(Phase 4)
(Phase 5)
'•'
(Phase 6)
(Phase 7)
(Phase 8)
(Phase 9)
writing
*M = Monitoring: Blood Lead, Hand Lead, Interior and Exterior Dust Lead and Soil Lead.
EM = Environmental Monitoring: Hand Lead, Dust Fall Interior and Exterior Dust Lead.
CP/W M: Pant and Water Monitoring.
In other areas sampled during the initial soil surveys but not selected for the study, necessary soil abatement in
public areas occurred during 1990.
3-2
-------
Figure 3-2.
BLOOD AND ENVIRONMENTAL SAMPLE COLLECTION
Sampling Phase
Sample 00
Blood
Hand Lead
Soil
Surface
Scraping X
Top 2 cm
Areas X*
Bottom 2 cm
Areas X
Exterior Dust
Neighborhood
House-Targeted
Interior Dust
Entry
Floor
Window
Dustfall
Mat
Paint
Water
01 02
X
X X
X
Xa
X
X X
X X
X X
X X
X X
X
X X
03 04
X
X
X
X
X
X
X
X
X
X
X
X
X
05
X
X
X
X
X
X
X
X
X
X
X
06
X
X
X
X
X
X
X
X
X
X
X
07
X
X
X
X
X
X
X
x
xb
xb
xb
08 09
X
X
X
X
xb
xc
x°
alnitial samples for Dandridge neighborhood of Area B were collected during Phase 02.
Collected but not analyzed.
"Thase 08 samples collected only from housing not sampled during Phase 04.
accomplished by selection of several areas within the potential study neighborhoods where
socioeconomic and ethnic backgrounds were similar. Other matching variables considered
included housing type, age of child, prior housing history, season, soil availability per block
(group match) and percent non-rehabilitated housing per block (group match).
In order to maximize the potential impact of soil-lead abatement on lead (PbB)
reduction, several features of the study area were sought. These included: a high percentage
of completely rehabilitated housing (largely free of lead-based paint) and a high percentage of
young children exposed to accessible, lead-contaminated, soil.
3-3
-------
3.2.2 Initial Soil Survey
An initial soil survey was completed in order to determine the concentration of soil-lead
in the proposed study areas. Soil-lead distribution was one of the criteria used in the final
selection of the project study areas. This information was also necessary for the
development of the abatement plans.
At the time of the initial soil survey, November 1988 through April 1989, there were
six neighborhoods being considered for inclusion as part of the Cincinnati Soil Project's
proposed study areas. One of the original neighborhoods under consideration was dropped
after the census because of an insufficient number of children and another neighborhood was
added in the Spring of 1989. The six neighborhoods were combined to make up the three
study areas, Area A, Area B, and Area C.
The initial soil survey began with the identification of all soil sites in the neighborhoods
under consideration for inclusion in the study areas. This was accomplished by first defining
the neighborhoods on Sanborne maps. (These are large scale, 1" = 100', city maps which
show streets, building outlines and approxiniate property lines.) The next step was to visit
each study area to visually locate all soil sites and rehabilitated housing. Soil areas and
rehabilitated housing units were then coded on the Sanborne maps (see attached). Two
separate teams of individuals independently surveyed, on foot, each neighborhood for
housing and soil sites. Any discrepancies between the surveys results were field checked by
project managers.
At the conclusion of the field surveys, the locations of the soil sites on the Sanborne
maps were compared with plat maps from the Hamilton county Auditor's Office in order to
determine the book, page and parcel number of each property within the neighborhood. It
was necessary to make this determination of book, page and parcel in order to find the name
and address of the owner of each soil site and each property where other types of abatement
would occur.
The property owners consisted of two groups; private owners and public owners. To
obtain permission to sample soil on the privately owned sites, all owners, including
individuals and corporations, were contacted by letter (see attached). Approximately 15% of
the owners responded to the letter. The owners who did not respond were contacted by Soil
Project staff members. A small percentage of the owners who were contacted refused to
3-4
-------
grant permission to sample soil. In one case the property was involved in litigation. In
other instances, the owners were apprehensive of the consequences.
Verbal permission to sample soil was acceptable, according to the Risk Management
Office of the University of Cincinnati Medical Center, if we followed certain rules. Those
rules were:
Dates and times of phone contact were recorded.
Notes were made of phone conversations.
Only a small number of approved individuals could make owner contact.
A significant amount of property with soil, approximately 20% in the study areas, was
owned by the City of Cincinnati. A right of entry agreement between the City of Cincinnati
and the University of Cincinnati was required prior to sampling soil on city-owned property.
In order to process the right of entry agreements we had to make the initial request of the
city government department responsible for each parcel of property. Those departments
were the Recreation Commission, Parks Department, Engineering and Public Works, and
Neighborhood Housing and Conservation. Once the individual department approved, the
City Solicitor approved the document as to form, and finally the City Manager signed the
document. The right of entry document then had to be approved by five separate offices at
the University of Cincinnati and signed by a University contracting officer.
After the identification of the soil locations and necessary permission to sample was
obtained, a sampling plan was completed for each soil site. The technicians who were to
collect the soil were trained in the collection protocol and appropriate health and safety
issues.
The initial soil survey was begun in December of 1988 and completed by April 15,
1989, with the majority of sites having been sampled by March 1, 1989. Soil samples were
collected by the environmental monitoring teams and sieved at the temporary field office at
1400 State Avenue. The sieved samples were sent to the Kevex (XRF) Soil Lab at the
University of Cincinnati Medical Center, where they were analyzed to provide preliminary
data for characterizing the lead content in the soils in the potential study areas. Later in the
study, these samples would be reanalyzed, after the recalibration of the Kevex (XRF), for a
final determination of soil-lead concentration.
3-5
-------
The initial soil survey was completed and the data were used to formulate the abatement
plans for the soil abatement which occurred in the summer of 1989 and two subsequent
summers of the project.
3.2.3 Identification of Study Areas
To determine which neighborhoods were appropriate for the study, project staff
conducted an extensive door-to-door census to obtain information about age of children,
housing types and condition of housing. Two-person teams canvassed several neighborhoods
to identify potential study areas. After careful evaluation of the census data, three areas
were selected for the soil-lead abatement project. Each area had 40 to 60 children under five
years of age, with most living in rehabilitated housing. Areas were also similar with respect
to the percentage of surface area that consisted of soil (average = 24%; range = 20 to
27%). The three study areas, A, B and C, each consisted of about 5-6 blocks of buildings,
paved areas and soil areas. Area A was completely contiguous. Area B consisted of three
non-contiguous sub-areas, referred to as neighborhoods: Findlay, Back and Dandridge.
Area C consisted of two non-contiguous neighborhoods: Glencoe and Mohawk. (Soil
sampling data presented later in this report revealed that the areas were comparable in lead
concentration with Area B and C being practically equivalent and A only moderately higher).
Maps of the study areas appear in Figures 3-3, 3-4, and 3-5.
3.2.4 Participant Enrollment Procedures
In the Spring of 1989, individual letters were sent to prospective study families within
the identified neighborhoods. A fact sheet describing the study and the benefits of
participation was also included in this mailing (see Appendix B). Shortly after the mailing of
the study announcement, recruitment teams, consisting of two persons, visited each family to
personally invite them to participate in the soil-lead abatement project. At this visit, a brief
description of lead poisoning in children was given and why participation hi the project could
benefit the community in general and their children in particular.
In addition, if the family agreed to participate, a written appointment reminder for
blood collection was given to the caregiver, along with notification that a fixed transportation
3-6
-------
A - PEND
Figure 3-3. Schematic diagrams of the Pendleton (Area A) and Findlay (Area B)
neighborhoods.
March 31, 1993
3-7
DRAFT-DO NOT QUOTE OR CITE
-------
B2 - BACK
Figure 3-4. Schematic diagrams of the Back and Dandridge neighborhoods, both hi
Area B. \
March 31, 1993
3-8 DRAFT-DO NOT QUOTE OR CITE
-------
C1 -MOHA
^
V-r. :.- -•-•__-
LEROY
COURT
* ' B :
F-
!4 E h,i
C2 - GLEN
Figure 3-5. Schematic diagrams of the Glencoe and Mohawk neighborhoods, both hi
Area C.
March 31, 1993
3-9
DRAFT-DO NOT QUOTE OR CITE
-------
expense reimbursement would be provided by the project. Referral names of other potential
participants were solicited. :
Several days prior to the blood collection appointment, an appointment reminder was
mailed and/or a reminder phone call was made to each family to encourage compliance.
A second census and enrollment period was conducted in 1990 using the same
procedures as above. The first blood collection for these individuals was in Phase 05 (June-
July 1990).
3.2.5 Community Relations
The Soil Project was announced with a city-wide, press conference conducted by the
University of Cincinnati Department of Environmental Health and the City of Cincinnati
Department of Health. The press conference was well attended by all media. (Videos of the
television coverage were utilized in employee orientation sessions.) Subsequent
announcements were made to local community leaders and property owners through small
meetings, letters with an accompanying project fact sheet (Appendix C), and phone calls.
Initial contact with potential participating families was made by individual letters and fact
sheets and followed up with visits to individual homes to describe, in detail, project goals
and to answer any questions.
An inner-project communication form was used to alert staff to family questions or
concerns. Use of this communication mechanism allowed for greater staff/family rapport and
enhanced continued family participation.
Location of the field office within the study area provided the families with convenient
* access to project staff. Staff remained visible and available in the study neighborhoods as
well.
Prior to each abatement phase, letters were sent to community leaders and property
owners describing the abatement procedures that would be implemented and alerting them of
a potential community disruption that might occur as a result of the abatement (temporary
playground closure etc.). Personal visits were made to property owners or tenants if
appropriate.
Interaction with health and public works agencies was maintained throughout the
project. Blood lead results were provided with subject consent to physicians and/or health
3-10
-------
care facilities to eliminate duplication of blood lead testing; i.e., blood lead screening efforts
by the Cincinnati Health Department clinics. Further, close communication with the
Cincinnati Lead Screening Program was maintained throughout the project, alerting them to
children requiring more aggressive follow-up as indicated by Centers for Disease Control
(CDC) guidelines. Project staff provided assistance to the Health Department in the tracking
of study participants with elevated blood lead.
Project presentations were made to several interested community groups describing
project plans and goals as well as providing general lead poisoning prevention information to
the community.
Quarterly newsletters were sent to participating families to keep them informed of
project activities and timelines. These newsletters also included general community
information and activities as well as other timely articles (see Appendix D).
With few exceptions, the soil-lead demonstration project was well accepted by the study
neighborhoods.
3.2.6 Subject Retention Plan
Several techniques were employed to maximize continued family participation. Caring
and attentive staff were key factors in continued family participation. Because of the
intrusive nature of the project, families often needed encouragement by project staff to keep
blood collection, ulterior sample collection and interior abatement appointments.
Continual family communication was maintained throughout the project. Special family
events were recognized such as family birthdays, new births and special holidays. Quarterly
newsletters were sent to individual families and caregivers were encouraged to contribute
articles or favorite recipes to the newsletter. Caregivers were encouraged to share photos for
the "Kids Photo" door at the project office. Periodic phone calls and home visits were made
to the caregivers simply to provide individual attention to the family.
To ensure families would not incur any expense associated with project participation, a
fixed transportation reimbursement ($10) was provided each time a family came to the
project office for blood collection. Also at the five blood collection times, each child
received a modest gift. Sunglasses were a favorite choice.
3-11
-------
Family mobility posed the greatest retention challenge. Often times, by moving,
families were able to improve their standard of living. Moving meant further project
participation was not possible based on project study design. Families moving during the
course of the study accounted for most (98%) of the attrition. Three families were dropped
from the study for repeated failure to keep scheduled appointments.
The new carpets and furniture provided as part of the interior dust abatement also
served as an incentive to remain in the study for the control group (C). However, since
these new furnishings were distributed to families in Areas A and B early in the study
(August-September 1989) their value as an incentive in these two areas was primarily as an
incentive* to join the study.
3.2.7 Human Subject Research Review Process
The University of Cincinnati and federal and state statutes require that all research
projects involving humans be prospectively reviewed and approved by the Internal Review
Board (DRB). The ERB's role in human research is to protect the rights of research
participants; approve research protocols that do not pose undue risk to subjects; ensure
subject understanding of the nature of the research and finally, ensure that the subjects's
participation was voluntary (consent form signature).
Therefore in early 1989, a detailed research protocol and consent statement was
submitted to the University of Cincinnati 1KB for review. Approval to conduct this human
research project was obtained prior to the recruitment of participants in May 1989.
Subsequently, annual progress reports have been submitted to the IRB as required.
Copies of the approved consent, medical release of information and withdrawal forms are
included in Appendix D.
3.3 ABATEMENT PROCEDURES
The effects of three lead abatement strategies: soil abatement, exterior dust abatement,
and interior dust abatement, were investigated as part of the Cincinnati Soil Project (CSP).
At the beginning of the project in 1988, there was limited technical information available on
the conduct of such abatements. Then, as now, the only cost-effective method!for the
3-12
-------
abatement of lead-contaminated soil was removal and replacement. Consequently, that was
the method selected.
Very little reliable data was available on the effectiveness of exterior and interior dust
abatement procedures in use at that time. This lack of information required a significant
effort to be expended on the development of effective protocols suitable to achieve the
desired reduction in lead exposure. The abatement procedures to be used also had to be
compatible with the specific requirements of a research project. Developing the interior and
exterior abatement methods was the first major focus of abatement-related activities at the
beginning of the Cincinnati Soil Project.
The second major focus of activities was the development of procedures necessary to
perform the abatement work through the University of Cincinnati, a state institution. Certain
difficulties were anticipated because some of the abatements were significant, invasive
procedures performed on both private and public properties in an urban environment. The
problem was compounded even further by the fact that the majority of the study population
was expected to reside in rental property. Thus it would be necessary to solicit participation
from families living in the rental units and additionally solicit the cooperation of the owners
of that rental property. This also compounded the liabilities associated with the abatement.
The University of Cincinnati would be responsible for damage both to the real property of
the landlords and the personal property of the renters, the participants in the research project.
Since the abatement was to take place in an urban environment other liabilities included
health and safety issues associated with handling lead-contaminated materials in populated
areas and operating heavy construction equipment in the same environment. Additional
concerns resulted from the uniqueness of the project. It was very unusual for a public
institution to make significant improvements such as the reduction of lead, on private
property.
It is necessary to define abatement as it was performed by the Cincinnati Soil Project.
The overall objective of the Soil Project, as stated elsewhere in this report, was to determine
if reducing exposure to soil lead, exterior dust lead, and interior dust lead would result in a
reduction in blood lead and/or hand lead in young children. Since there was no legal
requirement to perform the abatement, all participation by families and property owners was
voluntary. This resulted in some restrictions to performing a complete lead abatement.
3-13
-------
Additionally, budget constraints also prevented total abatement of the lead in the
neighborhoods where the research took place. Because of these constraints all of the
environmental lead was not completely abated, exposure to lead was only reduced. The
official name of the project suggests that it was not an abatement project. The Cincinnati
Soil/Lead Abatement Demonstration Project was in fact an abatement demonstration project
not an abatement project. That concept helped guide some decisions made during the
planning and execution of the project.
3.3.1 Interior Dust Abatement Methods Development
Lead dust exists in residential buildings in a variety of locations. Lead can be found in
the dust on floors and carpets. It can be distributed in the heating, ventilation, and air
conditioning (HEPA) system, attics, basements, furniture, and closets. In 1989, the
distribution of lead in residential units was investigated in an interior dust abatement project
in the South Riverdale neighborhood in Toronto, Canada where a secondary lead smelter was
the major source of the contamination. A pilot project (involving eight houses) was
conducted prior to the abatement of lead which occurred in over 800 homes. The lead
removed from the eight houses was distributed in the following manner:
Floors (dry removal) 42%
Surfaces (wet removal) 8%
Furnace Ducts 30%
Other Sources (dry removal) 16%
Other Sources (wet removal) 4%
The majority of the housing in the Cincinnati Soil Project was expected to utilize either
steam or hot water healing systems, thus there would be no heating (furnace) ducts. Based
upon this information and the overall goals and limitations of the Soil Project, an early
decision was made to concentrate the abatement activities on the floor dust and that dust
which accumulates on ledges, window sills and window wells. Interior methods development
focused on determining effective and efficient methods for reducing dust on the surfaces.
Because the floors in rehabilitated housing in urban Cincinnati were expected to be wood,
vinyl, or carpet, an effort was made to develop an efficient and effective procedure to abate
lead dust on wood and vinyl floors and carpet. I
3-14
-------
3.3.1.1 Development of Abatement Methods for Wood and Vinyl Floors
In order to test the effectiveness of proposed floor cleaning procedures, homes with
high levels of lead in the dust were needed. Such homes were located through other research
projects being carried out by the Principal Investigators of the Soil Project and other
investigators within the Department of Environmental Health. These projects involved
studying the effects of lead exposure in young children. The records from these projects
included data on both lead levels in the interior dust and data on floor type. From those
records it was possible to obtain a list of individuals meeting the following criteria:
(1) currently or recently active in other lead research projects;
(2) high lead concentrations in the interior dust;
(3) either carpet, vinyl, or wood flooring or a combination of those flooring types.
Because of these subjects' positive associations with University of Cincinnati researchers in
the past, it was generally easy to obtain permission from those study participants to test
cleaning methods on different wood and vinyl floors.
The questions to be answered in these homes were:
(1) In order to reduce dust lead to an acceptable level with a high efficiency particle
accumulation (HEPA)-equipped vacuum cleaner, what rate of speed should the
operator use and how many times should the floor be vacuumed?
(2) Would wet washing remove additional lead?
The testing procedure was as follows:
(1) Three separate squares, one meter by one meter, were delineated on the surface
of the floor type being tested.
(2) A separate dust sample was collected using the dust vacuum method (DVM) (for
a description of the DVM see Interior Dust Collection in the Mid-Term Project
Update) from a 25 x 25 cm area from within each of the one meter squares.
(3) A bag for a HEPA-equipped vacuum cleaner was tare-weighed and installed. (In
this study a Nilfisk GS80 was used.)
»
(4) The three squares were then vacuumed with the Nilfisk at the rate of 60 seconds
for each square meter.
(5) The vacuum bag was removed and weighed after this cleaning.
3-15
-------
(6) A second DVM sample was collected from each of the three squares and the
squares were cleaned again with the vacuum cleaner.
(7) The process was repeated until the squares were cleaned a total of five times and
sampled a total of six times.
(8) Each of the squares was then washed with 1,500 ml of tap water from the
residence. The washing was performed by a technician using a new sponge and
wearing rubber gloves. An aliquot of 500 ml was taken from each of the wash
buckets. A 500 ml sample of clean water from a wash bucket was; also collected
to provide data on background lead levels in the tap water.
(9) A final DVM sample was collected from each square.
All of the dust samples and water samples were analyzed for lead concentration. The final
result of this series of tests indicated that the lead concentration in the dust dropped from an
average of 1,121 ppm to below detection and the loading dropped from 178 mg/m to below
detection after the first vacuuming.
3.3.1.2 Development of Abatement Methods for Cleaning Carpeting
To determine the feasibility of cleaning carpets with a vacuum cleaner, three questions
were asked:
(1) Is there a difference among industrial type, HEPA-equipped vacuum cleaners in their
ability to remove lead dust from carpeting?
(2) How many passes at a speckled speed with an industrial-type, HEPA-equipped,
vacuum cleaner will reduce the surface lead dust to an acceptable level?
(3) Is wet carpet cleaning alone or in combination with dry vacuuming effective in
reducing surface lead dust?
In order to test the effectiveness of various industrial-type vacuum cleaners, manufacturers
were asked how vacuums differed. Vacuum cleaners are rated differently based upon two
characteristics:
(1) cubic feet per minute (cfm) of air moved by the machine;
(2) the ability of the machine to lift a column of water a specified number of inches.
3-16
-------
Higher cfm of air moved and higher water lift were said to result in better performance.
Were these differences significant? Three different machines which varied in these two
characteristics were selected for testing (Table 3-1).
TABLE 3-1. COMPARISON OF STREET SWEEPING MACHINES
. CFM Waterlift
1) Nilfisk GS80 87 75"
2)Wap767 100 90"
3) Euroclean UZ930 77 85"
In order to test these cleaners, it was decided to use the ASTM F607-86 Standard Laboratory
Method for Evaluation of Din Removal Effectiveness of Household Vacuum Cleaners. This
test procedure specifies "seeding" carpeting with known quantities of sand and talc,
embedding the material into the carpet, vacuuming the carpeting, and measuring the amount
of material recovered. Analysis of the data resulting from testing the three machines listed
above indicated the following:
(1) There was no significant difference between operators.
(2) There was no significant difference between vacuums.
In order to determine the number of passes required to reduce the surface lead dust
available to children, lead-contaminated carpets were needed. That need was again met by
families who were participating in other lead-related research projects. Families living in
high-lead environments, with carpets were contacted and asked if they would like to
exchange their existing carpet for new carpet. Those who agreed, all of those asked, were
offered a choice of carpets from three specified types and colors that a local vendor had in
stock. After the selection was made, the new carpet was ordered with edges bound.
The exchange was scheduled with the participant. A crew of three performed the
exchange which began with the collection of dust samples (dust vacuum method) from the
old carpet and the new carpet. A sheet of 6 mil Polyethylene was placed over the old carpet
for the purpose of not contaminating the top of the carpet with dirt or dust from the floor or
3-17
-------
bottom of the carpet. Enough plastic was left at the edges and the ends to seal the carpet in
the polyethylene hi order to prevent contamination of workers, transport vehicles and the
laboratory where the research would take place. After the carpet was removed, the floor
was vacuumed with HEPA-equipped vacuum sweeper and the dust was saved for later
analysis. The new carpet was unrolled and the furniture was put back in place. Photo and
time records were made of the process for use in the development of specifications for the
abatement contractors.
The lead-contaminated carpets were then taken to a research laboratory at the
University of Cincinnati where graduate students vacuumed one-meter square areas at fixed
rates of speed. Dust samples were taken between each vacuum treatment. These tests
indicated that even after ten passes at the rate of one minute per square meter with a HEPA-
equipped vacuum cleaner (Nilfisk GS80), significant surface lead dust remained.
In order to test the effectiveness of wet cleaning, a Rug Doctor steam cleaner ;
manufactured by the Rug Doctor Co. of Fresno, California was used to perform a series of
tests in which carpets were first dry vacuumed and then steamed cleaned. Additional tests hi
which carpets were steam cleaned and then dry vacuumed were also performed. The results
indicated that neither sequence was adequate to remove surface lead dust to an acceptable
level.
The final conclusion reached from this series of tests on carpets suggested that it is
more economically feasible to replace carpet than to clean carpet.
Earlier data from other investigations carried out within the Department of
Environmental Health have demonstrated that just as it is difficult to remove lead from
carpet, it is also difficult to remove lead from upholstered furniture. Because of this, and
because of our recent data on carpets, it was decided not to do any further research oft the
removal of lead from furniture.
3.3.1.3 Interior Dust Abatement Procedure Summary
Interior dust abatement followed the completion of exterior dust abatement in an area if
both occurred during the same time period. Interior dust abatement consisted of a
combination of vacuuming with a HEPA equipped vacuum and wet cleaning non-carpeted
surfaces (including window sills and wells) with a detergent. The vacuuming of floors was
3-18
-------
timed; carpets were vacuumed a total of three times at a rate of one minute per square yard
each time.
Selected carpets and upholstered furniture were also replaced. Residents with one child
in the study were provided with replacement carpeting for one common room, for example
the living room, plus the child's bedroom. Residents with two or more children in the study
received an additional carpet for a total of three carpets. Residents with a child in the study
were also eligible for the replacement of one standard sofa and one standard chair.
Interior dust abatement procedures were selected after extensive methods development
which revealed that chronically-contaminated carpets were not able to be effectively cleaned
even after numerous vacuuming effort with a HEPA-equipped cleaner. More details on
methods development and the final protocol are presented in Appendix F.
3.3.2 Exterior Dust Methods Development
Lead-containing dust exists in the exterior environment on a variety of surfaces. The
most common of these surfaces are streets, alleys, sidewalks, parking lots, steps and porches.
The largest surface areas which were to be cleaned as part of the abatement, were the streets,
alleys, and parking lots. Street cleaners would be the obvious equipment for use in cleaning
these large areas. Small walk-behind units or vacuum cleaners appeared to be the most
efficient equipment for sidewalks and other small areas.
3.3.2.1 Types of Pavement Cleaners Available
The first step in determining the most efficient street cleaners available for abatement of
exterior dust was to determine the scope of available equipment. An initial survey of all
street cleaner manufacturers marketing products hi the United States indicated four different
types of equipment. The four types differ in the particular mechanism used to remove dust
and debris from paved surfaces. These mechanisms are as follows:
(1) Broom sweepers
(2) Vacuum-assisted broom sweepers
3-19
-------
(3) Vacuum sweepers
(4) Regenerative air machines
The broom-type sweeper operates by one or more rotating brushes which move the^ dirt and
debris from the pavement to a conveyor, which then carries the material to a hopper. This
represents the first generation of mechanical street sweepers and has been used to clean city
streets for decades. These sweepers generally use water spray to control dust.
Vacuum-assisted broom sweepers represent a further advancement of this concept.
These machines operate by means of one or more rotating brashes which loosen diit and
debris from pavements and move it toward a hopper. At this point the process is assisted by
means of moving air or a vacuum which pulls the dirt and debris into the hopper. These
machines sometimes use water spray to help control dust. The air is filtered by a variety of
filter types, some of which are effective in removing particles as small as 2-3 microns.
Vacuum sweepers operate principally by means of an extremely powerful vacuum
k
system. Some of these cleaners will also use a rotating brash to move debris and dust from
the curb to the head of the vacuum apparatus. In order to control dust generated by the
rotating curb brooms, a fine water spray is often used. The exhaust air from the vacuum
system can be filtered through a wide range of filter size. Most manufacturers provide a
final filtration range between 2-8 microns.
The final type of pavement cleaner is the regenerative air machine. This machine
operates by means of recirculating air system. The blower part of system directs a blast of
air onto the paved surface, and the vacuum part of the system pulls the same air back up into
the hopper. The blast of air loosens material on the paved surface and the vacuum pulls that
material back into the hopper. No filtration system is needed for this machine because the
air is recirculated through the system. Brashes are available to move material from curb
areas into the path of recirculating air system.
3.3.2.2 Exterior Dust Testing
Because of the large number of manufacturers and the variety of machines made by
each manufacturer, it became apparent that we would not be able to test all of the machines.
It was decided to test the different classes of machines. Because of the lack of dust-control
3-20
-------
system on the broom-type sweepers, this class of machines was eliminated from
consideration. It was also readily apparent from observing the operation of these machines
and reviewing the literature that significant quantities of dust and dirt remain on the surfaces
cleaned by these machines.
The types of equipment to be tested were vacuum sweepers, vacuum-assisted broom
sweepers, and regenerative air machines. In order to see if one class of machines cleaned
better or more efficiently than others, several different machines within each sweeper type
were selected for evaluation.
Each test of a particular machine required 2 to 3 people and 3 to 4 h of time. The test
consisted of finding a "worst-case" section of pavement, i.e., a section of pavement outside
of the designated potential study areas, but similar to the worst kind of pavement within the
study areas. The "worst-case" was thought to be the brick-paved alleyways. By testing the
machines on these surfaces it would be possible to determine any differences that exist
among the machines.
The test consisted of delineating four 1-meter square areas in a row on the designated
pavement surface. The 1-meter squares were divided into four quadrants, designated A,B,C,
and D. Quadrants marked "A" were then cleaned by scraping and vacuumed with a small
portable vacuum cleaner in order to determine the original loading of dust on the surface.
The material removed from each "A" quadrant was weighed. The street cleaner being tested
was then driven across the four test squares. The quadrants marked "A" were recleaned with
our sampling vacuum and any dirt removed from those quadrants was weighed to
determining the amount of dirt/debris redistributed during the street cleaning process. The
four quadrants marked "B" were cleaned by the portable vacuum sampler and the dirt/debris
was weighed to determine, by subtracting from the amounts originally on "A", the amount of
material picked up by one pass or the street sweeper. The sweeper was then driven across
the four squares for a second time. After this vacuuming quadrants marked "A" were
brushed and cleaned with our small sampling vacuums and the material was weighed again to
determine any redistribution of material. The quadrants marked "B" were then cleaned, and
the material weighed as before, and finally the quadrants marked "C" were cleaned, and
material was weighed as before.
3-21
-------
This test was repeated three times on one machine from each class of street sweepers.
Two or three machines from each class were tested at least one time with this described
method. I
Analysis of the data collected from these tests indicated that several machines operated
at a level of efficiency so that 98% of the dirt was removed from the surface after two
passes. Selection of street sweeping equipment suitable for exterior dust abatement was
based upon the results of this testing. Any machine capable of removing above 95 % of the
dust loading after two passes would be acceptable for the abatement work.
The results of the exterior methods development efforts are presented in an appendix.
Similar tests were carried out on walk-behind type pavement sweepers and industrial
type vacuum cleaners in order to determine the most effective machine available for cleaning
smaller paved surfaces.
3.3.2.3 Exterior Dust Abatement Summary
Exterior dust abatement occurred immediately after soil abatement was completed in a
particular area. Streets were generally abated one side at a time, principally because of the
requirement that the street be free of parked cars for complete abatement to occur. The
study areas generally have more on-street parking than off-street parking and it was not
practical to restrict parking on both sides of a street at any given time. Therefore, abatement
occurred one side per day with the alternate side abated the following day. The paved
surfaces were cleaned with vacuum equipment capable of removing greater than 99% of the
dust after two passes as determined under test conditions. In some situations, hand tools
were needed to loosen material in cracks and crevices prior to vacuuming, particularly along
alleys and on sidewalks.
3.3.3 Soil Abatement Methods
At the beginning of the Soil Project, the soil abatement was expected to be
accomplished in one of three ways depending upon the concentration of lead in the soil
column (Figure 3-6). Some soil required resodding or, in some cases, the addition of topsoil
to a site or part of a site prior to resodding. The purpose of the topsoil was to provide a
base for the sod and/or to fill in the depressed areas, so that proper drainage could occur.
3-22
-------
E
Q.
c
o
1
.*_•
i
I
CM
i
CD
1,250
1,000
750-
500-
250-
Remove, Replace,
and Resod
Reso
if Cove
Inade-
quate
Cultivate
and
Resod
300
500 750 1,000
Top 2 cm Concentration, (ppm)
1,250
Figure 3-6. Decision criteria for soil abatement. The original plan for soil parcels with
the top 2 cm between 500 and 1,000 jig/g and the bottom 2 cm between 0
and 500 jtg/g was to cultivate and resod. Because a satisfactory method for
cultivating was not available, this soil was removed and replaced instead.
On some sites tilling was attempted as the appropriate method of abatement. Preliminary
testing suggested that mixing was not thorough, and delays in testing the adequacy of mixing
made the procedure unpractical as we implemented it. A more thorough evaluation of this
method is needed. We elected to discontinue this method on this project and to remove and
replace the soil instead.
On those sites where the concentration of lead required the removal of soil, the soil was
excavated to a depth of six (6) plus inches and replaced with topsoil having a lead
concentration of less than 50 parts per million (ppm). Where excavation was required the
soil was removed by mechanical equipment either front-end loader or backhoes or hand
3-23
-------
tools. It was loaded into hoppers or dump tracks and transported to an appropriate dump
site. During transport, hoppers or dump tracks were covered with tarps. Prior to loading of
hoppers or dump tracks, 2 mi-thick polyethylene was placed in the hoppers of dump tracks
to prevent soil from being spilled through cracks in the tailgate or beds of the transport
vehicles. In order to ensure that the soil was transported to the appropriate dump site, all
i
trucks were logged off of the abatement site by the Site Inspectors and logged into the
appropriate dump site by officials at the dump site. Appropriate techniques were used to
prevent spills of contaminated solid at the abatement site along with techniques for the
containment of dust. Excessively dry soil was moistened with water to control dust.
Replacement of soil was accomplished with a different set of equipment in order to prevent
contamination of replacement soil by soil or dust remaining on equipment used for ,
excavation. A program for the watering of sod began immediately after its placement and
remained in place during the summer months. Barriers were used when necessary during
soil abatement in order to protect people from injuries resulting from construction equipment.
Barriers were also necessary to prevent potential additional exposure to the residents in the
area.
i'
The three abatement strategies for soil differed, depending upon the concentration of
lead in a 15 cm core. Those strategies were removing and replacing, followed by sodding;
cultivating, followed by sodding; and resodding. If the average lead concentration in the soil
in the 15 cm column was greater than 500 ppm, regardless of the adequacy of the grass
cover, the soil was removed and replaced and the area resodded. If the average lead
i;
concentration in .the column was less than 500 ppm but the concentration in the top 2 cm was
500 ppm or higher regardless of its grass cover, the area was cultivated, to reduce the
concentration in the top 2 cm to less than 500 ppm in both the top 2 cm and in the column
and resodded. For areas where grass cover was adequate and the lead concentration was less
than 500 ppm in both the top 2 cm and in the column, no abatement occurred. If the soil
lead concentration in the top 2 cm was 300 ppm but less than 500 ppm and the average
concentration in the column was less than 500 ppm and the grass cover was inadequate the
area was resodded. No soil abatement occurred in areas where the grass cover was
inadequate but the concentration in the top 2 cm was less than 300 ppm and in the column
the average was less than 500 ppm.
3-24
-------
In order to assign the specified abatement strategies to the sites in the study areas, the
preliminary soil lead concentrations from the initial soil survey were superimposed over the
original site maps that were used for sampling. The specific lead concentrations were written
on the sampling lines shown on the site maps. Since parcel boundaries were also indicated
on the site maps, we were able to tell which particular parcels required abatement by each
specified method. On sites where there was only one parcel it was clear what form of
abatement was necessary. On sites consisting of more than one parcel we used the following
criterion:
If all soil-lead concentrations on all parcels indicated the same method of
abatement, the entire site was abated in that manner.
If the lead values in the soil column indicated different methods of abatement for
individual parcels on a multi-parcel site, then we used the most extensive method
of abatement for any parcel requiring that method of abatement. In addition, any
parcels immediately adjacent to a parcel requiring the most extensive method of
abatement were abated in that manner. For example: A site containing five
parcels numbered 1, 2, 3, 4, and 5, where parcel 3 had lead values in the column
requiring removal of the soil and parcels 1, 2, 4, and 5 required resodding of the
area as a form of abatement, then parcels 1 and 5 would be resodded and parcels
2, 3, and 4 would have the soil removed and replaced.
3.3.4 Sequence of Abatement
The second important factor in the determination of abatement neighborhoods was the
presence of lead-contaminated soil. An initial soil survey was carried to characterize the lead
concentration in the soil in the potential study areas.
Abatement was carried out in a specified sequence that was integrated with the
environmental monitoring in the areas where abatement occurred. The preabatement
environmental monitoring was completed in an area prior to the beginning of abatement.
The particular sequence of abatement was soil abatement, exterior dust abatement and finally
interior abatement.
Exterior dust abatement occurred immediately after soil abatement had been completed
in a particular area. This work was carried out by contractors possessing the specified
equipment and expertise in cleaning large paved areas. Soil project personnel coordinated
the schedule of abatement work so that areas scheduled for abatement could be posted to
prohibit parking. Streets were generally be abated one side at a time, principally because of
3-25
-------
the requirement that the street be free of parked cars for complete abatement to occur. The
potential study areas generally had more on-street than off-street parking and it was not
practical to restrict parking on both sides of a street at any given time. We therefore
required abatement on one side the first day and the alternate side the following day.
Interior abatement followed the completion of exterior abatement in an area, interior
abatement occurred after the residents were oriented by Soil Project personnel. This
preparation consisted of an explanation of what would occur and at the same time, the
resident was asked to select the color and texture of the carpeting that was to replace existing
carpeting. Soil project staff were responsible for insuring that the carpeting and furniture
was available at the time of the interior abatement. The interior abatement contractor was
responsible for vacuuming the dust from all ledges including baseboards, window sills, door
frames, etc. The contractor was also responsible for vacuuming all of the floors at a
specified rate, removing the carpet and furniture that was replaced, and installing the new
carpeting and furniture. These contractors were also monitored or inspected by the site
inspectors and the Cincinnati Health Department personnel working full-time with the
Cincinnati Soil Project.
Because exterior abatement frequently occurred on public property, it was necessary to
coordinate our activities with other agencies and utilities who commonly do work on public
areas. These agencies include the Sewer Department, Water Department, Highway
Department, Cincinnati Gas and Electric Company, Cincinnati Bell Telephone Company, and
private contractors who occasionally will do work on public property. In addition, we
wanted to be aware of any private construction projects planned or anticipated in the
abatement area. The reason for this was that we did not want to be in the middle of an
abatement procedure and have that process interrupted by private or public construction
crews who planned to excavate an area that was just abated or who would block access to a
street which we needed for abatement.
The major part of the coordination activity was performed by personnel assigned to the
Abatement Soil Project by the Cincinnati Health Department. These two individuals, a
Administrative Assistant in and a Senior Sanitarian had primary responsibility for this
coordination. This coordination was initiated prior to the onset of abatement. In addition to
planned public works/utilities work with which we had to coordinate, there were also short-
3-26
-------
term emergencies resulting from water main or gas line breaks or emergency road repairs.
These problems occur for many reasons and we had no control over their occurrence.
A procedure had to be established whereby we were notified as soon as possible, so that we
could alter our abatement procedures if necessary.
More detailed descriptions of the exterior dust and soil lead abatement methods are
presented in Appendix G.
3.3.5 Monitoring and Supervision
Contractors performed their work under the supervision of their own superintendent,
who was responsible for the day-to-day operation of the abatement work. All of the
abatement was inspected by soil project personnel responsible for such inspection, i.e., Site
Inspectors. The Site Inspectors had the responsibility of informing the contractors' field
superintendent of the time at which abatement would begin in any area. The Site Inspector
also had the authority to stop work in any given area for just cause. Abatement work could
be stopped in the event of physical hazard to property or personnel or failure to perform
work according to specifications. The Site Inspector was not responsible for the day-to-day
supervision of the contractors' employees.
3.3.6 Contract and Specification Development
Separate contracts were necessary for the three types of abatement; soil abatement,
exterior dust abatement, and interior dust abatement. The three contracts were written with
some parts in common, such as bidder notices, form of proposal, bid guarantee, contract
bond, and contract form. Since the University of Cincinnati is a state institution, the format
of some of these sections was determined by the State of Ohio. The contracts differed
mainly in the section referred to as the "specifications". Specifications for the abatement
work were determined by the various abatement methods development work completed prior
to contract development.
The development of the abatement contracts began with the initial writing of the
contracts and specifications. As with other contracts, the abatement contracts were reviewed
by personnel in the Business Office in the Department of Environmental Health, the Risk
Management Office of the Medical Center, and the Legal Advisory Services of the
3-27
-------
University of Cincinnati. Revisions were made based upon the needs expressed by each of
the departments reviewing the documents. Once the format of the bidding documents was
approved, they were sent to the Purchasing Department of the University which has an
established bidding process for work contracted by the University.
3.3.7 Bidding and Contract Letting
The selection of the abatement contractors was made through an established University
of Cincinnati policy. This policy was implemented through the Purchasing Department.
Some University policies may serve to eliminate some contractors because they may not be
able to meet established regulation. Some of these regulations that may serve to restrict the
number of contractors bidding on the abatement work are:
- Any contract over $4,000 required the payment of prevailing wages.
- Bidders were required to post a ten percent (10%) bid bond.
- The contract required the posting of a one hundred percent (100%) performance bond
along with separate material and labor bonds.
- Contractors were required to conform with Ohio State Affirmative Action regulations.
An additional requirement that was imposed by the management of the Soil/Lead Project was
a liquidated damages clause in the contract. This also served to limit the number of
interested and qualified contractors.
Contractors who could generally meet the requirements established by the University
had equal opportunity to submit the low bid and be awarded the contract. Soil Project
management established some policies which had the potential to influence the final selection
of the abatement contractors. We were able to construct the bidding documents such that we
retained the right to reject a low bidder based upon that bidder's reputation or previous
experience with the University of Cincinnati. Management was also able to influence the
selection of contractors by requiring that all contractors bidding on the abatement work
attend a pre-bid conference. This conference served to assure that the contractors totally
understood the requirements of the project. Soil Project Management evaluated all
contractors based upon previous performance on jobs performed for the University of
Cincinnati, other institutions, companies, and individuals.
3-28
-------
Because of the unavailability of acceptable low bidders, the Soil Project served as the
general contractors for the soil and exterior dust abatement in 1989 and for interior dust
abatement in 1991.
3.4 BLOOD COLLECTION AND ANALYSIS
3.4.1 Blood Collection
Blood samples were collected five times over the course of the study with each period
falling approximately six months apart. Table 3-2 shows the begin and end dates of
sampling during each phase.
TABLE 3-2. BLOOD COLLECTION SAMPLING PHASES
Phase #
01
03
05
07
09
# of Samples
Collected*
410
244
284
230
199
Date Sampling
Began
6/07/89
10/16/89
6/18/90
11/07/90
6/10/91
Date Sampling
Ended
8/10/89
12/7/89
7/24/90
12/10/90
8/08/91
Includes subjects, adults and siblings.
The clinic used for the collection of samples was within the field facility located in the
downtown area of Cincinnati near the study area. The room for sample drawing was totally
separated from any other field activities and was off limits to all those not part of the clinic
staff. The total clinic operation consisted of registration, an interview obtaining
demographic, housing and secondary residence information, signing of an informed consent
form and then finally the collection of the blood samples.
Blood samples were collected by phlebotomist trained in the proper techniques for
obtaining samples for lead analysis. They were instructed in a thorough cleaning technique
of the venipuncture site and also to avoid contamination of the sample or blood collection
3-29
-------
supplies. The Centers for Disease Control assumed the responsibility for providing the
pediatric Vacutainer™ tubes and the butterfly apparatus for sample collection. All supplies
were randomly screened by them for lead contamination.
Blood was collected for the analysis of lead, free erythrocyte protoporphyrin (FEP),
hematocrit, hemoglobin, serum iron (SFe), and total iron binding capacity (TE8C). In the
event a venous sample could not be obtained a fingerstick sample was collected. :Of the
1,367 blood samples obtained, 65 (5%) were fingerstick collections. For venous samples,
4 ml of blood was drawn into a 6 ml disposable syringe attached to a 23 gauge butterfly
apparatus. Approximately 2 ml of blood was immediately dispensed into a potassium
ethylene diamine tetracetic acid (K3EDTA)-containing pediatric Vacutainer™ tube to be used
for PbB, EBP, hematocrit and hemoglobin. The remainder of the blood was placed into a
second tube that contained no anticoagulant or additives. This was used for SFe and TIBC.
If a fingerstick sample was required the capillary blood obtained was used for PbB and
hematocrit only.
After collection, all K3EDTA tubes were thoroughly mixed on a Nutator™ and both
tubes refrigerated. Clinic personnel made deliveries of the samples to the lab within 4 h of
drawing and twice a day if needed. Once received by the lab staff, aliquots were taken for
PbB analysis, hematocrits and hemoglobins performed, and the remainder of the whole blood
sample refrigerated until analyzed for FEP. The clotted sample was centrifuged, the serum
extracted and frozen for future SFe and TTBC analysis. A more complete description of
sample collection and handling can be found in the Midterm Project Update.
3.4.2 Analytical Procedures
Blood for lead analysis was determined by Anodic Stripping Voltammetry (ASV) on the
ESA Model 3010 A Trace Metal Analyzer according to the method of Roda, et al., Clinical
Chemistry, Vol. 34, 563, 1988. The method involved aliquoting 100 jtl of well mixed,
whole blood into a metal decomplexing agent (Metexchange Reagent). This solution was
placed on a mercury coated graphite electrode whereby using controlled potentials, metals
were plated, and stripped. The current generated was picked up by a recorder. The peak
heights which are proportional to concentration were measured and the concentration of Pb
calculated from standard curves. All samples were analyzed in duplicate.
3-30
-------
The analysis of FEP was performed by the method of Chisolm and Brown (1975). It is
a double extraction method using ethyl acetate/acetic acid and hydrochloric acid. The
fluorescence intensity of a sample was measured on a fluorometer and the concentration of
the sample determined from a protoporphyrin standard curve. A 20 pi aliquot of whole
blood was used for the analysis.
Hematocrit or packed cell volume is the total volume of the red cell mass expressed as
a percentage of the whole blood volume. It is determined by centrifugation. The
Cyanmethemoglobin method is a spectrophotometric technique used for the measurement of
hemoglobin in a whole blood sample. The method was calibrated against a certified standard
and the absorbance of 20 fd of whole blood in a reagent measured at 540 nm.
Serum iron and TEBC were measured by a technique called controlled potential
coulometry on an ESA Ferrochem n Analyzer. The current resulting from reactions at the
surface of the test electrode in the instrument is proportional to the concentration of iron (Fe)
in the sample. The instrument automatically calculates and displays the unknown Fe and
TIBC concentration in jtg/dL. For Fe, 25 /ttl of serum was used while 100 /til was required
forTIBC.
Additional information and detailed description of the analytical procedures can also be
found in Appendix .
3.4.3 Quality Assurance/Quality Control
Validation of the quality of PbB and FEP measurements was accessed in part by
participation in several proficiency programs. One program sponsored by the Centers for
Disease Control provides participants with three unknown bloods per month per analyte.
Results were compared to the consensus values established by reference and participating
laboratories. Another proficiency program that we participate in conies through the College
of American Pathologists. Blood samples were sent by this group for blood lead
determination on a quarterly basis.
For the Urban Soil Lead Abatement Demonstration Project, additional procedures were
used to validate the quality of the PbB results. Field duplicates in the form of human blood
samples drawn from the same individual were split, given two different names and submitted
to the lab for the complete series of analyses. They were collected at a rate of at least 2
3-31
-------
individuals and 4 samples per week. In addition, 4 blood samples of different concentration
were supplied by the Centers for Disease Control to test for any analytical or procedural
variabilities within or among the three Soil Project cities. These samples were repeatedly
analyzed by each city prior to the beginning of the project. During the course of the study at
least two of the four were analyzed in duplicate with each set of unknowns. Furthermore,
one of the two was analyzed with its identity unknown.
Internally, the laboratory employed several additional techniques to monitor and control
the accuracy and precision of its PbB determinations. Human blood samples whose lead
values have been previously determined by isotope dilution mass-spectroscopy (IDMS), the
definitive method for lead determination, were incorporated into the analysis of the
unknowns. In these instances, IDMS bloods were analyzed by giving them a fictitious name
so as to be unknown to the analyst when concentration was determined. Duplicate samples
of two different IDMS determined reference bloods were also analyzed with each set of
unknowns. The two levels used during the Soil Project had lead concentration of 5.3 jig/dL
i
and 40.4 j^g/dL. Control charts documenting performance over extended intervals were
maintained for these samples. It should be noted the calibration of the instrument also
consisted of using whole blood samples whose lead concentrations were previously
determined by IDMS i.e., the instrument was calibrated with primary standards. As
mentioned, precision of FEP analysis was assessed by participation in the CDC proficiency
program. Assurance that the same instrument sensitivity was maintained was demonstrated
by periodically running a new standard curve. Another check was made by including in each
weekly run, blood samples of known porphyrin concentration.
Hematology controls supplied by Fisher Scientific Company were assayed at 6 mo
intervals to monitor the accuracy and precision of hematocrit and hemoglobin concentrations.
Serum Fe and TBBC determinations were evaluated by checking calibration against a
commercial control. In addition, to check reproducibility and instrument drift, repeated
analysis of aliquots drawn from the same human blood sample pool were analyzed over time.
Results for the QA/QC techniques discussed for all sample analyses can be found in
Section 4.4.
3-32
-------
3.5 ENVIRONMENTAL SAMPLES
3.5.1 Environmental Sample Collection
Following are brief descriptions of environmental sample collection and preparation
procedures. Detailed descriptions of these methods are contained in appendices to this
report. A training manual for paint and water sampling is presented in Appendix .
3.5.1.1 SoU Sample Collection
A. Soil Sample Collection Protocol
". 1. Site identification
Aerial photographs with ground verification were used to delineate the potential
abatement areas and the soil sites within the abatement areas.
2. Sampling Patterns and Sample Types
The system of book, page, and parcel used by the Hamilton County Auditor
provided a working system for uniquely identifying a unit of land. In addition,
because parcels of urban land are generally of regular and consistent size and
shape, and large areas of soil are made up of adjacent parcels, the parcel served
to delineate the boundaries of a soil sampling area.
The four types of soil sampling patterns that have been established and the
situations where they are used are as follows:
Pattern Where Used
Line (source)* Parcels adjacent to buildings or other painted structures
Line (area)* Parcels removed from sources such as buildings
Targeted** Bare areas or play areas within parcels
Small Areas Soil areas with dimensions less than about 10 ft.
*Line segment length not to exceed 40 ft.
**Targeted areas and line segments may lie in the same parcel.
Two types of soil samples were collected:
Surface Scraping: A composite of from 5 to 10 subsamples collected along
a line segment, or in a small area, or in a targeted area;
scraping depth = 0.5-1.0 cm.
Core Samples: A composite of from 5 to 10 subsamples collected along
a line segment, or from a small area; top depth =
0-2 cm; bottom depth = 13-15 cm.
3-33
-------
SOT! in a parcel which was adjacent to a potential source of higher lead
contamination, such as a building, received more intense sampling (line [source]
pattern) than soil which was relatively isolated from potential additional sources
of lead. The source pattern was believed to be the most appropriate to use in an
urban area, where most soil was near building foundations or streets, or was
previously built upon.
Sample line segment spacings are summarized as follows:
Line (source) Spacing: Line 1.1 = 0.5 m from boundary- The most typical
boundary is a building or a sidewalk. Line 2.1 =
10 ft. from boundary. Line 3.1 = 15-20 ft, from
boundary, depending on the width of the parcel.
NOTE: if a source exists on both sides of a parcel, the
numbering of line 1.1 is arbitrary.
Line(area) Spacing: Single line 1.1 along parcel centerline; or lines spaced
about 20 ft. apart in the case of an unusually wide
parcel (more than 30 ft.)
Soil areas which, were too small to support a line pattern were randomly sampled,
following the guidelines for the small area pattern.
Targeted samples were also taken in appropriate areas such as bare areas or near
play equipment. Since the targeted samples were intended to reflect potential
higher exposure of children to lead in soil, only composited surface scrapings of
the soil were taken.
5. Soil Sample Preparation
Samples were air dried for a period of at least five days.
Each sample was sieved twice; once with a number 10 sieve with a mesh size of
2 millimeters, followed by a number 60 sieve with a mesh size of 250 microns.
Soil passing through the number 10 sieve was called the "total sample". The soil
passing through the number 60 sieve was called the "urban sample". (Most of
the analyses are performed on the "urban" sample.)
3.5.1.2 Exterior Dust Collection
A. Exterior Dust Sample Collection Protocol
1. Collection Apparatus
The exterior dust samples were collected on paved areas with a battery-operated
portable vacuum of the type used to vacuum automobiles.
3-34
-------
A minimum of three sub-samples of dust from 0.5 ft. x 2 ft. areas made up a
single composite exterior dust sample.
Three exterior dust sampling strategies were used.
a. Neighborhood-wide sampling of abatement area
Composite samples were taken along designated blocks which did not exceed
500 ft. A tempkte area was sampled no less than every 100 ft. along this
length. For smaller areas, no less than four template areas made up the
composite.
Two areas were sampled which were defined as follows:
(1) Street gutter: the interface of the street surface and the curb.
(2) Sidewalk: the edge of the sidewalk farthest from the street; or alternately, when a
building abutts the sidewalk, the interface of the sidewalk surface and the building.
Since streets and sidewalks were not assigned parcel numbers on the Hamilton
County Auditor's plat maps, an extension of the book, page and parcel system
were used to assign a numerical identifier to the streets, sidewalks, and alleys to
be sampled.
Alleys were also sampled. For sampling and "parcel" designation purposes, the
alley were treated as if it was one side of the street; one composite sample from
the street gutter on both sides of the alley were collected, as well as a single
composite sample from each sidewalk or similar area if it exists along each side
of the alley.
In addition to sampling within the abatement area, the same type of sampling was
done in an approximately 200 ft. "buffer zone" which surrounded the abatement
area. [This "buffer zone" also received exterior dust abatement of streets,
sidewalks, and alleys and was therefore sampled to determine abatement
effectiveness and recontamination rates.]
b. Partially or completely paved parcel sampling.
Exterior surface dust was collected from paved areas within soil parcels such as
paved walkways. Exterior dust was also collected from completely paved parcels
such as parking lots. A composite sample of at least four subsamples were
collected from an area consisting of 4,000 to 8,000 square feet.
3-35
-------
c. Samples targeted to subject residences
Exterior dust samples were collected around the building where study subjects
reside. Two samples were collected around each building. One sample was a
composite of no less than two template areas from the sidewalk area adjacent to
the main entry of the building. The second sample was a composite of areas
from the other three sides, if paved.
Exterior dust sample preparation was similar to that for soil.
If a sample was larger than 200 grams, it was split to reduce the size to
approximately 100-200 grams. The following guidelines were followed in splitting
samples.
1. Samples weighing between 200-400 grams were split one time.
2. Samples between 400-800 grams were split two times.
3. Samples between 800-1600 grams were split three times.
3.5.1.3 Interior Dust Collection
A. Surface Dust Collection Protocol !
1. Collection Apparatus for Vacuum Method
The apparatus used to collect surface dust was a personal air monitoring pump,
an air monitoring cassette containing a 37 mm diameter 0.8 micron polycellulose
acetate filter, and a collection attachment.
Interior dust was sampled from areas such as the floor adjacent to the entry; from
carpeted or bare floors; from window sills and window wells.
A sample of dust was collected over a measured area, or composite of measured
areas so that three measures were obtained:
2
Dust Loading = mg dust/m
Lead Loading = /tg lead/m
Lead Concentration = fig lead/g dust or ppm lead.
2. Sample Areas
Entry (E): A floor area inside the residence directly adjacent to the main entry
to the residence.
Floor (F): A composite sample of at least 3 floor areas which includes but is not
limited to a sample from a high-traffic area in the main living area and a sample
from the child's bedroom.
3-36
-------
If carpet was present in the residence it was the first choice of sample area. If
carpet was not present, a mixture of non-carpet floor areas were sampled.
Window (W): A composite sample of at least three window areas (window sills
and window wells), including but not limited to a window in the main living area
and a window in the child's bedroom. The window sill was the preferred area
for sampling; window wells were sampled if an adequate amount of dust was not
available on the window sills.
Mat (M): The floor mat was sampled when it is put in place in the interior entry
area in the home at the post blood draw environmental visit and was sampled
again after the year's abatement activities are completed. At this time it was
replaced and sampled at each subsequent environmental visit occurring over the
next year.
B. Dustfall Collection Protocol
Dustfall samples were collected in polypropylene containers which had snap on lids.
The containers had the dimensions 10 1/8 in. x 9 3/4 in. X 2 1/2 in. deep. They
were "Tupperware"-type containers.
The dustfall containers were placed in the residence at the time of preabatement
visits in 1989 and 1990. In practice, it was best to place the container above floor
level in a relatively inconspicuous spot so that no one interfered with it. The first
choice for placing the container was placed on top of the refrigerator in the kitchen.
If this was objected to by the family, a nearby location was found.
3.5.1.4 Handwipe Sample Collection
A. Sample Collection Protocol for Handwipes
Collection of the hand-lead samples were done at the conclusion of each visit to a
residence.
The person collecting the hand-lead samples wore disposable gloves. The person
collecting the hand-lead samples cleaned his/her own hands with a disposable wipe
from a separate container of wipes kept for this purpose before touching the gloves
or other equipment. Once the gloves had been put on they were also cleaned well by
using additional clean wipes. For each residence, where one or more child's hands
was sampled, a field blank was taken. This was done in the following manner. Six
wipes were removed from the container, handled to simulate wiping a child's hands
and then placed in a single bag and submitted for analysis.
Lead in dust on children's hands was sampled by wiping each hand of the child with
three separate commercial Wet-Wipes. All surfaces of the hand, front and back, up
to the wrists, were wiped thoroughly with each of the three wipes. The wipes from
3-37
-------
each child were composited in a single scalable bag for transport to the laboratory.
The total quantity of lead was reported in /ng lead/pair of hands.
3.5.1.5 Water Sample Collection
Two 125 ml water samples were collected from each family participating in the
Cincinnati Soil Project. Those were the W-l sample which was a 30-min stagnation sample
and W-2 which was an overnight stagnation sample. Stagnation samples were collected hi
order to provide some uniformity of samples and to determine the amount of lead which was
dissolved in the water over a fixed time period. Collection of the water samples occurred
during Phases 4 and 8 in the springs of 1990 and 1991, respectively.
The 30-min stagnation sample was collected by the Soil Project staff of environmental
monitors during visits to the residents of participating families. Lead-in-paint monitoring
was also completed during this visit. At the time of the visit directions and a collection
vessel for the overnight stagnation sample, W-2, were supplied to a responsible adult. The
sample was picked up the following morning from the residence by environmental monitors.
3.5.1.6 Paint Collection by Portable X-ray Fluorescence
The concentration of lead in paint was determined by using an X-ray fluorescence
analyzer. Two types of instruments were used, the XK-2 and the XK-3, both manufactured
by Princeton Gamma-Tech, Inc. The XK-3 with a range of 0-10 mg of Pb per cm was the
primary instrument used. The XK-2 was a backup and also used in the event a reading on
the XK-3 exceeded 10 mg/sq cm2.
In each residence two surfaces, a painted woodwork and a painted wall in each of three
rooms or areas most frequently occupied by the subject child were evaluated (e.g. child's
bedroom, kitchen, living room). One reading was taken at three different locations on each
type of surface. >
A wall and/or trim immediately exterior to the dwelling unit entry was also sampled.
The sample paint sites on the exterior of a building was only examined with the owner's
permission.
3-38
-------
3.5.2 Environmental Sample Analysis
3.5.2.1 SoU
Soil Analysis
Soil samples were analyzed by XRF on a Kevex Delta Analyst Energy Dispersive X-ray
Spectrometer. Prior to analysis samples were air dried on a plastic plate to constant weight.
Each was then sieved by first passing the sample through a 2 mm sieve (Total Soil Fraction)
and after removing a portion passing the remainder through a 250 /*m sieve. This sieve
sample was termed the Urban Soil Fraction and is that which was analyzed by XRF.
For determination of lead, two grams of the sieved soil (Urban Soil Fraction) were
placed into labeled XRF sample cups. Operating conditions for the instrument were
originally set at 30 KV, 0.5 mA, 100 lifetime seconds acquisition time, and using a Mo
secondary target. The analytical conditions removed escape peaks but not background and
used the intensity ratio of the Pb (La) peak and Mo (Ka) to quantify the lead against a
quadratic calibration curve. After careful examination of the method it was determined that
to increase accuracy and precision it was necessary to operate at the following conditions:
30 KV, 0.5 mA, 200 lifetime seconds and the Mo target. Analytical conditions removed
escape peaks and background and now would use the L0 Pb peak instead of La. This was
ratioed against Mo Compton-Raleigh peak. Calibration no longer used one quadratic
formula, but through statistical evaluation of significance of standard curves, 2 different
linear equations with a break point of 4,500 ppm were used. Also, opposite to what was
done previously, concentration became the independent variable and intensity ratio the
dependent variable in establishing the concentration of Pb from the standard curve formulas.
The change in methods occurred after Phase 00 starting with sample #4608. The lead
content of all samples analyzed prior to that were recalculated using data accumulated from
reanalysis of a portion of samples and controls. (Additional information for collection,
sieving and analysis can be found in the midterm project update).
Soil Quality Assurance/Quality Control
Field lab blanks and field duplicates were collected and used to evaluate the quality and
precision of all aspects of soil sample handling and analysis. Field lab blanks consisted of
3-39
-------
material obtained from a sand/gravel quarry in Cincinnati. There were three types of native
soils provided:
1. Lake beds clays possibly Wisconsin glacial stage, 20-25,000 years old.
2. minion Till, 125,000 years old.
3. Pre-glacial Heuvial (river) sand, 150,000 years old.
Portions of these samples were bagged, so as to resemble samples collected in the field,
taken out into the field, numbered to occur randomly among the unknown samples and
returned to the field lab to be sieved with the unknowns. Analysis also occurred [randomly
with the rest of the samples.
A field duplicate was obtained within 6 mo of the original sample. These were
collected for one out of every 10 unknowns. Because lots or yards around residences are not
homogenous and man-made changes may occur over time, the field duplicates were actually
only a rough estimate of precision and reliability of sample collection.
Four different concentration soil samples were prepared by the U.S. Environmental
Protection Agency (EPA)-Environmental Monitoring Systems Laboratory (EMSL) of Las
Vegas, Nevada. These samples were numbered and inserted as unknowns into the Soil
Project sample stream. Thus, at the time of analysis their identity was completely unknown
to the XRF lab technician. The lead concentration of these samples was established through
a round robin comparison by the 3 cities involved in the Soil Projects.
Laboratory QC consisted of weekly instrument energy calibrations and the analysis of a
lo and hi control to validate each run of 14 samples. Results for the soil QA/QC can be
found in Section 4.1.2.
3.5.2.2 Exterior Dust
Exterior Dust Analysis
Exterior dust samples were handled, sieved and analyzed the same as the soil samples.
The XRF method change (discussed above) occurred starting with exterior dust sample
#2497. Detailed descriptions for collection and analysis were present in the midterm project
update.
3-40
-------
Exterior Dust Quality Assurance/Quality Control
The types of QA/QC that were applied to soil analyses were also applied to the exterior
dust samples. In fact, the field blanks and EPA-EMSL Las Vegas QC samples were the
same materials as used with the soil sample analyses. The lo and hi controls were also soil
and also the same as used for soil analysis. See Section 4.1.3 for QA/QC results.
3.5.2.3 Interior Dust
Interior Dust Analysis
The interior dust collection method which yields small quantities of dust precluded
analysis of the samples by XRF. Consequently, all interior dust samples were digested and
analyzed by flame atomic absorption spectroscopy (AAS). The process began with the
removal of the dust from the plastic air sampling cassette. This-was accomplished by rinsing
the entire inside of the cassette with distilled/deionized water into a pre-weighed, acid soaked
beaker. The sample was oven-dried and the weight of the sample determined. Acid
digestion occurs in 7 M nitric acid (HNO3), the sample was filtered and after heating brought
up to volume in.l M HNO3. Flame AAS analysis of samples was performed on a PerMn-
Elmer Model 2380 instrument. Appendices I and n give a more complete description of the
procedures.
Interior Dust Quality Assurance/Quality Control
Overall monitoring of interior dust procedures was evaluated by using dust samples
supplied by EMSL. These samples, prepared by EMSL, consisted of a known amount of
dust added to an empty sample cassette. In order to disguise these samples, 28 fictitious
family, subject and residence identities were used. (These were the same identities as given
the blood field duplicates.) The assignment of samples to each identity was prepared at the
field office and then sent to the laboratory as having been collected from a real family
residence.
One in every 25 interior dust samples was collected in duplicate. These were collected
adjacent to the spot where the actual study sample were taken. Field blanks were also
collected for interior dust. This involved attaching an empty sampling cassette to the pump,
setting the pump on a table with the nozzle pointed up, and running the pump, collecting
3-41
-------
1
only air for 3 min (average amount of time required for the collection of an interior dust
sample).
Quality control was a part of each phase of sample handling in the laboratory. For
each set of samples a method blank was inserted at the beginning of the preparation step and
a reagent blank both before preparation and digestion. A National Bureau of Standards
(NBS) standard was incorporated into each days preparation of samples as well as duplicates
of a reference dust control. Another NBS standard was inserted prior to digestion.
i
Calculation of percent recoveries and duplicate analyses were also a part of the AAS
procedure.
Control charts and limits were kept to evaluate weight differences in method blanks as
well as changes over time in standard concentrations. Values for blanks outside the limit of
the determined balance error required corrections to be made to all other weights. Samples
with weights less than 2 mg were considered insufficient for analysis.
If more than three of the five QC samples were outside the established limits the entire
set was considered invalid. For the Soil Project 3 sets of 24 unknowns were not acceptable
according to the QC criteria. Results for interior dust QA/QC are in Section 4.1.4.
3.5.2.4 Interior Dustfall
Interior Dustfall Analysis
The analysis of the dustfaH samples was the same as previously described for interior
dust. The preparation of these samples, however, was different. The dust itself was
collected in plastic tupper-ware like containers with lids. The first step after opening the lid
was to remove any obvious foreign objects such as insects, leaves, pins, etc. The sample
was then quantitatively transferred into a 250 ml pre-weighed beaker, dried and weighed.
Digestion occurred using 7 M HNO3 and analysis was performed by Atomic Absorption
flame Spectroscopy.
Interior Dustfall Quality Assurance/Quality Control
The only type of field QA/QC sample that can be acquired for dustfall is a field
duplicate. In our study 1 in every 25 dustfall containers placed in a home had a second one
3-42
-------
place beside it. There were no EMSL - Las Vegas controls supplied for insertion into the
dustfall sample stream.
As was the case for the interior dust analysis, lab method blanks, reagent blanks, lab
duplicates and controls were a part of the laboratory's standard operating procedures for
dustfall. In fact, after the original extraction of the sample from the container was weighed
the method proceeded exactly the same as that for interior dust. Therefore, all the QC was
combined and evaluation of the analysis consistent for both the interior dust and dustfall
samples. There were no dustfall daily sets that were rejected under the QC criteria.
3.5.2.5 Hand Dust
Hand Dust Analysis
Hand dust samples were comprised of a total of six wipes — collected from each hand of
the child. They were placed in a plastic zip-lock bag and submitted to the laboratory with
the other interior samples from the residence. There was also a field blank associated with
each residence taken at the time the child samples were obtained. This sample was used to
assess whether the environment or collection procedure contributed any contamination to the
study samples.
For the accuracy of the handwipe Pb concentration it was important that the supplier lot
number of the wipe material or "Wet-Wipes" be documented. From experience we have
found that the Pb concentration may vary from lot to lot by as much as 2.5 fig of Pb per
individual wipe. Not only was it important that the child's sample and field blank be taken
from the same lot but also the laboratory blank and controls. Thus, wipes were purchased in
case lots only, individual bottles were labeled with the lot number, bottles were supplied to
the lab, and records kept so that field and lab material would correspond.
Samples were digested using 7 M HNO3 heated on a hot plate at 120°C for 2 h. They
were filtered, brought up to volume with 1 M HNO3 and analyzed by flame AAS. Refer to
the Appendices for more detail.
Hand Dust Quality Assurance/Quality Control
Procedures instituted for the handling of the wipe material were very important to the
accuracy of hand dust Pb concentration. It was critical that field monitors were aware of the
3-43
-------
significance of lot number in the evaluation of the field blank and in the determination of the
analytical QC to be used for the sample. Field blanks were obtained prior to the collection
of the residence's handwipe samples and consisted of removing six wipes from the container,
handling them to simulate wiping a child's hands and placing them in a sample bag for
submittance to the laboratory. The concentration of these samples was used to evaluate the
potential for contamination in the unknowns. Field duplicates for hand Pb could not be
obtained since both hands of the child are wiped and constitute the sample.
The EMSL - Las Vegas laboratory supplied handwipe QC samples to monitor overall
quality of the handwipe data. They had placed six wipes in a sample bag and spiked them
with a known Pb standard solution. The supplies required for these samples were provided
by us and again the importance that lot numbers be considered in the preparation of the
samples was communicated. In order to disguise the samples, the 28 fictitious families
instituted for QC purposes were used.
Within the laboratory, lab method blanks consisted of six unused wipes from the same
lot number of material as the samples, digested and analyzed in the same fashion as the
unknowns. Two lab method blanks and 1 reagent blank were analyzed with every
20 unknown samples. The mean of the method lab blanks produced an average value for
that day's analysis of unknown samples. That value was subtracted from each concentration
to produce a corrected j*g Pb;hands. Lab controls were prepared by pipetting standard lead
solutions (4, 20, 40 and 100 /*g Pb/mL) onto six clean wipes in a beaker. These are also
from the same lot number of material as the unknowns and digested and analyzed
comparably. To validate their preparation each new set of solutions prepared were analyzed.
AAS duplicates and percent recoveries were also determined on final extracts. Quality
assurance/quality control results are in Section 4.1.6.
3.5.2.6 Water
Water Analysis
The water samples collected in the Cincinnati Soil Project were not analyzed by us as
originally proposed but by Environmental Health Testing, Inc. (FJHRT) of Cincinnati. The
samples presented to them were aliquots of the original 125 ml acidified water. The
analytical method employed by EHRT was a direct analysis of the sample using Mg(NC^)2
3-44
-------
and NH4HP04 as matrix modifiers and on a Perkin-Elmer Graphite Furnace Atomic
Absorption Spectrometer (GF-AAS) with Zeeman background correction. The instrument is
described as being comprised of a microcomputer-controlled spectrometer, a graphite furnace
with Zeeman magnet, a microcomputer-controlled power supply for the graphite furnace and
a printer for automatic reporting of results. The injection site consisted of a paralytic coated
graphite tube with L'vov platform. Only 1 ml of sample is required for preparation and
25 /il aliquots of the prepared sample injected into the furnace. Calibration is from
0-35 jtg Pb/L and sample results are given in /tg/L. The method detection limit is 1 jig/L.
Water Quality Assurance/Quality Control
Field blanks and field duplicates were parts of the collection procedure for water
sampling. Field blanks involved taking a 125 ml bottle of distilled/deionized H20 into the
home and allowing it to sit opened on the sink during the time the comparable study sample
was being collected. These were collected at 5% of the residences. Duplicates were
collected at 10% of the residences and done immediately following the initial visit and
sample collection at the residence.
In the laboratory distilled/deionized water samples and controls were incorporated into
the samples sent to FJHRT. It was specified to that lab that samples of a control should be
analyzed with each set of samples. In addition, all samples were analyzed in duplicate and
percent recovery determined on 100% of the samples submitted. Results are given in
Section 4.1.7.
3.5.2.7 Paint
Paint Analysis
The concentration of lead in paint was determined using two types of X-ray
fluorescence analyzers both manufactured by Princeton Gamma-Tech. The XK-3 instrument
has a range of 0-10 mg/cm2 and was the primary instrument used. The XK-2 model was
used as a backup and because of its extended range (0-75 mg/2) in the event the reading
exceeded 10 mg/cm .
The measurement of the lead concentration in the paint is fully automatic. The
instrument is placed on a designated surface and the handle depressed. Usually within
3-45
-------
25 seconds the lead content of the paint will appear on the digital display of the instrument.
However, prior to taking any concentration reading the instrument must be calibrated (XK-2)
or the calibration must be checked (XK-3). This is accomplished by using various levels of
lead film supplied by National Institute for Standards and Technology (NIST).
A very detailed description of the operation and calibration of the instruments can be
found in the appendices.
Paint Quality Assurance/Quality Control
Proper calibration and instrument operation were important factors contributing to the
quality of the XEF concentration readings. The instruments were calibrated and checked
using NIST lead film. These lead paint reference materials were developed by NIST for the
Department of Housing and Urban Development. At each residence single paint calibration
checks were made at the beginning and end of all measurements. For calibrating the XK-2,
*y i
readings were taken with the zero lead standard, the 1.5 mg/cm standard, and the
f\
2.99 mg Pb/sq cm paint standard. The XK-3 was checked by using the 0.6 mg/cm lead and
the 2.99 mg/cm2 standards.
To avoid errors in registering an incorrect value, the operator read the number for the
other team member to record. This was then read back to the operator. Also to reduce the
precision error of a residence room one reading was taken at three different locations on each
of the two surface types and in each of the three rooms tested.
In addition to the XRF readings, a semi-quantitative evaluation of the condition of the
painted surface was made. Using a scale of 1-3 the surface tested was rated as intact, tight
paint or deteriorated. The paint condition and lead content was combined to give a weighted
paint hazard score that has been shown to correlate with blood lead in children.
3.6 HEALTH AND SAFETY PLAN
A Safety and Health Plan (SHP) was written and implemented for the Cincinnati Soil
Lead Abatement Demonstration Project in November 1988. The purpose of the plan was to
have a written set of work procedures and safety standards to protect workers' health and
safety. The plan was written by the projects Safety and Health Officer (SHO) and reviewed
3-46
-------
by a team of safety and health experts. Revisions and updates to the plan were made
annually or as necessary.
Contents of the Safety and Health Plan included:
Project Description,
Site Descriptions,
Organization and Coordination,
Site Characterization,
Training,
Employee Protection,
Medical Surveillance Program,
Site Control,
Decontamination,
Emergency Response, and
Standard Operating Procedures.
A copy of the Safety and Health Plan is attached as Appendix L. A publication describing
the development of Safety and Health Plan and its initial implementation appears in
Appendix M.
Monitoring and controlling worker exposures to hazardous elements, including lead
dust, at the worksite were priorities of the Safety and Health Officer. Worker training,
medical surveillance, field and laboratory audits and personal air and sound level monitoring
were methods used to limit or prevent worker exposure to environmental hazards.
3.6.1 Training
All personnel potentially exposed to hazardous materials received basic safety and
health training before working with the project. All training sessions for new project
employees or contractors consisted of a minimum of the following topics.
1. Rights and Responsibilities
a. Safety and Health Plan
b. Occupational Safety and Health Administration (OSHA) lead exposure levels
c. Medical Surveillance Program
2. Hazard Recognition and Control
a. Site Characterization
b. Health Effects
c. Radiation
d. Hazard Communication
3-47
-------
3. Work Practices
a. Standard Operating Procedures
b. Fire Prevention
c. Equipment Use
d. Acquired Immune Deficiency Syndrome (AIDS) Awareness
4. Personal Protective Equipment
a. Protective Clothing
b. Respiratory Protection
c. Decontamination
5. Emergency Response
a. "Buddy System"
b. Contacting Emergency Resources
c. First Aid
Initial and annual training was conducted by the project's Safety and Health Officer
(SHO) and at times by other safely professionals. Training was in accordance with 29 Code
of Federal Regulations (CER) 1910.1200 and recommendation for personnel exposed to
blood borne pathogens.
Training sessions were usually between two and 6 h in length and were modified for
specific job tasks; An example of an agenda for contractor training is attached as Appendix.
3.6.2 Medical Surveillance Program
The MSP was designed to assess and monitor workers' health and fitness before
employment and during the course of work, provide mechanisms for emergency and other
treatment as needed and detail methods for accurate record-keeping for future reference.
Medical surveillance was provided for all personnel who were exposed to hazardous
substances or health hazards. All expenses related to medical surveillance were borne by the
project. Medical surveillance included the following:
Pre-employment screening to assess the workers' health and fitness for his/her job was
provided to designated employees. The employer provides the examining physician with a
description of his/her job duties. The SHO provided a copy of 29 CFR 1910.120, 29 CFR
1910.1025, 29 CFR 1910.95 with appendices and other applicable OSHA standards or
guidelines, occupational health clinic for use by the examining physician.
3-48
-------
a. Occupational and medical history, particularly with regard to cardiovascular or
respiratory disease; exposure to lead or blood-borne pathogens; and adverse
outcomes especially neuromuscular, reproductive and gastrointestinal dysfunction;
musculoskeletal disorders or impairments; and atopic diseases.
b. Comprehensive physical examination of all body organs (excluding pelvic and rectal
systems)focussing on hepatic, pulmonary cardiovascular and musculoskeletal
systems.
c. Baseline determinations by urinalysis, with differential, and Chem-23.
d. For personnel potentially exposed to contaminated dust or soil, baseline testing of
blood lead level and zinc protoporphyrin (ZPP) (or FEP).
e. For personnel whose exposures to noise equaled or exceeded an 8-h time-weighted
average of 85 decibels.
f. For personnel who wore respirators, a determination of fitness to wear a respirator
was made. (See Respiratory Program, Appendix L).
g. For personnel involved with environmental monitoring, physical capacity screening
evaluations were conducted to determine fitness for sample coring, if recommended
by the examining physician.
h. For personnel handling blood products, determination of the baseline hepatic profile
and hepatitis B antibody status is recommended; Hepatitis B vaccinations were made
available, as appropriate.
i. Environmental personnel were vaccinated against tetanus, as appropriate.
Periodic medical examinations were provided to monitor workers' health and fitness
during the course of work. These were compared to baseline data to determine biologic
trends that may mark early signs of adverse health effects, and thereby facilitate appropriate
protective measures. Included are:
a. Annual medical examinations.
b. Biologic monitoring for personnel exposed to potentially contaminated dust.
i) Blood lead every 6 mo.
ii) ZPP (or FEP) every 6 mo.
3-49
-------
c. Audiometric monitoring for personnel whose exposures equaled or exceeded time-
weighted, average of 85 decibels.
i) audiometric exam annually.
Medical examinations were provided as soon as possible upon notification that the
worker had developed signs or symptoms indicating possible over-exposure to hazardous
substances or health hazards or that an unprotected employee had been exposed in an
emergency situation. Treatment was provided at the discretion of the examining physician.
If the cause was determined to be occupational, the SHO was notified.
Any employee who experienced a needle stick, cut, human bite or mucous membrane
exposure which created the risk of exposure to blood borne pathogens was subject to the
procedures contained in a Needlestick Policy.
For personnel exposed to potentially lead-contaminated soil or dust, medical
examinations were provided at termination of employment or job reassignment. Blood lead
levels were assessed at the same time unless the employee had a sample(s) checked within
the previous thirty days.
Medical examinations were performed by occupational medical physicians of the
University of Cincinnati Occupational Health Clinic. Laboratory analysis of blood for lead
was performed by accredited labs under contract with the clinic or the University of
Cincinnati laboratory affiliated with this project. All other biological samples were analyzed
by an accredited laboratory selected by the physician. All employee health and medical
records were maintained in the Occupational Health Clinic in accordance with the
requirements of 29 CER 1910.120(f)(7). Employees received the physician's written opinion
in accordance with 29 CFR1910.120(f)(6).
Emergency and Non-Emergency Treatment
Provisions for emergency treatment and acute non-emergency treatment \yere made for
each site.
1. Emergency first aid:
a. Team leaders were trained in first aid by the American Red Cross.
b. A standard first-aid kit was available in laboratories and cars used to transport
worker to field sites.
3-50
-------
c. Eye-wash units were available in all laboratories and checked at least weekly by
a designated lab tech and quarterly by the SHO or representative for adequacy
of operation.
d. Team leaders notified "911" or transported an injured or affected employee to
the University Hospital if the injury required more than first-aid.
2. Record-keeping
Proper record-keeping was essential to protect worker health and safety.
a. Medical records will be maintained by the Occupational Health Clinic and preserved
on exposed workers for thirty (30) years after termination of employment (29 CFR
1910.120).
b. Reports by the Occupational Health Clinic were maintained for each employee by
the SHO. A copy of each report was sent to the employee upon receipt from the
clinic.
c. Records of occupational injuries and illnesses were maintained and posted yearly by
the SHO.
The SHO at least annually ascertained that each accident or illness was promptly
investigated, evaluated the efficiency of the medical surveillance program, reviewed potential
exposures and site safety plans and reviewed emergency treatment procedures.
3.6.3 Workplace Audits
Assessments of potential hazards in the work sites were conducted before work
commenced hi accordance with 29 CFR 1910.120(c). This assessment included a
determination regarding the applicability of 29 CFR 1910.1926 (Construction Standard),
29 CFR 1910 (General Industry Standard), and other OSHA, National Institute for
Occupational Safety and Health, or American Conference of Governmental Industrial
Hygienists guidelines, as appropriate. It was anticipated that employees would not be
exposed to airborne lead that exceed the action level defined in 29 CFR 1910.1025.
3-51
-------
3.6.3.1 Laboratory Audits
Laboratory facilities were initially reviewed by the SHO or designate to determine
potential exposures to lead or other hazards. Appropriate controls and the use of personal
protective equipment were instituted as necessary.
On-site reviews of laboratories were conducted on a regular basis. Copies of the
SHO's report were given to the laboratory supervisor, project managers, and administrative
manager.
3.6.3.2 Interior and Exterior Field Audits
The SHO or designate monitored field workers on a regular basis to assure compliance
with guidelines as stated in the SHP. When hazards existed, the SHO or designate alerted
the employee to the problem, addressed the issue with the field supervisor and recorded the
findings in their report. Follow-up audits were conducted to insure the problem or safety
hazard was corrected. Follow-up audits were performed within a two week period when
recommendations by the SHO or designate required any safety improvements. Response to
safety recommendations were implemented and enforced. As abatement commenced each
season, safety concerns were discussed with supervisors and reinforced to abatement
workers.
3.7 DATA MANAGEMENT
3.7.1 Data Management Objectives
To ensure data integrity, accuracy, and completeness.
To maintain the accountability and security of all project data sets.
To provide timely support to project day to day operation.
To provide quality data for data analysis.
To protect study family/subject confidentiality.
3.7.2 Data Management System Development
A data management system (DMS) based on Apple Mackintosh computers was
developed by following these steps: users' needs survey, system analysis, system design,
system programming, system implementation, installation and training, and system operation
and maintenance.
3-52
-------
3.7.3 Users' Needs Survey
We surveyed the users' needs, users computer experience and training, and their
familiarity with the existing data management system. We also studied the system which had
previously been used by our staff, the needs of this project as stated and the project data
management objectives.
3.7.4 System Analysis
We analyzed users' needs in terms of software, hardware, personnel, office facility;
and defined the system specifications to meet these needs within budget constraints. Based
on our system analysis, this project needed a personal computer based relational data
management system with user friendly interface, flexible computer entry screen and powerful
report generation capability. This is a portable system and users can share data easily.
3.7.5 System Design
Design Principals
Independent Programming For Data Management And Form-Based System
Development
All data and file management routines, user interfaces, and report generation routines
were compiled as the data management engine. A system of generic data set names was
defined, so any new data sets could be added to this system as long as the names were
defined by following the established rules. Therefore the form-based system can be
programmed independently of the programming of the data management engine. It gave us
flexibility in managing our database.
Form-Based Modular System Design
Every data entry form was considered as an individual module in the system and a
designated sub-directory is assigned for each data form. All the data sets associated with a
given data entry form were saved in its designated sub-directory.
3-53
-------
Graphic User Interface with Both Mouse and Keyboard Support
A self-contained user interface was programmed. Any user can follow the screen
instructions to perform data entry or data management. If any question is raised, the on-line
help is also available.
3.7.6 Design of the Components of Data Management Engine
Pile Structure
Hierarchical File Structure
Files/data sets are saved within a form sub-directory. Form sub-directories are created
within a. phase sub-directory, where phase stands for the study phase. In this project, we
have ten study phases and within each study phase we usually have many sample collection
forms, sample chemical results forms, and questionnaire forms. Within a given form of a
given phase, we have data set files, entry screen files, report files, and several other files for
maintaining data integrity. These files are saved in the designated form ^-directory.
Coherent and Consistent File Structure
Within a. form sub-directory, the file structure is the same, so automatic file
management can be programmed. The following is a brief list of the kind of files in a form
sub-directory.
Project's data sets for each data form are entered as the batched data sets initially and
den appended as a master data set. Data form information data set saves the key
information of this data form. Data definition dictionary has the field definitions, field
coding and the units of all the measurement variables in a data set. Three kinds of computer
data entry screens are programmed to provide data entry, update and retrieval. Report
generators are used to generate reports. Each report generator has consistent layout to
ensure its readability. All project data sets have the same data structure. Each project data
set has three kind of fields: linkage identifiers, data fields, fields for data management. The
linkage identifiers are defined at the beginning of each data set and are used to link all the
records in different data sets. The collected project data are entered into the data fields after
the identifiers. Several fields for data management use are assigned at the end of the project
data sets.
3-54
-------
Data Linkage System
A system of identifiers is defined by assigning an identification code to the following
entities: property, building on a property, apartment in a building, family in an apartment
and subject of a family. This system is the key to link all the data which were entered in
different forms of different phases together for data analysis and/or report printing.
User Interface
This system has a menu driven graphic user interface. It was programmed to support
both mouse and keyboard, for user's convenience. Data management subroutines, i.e., error
checking and record searching, are available as menu items. It supports on-line editing and
two-level data access: data editing with review or review only without editing.
Error Checking Routines
An error checking routine is programmed to compare the data which were entered by
two different individuals (double data entry). Range check and data coding validation
subroutines are also programmed.
Data Update Routine
An update subroutine with roll-back capability was programmed to ensure data
integrity. A data set was created to record the history of all the changes of a given data
entry form. Whenever an update was executed, the information before and after the update
would be recorded.
3.7.7 System Programming
System Engine Programming
The strategy that we took to finish the system engine programming was: We wrote the
generic routines for each function to be performed and tested them; combined the tested
routines for accomplishing a defined task as a module; and finally combined all the tested
modules as the data management engine. The required programming time for completing the
programming of the data management engine was equivalent to two seasoned programmers
3-55
-------
for 9 mo. The programs to provide the form-based data entry system were written when a
form had been developed and its data was ready for entry.
Programming for Each Form-Based Data Entry System
The programming of a form-based data entry system included the data structure
creation of all needed data sets, preparation of data definition dictionary, data entry screens
programming, report generator programming, etc. All of these programs and files were
saved in the given form sub-directory. Thus, we can consider the form-based data entry
system for each form as a peripheral: plug it into the data management engine when we
need to use it.
3.7.8 System Implementation, Installation, and Training
System implementation included computer software and hardware requisition, set-up
and installation, programmer training, programs compilation, and user training. The
software used to develop our data management engine is FoxBase+ for Apple Mackintosh.
Training was provided after the software and hardware were available or after our Data
Management System (DMS) was installed. This system can be installed by a trained staff
member.
Implementation for End Users
Personalized tutorial and system user's guide was available when our system was
installed for a user.
3.7.9 Data Base Management Operations and Maintenance
Data base management operations included routine form-based data entry system
programming, data entry and edit, data error checking and update, report printing, data
merging and data clean-up, etc. Data base management maintenance included hardware
maintenance, software bug fixes, and user technical supports.
3-56
-------
3.7.10 A Summary of Project Data Base
In the Appendix, several charts were used to present the file structure, data flow and
the data entry and update procedures of the Cincinnati Soil Project database. After that, a
summary list of all project data sets with the number of records in each data set was given.
3.8 GEOGRAPHIC INFORMATION SYSTEM
3.8.1 What is a Geographic Information System?
A geographic information system (GIS) is a computer system which can not only
integrate geographic data and tabulated attribute data together for further retrieval and
analysis but also support manipulation and presentation of spatially elated data. In this
project, the GIS that we used was called a generalized GIS. A generalized GIS is a GIS
which supports spatial statistical data analysis and modeling in addition to the available GIS
capabilities. The software used as our generalized GIS included: ARC/INFO, dBASE, SAS,
BMDP, and SPSS.
3.8.2 What Did We Plan To Achieve with Our Geographic Information
System?
(1) Identify the spatial distributions of environmental toxicants, e.g., lead, in the study
areas. (2) Investigate the spatial relations between environmental toxicants and human
exposure biomarkers. (3) Study the exposure pathways of human subjects to environmental
toxicants. (4) Study the migration pathways of environmental toxicants. (5) Propose
environmental remediation strategies and plans. (6) Manage and monitor the remediation.
(7) Measure the effectiveness of remediation and hence evaluate different remediation
procedures. (8) Study the rate of recontamination after remediation. (9) Retrieve and
present study outcomes, especially as maps.
3.8.3 Application of a Geographic Information System hi This Study
The geographic locations of the environmental sampling sites and buildings where study
subjects lived were digitized as different GIS layers. The sample collection information and
sampling results as well as subject questionnaire data were then merged to the associated GIS
3-57
-------
layers for future analysis. Thus the available data can be analyzed by their logical relations
and spatial relations. For example, we had the building layers, the street layers, the soil
layers and the property layers digitized for each study neighborhood and we merged the
house interior dust lead results to the building layer, the soil collection information and soil
lead results to the soil layer and the street dust lead concentrations to the street layer. The
spatial relation between the building house interior dust lead and the soil lead (or street dust
lead) could then be studied.
3.8.4 The Contribution of a Geographic Information System to This
Project
A GIS was essential for the full utilization of the soil and exterior dust lead data in the
Cincinnati project because of the area-wide nature of these data. Much of the soil in the
study area was not associated with specific housing units, but was in playgrounds, vacant lots
and at nearby housing units. A child therefore could have potential exposure to lead from a
number of these areas. Similarly the exterior dust sampling protocol involved collecting
composite dust samples from all of the streets, sidewalks, alleys, and parking lots hi the
study areas. The GIS provided a mechanism to determine soil and exterior dust lead
concentrations within specific distances of each child's residence.
3.9 DATA ANALYSIS
3.9.1 Data Analysis Objectives
1. Test project hypotheses,
2. Evaluate the quality of project's data,
3. Quantify the correlations between environmental lead measurements and human
lead, biomarkers,
4. Identify the key variables which can be used to quantify the changes hi human
lead biomarkers,
5. Conduct not only prospective and retrospective data analysis but also a sequence
of cross-sectional data analysis to understand the changes hi environmental leads
and human lead biomarkers,
6. Identify the spatial distributions of environmental leads in the study areas,
7. Investigate the spatial relations between environmental lead and human, exposure
bio-markers,
8. Study the exposure pathways of subjects to environmental leads,
3-58
-------
9. Study the migration pathways of environmental toxicants,
10. Measure the effectiveness of different remediation procedures,
11. Study the rate of recontamination after remediation,
12. Propose environmental remediation strategies and plans.
We accomplished these objectives by going through the following data analysis plan.
Due to the complexity of the data sets, we were not able to complete all data analysis within
the current available time frame.
3.9.2 Data Analysis Plan
The following were the key items in our data analysis plan:
Data conversion after completing data entry, error checking and updating;
Data clean-up and data coding-update;
Data list printing;
Preliminary data analysis: summary statistics and frequency distributions;
Correlational data analysis: correlation analysis, regression analysis (or analysis of
covariance);
Confirmatory data analysis: testing project hypotheses and the effectiveness of lead
abatement;
Prospective, retrospective and sequences of cross-sectional data analyses;
Modeling; and
Statistical methodology development.
3.9.3 Data Conversion
When data were entered and error checking was completed, several data conversions
have to be performed to ensure that comparable and accurate data were available for data
analysis. All project data were entered into Apple Mackintosh computers, through the
following steps, and then transferred to the University of Cincinnati's IBM mainframe
computer: First, convert project data base from Apple Mackintosh under FoxBase+/Mac to
IBM PC under dBASE. Second, convert this dBASE data base to PC-SAS data base with the
correct data type defined. Third, upload these PC-SAS data sets to mainframe computer.
3.9.4 Clean-Up Data and Data-Coding Update
(I) Check the linkage identifiers in each data set and make sure that all data in any
single data set can be merged with the associated data set, e.g., subject questionnaire data set
3-59
-------
should be merged with blood lead data set. (H) Check the data range and data coding to
ensure that the data to be analyzed were valid, e.g., outliers and missing values were
checked. If any errors were found, the master data set was updated by following the
established data update procedure.
3.9.5 Data List Printing
The data listing was printed after the clean-up of a data set was finished. This data
listing was visually reviewed and checked after it was printed.
3.9.6 Preliminary Data Analysis
Summary statistics:
Mean, standard deviation, geometric mean, geometric standard deviation, median,
percentiles, etc. were computed for each environmental sample, e.g., soil lead house interior
dust lead, exterior dust lead, etc., and human biomarker, i.e., blood lead and hand lead.
These summary statistics were computed for all the data in each individual phase and the data
which was collected in both preabated and postabatement phase. So, we can have a good
understanding about the collected data; and the comparisons between preabatement and
postabatement can be made easily.
Frequency Distributions
The frequency distributions of all the key variables were studied by the following four
approaches:
1. Single Data Form Based Analysis,
2. Multiple Data Forms Within a Single Study Phase,
3. Same Data Form From Several Study Phases,
4. Multiple Data Forms From Several Study Phases.
With a good understanding of the characteristics and the distributions of all the key
variables, we were able to propose more suitable statistical modeling and testing techniques
to analyze our data. i
3-60
-------
3.9.7 Correlational Data Analysis
Correlation Analysis
This was to study the correlations between human lead biomarkers and different
environmental lead concentrations, e.g., compute the correlations between blood lead, hand-
wipe lead,house interior dust lead, exterior dust lead, etc. Study the relationships between
human lead biomarkers and study subjects' mouthing behaviors as well as between human
lead biomarkers and their families' social economical status. We also investigated if lead
abatement had any impact on these correlations.
Regression Analysis and Analysis of Covariance
The regression analysis was used to predict the changes in human lead biomarkers from
the changes in the environmental leads, e.g., house interior dust lead, exterior dust lead and
soil lead. It was also be used to identify the key variables which have the significant
influence on the changes of human lead biomarkers. If at least one independent variable was
a categorical variable, e.g., the treatment group variable, then the regression analysis may
not be suitable, but the analysis of covariance can be used to achieve the above two goals.
3.9.8 Confirmatory Data Analysis
Test the Proposed Project Hypotheses
Study the Effectiveness of Lead Abatement
Several statistical procedures have been applied to study the effectiveness of lead
abatement: T-test, Analysis of Variance, Analysis of Covariance, etc. The nonparametric
statistics were considered when the assumptions of the classical procedures was not
appropriate.
3.9.9 Prospective, Retrospective, and Sequences of Cross-Sectional Data
Analyses
To study the changes in the human lead biomarkers and the environmental leads, we
propose to identify the key variances which had influences on these changes individually or
collectively through prospective retrospective and sequence of cross-sectional data analysis.
3-61
-------
3.9.10 Modeling
Modeling the Distributions of Environmental Leads
Where are the sources of environmental lead contamination? How do they spread? Or,
what are the distribution patterns? skewed or symmetric, uni-modal or multi-modal. We
also considered modeling the temporal distributions and spatial distributions of environmental
leads.
Modeling the Rate and Pathway of Recontamination of Environmental Lead
The understanding of the rate and pathway of environmental lead recontamination after
lead abatement will provide improved strategies for abating and reducing lead in the future.
Modeling Human Exposure Pathway to Environmental Lead Through Modeling the
Relationships Between Human Lead Bio-markers and Environmental Leads
We were interested in identifying and quantifying which and how study subjects were
exposed to environmental lead, e.g., soil lead, exterior dust lead, house interior dust lead,
paint lead, water lead, etc. A human lead exposure pathway model made it possible for us
to quantify the contributions of environmental leads to human lead body burden.
Establish A Model To Predict The Changes In Human Lead Biomarkers If
Environmental Lead Was Abated From One Level To Another Level, i.e., to quantify
the impact of abatement
We were also interested in quantifying the contributions of the environmental lead
abatement to the changes in environmental lead and to the changes in human lead
biomarkers.
3.9.11 Statistical Methodology Development
As mentioned before, many classical statistical procedures are not quite suitable for
analyzing our data due to the complexity of this study. Several statistical procedures were
investigated to provide satisfactory approaches to analyzing our data. We have some results
from these investigations, but more efforts are needed.
3-62
-------
Statistical Methods for Analyzing Familial Data
In this project, we have some study participants who came from the same families. In
some situations two families occupied the same apartment unit. The consequence to this was
that the needed assumption of many available statistical procedures may not be suitable. For
example, the usual Pearson sample correlation was not always appropriate to estimate the
"true" correlation between the blood lead and hand-wipe lead of our data, because it could be
biased. There are several other cases like tiiis. We have good progress in obtaining a
"good" procedure to estimate the correlation between hand-wipe lead and blood lead with
which these samples were collected from the families with varied sib sizes. However, before
we have bur methodology fully developed, we may have to use the not appropriate "usual"
statistical procedure to understand or analyze our data.
Statistical Methods For Analyzing Spatially Related Data
Our study was a neighborhood-wide study, so our lead variables are spatially
correlated. A generalized GIS system has been set up to facilitate analyzing our spatially
related data and several statistical procedures were investigated. However, because the final
report deadline was before a complete data analyses had occurred, we were not able to
complete this analysis. We will continue the work in this area as time and resources allow.
Statistical Methods For Analyzing Longitudinal Data
To analyze the changes of lead concentrations or loadings over time, e.g.,
postabatement vs. preabatement, the recognized approach is using longitudinal data analysis
techniques. Several methods are available and will be used in analyzing our data, e.g.,
structural equations with latent variables, linear models, and generalized linear models.
However, due to the covariance structure of our data, these methods needed further
modification to ensure that the relevant assumptions are fulfilled.
A Measurement Error Model With Random Coefficient Approach To Analyze Inter-
laboratory Calibration Data
We proposed a statistical model which can fully explain why the sample variance
increases as the chemical concentration increases. The statistical methodology development
is still under way and we have had our preliminary results presented in a national statistical
3-63
-------
conference. Currently, we obtained some results which can tell us if the concentrations of
the unknown standards from a lab is statistically the same as the consensus sample
concentrations.
3.10 PROJECT ADMINISTRATION
3.10.1 Financial Management
A budget and expense tracking system was used to financially manage the grant for the
Cincinnati Soil Lead Demonstration Project. The tracking system was developed to more
effectively manage the project budget and track expenditures by cost center.
The University, like most large organizations with a central accounting department, did
not provide detailed information necessary to manage an individual grant, on a timely basis.
Coir budget and expense tracking system provided a monthly review of performance against
forecast. The reports generated on a monthly basis included:
1. Comparison of Project budget vs. expenses/encumbrances. (Year-to-date and
Project-to-date)
2. Comparison of monthly budgeted payroll vs. actual expense and accumulated
monthly P/R expense for reviewing trends.
3. Cost Center reports (8) with budgeted vs. actual comparison. (Year-to-date and
Project-to-date)
4. Project summary report, detailed by category with budget vs. actual comparisons.
(Year-to-date and Project-to-date)
5. Cost Center summaries were created for a review of total expenditures specific to
each individual cost center. Such detailed and specific information may be used for
decision making purposes or determining unit cost within an individual cost center.
The following information describes the order of activities involved in utilizing our
budget and expense tracking system. A flow diagram (Figure 3-7) of the "Budget tracking
System" is displayed on the following page.
3-64
-------
CUFFS
Labor Distribution
Detail Report of Transactions
Adjustments - Memo
to Business Office
Hematology&
Environmental
Labs
Figure 3-7. Cincinnati soil project—budget tracking system flow diagram.
3-65
-------
Cincinnati Soil Project Budget and Expense Tracking System
The University of Cincinnati's financial system is called College and University
Financial System (CUFS). The accounting department at the University provided two
monthly reports: a budget summary report and labor distribution report. These reports
included each current month's expenses and encumbrances. We usually received these
reports 2 to 3 mo following the closing of the month.
Reports Generated by CUPS i
Two reports included in the monthly packet of financial information that were used for
our tracking system were the CUFS Budget Summary Report (Detail Report of Transactions)
and the Labor Distribution Report.
Creating the Cash Disbursements and Payroll Files for the Budget Tracking System
We used each month's current information from the CUFS budget summary report and
the labor distribution report as raw data to create the cash disbursements and payroll files for
our tracking system. The information was retyped into a format specifically designed for our
tracking system.
Current Expenses/Encumbrance Compared to Hard Copy Records
All encumbrances and expenses listed on CUFS were compared to hard copy records in
the project files. Copies of requisitions, invoices, contractual agreements, time sheets and
other documentation kept in our project files were checked and compared to confirm the
accuracy of the entry. A brief description of the expense and the name of the person
requesting the purchase were recorded to assist in determining the cost center, expense
category, and type of expense.
Error Checking and Adjustments of Inaccurate Charges to Project Grant
All entries were error checked and any discrepancies found in the financial information
were reported. Adjustments to the CUFS budget summary report and labor distribution
report to correct inaccurate entries were processed through the Department of Environmental
Health.
3-66
-------
Cost Centers Summaries Updated With Current Financial Information
Individual cost center summaries were updated with the current month's expenses and
encumbrances. Each cost center report included a comparison of actual to budget with a
variance that indicated the unobligated amount for each line item in a year-to-date and
project-to-date report.
Summary Report Updated With Current Financial Information
The summary report was also updated with the current's month's financial information
and served as a summarization totalling the financial data contained in the individual cost
summaries. The summary report included a comparison of actual to budget with a variance
that indicated the unobligated amount for each line item in a year-to-date and project-to-date
report.
Record Keeping
Each month's report was saved on diskette and a hard copy of the entire report filed
with other financial records of the Cincinnati Soil Lead Demonstration Project.
Financial Reports
In addition to the information provided through the budget tracking system, a monthly
financial report was presented to project investigators. The report contained a current
month's comparison of expenses to budget using the on-line capabilities of CUFS.
Additionally, projected spending through the end of the project period was included. This
report was informative and assisted in decision making based on current information.
3-67
-------
-------
4. RESULTS
4.1 STUDY POPULATION
One hundred and eighty-four eligible families with children less than five years of age
lived in the three study areas prior to study initiation. As a result of the high degree of
interest among caregivers of the children and enthusiastic project staff, 146 families
(225 children) agreed to participate in the study (Table 4-1). This represented 79% of the
eligible families. In June 1990 (Phase 5) an additional 45 families (66 children) were
recruited in areas B and C. They had moved into the study area after the initial recruitment
but prior to soil-lead abatement which was scheduled for those areas. At the completion of
the study (October 1991), 95 families with a total of 153 children were actively participating
in the study. A summary of project enrollment (subjects, new births, families and dwelling
units) by study Areas A (Pendleton), B (Findlay, Back and Dandridge) and C (Glencoe and
Mohawk) appears in Appendix N. New births represent infants born into study families.
The usefulness of their data depends on whether they entered the study before or after
abatement and will be described later. The families recruited in June 1990 who were living
in Area B had their interior abatements performed during the summer of 1990 along with the
exterior (soil and dust) abatements which had been scheduled for Area B during the summer
of 1990. Thus the subjects recruited in June 1990 (and living in Area B) were similar to
those in Area A in that all three abatement procedures were implemented during the same
period of time.
The initial data analysis will focus on the children who were recruited in June-July of
1989 and hereafter referred to as "initial recruits". As the study design stipulated, the focus
of the recruitment was on children living in the rehabilitated housing (about 90% of the study
population). The initial data analysis will, therefore, only include data for children living in
rehabilitated housing. Analysis of results from area-wide sampling, such as for soil and
exterior dust sampling on streets, sidewalks, alleys etc., will include all data regardless of
the type of nearby housing.
The age of the initial recruits at the time of the first blood collection is shown in
Table 4-2 by study areas.
4-1
-------
TABLE 4-1. CINCINNATI SOU, PROJECT ENROLLMENT OF SUBJECTS
Initial Recruitment11
Area
A
B
C
Total
Number Enrolled
(Phase 01)
56 (36)c
107 (68)
62 (42)
225 (146)
Total Recruited
Phase 01 & 05
Total New Births
Total Subjects
Number Remaining
(Phase 09)
23 (17)
47 (31)
30 (22)
100d (70)
291
16
307
Second Recruitment1*
Number Enrolled
(Phase 05)
not performed
29(19)
37 (26)
66 (45)
Number Remaining
(Phase 09)
not performed
16 (9)
21 (16)
37e(25)
Total Remaining (Phase 09)
Phase 01 Recruits 100
Phase 05 Recruits 37
New Births 16
153
'June-July 1989 (Phase 01).
bJunc-My 1990 (Phase 05).
^No. families indicated in parenthesis.
An additional 14 new births also participating in Phase 09.
°An additional 2 new births also participating in Phase 09.
TABLE 4-2. NUMBER OF STUDY PARTICIPANTS BY AGE RANGE
(Phase 01, Rehabilitated Housing)
Age (years)
Less than 1
1- <2
2- <3
3- <4
4- <5
5- <6
Total
Area A
9 (18.4)*
10 (30.4)
10 (20.4)
10 (20.4)
10 (20.4)
0(0)
49
AreaB
17 (21.2)
17 (21.2)
13 (16.2)
15 (18.8)
13 (16.2)
5 (6.2)
80
AreaC
11(25)
7 (15.9)
9(20.4)
7(15.9)
6 (13.6)
4 (9.1)
44
Percentage for area shown in parentheses.
4-2
-------
4.2 EXCLUSION FROM DATA ANALYSIS
The initial data analysis will focus only on children entered into the study during the
initial blood collection period (Phase 01, June-July 1989) and who were living in
rehabilitated housing. Therefore a number of children will not be included in the analysis
presented herein: children recruited from Areas B and C during Phase 05 (June-July 1990,
children living in non-rehabilitated housing, and children born in study families after older
siblings had already been recruited ("new births"). In addition, four subjects were excluded
who were identified as being chelated prior to entry into the study or pre-abatement or who
were victims of careless remodeling work.
Children Recruited During Phase 05
A total of 66 children were recruited during Phase 05 (June-July 1990), 29 in Area B
and 37 in Area C. The children recruited in Area B received all three types of abatement
(exterior and interior dust and soil) during the summer of 1990. Thus, their abatement
pattern is identical to that in Area A except that it was conducted one year later. In a future
data analysis we will explore combining them with Area A children for the purpose of
evaluating the impact of abatement.
Children Living in Rehabilitated Housing
Thirty-one of the children recruited in Phase 01 (June-July 1989) are identified as living
in non-rehabilitated housing: 2 in Area A, 18 in Area B and 11 in Area C.
The housing of most of these Area C children were initially misclassified as non-
rehabilitated housing when it actually was rehabilitated. Re-coding will be conducted in the
near future so that those children can be included in the analysis of children living in
rehabilitated housing and who were recruited during Phase 01. Five other children are
currently listed as living in "unidentified" housing. The correct housing type for these
children will be entered into the data set in the near future.
New Births
A number of children of study families were born during the study period. In many
cases (16) we were successful in recruiting them into the study when they were only a few
4-3
-------
months old. In many cases they were born after some abatement in their home environments
had already occurred. Data from those children will not be used as part of the primary data
analyses but will only be used in special case study analyses and other special studies. Four
of these children are in Group A, 10 in B and 2 in C.
Specifically Excluded Subjects
All data from four subjects who were either in housing undergoing renovation or who
were lead poisoned prior to their entry into the study were removed from the data analyses.
Subjects 058 and 059. The apartment in which these children were living was
undergoing paint removal by the owner resulting in excessive exposures to lead-containing
dust. Hand lead values were 1,366 and 1,386 /tg, respectively pre-abatement (June 1989).
These values are almost an order of magnitude higher than any others we have seen. The
family lived in Area B (Findlay sub-area) in a deteriorated non-rehabed 19th century
building. We have three blood samples from each child. The family moved April 1990.
Subjects 812 and 813. These subjects were each chelated for lead poisoning during the
summer of 1989 (July and August, respectively) prior to their entry into our study during our
second recruitment in June 1990 (Phase 05). Subject 812 had a blood lead of 46 in August
1987 as determined by the Cincinnati. Health Department.
4.3 SAMPLE COLLECTION AND ABATEMENT SCHEDULE
The overall abatement and monitoring time table for the Cincinnati project was
previously shown in Figure 3-1. The specific types and locations of the samples collected
during the ten sampling phases (00 to 09) were outlined in Figure 3-2. Certain reductions in
sample collection and analyses that were necessary due to time constraints are indicated in
Figure 3-2 for Phases 06, 07 and 09.
4-4
-------
4.4 BLOOD AND ENVIRONMENTAL SAMPLE QUALITY
ASSURANCE/QUALITY CONTROL RESULTS
Prior to the beginning of our study the precision and accuracy of each type of sample
measurement was evaluated. Goals were established in terms of completeness and
representativeness of study design. Coordination efforts among the three cities involved hi
the study resulted in the development of similar methods and protocols so that ultimately data
comparability could be realized. However, the validity of the Cincinnati Soil Project results
still depend in large part on the quality of sample collection and analysis. The results of the
Quality Assurance/Quality Control (QA/QC) procedures implemented in our collection and
analysis protocols are given in the following sections. Table 4-3 lists the different types of
samples collected and analyzed, and the quality control associated with each type.
At the beginning of the study 28 fictitious family identities were developed. Identifiers
were consistent with those used for the real study families. Except for soil and exterior dust
all field QC samples were given study sample numbers and identified as being collected from
one of the fictitious families. In this way, QC samples were submitted to the lab and
analyzed with their true identity unknown. Soil and exterior dust QC samples were, on the
other hand, distributed to the analyst without their associated location, building and family
identifiers. Therefore, the true identity of those samples was also protected.
4.4.1 Quality Control Evaluation for Blood
The hematology QC program consisted of the following elements:
1. Field duplicates - fictitious families
2. Blind control samples supplied by Centers for Disease Control (CDC) - lead (PbB)
3. Low and high isotope dilution mass spectrometry (IDMS) determined
reference samples
4. College of American Pathologists (CAP) PbB proficiency program
5. CDC, PbB and free erythrocyte protoporphyrin (FEP) proficiency program
6. Bi-annual analysis of hematology controls for Hematocrit and Hemoglobin
7. Human reference sample for iron (Fe) and total iron binding capacity (TIBC)
The analytical methods used for each analyte were as follows:
1) PbB - Anodic Stripping Voltammetry
2) FEP - extraction followed by spectofluorometry
3) Serum Fe and TIBC - electrochemical
4-5
-------
TABLE 4-3. QUALITY CONTROL EVALUATIONS FOR EACH SAMPLE TYPE
1) Blood
a. Field duplicates - fictitious families
b. Blind control camples supplied by CDC - PbB
c. Low and high ID-MS determined reference samples
d. CtoUege of American Pathologist* PbB proficiency program
e. CDC PbB and FEP proficiency program
f. Bi-annual analysis of hematology controls for Hematocrit and Hemoglobin
g. Human reference sample for Fe and TXBC
2) SOU
a. Field duplicates
b. Field lab sieving blank
c. Blind control samples supplied by EMSL - Las Vegas
d. Low and High reference samples
3) Exterior Dust
a. Field duplicates
b. • Field lab sieving blank
c. Blind control samples supplied by EMSL - Las Vegas
d. Low and High reference samples
4) Interior
a. Field duplicates
1. interior dust
2. dust&ll
b. Field blanks
1. interior dust
2. handwipes
c. Blind control samples supplied by EMSL- Las Vegas
1. interior dust
2. handwipe
d. Lab method and reagent blacks
1. interior dust :
2. handwipe
3. dustfall
e. Lab duplicate
1. interior dust
2. dustfall
f. Lab controls
1. interior dust
2. handwipe
3. dustfall
g. AAS analysis-duplicates and % recovery
1. interior dust
2. dustfall
3. handwipe
S) Water
a. Field duplicates
b. Field blanks
c. Lab blank
d. Lab control
c. Lab duplicates
f. % recovery
6) Paint
a. Instrument blanks
b. Instrument-calibration checks
4-6
-------
4) Hematocrit - centrifugation
5) Hemoglobin - calorimetric
A field duplicate for blood was one half of a total sample drawn from an individual
volunteer. Thus, approximately 8 ml of blood was collected from 1 person and dispensed
into 2 EDTA-containing pediatric Vacutainer™ tubes and 2 vacutainers™ with no
anticoagulant. Each set of two different tubes were labeled with one of the 28 fictitious
i
identities and submitted to the lab for the complete series of analyses. Fresh field duplicates
were collected at a rate of 2 individuals of 4 samples per week. The fictitous family
indentifiers were repeated in each phase. This type of QC sample can only evaluate
precision. Table 4-4 shows the results of the differences between duplicates for each test
when samples were analyzed blind.
Four different concentration blood samples were supplied by the Centers for Disease
Control to test for analytical or procedural variabilities within or among the three cities in the
soil projects. Prior to the start of the study these samples were repeatedly analyzed by each
city. Our assays were carried out in 13 separate daily runs at a rate of one run per week and
over the time period from 1/31/89 to 5/9/89. A target value was then determined for each
city's analysis of the sample and also a consensus value for the sample itself.
During the course of the study, at least two of the four samples were analyzed in
duplicate with each set of unknowns. Of the four two were always blind to the analyst. The
two blind samples were those with target values of 3.5 ± 8.9 /tg/dL. Table 4-5 shows the
results accumulated on these samples over the two years of the study.
In addition to the samples supplied by CDC for this project we continued to employ the
techniques that are a part of the standard operating procedures of our laboratory. That
included analyzing duplicate samples of 2IDMS determined reference bloods with each set
of unknowns. Limits were established for these samples to evaluate the validity of the
analytical run, and control charts were kept to document performance over time.
If the concentration of two or more of the four samples fell outside the established
limits, the run of PbB samples was rejected and all samples reanalyzed. This did not occur
with any of our analytical runs. However, on two occasions the 5.3 jtg/dL control in
Phase 03 fell well beyond the 3 standard deviation or rejection criteria. Possible
contamination of these QC samples was suspected. Listed in Table 4-6 are the averages,
4-7
-------
CO
0
*a
g
s< CO
§s
O
CO.
3
fi
i
en
CO.
U
K
ca
•Sh
I
w
fi
ca
1
£
*
i
u- oo oo o\
in O
OS
o
"«n
O
es
vo
•<*
en
en"
«n vo. o
C5 O Ci
CO CO
^^. Ci^
CS
00 00
o o
c
l> Qv OO
«n.
vo vo m
o o o o- o:
co oo
CS »-<•
c* co.
O O
B
4-8
-------
o
E£I
r*l
05
1
w H
X §
PQ O
§ 5
! w
Erf
co O
W CO
il
1
ii
o §
vv H
e §
5§ 2
So
O'rt
p |r
go
BLE4-5. BLOOD I
DISEASE <
<
C^
VH
H
rr"1
f*M
o
1
1
-1
00
co ON
3$
o
in vo
CS ^H
ON CO
00 O
in co
CO* -*
Target Valuea
Consensus Value
DATE SAMPLE
COLLECTED
^
ON
ON
ON
O
CO
VO
ON
O
VO
CN
ON
O
ts
VO
n
ss
ON
C5
f-
co o cs '-' co c* in
in in ON in ON
CO O CS O VO O Tj*
vo co
ON oo cs vq o
oo oo oo *-* co vo T-*
cs ••* cs co o
23 oo
CN
1— 1
ON
oo
00
r— <
ON
o
r— (
vo
8
| S Q > d a
<"*j £*l *^* m \*3 ^j
Pv, Z S vi U S S
^
m
w s
^.Q -g
1 d -fl
" ^ s §
shed by U.C. lab prior to i
i Value; U.C.; CDC; ESA
lyzed as a blind control th
a § a
f> & 5
111
ID §
•i ^ *§,
5*3 i
y 3 *Q
£*£$
4-9
-------
»
^
-3
i
1
s
1
I
X
c>
M
O
1
C/3
§
^,
2
H
JS
§
O
o
23
*
*
ii
O CS
•^ ^ 'I ^ ^
T-H ^H ^H T~< 1^
fS^ |^J |^J l^J ^^
^^J ^^5 c^a ^H ^^
^^* ^j" ^x|" ^T "'sf
^OQQ cS'-i eso oo OOON
1-1
t*» CO VO ITS | C~^
o o c> o o
u
1 o' ' o o o ' o
ON VO C— OO O\
0 ' 0 ' 0 ' 0 ' 0 '
«.,*«, ^ , °°. , *-! ,
vo irJ n vo
vooo OON «nes oo OOON
T— ( CO V> C~~ ON
O 0 0 O 0
•8
•o
"o
X
o
2
3
O
*
4-10
-------
average differences between duplicates and the standard deviations for these IDMS samples
in the Soil Project analytical runs.
Participation in two proficiency programs is another ongoing process in our labs.
Samples are provided by the Center for Disease Control on a monthly basis for PbB and FEP
determinations. Lead samples are supplied quarterly by the College of American Pathologist
proficiency program. Our results are compared to consensus values established by reference
and participating laboratories for these programs. Figure 4-1 and 4-2 show our performance
in these programs for PbB over the two year interval during which this study was run.
Figure 4-3 illustrates the results for FEP in the CDC program.
Hematology controls were purchased periodically during the Soil Project and used to
evaluate hematocrit and hemoglobin concentrations. Purchased from Fisher Scientific
Company, these controls closely resemble patient's whole blood samples and so provide a
control for the two methods. There are target values assigned to the samples which were
obtained using a number of automated instruments and manual methods. Four sets of these
samples were analyzed and the comparison is shown in Table 4-7.
Serum Fe and TIBC analysis were monitored by analyzing the same human serum
sample during each day's analytical run. The serum was a pooled sample collected from one
individual, aliquoted into separate tubes and frozen. Each sample pool was only stable for
approximately 6 mo. Therefore, prior to its expiration date another sample pool was
developed. Thus, pools were overlapped to maintain continuity. These samples provide
measurement of precision and a check for instrument or calibration drift. Table 4-8 shows
data generated during the Soil Project study time frame.
4.4.2 Quality Control Evaluations for Soil X-ray Fluorescence
The soil QC program consisted of the following elements:
1. Field duplicates (co-located samples)
2. Field lab sieving blank
3. Blind control samples supplied by Environmental Monitoring Systems Laboratory
(EMSL) - Las Vegas
4. Low and High reference samples
4-11
-------
8
O> o>
cJ'd
I I
in
co
o
eo
CQ
£
§
in
to
CO
in
I
o
in
I
o
HP/Oil gqd O n
4-12
-------
8
N.
" &
0> O>
o o
in
CO
-g
to
CM
0)
JQ
Q.
I
u>
- U>
~T
o
o
in
in
04
o
CM
10
in
ClP/6tl) gqd O n
4-13
-------
X
8""" CO
o>
I-1 O
I I
-^
O
-- 8
§
8
O •£.
— co O)
ai
u_
O
O
4- 8
0
CM
0
CO
0
-------
TABLE 4-7. UNIVERSITY OF CINCINNATI PERFORMACE IN THE ANALYSIS
OF HEMATOLOGY CONTROLS
Lot#
HEMATOCRIT (%) HEMOGLOBIN (g/dL)
University of University of
Fisher (Target Cincinnati Fisher (Target Cincinnati Results
Value) Results Value)
X S.D. X S.D. X S.D. X SrD.
Lotl
(Phase 1)
Lot 2
(Phase 3)
Lot 3
(Phase 5)
Lot 4
(Phase 7)
34.3 3.0 33.6 0.5 11.7 0.8 11.4 0.1
33.9 3.0 33.3 0.5 11.8 0.8 12.0 0.2
30.5 2.0 28.9 0.6 12.3 0.5 12.4 0.4
30.5 3.0 29.3 0.7 12.3 0.8 12.5 0.4
TABLE 4-8. SERUM IRON AND TOTAL IRON BINDING
CAPACITY SERUM POOLS
Number
of
Identity Analyses
Lot-M4* 16
LotM5 16
Lot M6 18
Serum FE
X
57.9
110.4
80.6
S.D.
2.2
5.8
3.9
TffiC
X
305
309
301
S.D.
13
20
13
% Saturation
X
18.9
35.9
26.8
S.D.
1.0
1.3
0.8
Human samples; no target value.
As was to be expected, soil field duplicates (co-located samples) were at best a rough
estimate of the concentration of the original sample collected. Duplicates were collected for
one out of every 10 unknowns and not necessarily done on the same day. They were,
however, collected within 6 mo.
4-15
-------
Field Lab Sieving Blanks consisted of material obtained from a sand/gravel quarry in
Cincinnati. These materials were previously determined to be very low lead content (Que
Hee et al., 1985). There were three types of native soils analyzed:
Blank #1. Lake bed clays possibly Wisconsin glacial stage 20-25,000 years old.
Blank #2. Uliniosian Till, 125,000 years old.
Blank #3. Pre-glacial Muevial (river) sand, 150,000 years old.
They were sieved to <250 p particle size and randomly inserted in the unknown
sample stream. Results are shown in Table 4-9.
TABLE 4-9. SOIL HELD LAB BLANKS
Blank #
1
2
3
N
90
116
94
X "" S.D. Minimum Value
11.0
14.4
15.1
10.6
11.0
8.9
0.9
0.9
3.5
Maximum Value C.V.
54.8
65.2
53,9
96
76
59
Study control samples for soil analysis were provided by EMSL - Las Vegas and
provided to each of the three Soil Projects. The raw samples were collected in each city and
shipped to EMSL where they were sieved, analyzed, and redistributed to the projects. These
soils were randomly inserted into the study's sample stream and analyzed unknown to
laboratory staff.
An X-ray Spectrometer (XRF) and Atomic Absorption Spectroscopy (AAS) concensus
value for the samples was determined in a round robin exchange among the cities. Also
from these data a correction factor for each city was calculated. The factor when applied to
all soil results yielded comparable data among the three projects. The following table
(Table 4-10) gives the results of our analysis, the corrected concentration and the concensus
value determined by the round robin exchange for the particular QC samples used in our
4-16
-------
TABLE 4-10. UNIVERSITY OF CINCINNATI LAB PERFORMANCE
IN ANALYZING ENVIRONMENTAL MONITORING SYSTEMS
LABORATORY QUALITY CONTROL SOIL SAMPLES
BY X-RAY SPECTROMETER0
BalM
BosM
CinH
CinL
N
49
32
31
130
X (ppm)
1,016
6,654
14,890
301
S.D.
40
268
635
11
C.V.
4
4
4
4
Corrected
xa(ppm)
884
5,788
12,951
262
Consensus
Valueb (ppm)
-
6,175
12,499
335
a x 's were multiplied by the constant 0.8698 determined by the intercalibration study to adjust for "within"
reliance.
Consensus value for soil samples determined by XRF using the mi
°Data are from QC samples analyzed during phases 2, 3, 5, and 9.
variance.
Consensus value for soil samples determined by XRF using the multiplicative model with weight= 1.
study. Our labs digestion and AAS values for three of the samples was: Bos M=6,937,
Cin H=13,195 and Cin L=379.
Two reference control samples were placed within each set of 16 samples analyzed on
the KEVEX XRF instrument. Their average concentration as determined by repeated
analysis was 170 ppm and 1066 ppm of Pb. A 2 standard deviation limit was set as rejection
criteria and subsequently any XRF run of samples with one or more of these QC samples
outside of the established range was re-analyzed.
4.4.3 Quality Control Evaluation for Exterior Dust
The exterior dust QC program consisted of the following components:
1. Field duplicates (co-located samples)
2. Field lab sieving blank
3. Blind control samples supplied by EMSL - Las Vegas
4. Low and High reference samples
The Quality Control used for the exterior dust was the same type as used for the soil.
Duplicates were collected for one out of every 10 unknowns. Field lab sieving blanks were
the same material as used for the soil. The EMSL-Las Vegas QC were also soil material of
4-17
-------
the concentrations used with soil. These were all dispersed randomly throughout the exterior
dust samples. The reference controls were also the same samples. Table 4-11 shows the
concentrations of the blanks prepared with the exterior dust and Table 4-12 performance on
analysis of EMSL QC samples (compare with Table 4-10).
TABLE 4-11. EXTERIOR DUST FIELD LAB BLANKS
Blank #
1
2
3
N
44
44
45
24.5
28.0
22.7
S.D.
4.9
8.1
6.5
Minimum Value
6.1
13.0
9.6
Maximum Value
30.4
64.4
26.3
C.V.
20
29
29
TABLE 4-12. ENVIRONMENTAL MONITORING SYSTEMS QUALITY CONTROL
SOIL SAMPLES ANALYZED WITH EXTERIOR DUST SAMPLES
(Phases 1,2,3,5 and 9)
N
48
24
23
117
X (ppm)
1,015
6,630
14,582
314
S.D.
43
326
826
56
C.V.
4
5
6
18
Corrected Xa Consensus Value (ppm)
883
5,757
12,683
273
6,175
12,499
335 .
& X *s were multiplied by the constant 0.8698 determined by the intercalibration study for "within" lab variance
Consensus value for soil samples determined by XRF using the multiplicative model with weight = 1. The
bi-weight range is currently being determined.
4.4.4 Quality Control Evaluations for Interior Dust
The interior dust QC program consisted of the following components:
1. Field Duplicates
2. Field Blanks
3. Blind control samples supplied by
EMSL - Las Vegas
4. Lab method and reagent blanks
4-18
-------
5. Lab duplicates
6. Lab controls
7. AAS analysis - duplicates and % recovery
Interior dust duplicates were collected at one every 25 residences at a spot adjacent to
the study sample site. Table 4-13 shows the average difference between those results for
each sample type and within each collection phase.
TABLE 4-13. PERFORMANCE ON ANALYSES OF
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
QUALITY CONTROL DUST SAMPLES
BalML
BosH
CinL
CinMH
N
34
35
26
38
Xa(ppm)
1,727
24,104
259
2,683
S.D.
275
2,337
44
225
C.V.
16
10
17
8
Consensus Value
1,492
N/A
232
2,378
a x 's were multiplied by the constant 0.9839 determined by the Intel-calibration study for "within" lab variance.
Consensus value for soil samples determined by XRF using the multiplicative model for "within" lab variance.
Field blanks were collected during the last four phases of the project. These consisted
of setting the pump on a table with the nozzle pointed upward and collecting air into an
empty sampling cassette for 3 min. The concentration range of 44 blanks was 0.11 - 1.75 pig
Pb with an average of 0.48 /tig/Pb ± 0.35. Three samples had a Pb concentration greater
than 1.0 fjtg and were considered contaminated. The sample weights were all very low
- < 0.0009 g.
Samples prepared by EMSL - Las Vegas were disguised using fictitious family
identities and sent to the laboratory. The samples, shipped from EMSL, were sampling
cassettes containing a known amount of standard dust. They were inserted into the studies
sample stream at the field office.
The laboratory handling of the interior dust samples was a fully controlled process in
which validity of the data was determined by whether or not blanks and controls fell within
established weight or concentration limits. The limits were calculated using the results of the
4-19
-------
first 25 assays of each QC type. If more than three of the six QC samples in a run were
outside of the defined limits, the entire set was considered invalid. In this project 154 sets of
interior dust samples were analyzed. Three interior dust sets (2%), of 22 unknowns each,
were not acceptable according to these QC criteria.
The initial preparation of the samples was very important. Due to the small sample
size collected by our vacuum method, the accuracy of the weight measurement was critical.
Weight corrections were applied to all unknowns in a set if blank values were > 1 standard
deviation (SD) and <2 SD of an earlier determined average method blank weight nf both
blank weights were >2 SD of the average method blank weight, then the QC sample was
considered out of limits. All samples were digested and analyzed after weight measurements
were determined, however, samples that weighed less than 0.002 grams (n = 45; 1.8%)
were not calculated in ppm. Average weights for the Soil Project dust samples for each type
of interior dust in each phase are shown in Table 4-14.
TABLE 4-14. AVERAGE WEIGHTS OF INTERIOR DUST SAMPLES (Grams)
Phase
01
02
03
05
09
Entry X
0.0567
0.0227
0.0350
0.0750
0.0700
Floor X
0.0435
0.0279
0.0350
0.0672
-
Window X
0.1164
0.0690
0.0386
0.1859
'
MatX
0.0115
0.0479
0.0739
0.0596
-
(Total number of samples = 2,490).
There are several types of blanks in our preparation and digestion procedure for interior
dust analysis. These include the method blank, preparation reagent blank and digestion
reagent blank. The purpose of all of these samples are two fold: (1) to determine the true
amount of weight or lead that is present in the unknown samples, and (2) to detect the
presence of contamination in the collection, preparation, and digestion of study samples
unknowns.
4-20
-------
Method lab blanks consisted of blank cassettes loaded with filter and support pads and
from the same lot of cassettes as used for the unknowns. These blanks were inserted during
the evaporation or preparation stage of each day's set of samples. A reagent blank was also
inserted at this point in each set and was actually a pre-weighed beaker of 50 ml of
deionized/distilled H20. Another reagent blank was introduced during the digestion step.
This was done at a rate of one per day and was simply an empty beaker handled as a sample
beaker.
The amount of lead in the method lab blanks ranged from 0.14-4.31 jtg Pb found.
Eight of 128 were > 1.0 /«g Pb - the rejection limit. Preparation reagent blanks ranged from
0.13 to 10.0 /tg of Pb found with nine of 128 samples exceeded the rejection limit of
> 1.0 jug of Pb. The digestion lab blank had seven values of 128 not accepted and ranged in
concentration from 0.13 to 4.17 /tg Pb per sample.
There were two National Institute for Standards and Technology (MIST) standards used
to track the accuracy of the interior dust method. They were standard #1648 Urban
Paniculate with a known Pb concentration of 6,550 ppm and Standard # 1646 Estuarine
Sediment of concentration 28.2 ppm. One of the NIST Standards (#1646 or #1648 arbitrarily
chosen) and duplicates of either of the reference samples were chosen to be incorporated into
the initial preparation of the samples. Approximately 50 mg of each standard dust was
weighed into a clean cassette and reseated. Prior to digestion of the samples 50 mg of the
NIST standard, the same as was added to the prep stage of that set, was weighed into a
beaker and digested along with other samples of the set. These gave us the means needed to
track the handling and analysis of the samples. In addition, two bulk dusts taken from homes
in the Cincinnati area were used to track the precision of the method over the course of the
study. These latter QC samples were intended to be more valid urban dust standards than
the NIST standards (Table 4-15).
The performance of the AAS instrument was also subject to QC monitoring. One in
every 25 samples was analyzed in duplicate. Every 20th sample was used to determine
percent recovery. The average difference between 160 duplicates was 1.42 ± 1.85 /tg. The
percent recovery calculated from 189 spiked samples averaged 102.2% ± 5.6.
4-21
-------
TABLE 4-15. SOIL PROJECT INTERIOR DUST
QUALITY CONTROL STANDARDS
(Floor and Window)
Sample
NIST 1,646
(28.2 ppm)
NIST 1,648
(6,550 ppm)
REFERENCE LO
REFERENCE HI
Stage of Analysis
Prep Digestion
Prep Digestion
Prep Digestion
Prep Digestion
N
105
57
46
19
118
23
113
28
X (ppm)
28.2
28.2
7,052
6,946
279
281
2,802
2,766
S.D.
4.7
3.6
528
318
52
39
191
133
C.V.
16.6
12.8
7.5
4.6
18.5
14.0
6.8
4.8
4.4.5 Quality Control Evaluations for Interior Dustfall
The interior dustfall QC program (Table 4-16 and 4-17) consisted of the following
components:
1. Field duplicates
2. Lab method and reagent blanks
: 3. Lab duplicates
4. Lab controls
5. AAS-duplicates and percent recovery
TABLE 4-16. INTERIOR DUSTFALL FIELD DUPLICATES
Sample #
211
212
285
286
311
312
236
237
ppmPb
454
152
14
23
44
122
482
572
Sample #
260
261
160
161
136
137
186
187
ppmPb
210
74
47
305
353
356
103
40
4-22
-------
TABLE 4-17. QUALITY CONTROL STANDARDS ANALYZED
WITH DUSTFALL SAMPLES
Sample
MIST 1646
(28.2 ppm Pb)
Reference Lo
Reference Hi
Stage of Analysis
Prep Digestion
Prep
Prep
N
12
13
11
14
X (ppm)
34.5
28.9
314
2,960
S.D.
6.2
4.1
3.8
210
C.V.
18
18
12
7
Interior dustfall samples were collected during two phases of the Soil Project - Phase 02
and 06. Field duplicates but not field blanks, were obtained at one in every 25 residences in
Phase 06. The comparison of the resulting eight duplicates is listed below.
The laboratory quality control was identical to that for the interior dust samples. There
was a method blank or empty container prepared as a sample. Reagent blanks were inserted
at each stage of sample handling and NIST and reference standards were used to track the
accuracy of the method.
The concentration of all blanks was below the 1.0 j&g Pb rejection limit. The NIST
standard inserted into the analyses was Standard #1646 Estuarian Sediment containing
28.2 ppm Pb. Low and high Pb concentration reference samples were also used. Out of
13 sets analyzed no sets had to be rejected. The standards are tabled in Table 4-17 and can
be compared with the standards and performances reported in Table 4-16.
The average difference for AAS analysis of duplicates was 1 ± 0.5 /*g Pb. The
average percent recovery of spiked samples was 104% ± 4.9.
4.4.6 Quality Control Evaluations for Hand Lead
The hand lead QC program consisted of the following components:
1. Field blanks
2. Blind control samples supplied by FMSL-Las Vegas
3. Lab method and reagent blanks
4. Lab controls
5. AAS-duplicates and percent recovery.
4-23
-------
The acquisition and analysis of field blanks is very important to the evaluation of the
quality of the unknowns associated with it. Proper handling and documentation of the wipes
used is critical to that evaluation. In the Soil Project field blank limits were set at 0 ± 6 /*g
Pb based on using the 2 standard deviation value for the lab reagent blank. Thus, any field
blank concentrations greater than 6 fig indicated contamination and rendered its associated
sample or samples invalid. In addition, samples with a field blank less than 6 /ttg would be
invalid due to suspicions such as samples containing less than 6 total wipes or using the
wrong lot of material to collect the sample. There were 2,369 handwipe samples analyzed
including field blanks and of the samples 55 were invalid based on the field blank
concentration.
Hand lead control samples were prepared by the EMSL-Las Vegas lab and sent to
Cincinnati for incorporation into the Soil Project samples. These samples contained aqueous
lead solution of various concentrations aliquoted onto six wipes. The samples included clean
wipes for use as field blanks. Before being sent to the lab the samples were labeled with the
fictitious family identities. Table 4-18 shows the results of those samples.
TABLE 4-18. ENVIRONMENTAL MONITORING SYSTEMS
LABORATORY QUALITY CONTROL HAND LEAD SAMPLES
Concentration
Level (/tig)
O
L
M
H
Concentration* Added by
N
94
38
36
38
X
0.1
4.8
17.8
38.0
S.D.
2.7
2.6
4.0
9.7
C.V.
5,086
54
23
26
EMSL G*g)
0
5
20
40
The amount of lead added to the wipes is yet to be obtained from the EMSL-Las Vegas lab.
Lab reagent blanks were deemed invalid if <0 or >3 /Kg of lead was found. Only one
out of 140 blanks was found to be invalid. Method lab blanks were used to correct all
samples by the amount of lead present in clean wipes. The value subtracted was the average
Pb concentration for all samples by the amount of lead present in clean wipes. The value
substracted was the average Pb concentration for all method blanks determined frob the
same lot of wipes. There were different lots of wipes used in the Soil Project Study and
4-24
-------
their lab blanks ranged in concentration from 5-23 /tg Pb per 6 wipes (no. of wipes used per
child).
Lab controls were blank wipes spiked with a standard level lead solution. The
concentrations used were 4, 20, 40 and 100 /ig/ml. To check the quality and concentration
portions of the stqcksolution of each control were analyzed periodically by AAS. Limits
were established for all controls and used in the validation of the sample set. If 2 or more
controls or blanks were found to be outside of the established limits, the set was found to be
invalid. Eight out of 140 sets could not meet the QC criteria. Two sets of 22 unknowns
(including field blanks) were excluded entry from the data base (less than 2%) while six sets
were included, but qualified.
Analytical duplicates and spiked recoveries were determined by the AAS analyst. One
in every 25 samples or hand leads were analyzed in duplicate and the average difference
between duplicates was 0.37. The percent recovery calculated from 150 spiked samples
averaged 102 ±6.
4.4.7 Quality Control Evaluations for Water
, The water QC program consisted of the following elements:
1. Field duplicates
2. Field blanks
3. Blind lab controls
4. Lab duplicate and percent recovery.
Water samples were collected from the residences during Phase 04 and 08 of the study.
Excluding field blanks and duplicates there were 278 water samples analyzed. Duplicates
were collected at 10% of the homes with a total of 26 samples collected at 13 residences
during both phases. The following Table 4-19 lists the study samples and their duplicates.
Field blanks were collected at 5% of the residences and totalled 10 samples. All values
were < 1.6 /ig/dL. There were 2 concentration level controls incorporated into the unknown
samples, sent to the lab and analyzed blind. The results of these samples for both phases are
seen in Table 4-20.
All water samples were analyzed in duplicate and all were spiked and recovery of Pb
calculated. Since such a large percent of the sample concentrations fell below the detection
4-25
-------
TABLE 4-19. COMPARISON OF DUPLICATE WATER COLLECTIONS
Sample Type
Wl
W2
Wl
W2
Wl
W2
Wl
W2
•Wl
W2
Wl
W2
Wl
W2
Wl
W2
Wl
W2
Wl
W2
Wl
W2
Wl
W2
Wl
W2
TABLE 4-20.
Known
Concentration
fog/L)
10
27
Apt # Study Sample Og/L) Duplicate (/tg/L)
008 <1.6 <1.6
008 <1.6 <1.6
063 <1.6 <1.6
063 <1.6
-------
limit the average difference between duplicates was calculated using all duplicates greater
than 1.6 /tg/L. That difference was 0.4 ± 0.4 for 71 samples with a mean percent recovery
of these samples of 98 % ± 4.
4.4.8 Quality Control Evaluations for Paint
The following QC program was used for determining the concentration of lead in paint
using a portable XRF:
1. NIST reference lead film calibration checks.
2. Triplicate readings at each sampling point.
4.5 SOIL LEAD CONCENTRATIONS
Soil lead concentrations by sampling phase, sample type (surface, top and bottom) and
study area are presented in Table 4-21 and Figures 4-4, 4-5 and 4-6; Geometric mean
values, upper and lower confidence limits, 95%-tile values and number of samples are
presented. Soil abatement occurred in Area A just prior to Phase 2 sample collection (Sept.-
Oct. 1989) and just after Phase 5 sample collection (June-July 1990) in Area B. As planned,
the final soil lead sampling in Area C (Phase 09, June-Aug. 1991) occurred just prior to
Area C soil abatement.
Abatement Impact
Pie- and post-soil abatement soil lead concentrations in Area A are represented by
samples collected in Phases 00 and 02, respectively. Soil lead abatement in Area A resulted
in a decrease in the geometric mean lead concentration of 140 ppm in the composite surface
scrapings, 146 ppm in the top 2 cm composite core samples and 113 ppm in the bottom 2 cm
composite core samples. The 95%-tile soil lead concentration in the top 2 cm decreased
from 2,695 ppm to 422 ppm. The decreases are statistically significant (p <0.05) as
indicated by the 95 % confidence limits which do not overlap. Soil lead concentrations did
not decrease between Phases 00 and 02 in Area B and C where soil lead abatement did not
occur between these sample collection phases.
4-27
-------
0
I
a
c?
S
^X
r-T o\
I
Vi
o
7
g-
O
5-
o
a.
; "
vi g
S
r-1
I1
M
a
I
0\
s-
7
I
•-(
a.
S
V?
§
"
I 1 I
2> £3 3t
tTi ^r CN
t
8
i
t § § 8 S
7 T 7 « v,
en
-------
200
Phase 0
Phase 2
Phase 3
Figure 4-4. Surface scraping soil lead concentrations by Area during the
preabatement (Phase 0) and early postabatement (Phases 2 and 3) periods.
The effects of soil lead abatement in Area B, which occurred during the Aug.-Sept.
1990 period can be seen by comparing top 2 cm composite core lead soil lead concentration
from Phase 05 samples (June-July 1990) and Phase 09 samples (June-July 1991). Soil lead
abatement in Area B was associated with a decrease in the geometric mean soil lead
concentration of 102 ppm.
Building Debris
Many of the soil areas hi this demonstration project are thought to be at locations where
buildings had previously been demolished. During sample collection each of the bottom
2 cm cores was visually examined to determine if building rubble could be visually observed.
The presence of such evidence would support the hypothesis that building debris was used as
4-29
-------
250
200-
I
150-
i
£100-
(9
Area A
AreaB
AreaC
50-
Phase 0
Phase 2
Phase 3
Phase 5
Phase 9
Figure 4-5. Soil lead concentrations in the top 2 cm by Area. Abatement occurred
between phases 0 and 2 hi Area A, and 5 and 9 in Area B.
part of the "soil" and that a building may have been located there. Soil lead concentrations
may tend to be higher in such locations because of the lead-based paint that would likely be
associated with some of this debris. Soil lead samples where "rubble was observed" had
higher lead concentrations hi Areas B and C but the difference was significant only hi
Area C (Table 4-22) for the pre-abatement sampling period, Phase 00. Rubble was observed
m 42% of the samples m Area A, 30% in Area Band 39% in Area C.
SampHng Pattern
As described in the Methods section, four sampling patterns were utilized: line
(source)-on lots adjacent to suspected sources of lead such as buildings; line (area)-in the
middle of large areas; small area; and targeted areas (play areas and bare soil areas). Parcels
4-30
-------
250
200-1
CD
Area A
AreaB
Area C
I
(0
g
'o3 150-
§
8
£100
Phase 0
Phase 2
Phase 3
Figure 4-6. Soil lead concentrations in the bottom 2 cm of a 15 cm core, by area.
Abatement of Area A was between Phase 0 and 2. Data are not available
after Phase 3.
sampled by the line (source) pattern were sampled along three parallel lines (in constrast to
one for line (area) parcels).
Soil lead concentrations by sampling pattern, sample type and study area are shown for
the pre-abatement sampling Phase 00 (Table 4-23). Also shown are soil lead concentrations
for all patterns combined (previously shown in Table 4-21). Soil lead concentrations for line
(source) sampling locations were higher than for line (area) locations for all sample types and
all study areas, as hypothesized. (These differences were statistically significant except for
the top and bottom 2 cm samples in Area C.) For the top 2 cm samples, line (source) soil
lead concentrations were 25 times higher than line (area) concentrations in Area A, 63 %
higher in Area B and 26% higher in Area C. Only surface scrapings were collected from
"targeted" sampling areas. Soil lead concentrations for "targeted" areas were generally
4-31
-------
vn
W
-S
•s
M
0
d
o"
£ S
\O
\o
O
•O
t=>
IT
•s
t>0
II
4-32
-------
u
M
G1
•* o
vo £
~ -
a. ©• 50.
CO Cl. O
-
S" i? <5- -
CJ. C. CJ. 52.
? 55 S
CO CO V>
P1 <5- _.
*»• Jo ^ -l 00
JL- s
S- C-
1
vo
o\
1
CO O\ ^ O
O CS 0\ CO
0\ CO
t^ »-(
Vp
M-
a-
8
I
1
>n
g
7
&
s\
0
f
3
? 0
OO v>
I «
1 o<
S. 12
*""* r^
1-* rH
«?
1 1
w H
J
OT
0
•3
O
d,
E
6
a:
o
-U'
1
1
«;
in1 S1
o\ n
»n \n
vo t^
of of
$" V?
7 7
S :g
C- G.
g CO
01 oS
.S 2
wO 3
S °
1 t
< a
Q.
vo
4
CO
'st^
O»
M
3
a -
tn1
^»x *» »-^
S « §•
« ^ 5-
— < VO rt
t- t- JS
c- e- ti
^H g>
4-33
-------
intermediate between those for line (source) and line (area) patterns. Soil lead concentrations
for "small area" were generally the highest of the area sampling patterns for Area A where a
number of these areas appeared to have existed for many years with little disturbance but
were the lowest in Area C. (There was only one small area sample in Area B.)
4.5.1 Grass Cover
One of four categories of "grass cover" were assigned to each soil sample: fully-
covered, >50% covered, <50% covered and bare. The soil lead concentrations hi the
surface scrapings in Phase 00 samples, presented in Table 4-24 and Figure 4-7, did not differ
in any systematic way for those grass cover categories. Lead concentration hi the two
partially-covered categories were lower than in the bare locations hi Area A; hi Area C,
concentrations in the locations with less than 50% grass were lower than hi any of the other
categories. The percent of sampling areas with > 50 % of the area bare, for Areas A, B and
C was 55%, 57% and 75%, respectively.
4.5.2 Component Neighborhoods Within Study Areas
Areas B .and C each included multiple non-contiguous neighborhood components:
Findlay, Back & Dandridge for Area B and Glencoe and Mohawk for Area C. Soil lead
concentrations for surface scrapings, top 2 cm cores and bottom 2 cm cores are shown in
Table 4-25 for Phase 00 and 02. The Dandridge neighborhood of Area B was not added to
the study until Phase 01, which did not contain soil sampling. Thus the initial soil samples
for the Dandridge neighborhood were collected hi Phase 02. It is unlikely that this 3-6 mo
delay in sampling introduced any significant bias to the resultant data. Soil lead
concentrations in the Findlay neighborhood were somewhat higher than those hi the Back
neighborhood while those hi Dandridge were more than twice as high as either of these two
for all sample depths. Concentrations hi the Mohawk neighborhood were about twice as high
as those in Glencoe for the surface scrapings and top 2 cm cores but only somewhat higher
in the bottom 2 cm cores.
Building Debris. In the Findlay neighborhood of Area B and hi the Mohawk
neighborhood of Area C the geometric mean soil lead concentration in the bottom 2 cm
4-34
-------
?O
u
M
ci
OS
S
d
o
\
^ oo
00 •-<
e* is £*
I I £J
CO CO I
s
O
-1
•*
oo,
oo to oo
*-M O\ *—I
2^
oo
B
§
I
o
oo
a &
00 i-H OO
•^ \o vo
^^X ^^ VK^
O\ O\ ^f
1
<0
a
«i oj rt
> t-< t-i
Soo
A v
infi
Oi
g-
g
•a
o
o
1
II
§
s
!
i
ii
I
4-35
-------
350
300 -
250 -
I
£
•o
CO
S
<3
Grass-Covered
> 50% Grass
< 50% Grass
Bare
Figure 4-7. Distribution of lead concentrations by ground cover type and area, prior to
abatement.
samples were higher in samples where "rubble was observed" (Table 4-26), as was
anticipated.
Grass Cover. In Area A neighborhoods lead concentrations in surface sampling were
lower in locations with more than 50% grass cover than in fully-covered areas (Table 4-27).
In the Dandridge neighborhood concentrations increased as the extent of grass cover
decreased. In the Glencoe neighborhood of Area C, concentrations were higher in bare
location than in those with less than 50% grass-cover.
4-36
-------
TABLE 4-25. SOIL LEAD CONCENTRATIONS BY AREA
COMPONENT NEIGHBORHOODS (ppm)
Sample/Neighborhood
Area B Surface Scraping
Total Area
Findlay
Back
Dandridge
Area B Top 2 cm
Total Area
Findlay
Back
Dandridge
Area B • Bottom 2 cm
Total Area
Findlay
Back
Dandridge
Area C Surface Scraping
Total Area
Glencoe
Mohawk
Area C Top 2 cm
Total Area
*•
Glencoe
Mohawk
Area C Bottom 2 cm
Total Area
Glencoe
Mohawk
Q.M.
101
124
80
not sampled
103
117
87
not sampled
62
60
66
not sampled
154
109
244
140
104
206
114
103
132
Phase 00
9556-tile
776
1,270
299
not sampled
780
1,184
256
not sampled
384
566
272
not sampled
1,653
602
2,547
1,200
580
2,547
848
510
1,346
n
273
149
124
not sampled
230
132
98
not sampled
230
131
99
not sampled
311
177
134
224
128
96
217
127
90
O.M.
155
85
73
303
148
122
93
257
103
48
70
144
126
(104-152)
95
(79-113)
218
(152-311)
163
(139-191)
115
(96-138)
251
(194-324)
114
(90-145)
100
(81-123)
147
(84-255)
Phase 02
95%-tile
1,572
1,085
205
2,071
1,192
1,437
247
1,617
1,162
438
181
1,503
1,759
465
2,192
1,347
804
2,666
1,044
424
1,677
n
281
89
53
139
369
134
103
132
203
45
26
132
179
118
61
241
133
108
89
59
30
Note: Soil abatement did not occur in areas B & C during above time periods.
GM = geometric mean.
n = number of samples.
4.6 EXTERIOR DUST
Exterior dust samples were collected from several types of locations: targeted samples
near entrances of study subject homes; street, sidewalk and alley samples collected
throughout the study area; and samples collected from parking lots and other paved areas.
Both lead concentrations (ppm) and lead loadings Og Pb/m2) were determined. (Only lead
4-37
-------
TABLE 4-26. SOIL LEAD CONCENTRATIONS (BOTTOM 2-cm SAMPLES)
IN NEIGHBORHOODS BY PRESENCE OR ABSENCE OF RUBBLE
OBSERVED AT DEPTH
(Initial Soil Samples [Phase 00])
Rubble Observed
Area and
Neighborhood
A. Pendeltoa
B. Kndlay
Back
Dandridge
C. Glracoe
Mohawk
G.M. (LCL-UCL)
210 (165-267)
83 (55-124)
63 (52-76)
157 (119-206)
112 (82-152)
204 (146-283)
95%-tile(n)
1,001 (81)
1,434 (40)
156 (29)
1,503 (83)
630 (34)
1,346 (50)
Rubble Not Observed
G.M. (T1CL-UCL)
207 (153-281)
52 (44-61)
68 (57-81)
125 (89-175)
101 (82-123)
76 (50-117)
95%-tile (n)
1,861 (101)
332 (91)
325 (70)
871 (46)
494 (92)
1,167(40)
Initial toil Mtnplti collected during Phase 02. .
OM •» geometric mc«n; UCL = lower confidence limit of geometric mean; UCL = upper confidence limit of geometric mean;
a •* number of timplej.
concentration data are available at this time.) Exterior dust samples were collected for all
sample locations during Phases 1, 2, 3 and 5 but only for targeted samples for Phase 9.
Exterior dust lead concentrations are shown in Table 4-28 for all locations and for
street, sidewalks and alleys (also shown in Figure 4-8) and in Table 4-29 and Figure 4-9 for
targeted areas and for parking lots and other locations. Geometric mean concentrations in
Area C are consistently about one-fifth to one-half of those in Areas A and B. For samples
from "All Locations", concentrations were lower in Area A than in Area B for Phases 1,3,
and 5 but were about the same immediately after abatement hi Area A, Phase 02. For
"street, sidewalk and alley" samples, concentrations in Area A and B were similar except for
Phase 03 at which time they were lower in Area A. Samples from "parking lots and other
locations" were lower than samples from other locations in all study areas. Lead
concentrations in targeted samples were similar for Area A and B and were much lower in
Area C, about 20-25% of those in Areas A and B. Exterior dust lead concentration were not
expected to be altered by dust removal, since the sources of the exterior dust (e.g., from
adjacent streets, sidewalks and other paved areas as well as paint chips and dust from lead-
based painted houses) were not removed.
4-38
-------
I
1
^
1
o
s
O
o
O
§11©
I I I I
c^
-------
o
I
00
o
a
oo
-
o\
5\
O
6
ts
8 "I
00 — «,
"
I
S
«
0
6
\o
oo
c
o"
jfi
T-Hi
I
s s
CO r-J
o
VO
S
&
0
a
vo
«o
t— !
vo
ft
°
s
vo
7
£
CO
vo
"
00
I
o\
0
OO
VO
CO
g§
00
58
O\
o\
S
i
I £
I
CO
vo
00
ts
oo
S\ c,
ts
oS
VI
R
°
o\
CO
1
OO
8
VO
VO
vo
CO
a
8
rf
i
I
8
§ S?
co, 53
7
00
vo
o
cj
d
g
03
d
OJ
d
g
o\
a
?
$0^
I
V
W1
cs
ON
1
CO
*>
1
1
•s
i
1
8
o a
% II
§ e
II i
IJ
B -5,
s I
O 5
4-40
-------
2500
Phase 1
Phase 2
Phase 3
Phase 5
Figure 4-8. Exterior dust lead concentrations in streets, sidewalks, and alleys by area.
Abatement in Area A, between Phase 1 and Phase 2 snowed no apparent
effect on lead concentration.
Exterior dust lead loading (mg Pb/sq. m.) data are presented for "all locations",
"streets, sidewalks and alleys", "targeted locations" (at the exterior entryways of subjects'
residences) and for "parking lots and other locations" in Table 4-30 and Figures 4-10 and
4-11. Data for "streets, sidewalks, and alleys" are further divided hi Table 4-31 by the
roadway areas ("streets", and "alleys") and by the sidewalks of each of these roadway types,
called "sidewalks (streets)" and "sidewalks (alleys)". The exterior dust abatement in Area A,
which occurred between Phases 01 and 02, did not produce any detectable reductions in dust
lead loadings in any of the sampling locations except the residence-targeted locations where
more than a 50% reduction was observed. This reduction was not evident about 3 mo later
in Phase 03 sampling. Dust lead loadings in alleys were lower post-abatement in Area A but
they also were lower in Area B where no exterior dust abatement occurred at that time.
4-41
-------
0\
60
PM
O
tf
0
S
CO
BATE
0
ON
Si, &7
a
T
8 §•
S
VO
cs
en"
s
8
1
oo
I
o
oo
I
§
en
8 6
*
00
vo
c-f
00
8
1
vo
en
en
a
^T
vo
vo
vo
en
ts
en
vo
en
1
r
O\
8
I
o\
e
1 1
0
e-
S
VO
VO
7
oo
en
vo
«s
v>
vo
o
c
I
ov
S5
N
* S
-------
3000
Phase 1 Phase 2 Phases Phases Phases
Figure 4r-9. Exterior dust lead concentrations at targeted locations by area.
Exterior dust loading (g dust/m2), Table 4-32, revealed similar patterns as Table 4-30 and
Figures 4-10 and 4-11.
4.7 INTERIOR DUST LEAD
The primary locations where interior dust samples were collected are: at the interior
entry, a composite of several interior floor samples, and a window sill/window well sample.
Data results are expressed as both concentration (ppm) and as lead loading Og Pb/m2).
Entry data are available for Phases 1, 2, 3, 5 and 9; and floor and window data for Phases 1,
2, 3 and 5. Dustfall and door mat data will be presented separately.
Geometric mean concentrations and loading results, along with sample size and upper
and lower confidence limits of the geometric mean (for dust lead loadings) are presented in
4-43
-------
0
h
4
S
o
Fh
I I I I
ill
V.> V>V>
sgggr
G'G'Citi
v5 co oo o
co c4 cn oo
i i i i
1 1 1 1 1 | 1 1
* *
a s
gifsss
,
1 1 1 1
1 1 1 1
^o^j-to"-*
o
cs-c-c^c.
vailabl
f
a
a
I
O
i
evel of
1
o
1
1
s §
e ^
8 §
M g
" 1
sS
4-44
-------
600
S- 500
o>
o> 400
3
£ 300
I
Q
'£ 200 -
S
100 -
Phase 1
Phase 2
Phase 3
Phases
Figure 4-10. Exterior dust lead loading in streets, sidewalks, and alleys by area.
Abatement in Area A, between Phase 1 and Phase 2 showed no apparent
effect on lead concentration.
Table 4-33 for samples from the housing of initial study recruits. The sample size decreased
throughout the study due to attrition. Data for entry and floor lead concentrations and
loadings are also presented in Figures 4-12 and 4-13 (entry) and Figures 4-14 and 4-15
(floor).
Interior Entry Dust. Loadings in Areas A and B decreased by 48 % and 71 %,
respectively, between Phase 01 and 02 (pre- and post-interior and exterior abatement for
Area A and pre- and post-interior abatement for Area B) compared to a reduction of 29% in
Area C where no abatement occurred. The same decrease persisted in the Phase 03 sampling
(Nov.-Dec. 1989) for Area A but for Area B levels increased somewhat (40% decrease from
4-45
-------
600
Phase 1
Phase 2
Phases
Kgure 4-11. Exterior dust lead loading at targeted locations by area.
Phase 5
Phase 01) and were somewhat lower still for Area C (34% decrease from Phase 01).
Interior entry dust lead loadings were highest in all areas during Phase 05 sampling
(Tune-July 1990). The increase from Phase 03 was 129% in Area A, 1,164% for Area B
and 242% for Area C. Interior entry loadings for Areas A and B dropped considerably
between Phase 05 and Phase 09, one year later (50% and 68%, respectively) but increased
for Area C (118%). Thus, it appears that the exterior abatements (soil and dust) that
occurred in Area A and B may have had an impact on reducing annual increases in dust lead
that affected Area C.
Floor Dust. Floor dust lead loadings in the Phase 01 and 02 samples followed the same
general pattern as the interior entry samples with decreases from Phase 01 to 02 in Areas A
and B (63% and 81%, respectively) and 19% in Area C. Loadings in Phase 03 remained
less than in Phase 02 in Areas A and C but rebounded somewhat in Areas B (net reduction of
4-46
-------
h
X.
o
M
o
o
iii
fftf
*3s§
^0^^-
So' of So'
comets
^-c>
il!
ooooo
in o cs
I I I
1 1 1
-ci.
O O\ TJ-
O* *O *—I
^-»> m vi
0^00
vjcnt^
m m m
iir
Itlf
coi^-'«-*
3
S
occu
•o
8
1
II
e
level
d
00
"8
II -g
*-4 *
as
II
c e<
" §
CD .5
4-47
-------
TABLE 4-32. EXTERIOR DUST LOADINGS
(g/m2)
Streets, Sidewalks
and Alleys
Parking Lots and
Partially Paved
Parcels
*
Residence-Targeted
Phase
01
02
03
05
01
02
03
05
01
02
03
05
Area
(n)
(HO)
(198)
(108)
(108)
(79)
(122)
(78)
(78)
(33)
(39)
(30)
(26)
A
G.M.
127
122
138
178
494
242
719
336
97
61
169
143
Area
(n)
(192)
(242)
(191)
(190)
(30)
(19)
(28)
(30)
(74)
(22)
(79)
(68)
B
G.M.
241
250
153
187
376
230
268
175
133
59
128
186
Area
(n).
(98)
(98)
(99)
(99)
(3D
n.a.
(3D
(31)
(56)
n.a.
(51)
(62)
C
G.M.
278
200
178
188
287
n.a.
165
301
194
n.a.
155
156
48% from Phase 01 to 02). Samples collected in Phase 05 revealed loadings in Area A that
were 31% below those in Phase 012 while Areas B and C showed increases over the same
period of 151% and 8%, respectively. Samples from the final sampling phase, Phase 09
(June-July 1991) showed a similar but stronger pattern for interior entry dust loadings With
the geometric mean in Area A being 40% less than that in Phase 01 (two years earlier) while
those in Areas B and C were 142% and 390% higher, respectively. Exterior lead abatement
(soil and dust) occurred in Area B in the 2-mo period after Phase 05. Although the
immediate effect of interior abatement was higher in Area B than in Area A (where both
interior and exterior abatement occurred), the decrease persisted in Area A in Phase 03
(4 mo post-abatement) while there was an apparent rebound in Area B. A net decrease from
Phase 01 in floor loading was even evident in Area A in the Phase 05 samples (10 mo post-
abatement) while a major increase occurred in Area B and a slight one in Area C. The
loadings in all post-abatement sampling phases in Area A were lower than the pre-abatement
level. Non-abatement-related factor(s) apparently have major impacts on floor lead loading
levels as indicated by the increases (in the Phase 05 to 09 levels hi Area B).
4-48
-------
en
T—(
en
I
ti
nc
en
"3
S3
ed
03
ON i^
58
*
*
I
1
I
An
en
OO
e
Wl
s
in
vo
V-5
oo
S3
o
en
cs
1
oo
oo
•<*
en
ON
I
en
S3
^
VO OO
en en
a
en
S 8
OO
1
en
o
x-v O
en oo
en, m
a 2
CN
en
en
S
en
en
7
oo
V)
S3
ON
*T>
cs
t>
I
en
oo
e
s
e^» j^
2 i
e
o
00
es
& <*
1
x-V "
«n cs
cs «n
«n
es
vo
e
f
3 «
o
en
1
o
x^
O
ON
7
en
CN
o
en
ON
en
ON"
c
S
g
en
ON
4-49
-------
3
es
&
x3
wS
I
*
VO
I
en en
VO
00
oo
|f
oo S-
en S
!
en
o
en
n.
OO i-H
$2, cn
s i
VO
n
I
«n
I
es
a
ON
cs
I
gi
§
en
o
»n
o
en
o\
fc
vo
ON
I
i—i
en
S
VO
0
4-50
-------
u
m
I
c
u
tion
Concen
~"' I
cj o
•n D?
CO
ON *-*
^ vo"
ON VO
CM CS
8
g.
^
vo
co"
a
OO
S- o
•* ^
O vo
/-v ON
VO CO
00 1
s •<*»
ON n
CO
1
ON
cS
O
°0
1 -
ON
§
8
VO
0
o
T—(
co"
§ I
00 I
CO •<*
5 o
1
0
vo
oo
<~^ CO
si
cs" ">.
eQ
o V
CO O
~ CO
7-
•a
$
I
C3
d
8_r
4-51
-------
800
700
a 600
£
• Area A
H AreaB
AreaC
600
400
S 300
•E 200
100
Phase 1 Phase 2 Phase 3 Phases Phases
Figure 4-12. Interior dust lead concentrations at the entry way.
1400
Phase 1
Phase 2
Phases
Phases
Phase 9
Figure 4-13. Interior dust lead loading at entryway by area. Abatement occurred in
Areas A and B between Phase 1 and Phase 2.
4-52
-------
500
o
Phase 1 Phase 2 Phased
Figure 4-14. Interior dust lead concentration on the floor.
Phases
250
Phase 1
Phase 2
Phase 3
Phase 5
Figure 4-15. Interior dust lead loading floor by area. Abatement occurred in Areas A
and B between Phase 1 and Phase 2.
4-53
-------
Window Dust. Window dust loadings decreased between Phases 01 and 02 in Areas A
and B by 56% and 78%, respectively, while in Area C a decrease of 26% was observed. In
Phase 03 further reductions (from Phase 02) in Areas A and B of 37% and 19%, respectively
occurred. In Area C, where no abatement occurred an 87% reduction was observed. For
samples collected in Phase 05, very large increases occurred in all areas with loadings being
6.4, 6.4 and 4.6 times those in Phase 01 for Area A, B and C, respectively. Since these
increases occurred in all areas they are not associated with abatement and are evidently due
to annual variation or yet unidentified other factor.
Reduced Data Set. In order to examine changes between phases for the same s&t of
housing, a reduced set of the data was analyzed. For a given pair of sampling phases, say
01 and 02, only the data from housing sampled in both phases was used. Interior entry dust
lead values (concentrations and loadings) are presented in Table 4-34 (and Figures 4-16 to
4-19), interior floor dust in Table 4-35 and Figures 4-20 and 4-21 (Area A) and Figures 4-22
and 4-23 (all areas) and interior window dust in Table 4-36. Patterns of change are similar
to those noted in the total data set but some differences are evident. Interior dust loading
decreases were still observed in Area A and B between Phases 01 and 02 and between 01
and 03. For a comparison of Phases 01 and 05 samples and 01 and 09 samples,loadings
were slightly lower in the later phase samples for Area A but were much higher for Areas B
and C. This pattern of changes may be due to overall annual increases which are meliorated
by the simultaneous interior and exterior abatement activities in Area A.
Interior floor lead loading reductions coinciding with abatement were evident in Area A
and appeared to persist through Phase 05, some 10 mo post abatement (Table 4-35).
A decline in Area B was only evident through Phase 03, 3 mo post-abatement; and in Area C
little change in loadings were observed.
Window dust lead loadings followed the same pattern as for the whole data set with
large increases between Phases 01 and 05 in all areas and a large reduction in Area C
between Phase 01 and 03.
Changes in geometric mean interior dust lead loading changes are summarized hi
Table 4-37. Changes are also expressed adjusting for the changes observed in the control
area (C) (Table 4-38).
4-54
-------
et
U
m
g
u
1
I
U
OO »—I
r>^ '
10 5?,
! So-
1
1
o
§f
en 1-4
CM CS
en CM
oo
0 CM
«n
00 CS
1
oo
CM,
ON
CM
00
Pen
o
JO cs
oo
OO ON
Si
vo
38
P
I
/-v
IT! -
^ I
O ON
s
le
OO
I
VO
<=>
00
CN
o o
i
0
.
co en
s
§s
<* ^1
»—i es
-
vo
22' §8
S 00
»n o
>~' I
^H ON
ga
f} —j
o 1
vo ts
CM 0
•T
ON
T3 -0
4-55
-------
800
Area A
AreaS
AreaC
figure 4-16. Comparison of interior entryway dust lead concentrations between
Phases 1 and 2. Dust abatement occurred between Phases 1 and 2 in
Areas A and B.
Interior dust loadings (mg dust/m2), Tables 4-39 and 4-40, show similar patterns for
dust lead loadings. About 10 mo after abatement, floor dust loadings in Area A were still
43% below preabatement levels while in Area B they were 65% higher than preabatement
levels.
4.7.1 Net Change in Lead Loading
Entry Dust
For the entry dust loading (Table 4-34) there is a significant reduction between
Phases 01 and 02 (before and after abatement) in both Areas A and B (where abatement
occurred) and an insignificant decrease in control Area C where no abatement occurred.
Comparing Phases 01 and 03 (before abatement and about 3 mo post-abatement) loadings are
about 50% lower in Phase 3 than in Phase 1 for Areas A and B but the confidence intervals
overlap. Comparing Phases 01 and 05, Area A showed no differences but Phase 05 levels
4-56
-------
500
<3
Area A
AreaB
AreaC
Figure 4-17. Comparison of interior entryway dust lead loading between Phases 1
and 2. Dust abatement occurred between Phases 1 and 2 in Areas A
andB.
were almost as high in Area B and more than 2X as high in Area C. Evidently conditions
across study areas, independent of abatement, occurred to increase dust lead loadings.
Either these conditions did not reach Area A (doubtful since it is adjacent to part of Area B
and within one-half mile of other areas) or the exterior abatement that occurred along with
interior abatement in Area A reduced its impacts.
Comparing entry dust loading results for Phase 09 with those for Phase 01 revealed
loadings about 10% higher in Area A, over twice as high in Area B and 3.5 times as high in
Area C.
Floor dust loadings (Table 4-35) decreased between phases 01 and 02 for both Area A
and B by about 60% and 80%, respectively, but less than 5% in Area C where no interior
abatement occurred. For Area A the reduction was about the same for Phase 01 to 03 but
for Area Bit was less than 50%. When comparing Phase 05 results with those from
Phase 01, loadings were almost 40% lower in Area A, but had increased more than 2-fold in
Area B and by about 10% in Area C.
4-57
-------
700
Area A
AreaB
AreaC
figure 4-18. Comparison of interior entryway dust lead concentrations between
Phases 1 and 5. Dust abatement occurred between Phases 1 and 2 in
Areas A and B.
Window Loadings
Results for window lead loadings (Table 4-36) revealed that significant decreases from
Phase 01 to Phase 02 in Areas A and B indicating an apparent abatement effect. Comparing
Phases 01 and 03 revealed significant decreases in all areas. Comparing results from
Phase 01 and 05 revealed large increases (3-6-fold) in all areas: Evidently, events
independent of abatement had a major impact on window dust loadings. Geometric mean
window dust concentrations exhibited only relatively minor variations.
Percent changes hi window dust, entry and floor dust lead loadings between phases for
the matched sets of housing are shown hi Table 4-37. If factors involved hi changes
occurring hi control Area C are assumed to be operative hi Area A and B, then an
adjustment hi changes hi Areas A and B may be able to be made to estimate the "net
abatement effect". Results shown hi parenthesis hi Table 4-37 are one attempt at such
4-58
-------
Area A
AreaB
AreaC
Figure 4-19. Comparison of interior entryway dust lead loading between Phases 1 and
5. Dust abatement occurred between Phases 1 and 2 in Areas A and B.
adjustment. If such adjustments are valid, an abatement impact may exist up through the end
of the study.
4.8 BLOOD LEAD
Blood lead concentrations for the initial recruits is shown for the three study areas and
for the five blood collection phases in Table 4-41. The geometric mean ranged from 8.02 to
10.45 in Phase 01 prior to abatement. For most of the study the geometric mean PbB was
lower in Area C (the control area) than in the two other areas. However, results from the
last blood collection revealed that concentrations in all three areas were almost identical.
A comparison of blood lead levels between phases, using only data available from the
same children in both phases (Table 4-42) reveals the same patterns as for the larger data set
show in Table 4-41.
The geometric mean of the blood lead ratios between phases (Table 4-43) reveals that
geometric mean .of the ratio of Phase 03 blood lead level to Phase 01 blood lead was lower
4-59
-------
U
w
W)
a
8
bJO
03
o
U
d>
P
oo
a
ON
*
*
*
oo
A
&
£
«n
^2
&
§?
v-i en
en ts.
en,
vo
en
1
I
CS
vo
vo
I
00
oo
en
f,
S
en
en
en
T
§
en
oo
en
3
cs
en
s
en
ON" oo
cs
vo
vo
•<*
en
S
vo
I
S 55
s i
S
a f
$ »
en ^
ss
CM
es
en
en
T
ON
OO
ON en
^ I
& «>
If
en
%.> en
«n p
cs
oo
8
>S
4*
^3.1
'..1
If
o §
8
M
^* O
1
.3
8^
on"S
S -S to
a-§^
4^-60
-------
600
I
500-
Phase 1
Phase 2,3 & 5
Phase 1 & 2 Phase 1 & 3 Phase 1 & 5
Figure 4-20. Pre- and postabatement interior floor dust lead concentrations in Area A.
in Area C (where no abatement occurred) than in Area B (interior abatement). Thus, blood
lead levels dropped more in Area C. The geometric mean of the ratio in Area A was slightly
lower in Area A than in Area B.
The mean decrease in blood lead level between phases (Table 4-43) revealed that the
decrease in Area C was 0.69 /ng/dL more than that in Areas A and B. Between Phases 01
and 09, blood lead levels decreased by 2.09 in Area B and 1.30 in Area C compared to an
increase in Area A. The difference between Areas A and B was statistically significant.
Since the subjects compared in Table 4-43 were active in the study in both phases
comparisons based on Table 4-43 should be given more weight than those from the data in
Table 4-41 which includes subjects in earlier phases who are not represented in kter phases.
Age. The.preabatement (Phase 01) blood lead levels of the children in the study was
highest for children between two and three years of age (Table 4-44 and Figure 4-24) and
tended to level off at about 8 jtg/dL at older ages. For children living in non-rehabilitated
housing, the peak, although higher by a geometric mean of 11 /tg/dL, was also between
2 and 3 years of age. The ratio of blood lead levels of Phase 05 to Phase 01, and of
4-61
-------
300
Phase 1
Phase 2.3 & 5
Phase 1 &2 Phase 1 &3 Phase 1 &5
Kgure 4-21. Pre- and postabatement interior floor dust lead loading in Area A.
600
Phase 1
Phase 2,3 & 5
1&2 (A) 1&3 (A) 1&5 (A) 1&2 (B) 1&3 (B) 1&5 (B) 1&2 (C) 1&3 (C) 1&5 (C)
Figure 4-22. Pre- and postabatement floor dust lead concentrations, all areas.
4-62
-------
300
Phase 1
Phase 2,3 &5
1&2 (A) 1&3 (A) 1&5 (A) 1&2 (B) 1&3 (B) 1&5 (B) 1&2 (C) 1&3 (C) 1&5 (C)
Figure 4-23. Pre- and postabatement floor dust lead loading, all areas.
Phase 09 to 01, are shown in Table 4-45 and Figures 4-25 and 4-26. For children at least
one and less than three years old at Phase 01, geometric mean blood lead levels were less at
Phase 05 for all three study areas.
4.9 HAND LEAD
Hand lead values by study group and study phase are presented in Table 4-46. Prior to
any abatement activity (Phase 01) hand lead values were lower in Area C than in either
Areas A or B with only the difference between Areas A and C being statistically significant.
Immediately post-abatement in Area A (interior and exterior abatement) and in Area B
(interior abatement) the difference in hand lead between areas was no longer statistically
significant. At about 3 mo post-abatement (Phase 03) hand lead values in Area C were
significantly lower than those in either Areas A or B. No further statistically significant
differences in hand lead occurred for the balance of the study.
4-63
-------
u
«
Rt
I
I
en
£ I
1 en
" jn
CS,
§
oo
en
OO
en
en
o\
«n
oo
en
cs
oo
en
00
TH
VO
/ s
in en
C^ °\
s 7
c^ «n
*— I OO
en
0
»n
cs
oo Y
en I
O ON
en
vo
a\
o
ON
I
P
in,
en
«n
m
00
cs
ON
a
t>
es en
» f
O ON
8
00
o
en
•«**
ON
.
00 ,
vo I
T-I VO
- 6
O
CN
cs
00
VO
ON
00
en
^ °
s T
O ON
"
/-v OO
CS T-<
in
'
vo
ON
vo
es
T-I
I
o
0
en
7
1
m
en
oo
es
I
£
ON
^
en
'
i>
en
-------
o
«
I
8
1
Window
&
Window
S
1
co en
oo in
oo
00
cs
S
VO
oo
en
O
c-
T-* en
en
o
en
cs
en
cs en
+ ^
oo cs
en en
S
00
o ^
en
=
S
xTX
00
8
.S
II
o
H
8s
'•3 M
oo w>
"§ S
P o
5 S
1
5 §3
-S SP
4-65
-------
No Abatement
2% Reduction
erior Abatement
8
o
o
I
if
m §
8-S
SI
S
A
Reduction
*
oo
•I
3
S
^*
1
r§
H-<
^
£n
Reduction
^
vo
cs
1— (
a
•1
a
2
^^
OO
vo
!
13
P^
^
vo
g
A
T— ^
o
§ ^
o '"'
c£ ft
ra
*^
•r^
cs
©
A
r— (
O
v
* *
4-66
-------
TABLE 4-39. INTERIOR DUST LOADINGS
Entry
Floor
Phase
01
02
03
05
09
01
02
03
05
Area
G.M.
1,240
340
410
1,130
444
490
200
180
280
A
00.
(32)
(33)
(3D
(25)
(18)
(32)
(32)
(31)
(26)
Area
G.M.
440
90
180
2,050
595
310
60
130
510
B
(n)
(64)
(64)
(60)
(42)
(32)
(66)
(66)
(60)
(45)
Area
G.M.
610
380
270
840
1,208
200
160
160
250
C
(n)
(41)
(39)
(36)
(32)
(22)
(42)
(40)
(35)
(33)
TABLE 4-40. INTERIOR DUST LOADINGS
Window
Mat
Phas
01
02
03
05
01
02
03
05
Area
G.M.
740
550
260
3,110
100
550
580
570
A
(n)
(32)
(30)
(29)
(26)
(31)
(33)
(28)
(25)
Area
G.M.
1,340
470
270
6,410
100
300
530
1,320
B
(n)
(64)
(60)
(60)
(45)
(60)
(64)
(57)
(39)
Area
G.M.
1,640
1,470
420
5,510
90
210
380
360
C
(n)
(42)
(39)
(35)
(33)
(40)
(40)
(36)
(32)
Over the course of the two year study, hand leads changed from being lower in Area C
to being similar in all three areas. The abatements implemented in Areas A and B may have
had a role in "equalizing" the lead exposures among the three study areas. In Table 4-47
and Figures 4-27 and 4-28 a reduced set of data is used to compare hand lead values between
pairs of phases using only data that are available for children active in both phases. In this
paired phase analysis, the geometric mean hand lead for Area C remained significantly lower
than that for both Areas A and B through Phase 05; for Phase 06 it was lower than that of
Area A, for Phase 07 it was lower than that for Area B and for Phase 09 it was lower than
4-67
-------
3
O
ON
8
VO
S
vi
tn
CO
I
VO
CO
f
CO
I
00
CO
ON
»n
vo
s
CO
^, .2
w 2
CQ
CO
a-
Abate
S
o\
8
CO
0
»-I
1
0
I
8.
c>
o\
vo
00
I
5?
ON
ON
3
g
§
s
00
in,
0
^— c
1
T-4
VO
ON
o
T<^
1
5
s
s
e>
ON
CO
00
ob
CO
s
CO
CO
o\
ON
CO
ON
:
1
8
00
S
ON
0
4-68
-------
o 3
IS
I
s
o
o\ vo
I I
§§
r< o
•~t »-<
I I
S ~"
50-52.
a?
00 p^
O» 00
I I
ci
-
sS-ss.
7
c*i
•* CJ
o oo
'?
gg
I I
Q-
fsr
o o
*— 1 *— 1
1 I
S
t— 00
I I
CO CO
vxjd.
CO «0
CO CO
Ov Ov
I I
iff
Sg!
t~ 00
SS
Ov O\
I I
VJ O\
ui vd
ss
t~ oo
?!?
O\ O\
j^ i
Ov Ov
5S.C.
32
I 1
2g
t-c.
J
1
I i
I I
"° •§
I S
II
g
S3
m
M
1
<3
_o
i
>j s
I i
s 1
o iS
4-69
-------
S
»-< Tf- in vo
oq o o O
O t-H »—I t—I
I I I I
O
CD
»-< VO t~- OO
^- OO 00 OO
d d o" d
O tr\ /[ ^.
°° t~- c^ i>
P. d d d
0
VO VO O "*
OO 00 OO OO
d d d d
O\ CO CO CO
• • • •
O T-H T-< T-H
1 1 1 1
•— *
|
S
|
3
s
V— '
i
s
s
J
0<
^" ^J" ^"" O\
^" Vi ® O
t d 9 ®
2 2 2 S
cp cjl c^ to
S-S.52.Ci
V( vo O
«?9TV
x-,^ ^
•f 9-r 9
2222
—i i-c vo VI
VO 00 f; tO
*?'?'?'?
g--% -— \ /.-^
5£- %Q- S-
§^^ vo •*
VI TT O
*r *r ^ «?
9 c-i c4
°3 OO C*- vo
T d o o
~ ^< _ rf
2222
tO Vj (""• O\
00
0)
1
s
•I
3
C
II
a
tJ
g
o
•s
6
o
-------
CL-UCL
d
»n ON
»n o CM o ON CM
c9 CM CO ^3" *O *sD
•5 V V V V V
2 2 S 2 3
i— i cs co rf >n
00 •
00 E^
i 7
00 fi.
CO CM
ii
OO
O
o i— i «n i— i
CM VO r}- vo
*-i '-i C*> OO
T— < CM T— ' «— '
o3 <3 13
•5 -S ^S
O O O
^ rg
O O
9 CM co ^t m vo
•S V V V V V
"22322
H-5 T-I CM CO rt in
4-71
-------
Rehabilitated
Non-rehabilitated
1to<2 2to<3 3to<4 4to<5 5to<6
Child Age (years)
figure 4-24. Effect of child age and house type on blood lead.
the geometric mean for both Areas A and B. Thus the overall conclusion is the same as that
previously stated for the more complete data set: hand lead values were initially lower in
Area C than for the other study areas but at the end of the study they were not statistically
significant than geometric mean values in the other areas.
Hand lead comparison by study group between specific pairs of phases (e.g., 1 and 2),
for children active in the study for both phases, are shown hi Table 4-47 and Figures 4-27
and 4-28. Only Area A showed a decline in mean hand lead between Phases 01 and 02
(Figure 4-28); all areas showed decline between Phase 01 and 03.
4.10 INTERCORRELATIONS
Intercorrelation between blood lead and hand lead and environmental dust, paint and
age variables are presented in Tables 4-48 to 4-51 for Phases 01, 03, 05 and 09,
*
respectively. Only correlations significant at p < 0.05 are shown. The number of
4-72
-------
o
.3
CJ
0
*o
4f
V
I
CO CO
en «-<
1 1
en t- T}-
o
vo
Tf CO VD O\ •*
oo oo vo t- vo
'-i O O O O O
»-< r-t O\ O\
^-i .-i rt o O O
1 1 1 1 1 1
O\ •* o ts Os
o\ vo vo t- vo N
*n ^ G" ,~>
«-( VO TM OO
»-« o o o o o
en «A o vo
•* vo o\ oo
OS CO O\ CO O\
o\ •<*• o\ oo °
vo oo
o\ p
o o c>
1
cs cn
V V
•* vo vo vo •*
5 OO >-i t— OO V5
r4 o ^ C> C3 0
O t-; r} O\ f;
Ti- TH »-^ O *"-<
CM O\ IS C< O\
cn ^* vo vo Vi
(S> T-< O\ VO ^H
en o\ «-i i> o
ts o
o >-<
J2 «s en •*
-------
Area A
H AreaB
AreaC
1to<2 2to<3 3to<4 4to<5 5to<6
Child. Age (years)
Figure 4-25. Impact of abatement on children living in rehabilitated housing; Phase 5
versus Phase 1.
<1 1to<2 2to<3 3to<4 4to<5 5to<(8
Child Age (years)
Figure 4-26. Impact of abatement on children living in rehabilitated housing, Phase 9
versus Phase 1.
4-74
-------
PQ
* $ s
O
CJ
o"
s
o
fi
tn
8
d
00
I
00
o\
!
CO
ON
e
VO
00
c
00
Jl^
<0
s
vd
VD
I
VO
o\
C*
o\
o\
VO
3
PX
52-
00
o\
8
vo
I
o
0
o\
"1
VO,
10
cs
en
10
oo
I
%
VO
vd
00
v
. 00
<
8-
vo
O
vo
e
a
a
a
0\
tn,
-
tn,
00
en
vd
3
6
o\
vo
•N
00
T
00
o\
ON
s
:e li
S
d
4-75
-------
o
«
O
s
o"
o"
fi
5,5,
en f-
ff
T-I 00
o
CN
en
2o?
Tf C*
O\ 00
f- VO
O CO
00 VO
1 1
eo en
es tn
52,!
tn M
1 1
tn es
en eo
en
00*00
yxH">.
S3
O\ vo
I I
VO T-I
yS. •*
,,
cj en
o\ vo
1 1
o\ vo
c*
-------
Area A AreaB AreaC
Figure 4-27. Comparison of hand lead loading between Phase 1 and Phase 3.
Area A Area B Area C
Figure 4-28. Comparison of hand lead loading between Phase I and Phase 2.
4-77
-------
TABLE 4-48. ESTERCORRELATIONS BETWEEN BLOOD LEAD AND HAND
LEAD AND ENVIRONMENTAL LEAD MEASURES AND AGE
(Pearson Correlations)
FOR OVERALL STUDY POPULATION AND BY AREA
(Phase 01)
(arithmetic)
PbB
PbH .
Interior
Entry Dust
Loading
Interior
Entry Dust
Concn.
Interior
Floor Dust
Pb Loading
Interior
Floor Dust
Pb Concn.
Interior
Window
DustPb
Loading
Interior
Window
Dust
Concn.
Interior
Mat Lead
Loading
Interior
Mat Concn.
Exterior
Dust Max
Cone.
Overall
1.00
0.47
0.0001
202
0.27
0.0003
181
0.21
0.003
190
0.18
0.01
186
0.27
0.0002
188
0.21
0.004
180
Blood Lead
ABC
1.00 1.00 1.00
0.52 0.44 0.28
0.0001 0.0001 0.04
51 95 56
0.55 0.21 0.36
0.0002 0.05 0.01
40 91 50
0.45 0.37,
0.002 0.006
44 57
0.55 0.34
0.0001 0.01
44 52
0.33 0.27
0.03 0.01
45 90
Hand Lead
Overall
0,47
0.0001
202
1.00
0.41
0.0001
193
0.39
0.0001
186
0.45
0.0001
195
0,48
OiOOOl
195
0.35
0.0001
191
0,34
0,0001
193
0.35
0.0001
185
A
0,52
0.0001
51
1.00
0,62
0.0001
40
0.85
0.0001
44
0.93 ,
0,0001
44
0,69
0.0001
45
0.73
0.0001
43
I
B
0.44 |.
0,0001
95
1-00
0.50
0.0001
95
0.57
0.0001
94
0.54
0.0001
96
0.31
0.002
96 ,.
0.26
0,01
92 :
0.34
0.001 !
93
Q.31
0.003
89
C
0,28
0.04
56
1.00
i
4-78
-------
TABLE 4-48 (cont'd). INTERCORRELATIONS BETWEEN BLOOD LEAD
AND HAND LEAD AND ENVIRONMENTAL LEAD MEASURES AND AGE
(Pearson Correlations)
FOR OVERALL STUDY POPULATION AND BY AREA
(Phase 01)
(arithmetic)
Blood Lead
Age
Age-
Squared
Paint-
Interior
Trim (Max)
Paint-
Exterior
Trim
Overall A
0.34
0.0001
202
0.25
0.0003
202
B
0.44
0.0001
95
0.36
0.0004
95
0.29
0.01
70
0.26
0.03
70
C
0.41
0.002
56
0.30
0.03
56
Overall
0.34
0.0001
202
0.31
0.0001
202
Hand Lead
A
0.35
0.01
51
0.36
0.009
51
B
0.34
0.0006
95
0.29
0.004
95
C
0.38
0.003
56
0.33
0.01
56
Top number is correlation coefficient, middle is significance level and bottom is number of observations. Only
correlations with, statistical significance <0.05 are shown.
Phase 01 is pre-abatement.
significant correlations decreased markedly after the abatements that occurred after Phase 01.
Table 4-52 contains all correlations for Phases 01-09, regardless of their statistical
significance, and using log transformed data. Correlations between blood lead and hand lead
are summarized in Table 4-53.
A comparison of preabatement intercorrelations among environmental and blood lead
data between the Cincinnati prospective study (Clark et al., 1991) and those of the soil
project (Table 4-54) revealed similar patterns except for paint lead (XKF, mg Pb/cm2). The
prospective study invalues a wide range of housing types from rehabilitated housing to 19th
century dilapidated housing while the soil project included primarily rehabilitated (lead paint
free) housing. As expected intercorrelation involving paint lead were not statistically
significant for the soil project while they were for the prospective study.
4-79
-------
TABLE 4-49. INTERCORRELAHONS BETWEEN BLOOD LEAD AND HAND
LEAD AND ENVIRONMENTAL LEAD MEASURES AND AGE
(Pearson Correlations)
FOR OVERALL STUDY POPULATION AND BY AREA
(Phase 03)
(arithmetic)
Blood Lead
Blood
Lead
Hand
Lead
Overall A
1.00 1.00
0.51
0.0001
B C
1.00 1.00
0.62
0,0001
Hand Lead '
Overall A
0.51
0,0001
151
1.00 1.00
B
0.62
0.0001
75
1.00
C
1.00
151
75
Interior
Entry
DustPb
Loading
Interior
Entry Dust
Pb Concn.
Interior
Floor Pb
Dust
Loading
Interior
Floor Dust
Pb Concn.
Interior
Window
DustPb
Loading
Interior
Window
DustPb
Concn.
Interior Mat
Pb Loading
Interior Mat
Pb Concn.
Exterior
Dust Pb
Concn.
Max
0.33
0.05
35
-0.23
0.05
76
0-23
0.009
130
4-80,
-------
TABLE 4-49 (cont'd). INTERCORRELATIONS BETWEEN BLOOD LEAD
AND HAND LEAD AND ENVIRONMENTAL LEAD MEASURES AND AGE
(Pearson Correlations)
FOR OVERALL STUDY POPULATION AND BY AREA
(Phase 01)
(aritfametric)
Blood Lead
Paint Pb
Interior
Trim Max
Paint Pb
Interior
Wall Max
Paint Pb
Exterior
Trim
Paint Pb
Exterior
Walls
Age
Age-
Squared
Overall A
0.22
0.01
133
0.28
0.0006
151
0.22
0.006
151
B C
0.37
0.003
62
0.31
0.01
62
f
0.31
0.02
59
0.38
0.0009
75
0.33
0.004
75
Hand Lead
Overall A
0.46
0.0001
134
0.22
0.01
135
0.37
0.0002
135
0.34
0.0002
119
0.32 0.38
0.0001 0.02
151 35
0.28 0.35
0.0005 0.04
151 35
B
0.61
0.0001
63
0.58
0.0001
63
0.57
0.0001
63
0.57
0.0001
60
0.28
0.01
75
0.24
0.03
75
C
0.55
0.0002
41
0.49
0.001
41
Top number is correlation coefficients, middle is significance level and bottom is number of observations.
Only correlations with statistical significant 0.05 are shown.
Phase 03 is about 3 mo after abatements hi Area A (exterior and ulterior dust and soil) and in Area B (ulterior
dust only).
4.11 MODELING
4.11.1 Modeling the Difference of Phase 1 and Phase 5 Blood Lead for the
Initially Recruited Families Who Lived in the Rehabilitated
Housing Units
We evaluated the effectiveness of soil lead and dust abatement through analyzing the
data collected before and after initial abatement, e.g., Phase 1 and Phase 5. In Area "A",
exterior dust (lead concentration), interior dust (lead concentration) and soil lead were abated
4-81
-------
TABLE 4-50. INTERCORRELATIONS BETWEEN BLOOD LEAD AND HAND
LEAD AND ENVIRONMENTAL LEAD MEASURES AND AGE
(Pearson Correlations)
FOR OVERALL STUDY POPULATION AND BY AREA
(Phase 05)
(arithmetic)
Blood
Lead
Hand
Lead
Interior
Floor
DustPb
Loading
Interior
Floor
DustPfa
Concn.
Interior
Entry
DustPb
Loading
Interior
Entry
DustPb
Concn.
Interior
Window
DustPb
Loading
Interior
Window
DustPfa
Concn.
Exterior
DustPb
Concn.
Max
Age
Blood Lead
Overall ABC
1.00 1.00 1.00 1.00
0.34 0.51
0.0001 0.0001
149 68
6.17
0.04
147
j,
0.19
0.02
146
0.16 0.31
0.05 0.04
145 143
Hand Lead
Overall A B C
0.34 0.51
0.0001 0.0001
149 68
1.00 1.00 1.00 1.00
r
1;
'
0.28 0.32
0.02 0.04
68 42
1
!
0.17
0.04
145
0.27 0.33 0.25 0.44
0.0007 0.05 0.04 0.003
149 37 68 44
4-82
-------
TABLE 4-50 (cont'd). INTERCORRELATIONS* BETWEEN BLOOD LEAD
AND HAND LEAD AND ENVIRONMENTAL LEAD MEASURES AND AGE
(Pearson Correlations)
FOR OVERALL STUDY POPULATION AND BY AREA
(Phase 05)
(arithmetic)
Age-
Squared
Interior
Mat Dust
Pb Loading
Interior
Mat Dust
Pb Concn.
Interior
Wall Paint
PbMax
Interior
Trim Paint
PbMax -
Exterior
Paint Pb
Overall
0.18
0.03
145
0.41
0.0001
137
0.22
0.007
146
Blood Lead
A B
0.40
0.002
59
0.30
0.01
68
0.36
0.003
68
0.36
0.03
35
Hand Lead
C Overall ABC
0.28 0.38 0.45
0.0007 0.02 0.002
149 37 44
0.41 0.28 0.49
0.006 0.0008 0.0001
43 137 59
0.34
0.05
34
Top number is correlation coefficient, middle is significance level and bottom is number of observations. Only
correlations with statistical significance 0.05 are shown.
Phase 05 is about 10 mo past abatement in Area A (ulterior and exterior dust and soil) and in Area B (ulterior
dust only).
after Phase 1. In Area "B", interior dust was abated after Phase 1. In Area "C", no
abatement occurred during this time period.
Based on our data analysis, we found that the mean difference of blood lead between
Phase 1 and Phase 5 (DPbB) in Area "A" was greater than that in Area "C", e.g., DPbB in
Area "A" was 2.13 mg/dL more than in Area "C" (i.e., blood leads increased in Area A
relative to Area C). The mean difference of handwipe lead between Phase 1 and Phase 5
(DPbH) had no significant influence to the difference of blood lead between Phase 1 and
4-83 •
-------
TABLE 4-51. INTERCORRELATIONS BETWEEN BLOOD LEAD AND HAND
LEAD AND ENVIRONMENTAL LEAD MEASURES AND AGE
(Pearson Correlations)
FOR OVERALL STUDY POPULATION AND BY AREA
(Phase 09)
(arithmetic)
Blood
Lead
Hand
Lead
Blood Lead
Overall ABC
0.28 0.46
' 0.003 0.01
Hand Lead
Overall A B
0.28
0.003
109
>
! c
0.46
0.01
30
i
1
109
30
Interior
Floor
DustPb
Loading
Interior
Hoor
DustPb
Concn.
Interior
Mat Dust
Pb Loading
Interior Mat
DustPfa
Concn.
Interior
Window
DustPb
Loading
Interior
Window
DustPb
Concn.
Exterior
DustPb
Concn.
Max
Interior
Wall Paint
PbMax
0.41
0.03
29,
0.22
0.02
107
0.56
0.0001
50
4-84
-------
TABLE 4-51 (cont'd). INTERCORRELATIONS BETWEEN BLOOD LEAD
AND HAND LEAD AND ENVIRONMENTAL LEAD MEASURES AND AGE
(Pearson Correlations)
FOR OVERALL STUDY POPULATION AND BY AREA
(Phase 09)
(arithmetic)
Overall
Interior 0.36
Trim Paint 0.0001
Pb Max 107
Exterior 0.36
Trim 0.0001
Paint Pb 107
Exterior
Window
Paint Pb
Age
Age-
Squared
Blood Lead
ABC
0.50
0.0002
50
0.48
0.0005
50
0.53
0.0002
45
Hand Lead
Overall A B C
0.21 0.28
0.03 0.05
109 51
0.20
0.04
109
Top number is correlation coefficients, middle is significance level and bottom is number of observations. Only correlations with statistical
significance 0.05 are shown.
Phase 09 is about 22 mo post abatement in Area A (interior and exterior dust and soil) and about 22 mo post interior dust abatement in
Area B and about 10 mo post exterior dust and soil abatement in Area 6.
Phase 5. The mean difference of interior house dust lead loading between Phase 1 and
Phase 5 (DPbDIMD) significantly affected the difference of blood lead between Phase 1 and
Phase 5. In other words, in Area "A", DPbB increased 2.13 mg/dL for per mg/m2 increase
in DPbDIMD; in Area "B", DPbB decreased 0.000001 mg/dL for per mg/m2 increase hi
DPbDIMD; in Area "C", DPbB increased 0.000089 mg/dL for per mg/m2 increase in
DPbDIMD. Age and social-economic scores were also important factors in predicting the
difference of blood lead, e.g., DPbB decreased 0.082 mg/dL for per month increase hi age.
DPbB decreased 0.22 mg/dL for per unit score increase in Hollingshead social economic
score (SES). The initial blood lead affected the difference of blood lead between Phase 1
and Phase 5. For high initial blood lead children, DPbB was smaller than that for low initial
4-85
-------
I
£
ssl:
§<;
&
o
y
o
o
* * *
o o
, ^
°
- o >
S 8 2
i
1
i 8 °- "
o >^ o ^
t; S*
O>o
S
o
3
P
o
o
O\ ^s Q\ ^~^ O\ iTT^ T™1
r-
-------
o
s
tt
O
s
o
s
o
K
O
1
EL,
00 x— v C
j Si-o
y
8
/
o 8 S
^0-6
2 to S vTiS o?
. O\ . t— . vo
O *— ' O *~' O *Zr .
13 2 ^
^ S\ '-J V) »-< - o
oo c? 2; 0*00 f-^
»— i ON ^*
-------
o
©
§
3
M
co
1
3
•a
o
eo
°
S 8 S 8
1*1
S -a -3
Z* o
* x— v VO x-v O\
•*VOcST-«o
vo*o«r*--
.!iJ-O''i'O
o
s s .8
9 & d S- d & 9 52-
9 Sd O-d 2J--p
>-;•<*• ^ o
r P vi P &
S3 3 §
I
4-88
-------
g
s
H
S
4g
a§
o
M
OO /— s r-l
o\ ^~* O ON c*$ o\
x ^-* x ^ x *"*
<3 ^^/ ^5 \^^ ^3 ^^^
/ V
r- vo »-< oo o\ m
<-< oo ts •*>-<•* oJ
_• >-< • *-( • i— i
O **— ^ O ^^ O v— '
»-i en «o o\
o o o o
i §
'6 --
1 s
& Q
oo
»-< oo
en
S 8
S
at W)
'^ trt
Scd
Q
Is
O\ ^™v ^ ^* •* C**
i 55 ^ ^ *1
o ^ 9 -2^ o
x
o\ *-< «n '-' o\
So 52- o
*— v.
Tf JO c<1 of
en
o
o o
Q
2 •
s S
fe «
•§ rS
R U
4-89
-------
pEJ
§
S
*
»-£
sal
ISli
Bit1
1
S
o
o
pq
G-oCJ-
..>n 1 en o ^ q o
" §
8
S iR Cr m «
o in o co ^
Ot— < *-X t— ( |-*v
v_x tJ x_, O
en
O
0 -0
s
1
S ^*?3
^
eo
vo OT
-
9 C- 9 H o £, d
8 8
•i
.»
•i
£
s
4-90
-------
o
6 S- 6 & o ^ 9
PQ
f ^-v "^
C tri i-«
S
w
o
8x-v Jg /-v
»-i O 00
c> Sq> G-
> CO /•«,
i-i O\
m
2^2-l^£
O^—'O^^O^—'Q'—'
•o
v>
8
S
S
I
I
1-1
w &
1$
1
g
.a
Cb
C
M
ifi
C
•S
.S .5
4-91
-------
TABLE 4-53. SUMMARY OF BLOOD LEAD
(Pearson Correlation
AND HAND LEAD CORRELATIONS
Coefficients)
Phase
01 Coeff.
P
n
03 Coeff.
P
n
OS Coeff.
P
n
07 Coeff. .
P
n
09 Coeff.
P
n
Overall
Arith.
0.47
0.0001
202
0.51
0.0001
151
0.34
0.0001
149
0.28
0.0018
126
0.28
0.003
109
Log
0.51
0.0001
202
0.40
0.0001
151
0.41
0.0001
149
0.38
0.0001
126
0.37
0.0001
109
Area A
Arith.
0.52
0.0001
51
0.07
0.7
35
0.27
0.1
37
0.11
0.9858
28
0.36
0.06
28
Log
0.40
0.004
51
0.08
0.6
35
0.27
0.1
37
0.36
0.0583
28
0.33
0.08
28
AreaB
Arith.
0.44
0.0001
95
0.62
0.0001
75
0.51
0.0001
68
0.32
0.0116
63
0.22
0.1
51
Log
0.52
0.0001
95
0.53
0.0001
75
0.61
0.0001
68
0.40
0.0011
63
0.37
0.008
51
Area C
Arith.
0.28
0.04
56
0.16
0.3
41
0.15 j
0.3
44
0.38
0.0221
35
0.46
0.01
30
Log
0.48
0.0002
56
0.11
0.5
41
0.10
0.5
44
0.32
0.0548
35
0.36
0.05
30
blood lead children, e.g., DPbB decreased 0.41 mg/dL for per mg/dL increase in initial
blood lead.
To investigate the effectiveness of soil lead abatement, three methods were used:
(1) correlation analysis,
(2) regression modeling, and
(3) structural equation modeling.
Two biomarkers, handwipe lead and blood lead, and two environmental exposure
markers, exterior dust lead and interior dust lead, were used in the analysis. In this report,
we investigated the relationships of these four markers and other variables, e.g., treatment
area, paint lead, age, house age and SES.
Correlation analysis was used to investigate the relationship between two variables
regardless of the influences of other variables. Multiple regression analysis was used to
investigate the relationship between one marker variable, e.g., blood lead, and other
variables. The structural relationships between these four marker variables and other
variables are ignored in the regression analysis. A structural equation model was used to
investigate the structural relationship between these four marker variables, blood lead,
handwipe lead, interior dust and exterior dust, and other variables simultaneously.
4-92
-------
B
1 •*
O\ «-<
O ft
d d
8®.
dS
00 C
O CO
9 S d S
*oo ^
6 S- o
ppm
!
3
4-93
-------
Note that b = r*SD(y)/SD(x) for any two random variables, y and x, where b is the
slope of the simple regression, y=a+bx, and r is the correlation coefficient,
cov(x, y)/(SD[x]«SD[y]). This relationship between the correlation coefficient r and the
regression slope b was used. L
The findings of these analysis are:
(i) In the correlation analysis, the difference of handwipe lead between
Phase 1 and Phase 5 was correlated with the difference of blood lead
between Phase 1 and Phase 5 (p-value <0.1). The difference of blood
lead in Area "A" was greater than that in Areas "B" and "C", e.g.,
DPbB in Area "A" was 0.18*5.28/0.46=2.07 /tg/dL more than that in
Areas "B" and "C".
(ii) In the regression modeling, the difference of interior dust lead between
Phase 1 and Phase 5 was correlated with the difference of blood lead
between Phase 1 and Phase 5. The difference of blood lead between
Phase 1 and Phase 5 in Area "A" was greater than that in Areas "B" 'and
"C", e.g., DPbB in Area "A" was 2.87 mg/dL greater than that in Areas
"B" and "C".
•C.
(iii) In the structural equation modeling, the difference of interior dust lead
between Phase 1 and Phase 5 had a significant contribution to the
difference of blood lead between Phase 1 and Phase 5. The difference of
blood lead between Phase 1 and Phase 5 in Area "A" was greater than
that in Area "C", e.g., DPbB in Area "A" was 2.13 mg/dL greater than
thatinArea"C".
4.11.1.1 Correlation Analysis
To investigate the trends of these four markers, we used the difference of eaph marker
between Phase 1 and Phase 5 as a new marker. This positive value for the new marker
indicated that the value of the corresponding original marker was an increase from Phase 1 to
Phase 5, and vice versa.
For the data collected in Phase 1 and Phase 5, calculations were made of the difference
in blood lead(DPbB), the difference in handwipe lead(DPbH), the difference in the median
interior dust lead loading(DPbDIMD) and the difference in the median exterior dust lead
loadingCDPbDELD) between Phase 1 and Phase 5 respectively. Thus,
DPbB=PbB(5) - PbB(l),
DPbH=PbH(5)-PbH(l), !'
4-94
-------
DPbDIMD=PbDIMD(5)-PbDIMD(l), and
DPbDELD=PbDELD(5)-PbDELD(l).
The variables included in this data analysis are listed below:
DPbB(mg/dL): difference of blood lead between Phase 1 and 5,
DPbH(mg): difference of handwipe lead between Phase 1 and 5,
AGE_M5(month): age at Phase 5,
AGE_MS5(month2): square of age at Phase 5,
AGE_MH5(month): age at Phase 5 for handwipe,
AGE_MHS5(month2): square of age at Phase 5 for handwipe,
DPbDIFL(Kg/ni2): difference hi interior floor dust lead loading between Phase 1 and 5,
DPbDELD(mg/m2): difference of median exterior dust lead loading between Phase 1 and
PhaseS,
PbDELDl(mg/m2): median of exterior dust lead loading at Phase 1,
PbBl(mg/dL): blood lead at Phase 1,
PbHl(mg): handwipe lead at Phase 1,
PbDIMDlfttg/m2): median of interior dust lead loading at Phase 1,
PbPnMN(mg/cm2): mean of ulterior paint trim lead,
*y
PbPET(mg/cm ): exterior paint trim lead,
PbP(mg/cm ): paint lead,
PNT: paint lead be removed or not (l=yes, 0=no),
PNT_NR: paint lead be removed from nearby building or not (1 =yes, 2=no),
JOB: household job related to lead (l=yes, 2=no),
ACT: activities in the home related to lead (l=yes, 2=no),
4-95
-------
WORKS (numeric): a composite score of JOB, ACT, PNT and PNT_NR at Phase 5, [PNT:
paint was removed from apartment building or not (l=yes. 0=no),
PNT_NR: paint lead was removed from nearby building or not (l=yes,
0=no), JOB: member of household had job related to lead (l=yes,
0=no), ACT: activities in the home related to lead (l=yes, 0=no).],
A: Area A (interior and exterior dust lead were abated),
B: Area B (interior dust lead was abated),
C: Area C (no abatement),
SES: Hollingshead social-economic score.
The summary results from correlation analysis are shown in the following six tables.
Summary statistics are shown in Table 4-55. Summary statistics for DPbB, DPbH, DPbJJFL,
and DPbELD by area are shown in Table 4-56. The Pearson correlation between DPbB and
the other important variables are shown in Table 4-57. The correlations between DPbH and
the other important variables are presented in Table 4-58. The correlations between
DPbDIFL and the other important variables are presented in Table 4-59. The correlations
between DPbDELD and the other important variables are listed in Table 4-60.
From Table 4-56, we can draw following conclusions:
(1) DPbB and DPbH were higher in Area "A" than that in Areas "B" and
"C"; ie., the mean of PbB increased from Phase 1 to Phase 5 in Area
"A"; the mean of PbB decreased from Phase 1 to Phase 5 in Areas "B"
and "C". The mean of handwipe lead increased from Phase 1 to Phase 5
in all areas.
(2) DPbDIFL was smaller in Area "A" than that in Area "B" and "C"; i.e.,
interior dust lead decreased from Phase 1 to Phase 5 in Area "A";
however it increased from Phase 1 to Phase 5 in Areas "B" and "C".
(3) Exterior dust lead increased in all areas. But it increased more rapidly in
Area "B" than in Areas "A" and "C".
4-96
-------
TABLE 4-55. SUMMARY STATISTICS
Variable
DpbB
DpbH
AGE_M5
AGE_MS5
AGE_MH5
AGE_MHS5
DPbDIFL
DPbDELD
PbDELDl
PbHl
PbBl
PbDIFLl
PNT
PNT_NR
PBP
WORK5
H_AGE
SES
A
B
C
Number
148
119
148
148
126
126
128
127
138
136
148
138
148
148
148
148
148
148
43
61
44
Mean
-0.64
3.99
42.52
2120.86
43.11
2168:57
-17.30
97489
368077
10.18
10.50
411.62
0.27
1.75
2.05
0.17
110.16
19.47
0.29
0.41
0.30
SD
5.28
18.29
17.76
1574.39
17.66
1578.09
1193.0
615021
635934
12.94
5.13
910.66
0.45
0.43
2.22
0.21
16.84
7.31
0.46
0.49
0.46
Note: A=l for Area A, A=0 for other areas; B=l for Area B, B=0 for other areas; C=l for Area G, C=0
for other areas.
4.11.1.2 Regression Modeling
The results of correlation analysis presented in the previous section have one drawback.
The correlation coefficient computed between any two random variables were not adjusted by
other covariates, e.g., environmental variables. This can be improved by regression.
4-97
-------
TABLE 4-56. SUMMARY STATISTICS OF DPbB, DPbH, DPbDIFL,
AND DPbDELD BY AREA
DPbB
Area
A
B
C
Mean
0.87
-1.51
-0.91
Std. Dev.
5.60
5.17
4.90
DPbH
Mean
9.10
2.55
1.70
Std. Dev.
25.23
17.24
11.10
DPbDIFL
Mean
-451.13
181.77
51.06
Std. Dev.
1288.84
1437.75
427.79
DPbDELD
Mean
119280.1
164066.3
71.77
Std. Dev.
308629.2
922978.1
223152.8
The multiple regression was used to fit the following four regression models of DPbB,
DPbH, DPbDIMD and DPbDELD. The initial models are given below. j
DPbB = A+B+DPbH+ DPbDIMD
+DPbDELD+AGE_M5+AGE_MS5+
+PbBl+SES+WORK5+PNT+PbP+PNT_NR+H_AGE +
A* DPbDIMD + A*DPbDELD+B*DPbDIMD +
B*DPbDELD,
* ,
4.11.1.3 Multiple Regression Modeling ;
Multiple regression was used to fit the following four regression models of DPbB,
DPbH, DPbDIFL and DPbDELD. The initial models are given below:
DPbB = A+B+DPbH+DPbIFL+DPbDELD+AGE_MH5+PbBl+SES
+WORK5+PbP+H_AGE
DPbH = A+B+DPbDIMD+DPbDELD+AGE_MH5+AGE_MHS5+
PbHl+SES+WORK5+PbP+PNT+H_AGE+PNT_NR
+A*DPbDIMD+A*DPbDELD+B*DPbDIMD+
B*DPbDELD,
DPbDIMD = A+B+DPbDELD+SES+PNT+PbP+PbDIMDl+PNT_NR-f-
+H_AGE+A*DPbDELD+B*DPbDELD,
DPbDELD = A+B+PbDELDl+SES+PNT NR+PNT+PbP+H AGE.
The summaries of backward elimination procedures for these regression equations is
given below.
4-98
-------
TABLE 4-57.
CORRELATIONS BETWEEN BLOOD LEAD AND
OTHER IMPORTANT VARIABLES
Variable
DPbH*
AGE_M5*
*•
AGE_MS5
DPbDIFL
DPbDELD
*
A
*
B
C
*
PbHl
#
PbBl
#
PbDIMDl
PbDELDl
PNT
PNT_NR
PBP
H_AGE
WORK5
*
SES
Number
119
148
148
128
127
43
61
44
136
148
138
138
148
148
148
148
148
108
Cone. Coef
0.15
-0.37
-0.33
-0.02
-0.08
0.18
-0.14
-0.03
-0.15
-0.46
-0.16
0.04
0.09
-0.12
-0.05
0.056
0.092
-0.18
P-value
0.099
0.0001
0.0001
0.80
0.37
0.025
0.09
0.69
0.08
0.0001
0.055
0.63
0.28
0.15
0.53
0.50
0.27
0.07
Denotes variables significantly correlated (alpha ^0.1) with DPbB.
The following conclusions can be derived from the Table 4-57:
(1) flie change in handwipe lead affected the change hi blood lead. The less change in handwipe lead, the
less change in blood lead. In other words, DPbB will increase 0.15*5.28/18.29 = 0.04 mg/dL for per
mg increase in DPbH because corr(DPbB, DPbH) = 0.15;
(2) the child age. factor affects the difference of blood lead between Phase 1 and Phase 5. For older
children, the difference hi blood lead between Phase 1 and Phase 5 was smaller than that for younger
children. In other words, DPbB decreased 0.37*5.28/17.76 = 0.11 mg/dL for per month increase in age
because corr(DPbB, AGE_M5) = -0.37;
(3) treatment (Area) affected the difference of blood lead between Phase 1 and Phase 5. In Area "A", the
difference of blood lead between Phase 1 and Phase 5 was greater than that hi Areas "B" and "C". In
other words, DPbB has 0.18*5.28/0.46 = 2.07 mg/dL more hi Area "A" than that in Areas "B" and
"C";
(4) the initial (Phase 1) blood lead affected the difference of blood lead between Phase 1 and Phase 5. For
the high initial blood lead children, the difference of blood lead between Phase 1 and Phase 5 was smaller
than that for the low initial blood lead children. In other words, DPbB decreased 0.46*5.28/5.13 =
0.47 mg/dL for per mg/dL increase hi PbBl because corr(DPbB, PbBl) = -0.46.
4-99
-------
TABLE 4-58. CORRELATIONS BETWEEN HAND LEAD AND
OTHER IMPORTANT VARIABLES
Variable
AGE_MH5
AG E_MHS5
DPbDIFL
DPbDELD
PNT
PNT.NR
PBP "
WORKS
H AGE
A*
B
C
PbHl*
PbBl
PbDIMDl
PbDELDl
SES
Number
116
116
108
105
119
119
119
119
119
31
51
37
119
119
109
112
119
Corre. Coef.
-0.008
0.023
-0.17
0.09
0.08
-0.07
0.087
0.09
0.08
0.17
-0.07
-0.08
-0.40
-0.026
0.045
0.07
-0.08
P-value
0.93
0.81
0.087
0.35
0.40
0.46
0.34
0.32
0.37
0.07
0.46
0.36
0,0001
0.77
0,64
0.46
0.36
In the above table, the variable marked with "*" are significantly correlated with DPbH. ;
From Table 4-58, the following conclusions are evident:
(1) The initial (Phase 1) handwipe lead affected the difference of handwipe lead between Phase 1 and Phase 5
(DPbH). For children with high initial handwipe lead, the difference of handwipe lead between Phase 1
and Phase 5 was smaller than that for the low initial handwipe lead children. In other words, DPbH
decreased 0.40*18.29/12.94 = 0.57 mg for per mg increase in PbHl because corr(DPbH, PbHl = -
0.40.
(2) The area factor affected the difference of handwipe lead between Phase 1 and Phase 5. In Area "A", the
difference of handwipe lead between Phase 1 and Phase 5 was larger than that in Areas "B" and "C". In
other words, DPfaH was 0.17*18.29/0.46 = 6.76 mg more in Area "A" than that in Areas "B" and "C"
because corr(DPbH, A) = 0.17.
4-100
-------
TABLE 4-59. CORRELATIONS BETWEEN DPbDIFL AND
OTHER IMPORTANT VARIABLES
Variable Number Cone. Coef. P-value
DPbDELD
PbDELDl
PbP
WORK5
A*
B*
C
PbDIFLl*
SES
114
119
128
128
32
56
40
128
128
-0.02
-0.06
0.02
-0.03
-0.21
0.15
0.04
-0.52
-0.007
0.83
0.54
0.81
0.74
0.017
0.096
0.66
0.0001
0.93
Denotes variables significantly correlated (alpha < 0.1) with DPbB.
The following conclusions can be derived from Table 4-59:
(1) Treatment (Area) affected the difference in interior floor dust lead between Phase 1 and Phase 5. In
Area "A", the difference in interior floor dust lead between Phase 1 and Phase 5 was less than that in
Areas "B" and "C". In Area "B", the difference in interior floor dust lead wasffreater man that in Areas
"A" and "C". In other words, DPbDIFL was 0.21*1193/0.46=544.63 |i|/m* less in Area "A" than
that in Areas "B" and "C"; DPbDIFL was 0.15*1193/0.49=365.20 pg/m *2 greater in Area "B" than
tharin Areas "A" and "C".
(2) The initial (Phase 1) interior floor dust lead affected the difference in interior floor dust lead between
Phase 1 and Phase 5. In the higher initial interior floor dust lead area, the difference in ulterior floor
dust lead between Phase 1 and Phase 5 was smaller than that in the lower initial interior floor dust lead
area. la other words, DPbDIFL decreased 0.52*1193/910.66=0.68 fig/m2 for per /*g/m2 increase hi
PbDIFLl because corr(DPbDIFL, PbDIFLl) = -0.52.
For DPbB,
step 1 AGE_M5 is eliminated with p-value=0.97;
step 2 DPbDELD is eliminated with p-value=0.85;
step 3 PbP is eliminated with p-value=0.65;
step 4 H_AGE is eliminated with p-value=0.62;
step 5 DPbDIFL is eliminated with p-value=0.43;
step 6 WOKK5 is eliminated with p-value=0.33;
step 7 B is eliminated with p-value=0.17;
step 8 A is eliminated with p-value=0.16.
4-101
-------
TABLE 4^60. CORRELATIONS BETWEEN EXTERIOR DUST LEAD AND
OTHER IMPORTANT VARIABLES j
Variable Number Corre. Coef. P-value
SES*
PbDBLDl*
PBP
H_AGE
WOBK5
A
B
C
127
127
127
127
127
35
50
42
0.17
-0.31
0.03
0.07
-0.12
0.02
0.087
-0.11
0.05
0.0004
0.74
0.41
0.17
0.81
0.33
0.21
la. the above table, DPbDELD was highly correlated with. PbDELDl, SES and PNTJNR.
From Table 4-59, the following conclusions can be made:
(1) The initial (Phase 1) exterior dust lead affected the difference of exterior dust lead between Phase 1 and
Phase 5 (DPhDELD). For the high initial exterior dust lead area, the difference of exterior dust between
Phase 1 and Phase 5 was greater than that in the low initial exterior dust lead area. In other words,
DPbDELD decreased 0.31*615021/635934 = 0.30 mg/m2 for per mg/m2 increase in PbDELDl because
corr(DPbDELD, PbDELDl) = -O.31.
(2) The social-economic status affected the difference of exterior dust lead between Phase 1 and Phase 5. In
the higher SES families, the difference of exterior dust lead between Phase 1 and Phase 5 was greater
than that in the low SES family. In other words, DPbDELD increased 0.17*615021/7.31 =
14302.8 mg/m2 for per SES unit increased because corr(DPbDELD, SES) = 0.17.
ForDPbH,
step 1 H_AGE is eliminated with p-value=0.87;
step 2 PbP is eliminated with p-value=0.81;
step 3 DPbDELD is eliminated with p-value=0.52;
step 4 WORK5 is eliminated with p-value=0.40;
step 5 SES is eliminated with p-value=0.36;
step 6 AGEJMH5 is eliminated with p-value=0.28;
step 7 DPbDIEL is eliminated with p-value=0.22;
step 8 B is eliminated with p-value=0.15;
step 9 A is eliminated with p-value=0.12.
step 10 AGE_MHS5 is eliminated with p-value=0.11.
4-102
-------
For DPbDIFL,
step 1 SES is eliminated with p-value=0.95;
step 2 PbP is eliminated with p-value=0.93;
step 3 DPbDELD is eliminated with p-value=0.89;
step 4 WORKS is eliminated with p-value=0.87;
step 5 A is eliminated with p-value=0.76;
step 6 H_AGE is eliminated with p-value=0.37.
For DPbDELD,
step 1 PbP is eliminated with p-value=0.43;
step 2 A is eliminated with p-value=0.47;
step 3 H_AGE is eliminated with p-value=0.12.
After the backward eliminations, four final regression models remained:
Model 1 DPbB = 8.52 + 0,038 DPbH - 0.00079AGE_MS5 - 0.17 SES - 0.43
PbBl,
Model 2 DPbH = 9.38 - 0.50 Pbffl,
Model 3 DPbDIFL = 169.05 + 297.89 B - 0.91 PbDIFLl,
Model 4 DPbDELD = -228137,95 + 13290.31 SES + 316397.88 B + 481704.88
WORKS - 0.35 PbDELDl.
From the final models the following conclusions were made:
(1) The change of the difference in handwipe lead between Phase 1 and
Phase 5 (DPbH) affected the change of the difference in blood lead. In
other words, blood lead decreased 0.038 jtg/dL for per /tg/dL decrease
in handwipe (see Model 1).
(2) Age, SES, and PbBl were important factors to interprete the difference
of blood lead. For the children with high SES (>30 unit scores) and
moderate high PbBl (> 18 /tg/dL), the blood lead would decrease from
Phase 1 to Phase 5. The older these children were, the more rapidly
their blood lead decreased from Phase 1 to Phase 5 (see Model 1).
(3) The difference in handwipe lead between Phase 1 and Phase 5 was
reversely proportional to the initial handwipe lead. For the higher initial
handwipe lead children (PbHl > 19 /ng), their handwipe lead decreased
from Phase 1 to Phase 5 (see Model 2).
(4) The differences in interior floor dust lead were different in different
areas. In Area "B" where the initial ulterior floor dust lead was higher
than 515 ^g/m , the ulterior floor dust lead loading decreased from
4-103
-------
Phase 1 to Phase 5. However, in Areas "A" and "C" where the initial
interior floor dust lead loading was higher than 186 jig/m2, the interior
floor dust lead loading decreased from Phase 1 to Phase 5 (see Model 3).
(5) The differences in exterior dust lead were different in different areas. In
Area "A" and "C" where household activities were not related to lead
and the initial exterior dust lead loading was moderate high (PbDELD >
870,000 Aig/m2), the exterior dust lead loading decreased from Phase 1 to
Phase 5. In Area "B" where household activities were not related lead
and the initial exterior dust lead loading was higher than
1,770,000 jig/m2, the exterior dust lead loading would decrease from
Phase 1 to Phase 5.
4.11.1.4 Structural Equation Modeling
The four dependent variables DPbB, DPbH, DPbDIMD and DPbDELD were mutually
correlated. The regression models shown above do not account for this fact since they were
fitted individually. When the equations in a system are interdependent, e.g., the response
variable in one equation appears as the regressors in other equation, the ordinary least square
estimates of the parameters in the system may be inconsistent. To solve this, we fit these
i-
four models simultaneously by a system of linear equations.
We re-analyzed the above models as a system of simultaneous linear equations, using
SAS SYSUN. The initial models here are similar to the initial regression models in
Section 2.
i
The final models were:
DPbB = 10.28 - 0.18 SES - 0.064 AGE_M5 - 0.46 PbBl,
DPbH = 5.78+ 0.002 AGE_MH5 - 0.62 PbHl,
DPbDDtfD = 291.87 - 0.91 PbDIELl, •
DPbDELD = 273191B.
The final structural models are shown in Figure 4-29.
4-104
-------
2731.818
p-O.003
figure 4-29. Structural equation analysis: relationship between blood lead and
environmental lead.
From the final SYSLIN models, the following conclusions are drawn:
(1) In the DPbB model, SES, AGE, and PbBl were important factors to
predict the difference of blood lead between Phase 1 and Phase 5. For
children with high SES (> 30 unit scores) and moderate high PbBl
(> 10 jtg/dL), their blood lead would decrease from Phase 1 to Phase 5.
The older the children, the more rapidly their blood lead decreased from
Phase 1 to Phase 5.
(2) In the DPbH model, the initial handwipe lead affected the difference in
handwipe lead between Phase 1 and Phase 5. For moderate high initial
handwipe lead children (Pbffl >26 /tg), the handwipe lead decreased from
Phase 1 to Phase 5.
(3) In the DPbDIFL model, the initial interior floor dust lead affected the
difference in interior floor dust lead between Phase 1 and Phase 5. In the
area with higher initial interior floor dust lead (PbDIFLl >325 /*g/m2),
the interior floor dust lead loading decreased from Phase 1 to Phase 5.
4-105
-------
(4) In the DPbDELD model, area factor affected the difference in |.
exterior dust lead between Phase 1 and Phase 5. In area "B", the
exterior dust lead loading increased by 273191 jKg/m2 from Phase 1 to
Phase 5; and the exterior dust lead did not change significantly in
Areas "A" and "C" from Phase 1 to Phase 5.
4.11.1.5 Comparisons of Treatment-Effects Among the Three Statistical Approaches
The treatment-effect comparisons from these three methods were as follows:
To use the correlation analysis to compare the treatment effects across areas, we rely
on the fact that the b=r*SD(y)/SD(x), where b is the coefficient of simple regression
y=/t-f-bx, r is the correlation coefficient between x and y. The positive value of r will give
the positive value of b which indicated that y significantly increased as x increased.
Correlation analysis
DPbB was significantly greater in Area "A" than that in either Area "B" or "C".
DPbH was significantly greater in Area "A" than that in either Area "B" or "C".
DPbDIEL was significantly greater in Area "B" than that in either Area "A" or "C".
DPbDELD was not significantly different between areas.
Multiple regression analysis
f
DPbB was not significantly different between areas.
DPbH was not significantly different between areas.
DPbDIEL was significantly greater in Area "B" than that in either Area "A" or "C".
DPbDELD was significantly greater in Area "B" than that in either Area "A" or "C".
i
Structural equation analysis
i
DPbB was not significantly different between areas.
DPbH was not significantly different between areas.
i
4-106
-------
DPbDIFL was not significantly different between areas.
DPbDELD was significantly greater in Area "B" than that in either Area "A" or "C".
From the above results, the following conclusions were made:
(1) The difference in blood lead between Phase 1 and Phase 5 in Area "A"
was significantly greater than that in Areas "B" and "C" by correlation
analysis; however, it was not significantly different across areas by the
other two methods. Also, the difference in handwipe lead in Area "A"
was significantly greater than that in Areas "B" and "C" by correlation
analysis; however, it was not significantly different across areas by the
. other two methods.
(2) The difference in interior dust between Phase 1 and Phase 5 was
significantly greater in Area "B" than that in Areas "A" and "C" by
correlation and regression analysis; however it was not significantly
different across all areas by the structural equation method.
(3) The difference in exterior dust between Phase 1 and Phase 5 was
significantly greater in Area "B" than that in Areas "A" and "C" by
structural and regression analysis; however, it was not significantly
different across Areas "A", "B", and "C" by correlation method.
(4) Since the SYSLIN method is superior to the other two methods in terms
of investigating their structural relationship, the results from SYSIIN
were thought to be more reliable.
4.11.1.6 Statistical Modeling Conclusions
In this study, the differences in blood lead, handwipe lead, and interior dust lead were
not significantly different across all areas. Age, initial blood lead, and socio-economic score
were the important factors in the prediction of the. change in blood lead. In other words,
DPbB decreased 0.064 /*g/dL for per month increase in AGE; DPbB decreased 0.46 /ng/dL
for per /*g/dL increased in PbBl. DPbB decreased 0.18 /tg/dL for per unit score increase in
Hollingshead social economic score (SES). From structural equation analysis, we found that
for the children with higher SES scores (> 30 units) and moderate high PbBl
(> 10 jttg/dL), their blood lead decreased from Phase 1 to Phase 5. The older these children
were, the more rapidly their blood lead decreased from Phase 1 to Phase 5. For children
with moderate high initial handwipe lead (> 26 ptg), their handwipe lead decreased from
4-107
-------
Phase 1 to Phase 5. For lower initial handwipe lead (< 9 /*§) children, their handwipe lead
increased from Phase 1 to Phase 5. The older these children were, the more rapidly their
handwipe lead increased from Phase 1 to Phase 5. In the areas where the initial interior
floor dust lead loading was higher than 325 /tg/m2, the interior floor dust lead loading
decreased from Phase 1 to Phase 5. Exterior dust lead loading increased by 273,000 /ig/m2
in Area "B" from Phase 1 to Phase 5, but it did not change significantly in Areas "A" and
"C". "
4.11.2 Cross-Sectional Structural Equation Models for Loading Data
Measurements were made on different sampling units during the soil abatement project.
Blood lead (PbB) and hand lead (PbH) were measured on individual children; interior dust
lead (PbD) and paint XPJE7^ on individual apartments; and exterior paint (XRFext) and
exterior dust lead on individual buildings and nearby paved areas. Analyses of these data
must take into account these design features, to avoid inflation of the degrees of freedom for
toting important effects and biased estimates of the postulated relationships.
Structural equation models including variables which quantify the nested design were
used to explore the relationships among the environmental data, each expressed as loadings,
and the outcomes of PbB and PbH. Each equation included indicator variables for different
buildings. The interior PbD and XRF, and PbH and PbB equations also included terms for
different apartments (within buildings); and the PbH and PbB equations also included terms
for different siblings (within apartments). Interaction terms, between the mediating
endogenous variables and these nested factors were created wherever possible. These
interactions, if significant, served as error terms for testing the endogenous variables' effect.
For example, a PbH* family interaction was introduced into the PbB model to test the
significance of PbH's effect on PbB. In all models, buildings variance served as the error
t
term for testing the effects of Abatement Area and Neighborhood (within areas).
Each of the endogenous variables were initially entered into equations further in the
i1
postulated "casual chain" (Exterior Measurements -> Interior Measurements (PbD^j^),
PbDcntiy) -> PbDfloor) -> PbH -> PbB. The covariates of housing type and age,
4 behavioral variables [JOB, ACT, PAINT and PAHSrrneighborilood]* were entered into each
structural equation).* Race, sex, child's age and SES were entered into the PbH and PbB
4-108
-------
models; mouthing behavior was used in the PbB model only. A number of interactions were
, also considered. Abatement area, age and mouthing was interacted with each environmental
- Pb variable in every model in which both of these main effects were initially used.
A strategy of backward elimination of insignificant effects was followed. Interactions
were removed before main effects and only terms compared directly to error were candidates
for removal.
Multiple imputation of any missing endogenous variables' data was performed. Any
bias due to imputation was removed from the model by introducing indicator variables for
these imputed data points into the structural model.
The final models are shown in Tables 4-61 through 4-63 and Figures 4-30 through
4-32.
*JOB = job in an industry where lead may be used.
ACT = hobby or other activity which may use lead.
PAINT == painting in apartment or building within past six months.
= painting in neighborhood within past six months.
4.11.2.1 Summary and Conclusions from the Cross-Sectional Structural
Equation Models
The preabatement model shows no effect of environmental lead loadings on PbB and
PbH (Figure 4-30). This may partly explain the difficulty that was encountered by this study
in attempting to significantly reduce these indices of exposure.
The postabatement models (Figures 4-31 and 4-32) appear to show both a pattern of
recontamination and differential effects by abatement areas. Different distal sources of lead
may be acting in each abatement area. These cross-sectional models are a prelude to even
more detailed and complex longitudinal models. We plan to use these results to help us
model the changes in PbB, PbH, and PbD over the course of the project.
4-109
-------
TABLE 4-61. RESULTS OF STRUCTURAL EQUATIONS MODELING
USING PHASE 1 LEAD LOADING DATA
Exogenous (Phase 1)
Estimate
tofF
P
Dependent: In (PbB)
Intercept
ln(PbH)
Age*Age
Mouthing
Areas
A
B
Neighborhoods
B
D
G
Families
1.375
0.093
0.046
-0.0006
0.217
0.044
-0.137
-0.040
0.251
-0.050
2.13
6.81
-5.78
2.45
F (2, 87) = 0.17
F (3, 87) = 0.50
F (87, 1,019) = 1.70
0.02*
0.0001
0.0001
0.008*
0.85
i
0.68
0.0001
Dependent: In (PbH)
Intercept
Age*Age
Areas
A
B
Neighborhoods
B
D
G
Families
0.720
0.069
-0.0007
0.188
0.175
0.353
0.012
-0.472
6.13
-4.15
F (2, 83) = 0.04
F (3, 83) = 0.33
F (83, 1,019) = 2.19
0.0001
0.0001
0.96
i
0.80
1
0.0001
Dependent: In (PbDfloor)
Intercept
In CPbEWdow)
Areas
A
B
Neighborhoods
B
D
G
Apartments
3.718
0.097
0.134
0.263
0.299
-0.610
-0.834
-1.018
2.07
3.43
F (2, 79) = 0.02
F (3, 79) = 0.37
F(79, 1,019) =5.14
0.02*
0.0001*
0.98
0.77
0.0001
4-110
-------
TABLE 4-61 (cont'd). RESULTS OF STRUCTURAL EQUATIONS MODELING
USING PHASE 1 LEAD LOADING DATA
Exogenous (Phase 1)
Estimate
tofF
P
Dependent: In (PbDwilldow)
Intercept
House Type
Areas
A
B
Neighborhoods
B
D
G
Apartments
5.968
1.175
2.253
-0.502
0.090
1.478
1.816
—
2.47
F (2, 78) = 0.58
F (3, 78) = 0.37
F (78, 1,019) = 8.96
0.008*
0.56
0.77
0.0001
Dependent: hi (PbDentry)
Intercept
1* (PbDex.)
House Type
Areas
A
B
Neighborhoods
B
D
G
*
Areas In (PbDexl)
A, In (PbDext)
B ln(PbDext)
Apartments
24.067
-1.637
1.836
-23.581
-20.238
-0.695
-0.085
-1.350
1.947
1.588
—
-4.80
3.86
F (2, 79) = 1.11
F (3, 79) = 0.12
F (2, 1,019) = 10.28
F (79, 1,019) = 8.53
0.0001*
0.0001*
0.33
0.95
0.0001
0.0001
Dependent: hi (PbDext)
Intercept
Paint
Areas
A
B
Neighborhoods
B
D
G
Dwellings
12.915
0.272
0.010
-1.829
1.852
2.291
-1.237
- —
2.69
F (2, 49) = 0.74
F(3, 49) = 1.07
F (16, 1,019) = 20.22
.._
0.004*
0.48
0.37
0.0001
4-111
-------
TABLE 4-61 (cont'd). RESULTS OF STRUCTURAL EQUATIONS MODELING
USING PHASE 1 LEAD LOADING DATA
Exogenous (Phase 1)
Estimate
tofF
P
Dependent: In (XRF^
Intercept
Areas
A
B
Neighborhoods
B
D
G
Apartments
1.086
-0.244
-0.335
-0.187
0.223
0.243
—
F (2, 73) = 0.25
F (3, 73) = 0.44
F (73, 1,019) - 2.57
—
0.7;8
0,72
0.0001
Dependent: In (XRFex()
Intercept
Paint
Areas
A
B
Neighborhoods
B
D
G
Dwellings
0.843
0.146
-0.282
-0.095
-0.011
0.222
-0.021
- —
2.11
F (2, 45) = 0.09
F (3, 45) = 0.09
F (45, 1,019) = 6.16
; *
0.02
0.92
0.97
-t
0.0001
4.12 HEALTH AND SAFETY
4.12.1 Workplace Audits
Laboratory clinic and field audit reports filed between 1989-1991 were maintained by
the SHO. During 1989, 14 lab reports were prepared; 18 in 1990, and 6 in 1991.
Between 1989-1991, 28 field audits were conducted. A table indicating the types of
audits, and dates performed and are included in copies of the Interior and Exterior Safety
Audit forms which were used for site safety evaluations Appendix O. In general these audits
t
revealed that appropriate health and safety procedures were being followed. Common
problems noted in some of the ulterior and exterior field audits included:
4-112
-------
TABLE 4-62. RESULTS OF STRUCTURAL EQUATIONS MODELING
USING PHASE 3 LEAD CONCENTRATION DATA
Exogenous (Phase 3)
Estimate
tof F
P
Dependent: hi (PbB)
Intercept
In (PbB)
Age
Age*Age
Areas hi (PbH)
A* In (PbH)
B* In (PbH)
Families* hi (PbH)
Areas
A
B
Neighborhoods
B
D
G
Families
1.639
0.119
0.044
-0.0005
-0.306
-0.048
0.180
-0.203
-0.093
0.379
-0.338
—
F (1, 42) = 1.70
5.63
-5.04
F (2, 42) = 1.69
F (42, 960) = 1.52
F (2, 79) = 0.48
F (3, 79) = 1.73
F (79, 960) = 0.92
#
0.20
0.0001
0.0001
0.20
0.02
0.68
0.17
0.63
Dependent: hi (PbH)
Intercept
hi (PbD^d^,)
Age
Age*Age
Job*
Age* hi (PbD^^)
Areas
A
B
Neighborhoods
B
D
G
Apartments
-1.253
0.119
0.291
0.072
-0.0005
0.735
-0.004
0.870
1.555
-0.236
-0.379
0.349
—
1.88
2.13
3.88
-2.72
2.15
2.18
F (2, 60) = 4.30
F (3, 60) = 0.36
F (60, 960) = 1.46
—
0.03*
0.02*
0.0002
0.008
0.02*
0.03
0.02
0.78
. 0.01
4-113
-------
TABLE 4-62 (cont'd). RESULTS OF STRUCTURAL EQUATIONS MODELING
USING PHASE 3 LEAD CONCENTRATION DATA
Exogenous (Phase 3)
Estimate
tofF
P
Dependent: In (PbDfloor)
Intercept
In (PbDentIy)
Areas
A
B
Neighborhoods
B
D
G
Apartments
3.992
0.301
-1.527
-0.422
-1.046
-0.978
-1.899
—
6.82
F (2, 70) = 1.09
F(3, 70) =2.16
F (70, 960) = 8.92
—
0.0001*
0.34
0.10
0.0001
Dependent: In (PbD^a^)
Intercept
In (PbDext)
In (XRFjjjt)
Areas
A
B
Neighborhoods
B
D
G
Apartments
-0.039
0.475
0.918
-2.907
0.201
-0.563
-1.372
-0.565
—
3.36
4.79
F (2, 70) = 1.77
F (3, 70) = 0.48
F (70, 960) = 4.63
— .
0.006*
0.0001*
0.18
.*
0.70
,
1
0.0001
Dependent: In (PbDentiy)
Intercept
In (XRFext)
House Type
Areas* In (XRFext)
A In
-------
TABLE 4-62 (cont'd). RESULTS OF STRUCTURAL EQUATIONS MODELING
USING PHASE 3 LEAD CONCENTRATION DATA
Exogenous (Phase 3)
Estimate
t of F
P
Dependent: In (PbDmat)
Intercept
Areas
A
B
Neighborhoods
B
D
G
Apartments
6.089
-1.099
0.022
-0.861
-0.105
-1.532
—
F (2, 68) = 1.02
F (3, 68) = 2.17
F (68, 960) = 5.64
—
0.37
0.12
0.0001
Dependent: hi (PbDext)
Intercept
House Type
Areas
A
B
Neighborhoods
B
D
G
Dwellings
10.924
1.288 '
0.496
0.612
0.818
0.638
-1.087
—
7.09
F (2, 40) = 0.12
F (3, 40) = 0.46
F (40, 960) = 22.57
- ' —
0.0001*
0.89
0.71
0.0001
Dependent: hi (XRF^
Intercept
Areas
A
B
Neighborhoods
B
D
G
Apartments
1.001
-0.159
-0.307
-0.114
0.396
0.304
—
F (2, 70) = 0.25
F (3, 70) = 0.82
F (70, 960) = 2.68
—
0.78
0.49
0.0001
Dependent: hi (XRFext)
Intercept
House Type
Areas
A
B
Neighborhoods
B
D
G
Dwellings
0.733
-0.172
0.026
0.026
0.191
1.156
—
F (2, 44) =. 0.06
F (3, 44) = 0.09
F (44, 960) = 5.62
—
0.94
0.97
0.0001
4-115
-------
TABLE 4-63. RESULTS OF STRUCTURAL EQUATIONS MODELING
USING PHASE 5 LEAD LOADING DATA
Exogenous (Phase 3)
Estimate
tof F
P
Dependent: In (PbB)
Intercept
ln(P5H)
In (PbDfloor)
In (PbDentiy)
In (PbDexl)
Mouthing
Paint
#
Areas hi (PbH)
A* In (PbH)
B*InjpPfaH)
Areas In (PbDfloor)
A, In
-------
TABLE 4-63 (cont'd). RESULTS OF STRUCTURAL EQUATIONS MODELING
USING PHASE 5 LEAD LOADING DATA
Exogenous (Phase 3)
Estimate
tofF
P
Dependent: hi (PbH)
Intercept
In (PbDfloor)
to (PMDwtadow)
to (PbDmat)
Age
Sex
Areas* In (PbDfloor)
\ to (PbDfloor)
B ln|PbDfloor)
Areas hi (PbD^,^)
A hi (PbD^d,^)
B hi (PbD^,^)
Areas
A
B
Neighborhoods
B
D
G
Families
2.318
0.418
-0.365
0.196
0.019
0.296
-0.520
-0.644
0.485
0.571
-2.880
-2.755
-1.148
0.068
-0.681
—
2.68
-4.04
2.83
5.09
1.91-
F (2, 874) = 5.71
F (2, 874) = 8.22
F(2, 51) = 1.17
F (3, 51) = 0.81
F (51, 874) = 2.10
—
0.004*
0.99*
0.003*
0.0001
0.06
0.003
0.0003
0.32
0.49
0.0001
4-117
-------
TABLE 4-63 (cont'd). RESULTS OF STRUCTURAL EQUATIONS MODELING
USING PHASE 5 LEAD LOADING DATA
Exogenous (Phase 3)
Estimate
tofF
P
Dependent: In (PbDfloor)
Intercept
In (PbD^yjndow)
In (PbD^
In CPbD t )
ln(XRFext)
House Age
Areas* In (PbD^dow)
*"* JUH, v Jt ^"^^ii/Tn^QTivx
JC* JL1JL. VJ* "^"^nvindo W'
Areas* In (XRFext)
A*ln(XRFexl)
B ln|XRFext)
Areas In(PbDext)
A* In (PbDexl)
B InCPbD^,)
Areas
A
B
Neighborhoods •
B
D
G
Apartments
13.946
-0.004
-1.669
0.142
-0.490
0.097
0.474
0.651
-2.059
1.251
2.023
1.363
-27.602
-24.225
2.408
2.370
-2.817
—
-0.04
-5.03
2.33
-1.59
3.01
F (2, 874) = 12.54
F (2, 874) = 11.41
F (2, 874) = 7.23
F (2, 53) = 1.45
F (3, 53) = 2.15
F (53, 874) = 7.18
*
0.52
0.99*
*
0.01
*
0.94
0.002*
0.0001
0.0001
0.0008
0.24
, -
0.10
0.0001
Dependent: In (PbD^dow)
Intercept
Areas
A
B
Neighborhoods
B
D
G
Apartments
10.768
-2.195
0.619
-4.524
-1.855
-0.253
—
F (2, 54) = 0,80
F(3, 54) = 1.12
F (54, 874) = 7.17
—
0.46
0.35
0,0001
4-118
-------
TABLE 4-63 (cont'd). RESULTS OF STRUCTURAL EQUATIONS MODELING
USING PHASE 5 LEAD LOADING DATA
Exogenous (Phase 3)
Estimate
tofF
P
Dependent: In (PbDentry)
Intercept
InCXRF^
House Type
House Age
Job
Paint
Areas
A
B
Neighborhoods
B
D
G
Apartments
-17.098
0.738
1.012
0.170
1.541
0.995
-0.895
4.724
-0.488
-0.484
1.500
—
3.66
2.50
4.01
2.85
2.34
F (2, 54) = 2.77
F (3, 54) = 0.25
F (54, 874) = 6.24
_
0.0002*
0.007*
0.0001*
0.003*
0.01*
0.07
0.86
0.0001
Dependent: hi (PbDmat)
Intercept
Job
Paint
Areas
A
B
Neighborhoods
B
D
G
Apartments
4.237
2.719
1.065
2.243
1.733
-0.548
0.861
0.154
—
5.47
2.91
F (2, 53) = 1.62
F (3, 53) = 0.46
F (53, 874) = 3.56
—
0.0001*
0.002*
0.21
0.71
0.0001
4-119
-------
TABLE 4-63 (cont'd). RESULTS OF STRUCTURAL EQUATIONS MODELING
USING PHASE 5 LEAD LOADING DATA
Exogenous (Phase 3)
Estimate
tofF
P
Dependent: In (PbDext)
Intercept
In(XEFext)
House Type
Areas
A
B
Neighborhoods
B
D
G
Dwellings
12.691
0.211
0.452
0.226
-0.170
-1.093
0.830
-2.099
3.07
4.12
F (2, 35) = 0.03
F (3, 35) = 2.03
F (35, 874) = 43.94
0.001*
0.0001*
0.97
0.13
0.0001
Dependent: In (XEF^t)
Intercept
Areas
A
B
Neighborhoods
B
D
G
Apartments
0.548
0.621
0.240
0.350
0.022
0.181
F (2, 56) = 1.19
F(3, 56) =0.34
F (56, 874) = 1.58
0.31
0.80*
i;
0.005
Dependent: In (XRFext)
Intercept
Areas
A
B
Neighborhoods
B
D
G
Dwellings
0.716
0.125
0.060
0.113
0.063
0.105
F (2, 38) = 0.06
F (3, 38) = 0.08
F (38, 960) = 1.79
0.94
0.97
0.003
One-tail p-value.
Interior Dust
employees not wearing proper foot protection
safe lifting practices not followed
emergency phone numbers not available at site
dust containers not equipped with covers.
4-120
-------
15
CD
•c
a
o.
I
W)
I
J>
i
§
a
OS
I
1
-------
J
CO
I
S>
CO
fi
0
•s
O)
c
CD
T3
03
CD
I
&
LL
O
4-122
-------
I
1
4-123
-------
Exterior Soil
• employees not wearing proper foot protection
• dust control not maintained
• first aid kit and emergency phone numbers not available at site.
In addition to regular laboratory audits, walk-throughs to identify area lead exposure
and to observe work practices in laboratories were conducted by a University of Cincinnati
M.S. industrial hygiene student (see Appendix P). These reports showed that exposures
were far below all currently accepted standards use by OSHA, or exposure guidelines from
NIOSH and the ACGIH.
4.12.2 Special Workplace Evaluations
4.12.2.1 Price Hill Facility
As with laboratory monitoring, personal and area air monitoring was performed by
University of Cincinnati industrial hygiene graduate students during abatement periods.
Early in the study a survey was conducted of a soil sieving operation at a temporary field
site. Personal monitoring of three abatement employees' breathing zones was conducted for
lead exposure as they performed their usual work tasks. The employees' breathing zones
were monitored for time period up to 6.5 h. The ventilation results show that the hood sash
should have been pulled down to the halfway position to achieve optimum protection. The
!"
samples collected indicated that the three workers were exposed to just detectable amounts of
lead-containing dusts which were far below the current OSHA action level for lead.
4.12.2.2 Noise Levels
A 1990 survey of the 12V portable exterior dust sampling vacuum equipment which
revealed noise sound levels over 90 dB prompted the installation of a hearing conservation
program which required employees to wear hearing protection when exposed to sound levels
of 85 dB or greater.
During abatement in Fall 1991, personal sound levels measurements were obtained
from three abatement workers: one backhoe operator, one bobcat operator and one randomly
selected student. Each of these individuals wore personal noise dosimeters for 7 h. Results
4-124
-------
of the three samples (68-77dB) were well below the OSHA 85 dB action level for an 8 h
day.
4.12.3 Manuscript on Safety and Health Plan
A manuscript on the development and implementation of this safety and health plan,
previously introduced as Appendix M in Section 3.1, summarizes the above and additional
material.
4.13 ABATEMENT COSTS
Abatement costs for soil, interior dust and exterior dust abatement for project years
1989, 1990 and 1991 are summarized in Table 4-64. The unit abatement costs are expressed
per area, per housing unit (interior dust) and per study child. These units are somewhat
arbitrary because there were many other children living in the study area, in addition to
children enrolled in the study. For example, since the exterior abatement (soil and exterior
dust) was conducted on an area-wide basis, there were many children living in the area other
than those who were enrolled in the study. In addition, there were many children in the
abatement areas who were too old to be eligible for the study.
The total cost of the abatement, excluding methods development, preparation of
contacts and inspection costs was about $712,000. The average total cost per study subject
was $3899 with 58% of the cost for soil abatement, 19% for interior dust abatement and
23 % for exterior dust abatement.
4-125
-------
TABLE 4-64. ABATEMENT COSTS
Soil Abatement
Total Cost $
Area, m2
# Subjects
$/M*
$/Subject
Interior Dust
Total Cost $
# Housing Unit
Area, Mt2
# Subjects
$/Housing Unit
$/M*
$/Subject
Exterior Dust
Total Cost$
Area, M2
# Subjects
$/M*
$/Subject
1989
130,712
3,708
56
35.25
2,334
112,746
93
6,867
163
1,212
16.42
692
37,366
38,464
56
0.97
667
1990
**
228,850
8,381
59
27.30
3,025
23,632
16
1,331
29
1,477
17.75
814
66,977
75,496
59
0.89
1,135
1991
65,537
2,079
52
31.52
1,260
46,082
41
3,531
52
1,124
13.05
886
***
OVerall
425,099
14,168
167
30.00
.2,244
182,460
150
11,729
244
1,216
15.56
748
104,343
113,960
115
0.92
907
Docs not include preparation and inspection costs.
t
-------
5. CONCLUSIONS
Conclusions will be presented in two formats: (a) as responses to questions asked
concerning whether or not our evaluation of descriptive statistics of the study data provided
evidence to support certain conclusions and (b) as general conclusion based on evaluation of
study results. These are not mutually exclusive formats.
5.1 RESPONSE TO QUESTIONS
1. Do we find evidence that soil abatement reduced soil lead concentrations?
Yes: Table 4-21, Area A, Phase 0 versus Phase 02
Area B Phase 05 versus Phase 09
2. Do we find evidence that soil and exterior dust abatement alone reduced interior house
dust Pb concentration?
Not part of study design in Phases 00 through 05 but can compare Phase 06 vs.
07 in Area B when data are available.
3. Do we find evidence that interior dust abatement reduced interior dust Pb
concentration?
No: Table 4-33; Area A and B; Floors; Phases 01 vs. 02.
4. Do we find evidence that interior dust abatement, alone, reduced interior dust Pb
loading?
Yes: Table 4-33; Area B; Floors; Phase 01 vs. 02.
5. Do we find evidence that soil and exterior dust abatement reduced exterior dust lead
concentration?
No: Table 4-28, Area A, Phase 01 vs. Phase 02
Area B, Phase 05 vs. Phase 09
loading?
No: Table 4-30, Area A, Phase 01 vs. Phase 02
(except targeted samples)
5-1
-------
6. Do we find evidence of soil recontamination?
No: Table 4-21; Area A; Phase 02 vs. 09.
7. Do we find evidence of exterior dust recontamination?
Can't test since initial reduction not demonstrated (except target samples).
8. Do we find evidence of interior dust recontamination?
Yes: Table 4-33; Area B; Floors; Phase 02 vs. 05.
9. Do we find evidence of the impact of abatements on hand lead?
No: Table 4-46; Area A or B; Phase 01 vs. 05 or 09.
10. Do we find evidence of the impact of abatements on blood lead?
Perhaps: Table 4-41; Area B; Phase 01 vs. 05
Table 4-44; Area B; Phase 01 vs. 05
Table 4-45; Area B; Phase 01 vs. 05.
•i
11. Do we find evidence that intercorrelations among measures were as hypothesized prior
to any abatements?
Yes: PbD (exterior) vs. PbD (interior)
PbD (interior) vs. PbH
PbH vs. PbB
12. Since the findings are largely negative, do we have evidence that they are not due to
sampling or analytical error?
Yes: extensive QC data confirms accuracy and precision of
measurements.
13. Do we have any evidence that abatements might have caused a transient increase hi
exposure?
No.
14. Was there evidence that the environment was stable over the course of the study?
No: In the control area environmental lead levels varied widely
during the course of the study.
5-2
-------
5.2 GENERAL CONCLUSIONS
(1) As a result of abatement geometric mean soil lead concentrations decreased in the
top 2 cm samples by 146 ppm in Area A and by 102 ppm in Area B.
(2) As anticipated, soil lead concentrations were higher when a building was located
near the sample location (line (source) sampling pattern) than for areas more
remote from such structures (line (area) sampling pattern).
(3) Soil concentrations were also greater as anticipated for soil areas were building
debris was observed in the bottom 2 cm of the 15 cm soil cores.
(4) Exterior dust abatement was not effective in reducing lead loadings, as measured
in .samples collected within weeks of abatement, except at locations near the entry
to subjects' houses where more than a 50% reduction was observed in the first
post-abatement sample.
(5) Interior dust abatement was effective in reducing interior dust lead loadings by
about 40% in Area A and by about 60% in Area B. Loadings in Area B, where
only interior abatement has occurred, showed evidence of some recontamination
3 mo later. There was some indication that the impact of the abatement persisted
10 mo later in Area A and perhaps as long as 22 mo post-abatement.
(6) There was no evidence that blood lead levels were reduced by soil lead or dust
abatement in Area A. There was a slight reduction (net reduction over control
area) of 0.6 pg/dL in Area B that may be attributed to interior dust abatement
(this difference was not statistically significantly).
(7) The inability to achieve a sustained reduction in exterior dust lead loadings was
probably an important contributing factor for the lack of blood lead level
reduction.
(8) The interior dust abatement in this project was performed under rather favorable
conditions: (1) it was performed in completely rehabilitated housing with a lack
of significant amounts of lead-based paint and with tight floors with no dust-laden
cracks); and (2) the protocol for abatement featured a substantial amount of
furniture and carpet replacement as well as wet and dry cleaning. Even under
these somewhat ideal conditions, reduction in interior entry dust lead loadings of
only about 50% were reached. Interior floor dust lead loadings were 37% lower
10 mo post-abatement in Area A. In other types of housing containing lead-based
paint, with housing in poorer condition and with a less rigorous protocol applied,
a single time interior dust abatement would not be as likely to achieve even these
modest results.
5-3
-------
-------
6. REFERENCES
Bornschein, R.L.; Hammond, P.B.; Dietrich, K.N. (1985) The Cincinnati Prospective Study of low level lead
exposure and its effects on child development: protocol and status report. Environ. Res. 34:4-18.
Bornschein, R.L.; Krafft, K.M.; Clark, C.S.; Peace, B.; Hammond, P.B. (1986) External surface dust lead,
interior house dust lead and childhood lead exposure in an urban environment. In D.D. Hemphill (ed),
Trace Substances in Environmental Health II. University of Missouri, Columbia,
Clark, S.; Bornschein, R.L.; Succop, P.; QueHee, S.S.; Hammond, P.B.; Peace, B. (1985) Condition and type
of housing as an indicator of potential environmental lead exposure and pediatric blood lead levels.
Environ. Res. 38:46-53.
Clark, S.; Bornschein, R.L.; Succop, P.; Peace, B.; Ryan, J.; Kochanowski, K. (1988) The Cincinnati Soil Lead
Abatement Demonstration Project. In B.E. Davies and B.G.Wixson (eds), Lead hi Soil: Issues and
Guidelines. Science Reviews Limited, Northwood, Great Britain.
Clark, S.; Bornschein, R.; Succop, P.; Roda, S. (1991) Urban lead exposures of children in Cincinnati, Ohio.
Chemical Speciation and Unavailability 3:163-171.
Chisolm, J.J. Jr.; Brown, D.H. (1975) Micro-scale photofluorometric determination of "free erythrocyte
porphyrin (protoporphyrin IX). Clin Chem 21:1669-1682.
6-1
-------
LIST OF APPENDICES
A Calculation of Required Sample Size
B Letter to Prospective Families and Project Fact Sheet
C Letter to Property Owners and Community Leaders and Project
Fact Sheet
D Project Newsletters
E Institutional Review Board Consent Forms
F Interior Dust Clean-up Methods Development Manuscript
G Training Manual for Paint and Water Sampling
H Worker Safety and Health Plan Publication
I Agenda for Contractor Training
3 Interior and Exterior Safety & Health Audits
K Door Mat Placement,Sampling and Removal Sequence
-------
APPENDIX A
Calculation of Required Sample Size
This appendix was Addendum A "Power Calculations" of the
September 30, 1988 Grant Agreement between the U.S. Environmental
Protection Agency and the University of Cincinnati.
Reference (5) in this document is:
Succop, P.A., O'Flaherty, E.J.,Bornschein, R.L., Clark, C.S.,
Kraftt, K., Hammond, P.B., Shukla,R. A kinetic model for
estimating changes in the concentration of lead in the blood of
young children. In: Lindberg, S.E. and Hutchinson, T.C. (Eds).
Proceedings of the International Conference: Heavy Metals in the
Environment, Vol.2 New Orleans, September 21987, pp. 289-291.
-------
DEMONSTRATION OF POWER AND SAMPLE SIZE CALCULATIONS
Estimates of the effect sizes to be expected are obtained from Kinetic
Model (5). Assume that we are sampling children whose ages are
distributed uniformly across the age range of interest (12 to 60 months),
half of whom have blood leads (PbB's) whose kinetic change with time will
resemble that of children having resided continuously in Rehabilitated
housing, the remaining half having PbB's whose kinetic change with time
will resemble that of children having resided continuously in Satisfactory
housing to their current age, L
From Kinetic Model (5), a child residing continuously in a single housing
category from birth is modeled
PbBt « .94»PbBprcnal8l«{ 1 -«p( -.063"tprcnal8l)}«(«p( -.063«t)}
+ ej»{l-exp(-.063«0},
where at is the estimated asymptote for the housing category in which the
child resides. For a child continuing to reside in this category for At
additional months, PbB is modeled
PbBl4At«.94«PbBprcnela1«{l-exp(-.063«lprcnalal)}«{exp(-.063«a*AU)} (2)
+ e,«{l-exp(-.063Mt*AU)],
at time t* At, while a child moving to a different category of housing has
an expected PbB of
+ i,«{I-exp(-.063M))»(expC-.063«lAU)} (3)
* a2*n-exp(-.063«[AU)}
where 82 fs the asymptote for the category Into which the child moves.
The prenatal effect in Eqs. (I) through (3) may be modeled assuming full
-term pregnancies (tpreoa^'S) and by using the observed Lead Program
Project's prenatal, maternal PbB mean of 8.2.
The expected mean PbB for children at the 3 sampling occasions
planned by the study design may be estimated by integrating the variable t
fn Eqs. (1) through (3) above in the limits [12,60] and dividing by 48, i.e.,
the range in ages of children to be sampled. Integrating ( 1 ) provides an
estimate of the geometric mean PbB for all children to be sampled just
117
-------
prior to abatement
60
(48)*' JjjPDBtdl • .495808123 4 .6505742348,. (4)
The design of the abatement study calls for repeated sampling of PbB's 6
and 12 months after abatement. Integrating (2) for At*6 and At* 12
provides
60
(48)~tj J2PbBl46dl«.340665394 4 .8973320788,, (5)
and
60
(48)''J12PWt+12« " .234062694 4 .9294586048,, (6)
respectively. Assume that soil abatement has an equivalent effect to that
of a move for a child from a Rehabilitated or Satisfactory house to a
newer (Post World War 2) residence. Then integrating Eq. (3) for At-6 and
&t« 12 provides,
60
(48)'!J PbBl45*«,340661394 4.5844152018, 4.31291687782. (7)
and
60
(48)' J.-PbB^t2dl-.234062694 4 .4015222548, 4.52791678282, (8)
as expected geometric mean PbB's for children living In abated residences
6 and 12 months, respectively, following abatement.
Using the average of the two asymptotes estimated for the
Rehabilitated and Satisfactory categories as representative of
pre-abatement asymptotes provides aj«HO7.l9* 19.97)- 18.58. The
asymptote for the Private Post World War II housing category provides
a2" 1426. Then the geometric mean PbB's for the 5 areas to be used in this
study at the 3 sampling occasions may be estimated from Eqs, (4) through
(8) and are listed in the following Table.
118
-------
AREAS
interior
AREA A AREAB AREA C nonafaated
April-Hay, 1989 16.299 16.299 16.299 16.299 16.299
October, 19B9 15.661 15.661 17.013 17.013 17.013
April-May, 1990 15.222 15.222 17.503 17.503 17.503
These estimates assume that removal of interior dust will provide na
additional diminishment in PbB.
Assuming an equal number of children will be sampled in each group.
the grand In(PbB) means are estimated to be 2.791 1, 2.8009, and 2.8065 at
these 3 sampling occasions. The effect sizes, T, expected to occur in
separate analyses of covariance of 1n(PDBt+6) and ln(PbBt4l2) for each of
the j areas are
Tj(t*6) * j(t*6) * (t*6) * &(j(t) " (t)}' and
where ?+t is tne mean of Jn^Pl)Bt4At^ and
-------
study have residual PbB's 6 and 12 months following abatement that are
akin to those observed in the Lead Program Project's cohort at 42 and 48
months. (Once again we use the assumption of sampling children whose
average age at the beginning of the study is 36 months.) The MS£ from
analyses of covariance of the Lead Program Project's data may then be
used to estimate the expected residual variance to be encountertd In this
study. Specifically, we estimate that
M5Elln(PbB42Hln(PbB36),Housing categories]} -.07628 (13)
and
M5E(1n(PbB4Q)|ln(PbB36),Housing categories]} - .05756. ( 1 4)
Housing category has also been removed from the residual variation since
the abatement study plans to follow only non-movers, which we assume
are, once the study begins, in a non-changing environmental "category".
The effect of housing over and above the prior PbB (s quite minor at these
ages.
Thus, the noncentrallty parameters that occur (n testing the overall
hypothesis of no soil abatement effect are
.07628
.05756
using all 5 groups in ANACOVA
120
-------
The power for detecting a significant effect due to abatement on PbB at
t+6 and t* 12 months is then estimated to be:
n poweKt*6) poweKM2)
10 .103 593
15 .136 , AM
20 .173 .585
25 514 .704
30 .253 .794
35 594 .861
40 .336 .909
45 .375 .941
50 .415 .963
BOX power for finding a significant effect of soil abatement on
ln(PbBt412) is expected to result if at least 31 children in each of the 5
areas are sampled. Other effects of abatement, especially those which are
captured in the environmental measurements, will have at least, and
probably much greater, power than that shown here for PbB, since our
published longitudinal study data has indicated much stronger
relationships among the environmental measures of housing, soil lead, and
dust lead, than those found between the environment and indices of body
burden such as PbB.
121
-------
APPENDIX B
Letter to Prospective Families and Project Fact Sheet
-------
University of Cincinnati Cincinnati Soil Project
1142 Main Street
Cincinnati, Ohid 45210-1936
Phone (513) 651-4774
May 26, 1989
Dear Resident:
Do you remember our census of young children that you participated in?
Well, you may be interested in knowing that partially as a result of that
census, your housing Is in one of the areas that are eligible to participate !ln a
project that the University of Cincinnati, with cooperation from the Cincinnati
Health Department, is conducting.
The project Is designed to determine If certain environmental improvements
are able to reduce the exposure of young children to lead from soil and dust.
You and some other residents in your neighborhood will be asked to bring your
young children (under five years of age) to the Soil Project Office to enroll in
the study and to have their blood measured for lead over the next two years.
Enclosed Is a fact sheet that provides some further information.
In the meantime, if you have any questions, please call JoAnn Grote or
Sherry Wilkens (651 -4774).
Sincerely,
JoAnn Grote Sherry wiIkens
Resident Coordinator Recruiter
-------
Cincinnati Soil Project
What is the Soil Project?
It is a three year research project funded by the U.S. Environmental Protection
Agency to the University of Cincinnati Department of Environmental Health with a
subcontract to the Cincinnati Health Department.
What is its purpose?
Its purpose is to determine if efforts to reduce the amount of lead in available
soil and dust in residential areas can lower the blood lead of young children (under
5 years of age) living in these areas.
How will available soil lead be reduced?
Two methods will be used: removal of soil to a six inch depth and replacing it with
fresh topsoil or cultivation of the soil if only the top layer is higher in lead.
New sod will be placed on areas where soil has been either removed or cultivated.
How will dust lead be reduced?
In exterior paved areas (for example, streets, alleys and sidewalks) vacuum
equipment will be used to remove dust and dirt. Dust from inside of homes will be
removed by use of special vacuum equipment and in some cases by a limited
replacement of carpets and furniture.
How were areas selected?
Areas selected had to have a large number of children less than five years of age
and a high percentage of FHA rehab housing.
What is involved in the project?
In addition to the soil and dust activities mentioned above, blood lead of children
will be periodically determined (a total of 5 times over three years) and extensive
environmental sampling (for example, soil and dust from both inside and outside
homes) will take place a total of 7 times in three years.
If/ill soil and dust clean-up occur the same time in all areas?
No, because this is a research project, clean-up activities will be spread over a
three year period.
Will public recreation areas be included?
Yes they will if they lie within areas being studied. We will work closely with the
Recreation Commission to make sure that disruption of their use is kept to a minimum
and that appropriate surface coverings are placed around swings etc.
Why was Cincinnati chosen?
It was chosen for several reasons (1) the existence of an established lead research
group at the University of Cincinnati, (2) evidence of high soil and dust lead in
residential areas, (3) an active lead screening program in the Cincinnati Health
Department, and (4) a history of cooperation between the city, the University and
the medical community.
(OVER)
-------
-------
APPENDIX C
Letters to Community Leaders and Property Owners Describing Study
and an Accompanying Project Fact Sheet
-------
Unlvetmlty of Cincinnati Cincinnati Soil Project
1142 Main Street
Cincinnati, Ohio 45210-1936
Phone (513) 651-4774
He would like to Inform you about a University of Cincinnati research project .
involving small portions of several Cincinnati neighborhoods. The Cincinnati
Health Department is also involved with us in this effort.
The project goal is to determine if reducing the amount of lead in soil arid dust
will reduce the exposure, of young children to lead. The focus 1s on those areas
where there is a high concentration of rehabilitated housing where the sources of
lead are generally thought to be from the routine entry of dust into the home from
outside paved and soil areas.
The environmental improvements will include soil surface clean-up in areas with
lead levels above a pre-determined value, extensive cleaning of paved areas, and
dust removal inside of homes of young children who are enrolled in the project.
As a part of the dust removal from the interiors of homes, a limited amount of
furniture and carpets may be replaced. Exterior clean-up activities will be
available for all properties within defined areas while interior clean-up will
occur only in homes where young children are enrolled in the project.
Because this is a research project not all clean-up activities are scheduled ti/
occur during the same year. The established schedule is as follows:
Neighborhood Area Year Clean-up
Portions of Pendleton 1989 ' Interior and Exterior
i
Portions of Findlay Market 1989 Interior
and Back St. area
Portions of Findlay Market 1990 Exterior
and Back St. area
Portions of Glencoe Place 1991 Interior and Exterior
and Mohawk area
The specific boundaries of these areas are shown on the enclosed sheets.
-------
An Initial soil sampling and a census of young children have already occurred as
part of our efforts to determine eligible areas. Actual recruitment of
participating families with young children 1s scheduled to start in the near
future.
Enclosed is an information sheet which provides some further information on this
project. Should you have any questions, please call Linda Conway-Mundew at the
Soil Project Office (651-4774).
Sincerely,
Linda Conway-Mundew Scott Clark
Project Administrator Principal Project
Investigator
JoAnn Grote Bill Menrath
Resident Coordinator Property Coordinator
LCM:SC:OG:BH/jo
Enclosures
-------
University of Cincinnati Cincinnati Soil Project
1142 Main Street
Cincinnati. Ohio 45210-1936
Phone (513) 651-4774
BACKGROUND INFORMATION ON CINCINNATI SOIL PROJECT
1.- This is a demonstration project requested by the U.S. Environmental
Protection Agency to provide answers to specific questions regarding
lead in the environment.
2". The specific questions to be answered are:
(a) Are any of the following methods effective in reducing exposure of
young children (less than 5 years of age) to lead? reduction of
lead in soil available to the children, reduction of lead in
exterior dust, and reduction of lead in interior dust.
(b) Which of the above methods, alone or in combination with one or
both of the other methods, is most cost effective in reducing the
exposure to lead?
(c) How quickly do soil and dust areas become recontaminated after
clean-up?
3. In order to answer the above questions in an efficient way it is
essential to select areas with a high density of young children and in
•which a high percentage of the children live in housing where lead in
soil and dust are the major immediate sources of lead to the children.
4. In Cincinnati houses rehabilitated under various HUD programs, where
lead paint availability on the interior and exterior of homes was
virtually'eliminated, represent the type of housing where the lead in
soil and dust are thought to represent the major sources available to
the children. •
5. The purpose of this demonstration project,therefore, is not:
(a) To reduce the blood lead levels of the children with the highest
blood lead (such children almost always reside in houses that have
not had their paint lead removed); or
(b) To remove the most lead-contaminated soil in the city (again, such
soil is likely to be adjacent to structures still having lead-
based paint on them).
6. The purposes of the demonstration project require that we select
portions of neighborhoods that will enable us to provide answers to the
US EPA's questions, within the constraint of time and money imposed by
the project, and within the research guidelines developed by the US EPA,
the Centers for Disease Control and by our research group.
(OVER)
An affirmative action/equal opportunity institution
-------
7, There are many factors that constrain what we do and how we do it.
Among these factors are the following:
(a) Season of the year influences the degree of lead exposure and was
considered in deciding when we will abate lead and when we will
measure blood level.
* (b) There must be appropriate comparison neighborhood areas that are
not soil or dust abated the same year; these comparison areas must
be very similar to the areas being abated.
(c) In order for our findings to be statistically significant, the
areas selected must have a minimum average of 50 children that
complete the two-year study. Based on our past experience we know
that only 80 % of those families we contact will participate and
that, for various reasons, only 80 % of those who join will
actually complete the study.
Therefore each study and comparison area must have an average of
about 75 children less than 60 months of age at the beginning of
recruitment.
8. The areas that meet the criteria for inclusion in the project are
portions of the following neighborhoods: Pendleton, Findlay-Back, and
Glencoe-Mohawk. Other areas considered include North and South portions
of Lower Price Hill, the John St. area of the West End,and portions of
Walnut Kills.
9. Through an agreement with the City of Cincinnati that was required prior
to sampling city-owned property, we agreed to abate any city property
found to contain soil with lead concentrations requiring some abatement
according to our guidelines, regardless of whether that area was
eventually^selected for inclusion in the study.
-------
QI^c~ P,~_
P
i
f
n
T
t
L
5?
?"
n
0
c
_J
L.«r*
z
GLENCOE NEIGHBORHOOD
The properties included in the Glencoe
neighborhood will be those properties
fronting on the north and south sides of
Glencoe Place west of Inwood Place and
View Court, properties fronting on Leroy
Court, properties on Adnored; and properties
on Deronda.
MOHAWK NEIGHBORHOOD
The Mohawk neighborhood boundary begins at
the intersection of Central Parkway and
Mohawk Place and proceeds east on Mohawk
Place to McMicken Avenue, southeast on
McMicken Avenue to Elm Street, south on Elm
Street to the parcel of property at 2035 Elm
Street along the boundary of that parcel to
Colby Alley, across Colby Alley to the parcel
of property at 2036 Dunlap, north on Dunlap
to Stark Street, northeast along Stark Street
to Central Parkway, north on Central
Parkway to the intersection of Central
Parkway and Mohawk Place. The Mohawk
neighborhood is enclosed within these
boundaries.
-------
THE BACK STREET NEIGHBORHOOD
The Back Street neighborhood boundary
begins at the parcel of property located at
1650 Hamer Street and proceeds north on
Hamer Street to the intersection of Hamer
and Elder, east on Elder Street to the
intersection of Elder end East McMicken,
southeast on East McMicken to the parcel of
property at 50 East McMicken, along the
boundary of that parcel to Hust Alley,
'southeast along Hust Alley, across Lang
Street to the parcel of property at 100 East
McMicken Avenue, along the boundary of that
parcel to East McMicken Avenue and across
East McMicken Avenue to Back Street,
northwest on Back Street to the parcel of
property located at 25 Back Street, along the
boundary of that parcel to the starting point
at 1650 Hamer Street. The Back
neighborhood is enclosed within these
boundaries.
-i
i
i
~>
'§
FINDLAY NEIGHBORHOOD
The Findlay neighborhood boundary begins at
the intersection of Elm Street and Sellew
Alley and continues north along Elm Street
to Forbus Alley, west on Forbus Alley to
Race Street, across Race Street to Addy
Alley, east on Addy Alley to the parcel of
land at 18 Findlay Street, west on Findlay to
Goose Alley, south on Goose Alley to Elder
Street, west on Elder Street to Race Street,
north on Race Street to Sellew Alley, west
on Sellew Alley to Elm Street, returning to
the starting point at the intersection of
Sellew Alley and Elm Street. The Findlay
neighborhood is enclosed within these
boundaries.
-------
PENDLETON NEIGHBORHOOD
1
1
r
1
I
1
1
1*1 J^
» f
•
^M «i^ ^MH* t1*
* J
j- J
-1
I
|
i
s
/
i— — «"- *"™l E. I1VW S.*r«.t Tl
1
I
I
f
f
F
I
•
?
y
3,
J
r
n
i
a n
»«
1
I
i
\
•R
*
1
I
1
1
i
1
'
The Pendleton neighborhood boundary begins
at the intersection of Housman Alley end
Broadway and proceeds north on Broadway to
East 12th Street, west on East 12th Street to
Bunker Alley, north on Bunker Alley to East
13th Street, east on East 13th Street to
Broadway, north on Broadway to 1320
Broadway, continuing east on that parcel of
property to Spring Street, including 1323 and
1320 Spring Street, continuing east along the
parcel at 1320 Spring Street, continuing east
along the parcel at 1327 Pendleton Street to
Pendleton Street, south on Pendleton street
to Levering Alley, west on Levering Alley to -
Spring Street, south on Spring Street to
Housman Alley, west on Housman Alley to
the starting point at the intersection of
Housman Alley and Broadway., The Pendleton
neighborhood is enclosed within these
boundaries.
-------
APPENDIX D
Sample Project Newsletters
-------
ti A HKST DEGREE PROBLEM
When people uc smoldng around you,
you smoie too.,,and U'i rurmfuL The iraoke (
thii rises off the end at a burning tigareue,
oiled sioeiueam smoke. is twice u high in
ur and nicotine u Uut which ihe actual smoker inhales. The
smoke is Tilled with hundred! of chemicals many of which tie
known u date cancer. Studies (how Uul secondhand smoke
can cause non-smokers to develop bean disease and lung
cancel.
U you smoke. you probably don't realize how your smoking
can cuue hum to those you love. When non-smokers like
young children are forced to breaihe air that a filled wilh
cigarette smoke, things happen thai you can't tee. Their heart
best speeds up, their blood pruture rises. And dangerous
carbon monoxide icep into their blood. Babiet of parentt who
smoke ai home have a much higher nt£ of lung rfi *"?"•* Rich
as bronchitis and pneumonia than bshirs of nonsmoking
parents. Cigarette (moke can actually aggravate and even
Irieeer symptoms in come asthmatic children. Parents should
only smoke outside the home or, better yet. quit s molting
completely. More than you may realize, you set an simple
for year children. Most teenagers who smoke on a regular
bails ara from families where one or both parents *m*^"
Here'* tow to quite looking today!
•Decide that quilting unokinj is one of the best things that
you could eb for younelf and your family.
•When you want to quit, you will quit. Keep working as ill
•Mate a list of reasons why you want to quit and poo ft
aoEaewhere readily noflcrp^r
•Try whatever seemi best to you. Taper off. Or quit all at once.
•Switch to t brand you *'iinriL
•Throw away your matches. If getting your cigarette lit is
difCcub, maybe you wont smnlrr.
•Remind younelf every morning bow imponant it is to quit.
Million! of people have stopped smoking. Are their families
more impottani than your family?
Contact Ihe American Lung Aiton'aiinn of S.W. OH at
751-3650 about program that can help you f^
SUMMERTIME FOOD SAFETY RULES:
Wilh warm weather comes picnics, barbecues and food
poisoning. If you could just throw the refrigerator under one
arm and take it wilh you, there wouldn't be any problem in
caring for food to jo.
leu's became Ihe best way to fight food poisoning is to keep
periihabk foodi-cipcctally meal and poultry-cold between
preparation and serving.
•Keep food cold. This is Ihe best way to fight baoeria.
•Keep bacteria on your hands out of food-Everyone in the
family should wash hands before preparing food.
•Dent spread bacteria from raw meal and poultry to other food.
Waihhawttaflereceuctwithrawiaeaiirrijouluy. Use*
fresh plaie-and utensil set for each food.
•Thoroughly cook raw raeal, poultry and fish.
•Dont use food from dsmryM cflfiUuynM!*hrfs? cans and glass
jats for (Seras. cracks or bulging lids; paper packages for leaks
•Make A
rhfr*1fr of your ipplii
tefritenior registers a. safe 40* For tower.
stt at 0' For lower.
ri' that your
Freezers should be
FINGER JELL-O
2 envelopes of unn*vored gelMlin
2 1/2 cup* wider
1 package (6 ounces ) or
2 packages (3 ouncei) flavored Jell-o
The kids will luve ill Dissolve unflavorcd ^eUun in one (1)
cup of cold waler. Set aside. In saucepan .bruit 1 cup of water
to a boil and aild Jell-o. Add diitolved unflavorcd muuure. Sur
and add 1/2 cup cold water. Pour into a lightly greased pan and
refrigerate until firm (about 2 hours). Cut into tquares or use
cookies cullers and sure in an airtight container in the
refrigerate.
HOMEMADE POPS1CLES
Four juice or Kool-aid (add slices of fruit if denied) into small
container! or paper cups. Place in freezer. When alighily
frozen, add a popsicle stick or straw. Place back in freezer
until frozen. Loosen popsicle by running a liule warm water
over container.
BANANA PUDDING
6 lo 8 medium s
1 large package of instant vanilla pudding
I oz container of cool whip (any store brand will due)
£ oz container of sour cream (if you like it, not necucaiy)
1 box of vanilla wafers
In m large bowl mix instant pudding using directions for
making pie on package. Add container of cool whip and sour
cream fif desired) to pudding mixture. Begin by starting with a
layer of pudding in bottom of t bowl, add a Uyer of vanilla
wafers, next slice bananas on top of wafers. Continue
layering, ending with a layer of crushed vanilla wafer. Enjoy!
BIKTH ANNOUNCEMENTS
The rinrinnaii Soil Project proudly announces the birth of Ihe
following babies !
MathiasG.
DonueC.
Christopher D.
Gwendolyn O.
David W.
II2W1
1/2W91
2/17/91
3/14/91
4/2U/91
Good Luck.in the future!!!!!!
SU1OIERT1UE WORD SEAKCH
s
*
o
f
K
C
J
i.
t
a
L
O
T
X
If
Z
E
O
O
T
W
1
"
I
f
U
M
W
F
U
°
SI
IE
C
T
S
a
»
M
it
a
UH7S
MUhADe
S
Y
N
a
s
§
i
c
A
t.
U
T
O
X
I
a
i
r
i
a
o
N
C
K
.M
S
C
1
D
t
U
A
r
r
N
1
T
W
'fc
At
£
0
S.
O
O
'I
s.
i. a
H i
a L
* ':'*
T S
O 1
H W
C 0
a t
Mnicu
SWIM
-------
CINCINNATI SOIL PROJECT
NEWSLETTER
No. 7, Summer 1991
1142 Main St., Cincinnati, OH 45210
PKOCRAM PROGRESS REPORT:
This summer will be a busy lime for our
fsmpfr* and sutff. Sherry will be lending
appointments to nil frniili*y to come to
our Main Slice! of lice in Sax fat ihe bat blood collection
Our monitoring teams will collect gfmriff for the final f""»
Also Ihii summer, our focus will be on Ihe Glencoe and
Mobawktreas. We will remove and replace lead contaminated
toil and clean Ihe streets and other paved areas around Banoah
Park and Glencoe Place in ML Auburn. Furniture and cupel
replacement in the filrnr^rr and Ivtobawk rt*fff**T*^MtnMMti is
also scheduled.
CALENDAR OF EVENTS:
JUNE 1 Kids Fest (Sawyer Point)
JUNE < Last day Cintl Public Schools
JUNE 14 Flag Day
JUNE 16 Father's Day
JUNE 21 1st day of Sumcr
JULY 4 Indcpendenca Dajr
•••*•••**•••••«••*•**
SUMMER LEAD ALERT:
OWttPMNTCWF*
As summer approaches, our children spend
more time py^i^ Children are pKpotfd
to more sources of lead outside playing
than they are during the winter months.
Lead is often found on painted porches.
fences, garages and windows. Lead can be in dust and din
around old buildings and houses where children play. Infants
and young children playing in these areas often put their
hands, toys, dirt or even food they drop on the ground in Ineir
pyMnhy These are Ihe most CPHUQOQ ways chili&cn ^yiTCTnft
exposed to lead.
LEAD POISONING PREVENTION TIPS:
•Pi scf) ur ftf [iiumb welkins
•Keep toys clean of din and du£
•Keep arcitt where painl is peeling or cracking
d"£Mn by dunp mopuic or vacuuming [daily if
needed]
•Good hind washing skills, especially before
every meal and suck
•Teach children not to eat dirt or paper
It is very important to have all children between the ages it I
and 6 years screened for lead at regular clinic visits, usually
every 6 moruhs. by* bloodiest. Eady detection will nuke ihe
prublem f
SUMMERTIME FUN PLACES:
OveMhe-Khine Communit
1715 Republic Street
381-1893
•IZandundcr
13 • 17
18-54
55 and over
•swimming
•tolthdl-boys and girls
•lunchu (free up to 17yn)
•skating
•tennis
free
$5X0 (Dec to Dec)
$6.00 (Dec 10 Dec)
$4.00 (Dec to Dec)
-iittcnucdiale
•cowpcuiive •
•Boys Scout and Cub Scout clubs
•arts and crafts
•weekly field trips (punned summer trips)
-Indianapolis Children's Museum
-Kooky Hark Sue Park
LeBlouU Bays and Girls Club
1620 Central Parkway
721-7600
Summer programs begin June 10 thru Aug 16
Monday - Friday 10 am to 5 pm
Membership feeiS.OO per year
•swimming
•game roooi
•gymnastics
•weekly day camps for children of all ages
•fcinrhrs served free to children 12 years and under
* KID'S FEST
to, a SATURDAY, JUNE 1
L '-^ 10 am la * pm
Lai Ceolral Riverfront Area
H'a the largest «"»jj» • day event for chiMren in the coiuuryl
Area Highlights
•Showboat Majciuc-View "The All American Melodrama".
•River Wharf-Enjoy ftcc riverboal rides.
•TmapiT]pi^rt^ Staiiocu-Join Ihe circus or tMMincc on a dragon.
•Dinosaur Dugout-Panicipaie in a grand slam of auinlies
with Spike Ihe "•'"""""
•Sport Court Challenge-Play tennis and sand volU ybalL
•Teen Scene-Make tr'ini'' and shoot some hoops.
•Wheels of Kiniinc-Rollerblade with the Cyclones.
•Kids Talk Radio-Join WCKY for a live broadcast.
•Walk of Fame-Meet Ibe 19X1X stars: Michael and llan
Simrsfwi
•Our Family of Animals-Become friends with ihe animals.
You get all of this and more. And. it's Free for children and
their families.
For more information call Ihe Cincinnati Talking Yellow
Pages and enter code 3150. Kids Feit is produced by Ihe
Cincinnati Recreation Commission and Media sponsors:
19XIX TV and WCKY Radio Corporate Sponsors: White
Castle. Hutch. LaRosx's, BB Riverboats and Frisch's
SUMMER WATER SAFETY:
Warm weather is betel With onset, our children are out and
about playing. While enjoying the outdoors, we need to leach
our children to think and practice water safety rules. Injuries
are Ihe leading health risk facing children today. However,
some simple safety measures could lower Ihe risk of an injury
or perhaps <*j"*1** Hcxe are ymr basic safety rules when your
rf»nrf goes swimming
•Never leave a young child alone in Ihe tub. not even to get a
towel (children can drown in less than 2 inches of water in
seconds). Beware uf backyard pools, children can wander off
and afridmially fall in.
•Teach your children to always swim with a friend, never alone.
•Do not push or jump on other swimmers.
•Always supervise children near water, even older children.
Older kids are al lilt when they overcainuue lhmr ability and
r riow deep the water is.
•Enroll children over age 3 in swim classes. Check with your
neighborhood recreation center for swim programs offered.
The Boys and Girls Club located al 1620 Central Parkway
offers a S3.00 per year individual membership fee which
includes swim rlavirt. Keep in mind that lemons do not make
your child "drown-proof*.
-------
CINCINNATI SOIL PROJECT NEWSLETTER
NO. 6, SPRING 1891 1142 UAIN ST.. CINCINNATI, OH 45210
PROGRAM PROGRESS REPORT: •.. ..>.
In the past several'monlha,alnca
our last visit to Soil Project
lamlllas, our .technicians have ~
been busily collecting «on*nd,.^
.street dust *araplea ;froa tpttiou,^- — ..
locstlont 'In :^be Jietudy •irnalghborhoo'da.Vit^ies*
aamplee will be added -to'/the thoussnds.pt other
handwlpe-samples tram
study. ••'The lnlo;»»Uon
P'.
Speaking ol collecting .samples,-we wttl again be
visiting some families to,about,* saonth. ..Just aa
we did last yssr at this '.-time, ••<*• will be
collecting water samples snd analyzing -the.-.paint
on the walla In some homes. We will cot rsvUH
those .homes where we hsve. jlready .collected
water and snalyiad the palaC^rlWe wU only visit
the homes ol new ismllles who were enrolled In
the Soil Project during 1990. The next.time we
will vlilt ill ol the -tsatlles In the -Soil Project,
will be In the summer .cl 1981 .and (fbat «IB • «nd
the project., ..........
We sre now In'rthe. process -of planning Jor fcth*
cleaning sctIvlUes*ior>1991.,.J.Our work this "year
will be la those tioroea.'Jn the.^Oleoco* ssod^»»ohiiwk
areas,-''lhe,-«res« ,-whIch ;?have ,.not j^ilrea'i/
received ..lurhlture, ...jarpetlng .and Viln'te'rlor
cleaning. .Those ttwo itelghborhoods will ' *'
receive -the -exterior ' ' -• -
eummar oi
••••••••«•
THE -TV BEAST AND YOUR CHILD:
Vi' "''^•••"•i->u.:.'.';''"
Has .the TV became • problem
in. your home? • -Have you ever <•
^welched your .-child Hitching • •
rTV.7,...Have you twtteed the trenc*-«ke expraialon
jfopvtils or her,h<*oe .end Basked youraell, -What'f
(jjnytlrtUd.jloUJirV.^Po ,jr««r-,«iae-.-JV,ao.fra.p your
&rtjl|d.-,oecuplad;i»»hl|j| -ijrou' ••»!». .fcusy dolno
«lse7'#1He sale* than questions ruc.u..
and t..ch.r
TV.'.,.How.v.r,
-•xeiaalv*'-
. .,!*» ,p»bbl»a U thai chlldr*n
isalj «nd 1o /b«
ra I t«n" : b» com.
miss ouii on
:B'y limiting ' Ib« araoont ot TV your child walch.i,
you esn h«lp cr4or icraatlva and Partormlng Art!
anca productton: . :
Kan Voltmtatr
Arbor Day (In Ohio)
Uamortsl Day " •
Kahn'a/Hlllihlia Farms Esstar Egg Hunt at:
Sawyer Point , • '
Sunday, Uarch 31
Reglatrellon 12:00 noon
Egg Hunt 1:00-3:00 p.m.
Chlldran up to 12 yrs old.
WINK radio disc Jockayl Prlzaal
.. ._
d jBarnsUIn". . •CJ:4- .'•.-* >"( -
For mora Intonaatlon, call tha achool boi olttea
Funl
-------
FROM THE DESK OF NURSE DONNA-
INFANT FORMULA VS. BREAST UILK
Because ol your baby'a rapid
growth, at no lime Is nutrition
more Important than during
infancy. During tha first
year, particularly during the
flrat 6 r .nlhs, m. ; U your
baby'a primary food. «uraan'"i"raast -milk • :or
infant formulae, which are-dealgnad to be •-
nutritionally as much • Ilka,-breast «llk «s
poaalbla, are noat desirable-4or baWea/^.v. "t- . -.
tluch *f the literature' . ,
breast Mttlc'4a .tha -auparlorVlaadlng'.forJijriSfaii:^
Tha •poaltlvee" -Jor braaatleading rrir'e' •'««'«:
lollowa: :..^breaef ',-«llk •i»--»tty-"rto''/-^tir,ftt'~>-tt^
contains; the right ,*u1rl«rta .Oaijha jijHi
.akotji -kui»B« _l.Vll> •*!Ijllb.^' \j. "-*1' »'^l.L.l!"S*a*Ji
A TOUCH OF HISTORY:
February, as we all know
was Black History Uonth.
A time when special
efforts are extended
to every Individual
regerdless of color to broaden their level of
^duoaUon ,>pn black historian*., -/Although -there
countleaa"'years of .education-'lost, the
••--rn-wmhould -.'be .carried with us
•^reierance '/to -the '••Afro-American
•'ia .ttae to ini1UI"1n --our young
!ran>PoiltJv»:MJe.«sodels. ^rrhere it much
'^i^KWducala.-W'.NcfatUlran In the
,4o p'rorabte positive
mothara '
apecUI -
iftSttWmoV^pJn*^,.
able-to *tumbl«'*«»*,,,r_,.r. ,
Wlii,6i,V ^Wng •ia'^rr^:i^)'.,
of mid«.^K^r/.§iiroi«»^lJJilrS . ^
can be axpraaa»d by .aaaana •it •• bra«»t.p'ump';«if
can faa atorad in-iha refrigerator or; treezer.
Breast teedlng';can'-alao be •upplaaianta'd «w!th
lomuU. Aa an added bonus, nuralng helps apeed
up tha contraction of tha uterus to its pre-
pragnancy alze.
ion '-fat
**nv.fe.
Whan a taothar • la not
commercially prepared infant^fomula v-'Js ./i
nuultteii.^^nllil^^a^^llix^ri^tt'^in'^
eonHdanU^na.'tea^aid ;^a^m.iaai)»K^w. tlS
doctor. -However •-afJthW^StSii-tfi^iSi.Wiiii'fiiiii. :-r?
doctor. -Hbwayar.^f
to -'
•mount of walar «ugga«t*d.VSn»arar||«:iSS65Uad-»oX-?
add watar - to '•"raady to ; «aadVrtp«Vl»l»h*?»lapy ^i
nothara chooaa -|o .-bottla -^JifSalacauaa7
braaatfaadlng maant faalng -«o
mothtrt want ;to ,botu»-laad ,'ao
can shara ln:;lhU,»a|>act of >
. .-..r ^ — ..* ". * "^''. '*• '• •'• ____ •-..
Whan dacldlng >wh*th«r to braaatfaad ^
laad, th. Important WoOo ; :MaMmbtr'^eyi|e <*uk*
your own ~eholcaa:^*oor approach '!io?l»adlig^ipay
not faa tha *ama aa your nalghbor'a ' tor --taat
irland'a. '-But both can faa lilting for .->oVr'
particular family. -^Patanta .today hava \aa»aral
«ptlona for providing' Ufa,' adaquat* nutriUon. :
• •••>. .<«•. ....... ... "
, • ^Jaota''-lor'^slr.^ c..nlu.ny
.-- —. -own -manufacturing icompsny end
le jsillllpns, in the process.
%S&Sf$i&t'iifi^g£:'".-*;.- ' '••
'»-aav'sd 'tboussndt ol lives
•because'tl tbt ettorti msdt to
-the eonsenraUon -of Mood plasau.
;' fe£v. ;**gl..._.
40 yrs to
'improvt working
C^'opposid to racist
-.blaek':'-»htrapranaar who
,' »l»»ary 'JJaya -by
' inn-chant. 'riH« wa*
1700'a1/
:~-:€-*'i~S?^:-J$&!±Xij.
^g.;^:gij!;i£<^A:iJ:
. ^S^p^JS
tiara's »
/tha .«amaa' of lhaaa
FR'OII?BHiRRrs'-KITCHEN:'.',
~ 'n Craao
..............
THE CINCINNATI SOIL PROJECT
1142 MAIN ST.
CINCINNATI, OH 45210
lened 'ice
rats .-container. ^riUiiJ^i
is In a pUstlc'-bVg «i»3
-at 'polling pin or by band.
• . - \ "iK^^"" ~".™>——.-»»——.—^ *o ice cresrn and CooJ IVWp
*-*-i:.-/v^1>*i1«:tii»a.^-'ipoon Mce cream into s freezer bowl
" ' ' "" : aad AV>>>>(
-------
THF "Fstc OF NURSg
AND NUTRITION.
Many of ut have heard a pregnant mother
«y * I'm eating for two now, ao I'll aat
Oil* extra bag of potato chip* and candy
our II la trua that during pragnancy
axtra cilorlaa ara required. However.
II la iha quillty of cilorlaa that count,
not Iha quantity. It la alao Irua that
during pregnancy, ona haa to aupply all
Iha nulrlanta for two. Both the
eipactenl molhar and har baby do naad
mora vitamins and milk than other
people. However, alnca the baby la
email, Iha number of extra caloric* '
required to accompllih thle le only 300
to 100 calorlaa • day. A healthy woman
cm gat moat ol Iha vitamin* and
mineral* needed from • well-balanced
diet Thla Include* aavaral aervlnge of
frulle and vagetablea; euch aa fruit
Juice*, melon*, orange*, broccoli and
graani, whole anrlched gralne; from
bread* and careala. prolalne; euch ••
beat, agga, and beam, and four la five
glim* ol milk a day. No clngl* food
contain* all the nutrient* you and your
baby need. Therefore, eating fooda from
*H ol Die food group* ara Important
titled below are «om* do'* and donta
regarding pragnancy and nutrition:
YOU 00 NEED
SALT - Salt food to tad*, unlea* your
doctor tell* you different.
LIQUIDS • While pregnant you need at
laaat *tx I ounca glaiit* "a day. Thia
could be water, milk, fruit Juice*-and
aoupi.
FATS AND OILS - At leatt 3 tablaipoon*
of fat* and gll* In addition to auggaatad
dial.
IRON AND FOLACIN - You may naid to
aupplemsnl theee nulrlenle with your
diet Check with your doctor about a
•upplamant Be aura to lak* vitamin*
and dietary •uppl*m*nt* a* your doctor
•uggcilil
YOU DOHT NEED
ALCOHOL - Don't drink. Alcohol can
cauaa your baby to be email and can
cauaa other problem* Including mantel
retardation.
DRUGS - Dont take street drug*. Some
drug* can ciuaa your baby to be born
with an addiction or birth datecta. Don't
take any medlclna* euch aa aaplrln
without your doctor* okay.
CIGARETTES • Smoking can caua* amaller
bable* and emallar. bablee are more
liktly to h«va hiiilh probltm*.
NOH FOOD ITEMS - Don't eat thinga like
clay, laundry ilarch or baking *od*. If
you faal like eating thing* that are not
food*, Ull your doctor.
•JUNK" FOOD - Your baby need* lota of
protein, vitamin* and mineral*, that ar*
not provided by many enack fooda.
What YOU eat can make a big difference
In the physical and mental development
ot YOUR BABY. It can alio determine
whether you will have a normal
pregnancy and delivery!
BIRTH
MICHAEL T.
VASUEEN P.
KIERAN G.
AARON B.
07/30/90
10/07/90
10/09/80
10/25/90
FROM
KITCHEN-
CHRISTMAS WREATHS a la
RICE KRISPIES
You can deviae many holiday treats
ualng the well-known Rice Kriapie*
bar reclpa, Including these mini-
Chrlatmaa wreath*.
6 cup* Rica Kriapiaa
4 cup* mini-marahmallowe (or 1 bag
of regular marshmallowe)
1/4 cup mergerine
green food coloring
rad cinnamon candies
toothpick*
Melt margarine in 3 quart aaucepan;
add marahmallowe and cook over
low heel. Stir conaUnlly until
•yrupy. Remove from heal. Add
green food coloring until mixture
becomea a dark green color. Add
cereal and atir until well coeted.
With buttered hands shape mixture
into a 'doughnut*. Cool. Decorele
with rad candle*.
-------
CINCINNATI SOIL PROJECT NEWSLETTER
NO. 6 WINTER 1880
1142 MAIN ST.. CINCINNATI. OH 45210
HAPPY HOLIDAYS FROM ALL OF US: BOB, SCOTT. UNDA. SANDY. JOANN. BILL M.. BILL H.. WINKEY. BELINDA. PENNY,
JOYCE RONNIE. ROSE PAT. JOHN. JIM, SHARON. RANGA, KATE. TRACI, JESSICA. CARLA. TODD. NORM, RAMONA,
TONIA. TANA, KRIS, KELLYE. DENIS, CHRIS, JASON, JIAN UANC, USA. DONNA, HERB, DENISE AND SHERRY
WE WISH EVERYONE THE BEST DURING THIS HOLIDAY SEASON AND IN 1891. WE APPRECIATE YOUR COOPERATION IN
MAKING THE CINCINNATI SOIL PROJECT A SUCCESS IN 1880 AND WE LOOK FORWARD TO CONTINUED SUCCESS IN 1891.
PROGRAM PROGRESS REPORT;
Early In October %»• completed the
removal and replacement of lead
contaminated coll end the cleaning of the
etreeta. perking lota, and alleys In the
area* around Flndlay Market, Grent
Park, and lath end 13th etreeta neer
Reeding Road. Pert of that cleanup work
Involved renovation in aome of the public
pe'rke. In Grant Park we replaced a lot
of eroded eoll, end we laid new eod and
repleced the ereaa under the play
equipment with ehredded berk. In the
Flndlay playgrounda, we replaced the eod
and Inalalled new eoft pley eurfacee
under the play equipment The final atap
In our cleanup process waa to clean the
etreete, aldewalke, alley* end other
paved eurfece* In those ereea where we
completed the eoll removal and
replacement.
Now thet we have completed the 1880
eoll removel end replacement end duet
cleenup, our monitoring etelf will be
reeponelble for collecting eoll end otreet
duet aamples. The purpoae of collecting
the eemplea le to check to eee If the leed
wee removed end to eee If the cleaned
area* become recontimlnated with lead
from other parta of the city. Because of
the need to do this, you will very likely
eee our eleff collecting those eamples on
day* whan weather permits through the
end of Jenusry. After that they will
work In the laba proceeslng the eemples
Betting them resdy for analysis.
In the eprlng end eummer of 1881 we
look forwerd to completing the
abetement end collecting the final
aemplee necessary to complete the eoil
project.
THE CINCINNATI SOIL PROJECT
HOLIDAY SCHEDULE
CHRISTMAS EVE •Mondey, December 24
CHRISTMAS DAY-Tueadsy, December 25
NEW YEAR'S DAY-Tuesaey. January 1
DR. MARTIN LUTHER
KING'S BIRTHDAY-Monday, January 21
CHRISTMAS TRADITIONS:
SANTA CLAUS
About 100 yeers ego, the Senta Clauc as
we know him today was created. H* is
ehort end plump with red cheeks. He
dresees In red clothing with while fur
trim end he travels with a team of
reindeer end e slsigh. This is Amsrice'e
contribution to the Santa Claus tradition.
Up until this time this old whit* baerded
man wee know as St. Nicholas. H* was
dressed In a black robe end rode on a
grey mare, end hi* day of celebration
wee December 6.
HOLIDAY SAFETY TIPS:
1. Kssp ths base of your Christina*
tree wet.
2. U«* only UL approved Christmas
light* or appliances.
I. Chsck Chrislmse lights for frsyed
or worn wires.
4. DO NOT leave your Christmas light*
on when you leave th* house.
S. DO NOT overlocd electricel cords or
outlsts.
6. NEVER run wires under rugs, behind
rediators, over doorwaye or ecross
welkweye.
WINTER SAFETY TIPS:
1. Keep portable heaters away from
curlaine end other flammabl*
materlale
2. B* cure your healer ie In good
working condition. All room
_ heeler* n**d frequent checkups end
cleaning. A neglected heeler I* a
critical fire hazard.
3. Avoid using electric spaca healers in
bathroom*, and do not touch one
when you ere wit.
4. Keep young children eway from
space heeler*, especially II they
are wearing e nightgown.
S. NEVER usa e gas rang* or en oven to
heel your kitchen.
i. DO NOT laava lit ov*n door* opan.
Children could burn lhamselv** on
the door or heating element.
7. II windows are emergency eiits in
your horn*, train your lamily how
to use them in ces* fire ehould
etrik*. Sea to it thet storm
windows opan and that lhay ere not
blocked by objacts.
ATTACHMENT A
-------
THE CINCINNATI SOIL PROJECT NEWSLETTER
NO. «, FALL 1991
1142 WAIN ST, CINCINNATI. OH 45210
PROGRAM PROGRESS REPORT:
1601 I* the year for Interior and
exterior clean-up activities In the
Gterec* and Uohtwk areas. As the
le*m begin to turn to the tea. we ara In the process ,
of completing our final clean-up activities In Hannah
Park. Irtfonnatlon o*thered during tha Soil Projtct
Study, hopefuHy. wl! U1I ut whathar our daanlng
SitflvWii ovar tha Uit thraa yaan ha» tower blood
laad levels.
Wa appreciated your participation In our projact and
want lo say thank you'. It wa» tarrttllc having
(amtfe* that ara committed to Improving chtd haalth
in oureftidy.
n wat • pleasure to aaa avaryona who attended tha
family matting hald at 8awyar Point In August.
pAPEP. PLATE MASK
You wffl mod:
paper plate
non-toxic felt-tip pan*
adstora
plastic apooni u -
tiaaue papar. contvuctlon papar. feathers, gltnar.
felt acrapt. ate
What to do:
1. Hold the plate to your face and ask a parent.
oMar brother or alder or friend to marie where
your ayes and mouth ara.
2. Cut out tha note where the mark* are.
3. Dadde what character you want to be and
design your plate to look that way. (a cat. dog.
down or scary witch)
GCIUEER
6-12
14'
18
24
31
CALENDAR NOTES
i« PREVENTION WEEK
COLUMBUS DAY
SWEETEST DAY
UNITED NATIONS DAY
HAUOnEEN
4. Glue the a spoon to the back of the plate so It
can be used as a handle to hold your mask up
or have a parent make a smaB hole on both
aides of your face you mada • put a string
through them that wll ft ovor your head so
you do not have to hold It up.
NOVEMBER
6 ELECTIONDAY
11 VETERAWSDAY
• 28 THANKSGIVING DAY
WORD SEARCH
KEEP YOUR GHOST AND GOBLINS SAFE
We aB know the enjoyment our
ftSe onei g«t from Halloween,
cVeulng up and going out to
collect their tricks and treats.
TOs KaSoween left keep safety
In mind while preparing to sand
our chldran out to take part In this fun.
1. Be sure the child's costume fits property, so
that the youngster has no problem with
weJWng or moving their arms about freely.
2. H wearing • mask, be sura the child has
proper eye and nose openings.
3. H the costume requires any decoration other
than on the clothing (things to carry) aaa
that b Is Sght weight and safe for your child
to wstk with.
4. H your area's trick or treat time will be
during dusk or dark, a small flashlight would
be helpful.
6. NEVER allow your child to Trick or Treat1
ALONE.
6. Talk to your children about the Importance of
staying with the group and avoiding
sUangtrs.
7, When your child returns home, go through"
htoher trick or treat bag and through away
anything which looks to have been tampered
with. (Anything that Is not Individually store
wrapped)
By folJowfcg these tips a SAFE and FUN Halloween can
A
C
1
U
o
0
H
0
R
8
E
A
Q
B
O
J
F
T
H
Q
0
N
E
A
E
H
E
F
K
L
A
X
K
1
N
0
a
H
A
V
N
A
C
O
H
0
s
T
Y
E
A
P
R
P
8
K
Y
A
H
E
R
E
O
B
N
A
S
H
C
W
L
O
W
1
T
C
H
1
C
R
L
A
E
L
E
R
C
L
B
E
K
S
N
1
L
B
O
G
H
K
R
X
B
P
E
O
a
B
R
W
H
B
O
V
H
N
H
C
U
T
V
P1
E
A
O
R
C
V
E
II
H
K
W
S
V
E
U
A
T
W
B
V
P
U
H
P
K
1
N
A
1
R
N
R
T
R
W
B
8
L
P
W
H
E
:T
H
A
L
F
O
T
A
U
Y
D
N
A
C
A
C
Y
C
B
N
M
O
0
A
T
T
H
F
O
BLACKCAT
CANDY
CARVE
OH06T
OOBLMS
HM1OWSN
MASKS
PUUPKW
PUMPKIN PtE
SCARY
TRICK CM TREAT
-------
WHAT YOUR CHILDREN AND YOU SHOULD DO
IN CASE OF FIRE:
Tho month of October ft designated as national fire
safety month. Fire fascinates children. Even If they
know It b dangerous, they are amazed by (tames.
T«*ch your children what to do In a fire:
1. GET OUT FAST, SECONDS COUNT. Phone
for help from a.neighbor's home, not from
Inside a burning building.
2. CRAWL LOW under the smoke.
3. TEST tie door. 0 tt Is hot or there Is smoke.
use another way out
4. ONCE OUTSIDE, STAY OUT. There Is
nothing more Important ki your home than you.
IF CLOTHING CATCHES ON FIRE:
Some children falsely believe that their domes will
protect them from fire. Show children, age 3 and
older, what to do If their dothas catch on fire.
Practice with them.
1. STOP. Running fans the flames, making the
fire bum faster. SHOUT for help. Oont run
for help.
Z. DROP to the floor. Cover your face.
3. ROLL back and forth to put out flames.
4. COOL a bum with cool water.
TEACH YOUNG CHILDREN THAT:
-Matches and lighten! are tools for adults, not
toys.
-Children who play with matches or lighters
can be badly burned and can hurt others.
-If they find matches or lighters, tell an aduK
the location right away.
THANKSGIVING._and HOW IT STARTED
We celebrate Thanksgiving as a cultural holiday and a
religious observance. Abraham Lincoln was the first
President to Issue a proclamation marking the last
Thursday In November our annual Thanksglvfng Day.
In 1941 the day was officially declared a national
public holiday. The traditional turkey and pumpkin pie
may have their beginnings In the Thanksgiving Day of
the early American settlers, but the Idea of
celebrating the completion of the harvest end
rendering homage to the spirits Is an andent custom
that Is em) practiced In may foreign lends. Although
some part of this American holiday Inds many men
watching football games and many women In the
kitchen preparing foood. families are united at the
dinner table for the traditional festive meal - a
gathering reaffirms family ties.
RECIPES FOR EVERYONE:
EASY PUMPKIN PIE
1 large can of pumpkin
1 can of evaporated milk
2tsp pumpkin pie spice mix (or to taste)
2 aggs
3/4 cup sugar
1/2 tsp *alt
1 frozen 0 Inch pie crust
Thoroughly mix pumpkin, canned milk, eggs and
pumpkin pie spice together. Add sugar and salt. Pour
into pie crust. Preheat oven to 425* and bake for 15
minutes, reduce heat to 350* and continue to bake for
another 45 minutes or until you can stick a knife In
the center of the pie, put! It out and the knife comes
out dean. Let pie cool completely before serving and
If desired put cool wrttp on top of pieces when serving.
ENJOYBII
JACK-O-LANTERN PIE
1 pkg orange Jollo (4 serving size)
1 cup boiling water
1 pint vanilla Ice cream
1 pkg chocolate crumb pie crust
1 small container of cool whip
candy com
Mack licorice cut into strips
Dissolve Jetlo ki the boiling water. Stir In ice cream
until thoroughly mixed. Place In freezer for 10
minutes. Remove from freezer and pour Into pie
crust, place In refrigerator for 2 hours. Before
serving cover with cool whip and use candy corn and
sconce to make a ]ack-o-lantem face on the pie. The
children will love it.
HAPPY HOLIDAYS!!!!
U.C. DEFT. OP ERVIB'. HEALTH
THE CIlfCIHKATI SOIL PROJECT
1142 KAIK STREET
CINCIHHATI, OHIO 45210
-------
-------
APPENDIX E
Institutional Review Board-approved Consent and Release Forms
-------
Family l.D.
Child l.D. _
INFORMED CONSENT STATEMENT
UNIVERSITY OF CINCINNATI MEDICAL CENTER
Soil and Dust Abatement Study
11. INTRODUCTION
Before you agree to participate in this study, it is important that the following
explanation of the proposed procedure be read and understood. It describes the
purpose, benefits, risks and discomforts, and precautions of the study. It also decribes
the right to withdraw from the study at any time. It is important to understand that no
guarantee or assurance can be made as to the results. It is also understood that refusal to
participate in this study will not influence standard treatment for the subject.
II. PURPOSE OF STUDY
I,. ., agree to participate in a medical study the
purpose of which is to determine the amount of lead in my blood or that of my child
(children), . — Also, my home will be surveyed to
identify possible sources of lead in soil, dust, paint and water. This information will be
used to examine the effectiveness of dust and soil lead removal methods.
III. PROCEDURES AND RISKS
I have been told that the blood sample obtained will be measured for lead. A small
amount of blood, about one teaspoon, will be drawn in order to permit measurement of
lead and related measurements such as iron. I know that my child may cry for a moment
when his/her finger or arm is pricked to obtain blood for the lead measurement. The risk
of simple venipuncture include: commonly, the occurence of discomfort and/or bruise
atthesite of puncture; and less commonly, the formation of a small blood clot or swell ing
of the vein and surrounding tissue, and bleeding from the puncture site. If I am a woman
and I am or should become pregnant, there is no risk to me or my fetus by participating in
the study. My blood will be drawn once. My child's blood will be drawn five times at six
month intervals. I have also been told that no risks are associated with the survey of my
residence for lead sources or with procedures used to reduce the amount of lead in soil
and dust around my home. I will be participating in the protocol for approximately 2
years or until I move from my current residence, which ever comes first. If there is a
significant variance from the stated time period, I will be notified.
IV. BENEFITS
Several benefits arise as a result of my participation on this study. First, I will be notified
of the blood lead results. Second, the health department will be notified if high blood
levels are found so that appropriate actions can be taken. Certain environmental
improvements, such as street cleaning and playground improvements may also occur in
my neighborhood.
-------
Family I.D.
V. CONFIDENTIALITY OF RECORDS
All information gathered will be kept private and confidential (my and my child's records
will be identified by a code number which will be available only to the study
investigators). This information will not be made available to anyone not connected with
the study without my permission.
VI. AVAILABILITY OF INFORMATION
Any questions that I may have concerning any aspect of this investigation will be
answered by Dr. Scott Clark or an associate at 558-1749
VII. COMPENSATION
The University of Cincinnati Medical Center follows a policy of making all decisions
concerning compensation and medical treatment for injuries occuring during or caused
by participation in biomedical or behavioral research on an individual basis. If I believe I
have been injured as a result of research, I will contact Dr. Scott Clark at 558-1749 or Dr.
John Vester, Chairperson, U.C. Medical Center Institutional Review Board at 558-5259.
VIII. FISCAL RESPONSIBILITY
Funds are not available to cover the costs of any on-going medical care and I remain
responsible for the cost of non-research related care. Tests and studies done soley for
the purpose of research will be paid for from research funds and I should not be billed for
them. Blood sampling will take place at a clinic near my home. I understand that I will
reimbursed $10 for expenses incurred in bringing my child to the clinic. If I have
questions about my medical bill relative to research participation, I may contact Dr.
Scott Clark.
IX. THE RIGHT TO WITHDRAW
I am free to withdraw from this survey at any time. Should I wish to withdraw, I have been
assured that standard therapy for my child's condition will remain available to me. I have
been informed of the probable consequence of my withdrawal from the study.
Withdrawal should be made in writing and a form will be given to me for that purpose.
X. Are you (or your child) currently participating in another study?
DYes
Investigator Title of study
QNo
Consent and witnessing:
Subject: Date:.
Caregiver (legal guardian): ; Date:.
Investigator: : Date:.
Witness: Date:.
Copies: Investigator's file, and subject or legal guardian.
-------
Release of Medical Information
I authorize The Cincinnati Soil Project to release the following Blood Collection
results: Blood Lead and EP, Hematocrit and Hemoglobin and Iron Studies to the
Women Infant and Children (WIC) office, physicians and health care agencies
involved in our medical care.
The following child/children and myself (Caregiver) shall be covered under this
release:
This authorization will remain in effect until such time as revoked in writing
by me.
Careglver's signature Soil Project signature
DATE DATE
-------
CONSENT TO WITHDRAW FROM STUDY
DATE:
, wish to withdraw my child,
, from the medical study.
I have been informed that at the present time my child's blood lead level is
Normal/Elevated/Unknown*. I have also been told that it is important that I
continue to take my child to a clinic for routine medical check-ups, and to have
his/her blood checked for lead at least every months.
Consent and Witnessing:
Date: ______
Mother/Caregiver
' Date:
Investigator
! Date:
Witness
Copy to: Mother/Caregiver
Investigator
Child's Chart
* INASMUCH AS YOUR CHILD'S BLOOD WAS NEVER DRAWN, HIS/HER BLOOD
LEAD LEVEL IS UNKNOWN.
-------
-------
APPENDIX F
Interior Dust Clean-up Methods Development Manuscript
"Clean-up of Lead in Household Carpet and Floor Dust"
A manuscript submitted for publication to the American Industrial
Hygiene Association Journal.
-------
Clean-up of Lead in Household Carpet and Floor Dust
Lynda Ewers1'2, Scott Clark1*, William Menrath1,
Paul Succop1 and Robert Bornschein1
Department of Environmental Health
University of Cincinnati, Cincinnati, OH 45267-0056
2PRESENT ADDRESS:
National Institute for Occupational Safety and Health
Cincinnati, OH 45226
Corresponding Author
-------
Abstract
Effective methods to remove lead-containing dust from
household carpets and other floor areas are necessary to reduce
exposure of young children to dust lead from household paint as
well as that entering the home from other sources. Since such
methods were not available in the literature it was necessary to
develop them for use in a Superfund-supported soil and dust lead
abatement demonstration project. Methods were tested on carpets
obtained from homes of children with high blood lead and on new
carpets artificially-contaminated in the laboratory using an ASTM
method. Carpets removed from these homes were not able to be
cleaned effectively in the laboratory by repetitive cleaning with
HEPA filtered vacuum cleaners. The lead concentration in the
removed dust remained about the same from the initial cleaning (1
min/sq.m.) to the final cleaning (total cleaning time of 10
min/sq.m.). The lead loading on the surface of the carpets often
increased during the cleaning process due to the action of the
vacuum cleaner in bringing to the surface lead from deeper in the
carpet. For bare wooden floors over 95% of the total dust removed
by the combination of dry vacuuming (5 min/sq.m.) followed by wet
washing was removed by the dry vacuuming. For linoleum, more than
75% was removed by vacuuming for 5 min/sq.m. However little was
removed in vacuuming after the initial two minutes and about 20%
2
-------
was removed in the final wet washing step. HEPA vacuuming of the
new laboratory-contaminated carpets revealed that two of the
commercially-available vacuum cleaners tested were essentially
equivalent and each removed significantly more dust than a third
vacuum cleaner during a cleaning process of a total duration of 10
min/sq.m. Cleaning for 6 min/sq.m. was necessary to remove more
than 70% of the embedded dust by the two more efficient vacuum
cleaners. Cleaning efficiencies were about the same for short
pile and sculptured carpets.
Based on this research and recent reports from others
developing similar methods, it was concluded that in many
situations it was more practical to replace the carpets rather
than to perform costly and only partially-effective cleaning
procedures. HEPA-vacuuming cleaning of carpets was shown to
increase lead-dust on the surface under some conditions.
-------
The reduction of young children's lead exposure is one of the
nation's most important environmental health goals. Many
investigators have concluded that lead-contaminated house dust is
a critical link in the exposure pathway, resulting in elevated
blood lead of young children. The lead contaminants in dust are
primarily derived from lead-based paints which were former widely
used on both the interior and exterior of housing. Lead dusts can
also enter housing from areas outside the home with lead-
contaminated dust and/or soil, as well as through lead-containing
dust carried home from the workplace. A typical exposure scenario
involves lead in exterior dust, from soil and paint, contributing
to lead in interior floor dust, which is then picked up on the
hands of children, where it is subsequently ingested resulting in
elevated blood lead. The childhood lead exposure pathway model
that has been developed for Cincinnati and other areas is
summarized in Figure I11"61 suggest that residential lead abatement
activities should include the cleaning of locations in the house
where lead-contaminated dusts are present, such as carpets and
other floor areas.
Because of the known relationships between interior dust
lead and blood lead, interior dust abatement was included in the
exposure reduction strategies of the Cincinnati Soil Lead
Abatement Demonstration Project. This project was one of three
-------
carried out under a provision in the Superfund Amendments and
Reauthorization Act (1986) which called for a "... pilot program
for removal, decontamination, or other action with respect to
lead-contaminated soil in one to three different metropolitan
areas". Interior dust abatement is also an important
consideration as an essential component of lead-based paint
reduction strategies. The U.S. Department of Housing and Urban
Development (HUD) has acknowledged the importance of lead-dust in
its Interim Guidelines for Lead-Based Paint Abatement*71 and has
recommended the removal of furnishings, including wall-to-wall
carpeting, prior to abatement. Recommendations for the cleaning
of furnishings such as carpets and furniture prior to their being
returned to the abated dwelling unit are not yet included in HUD
guidelines. These items may well have been contaminated prior to
the lead-based paint abatement process, and therefore placing them
back.in the housing will only reintroduce an exposure source to
the otherwise cleaned housing.
A search of the literature revealed only a single paper
systematically documenting efforts to remove lead-dust
contamination from carpets.181 Milar and Mushak181 concluded that a
two-part steam cleaning, initially using a sodium
hexametaphosphate commercial cleaning product, followed 24 hours
later by steam cleaning using a commercial detergent, removed
-------
about 60% of the lead-dust expressed as a surface loading basis
and reduced the lead concentration by 30 to 50%.
In a previous publication we reported results of a
preliminary evaluation of Milar and Mushak's181 procedures, with
the additional step of dry-vacuuming with a HEPA-equipped vacuum
cleaner. The test was carried out in several homes contaminated
by worker take-home of lead-contaminated foundry dust.<9) Although
these vacuum and shampooing procedures can remove some of the
lead-containing dust from carpets, it is important to know whether
these methods reduce the amount of lead dust at the surface (mg
Pb/sq.m.) where it is most available to children. (Normal
vacuuming will tend to redeposit below surface dust through
bringing it to the surface from deeper in the carpet and through
non-filtration through vacuum cleaner bags of high porosity.)
These methods were found to either have only limited impact on, or
even increase surface loading as determined by an exposure-related
sampling method developed by Que Hee et al_. Repetitive dry
vacuuming was also previously evaluated on three carpets in a
routinely-cleaned suburban home where exterior paint sanding had
occurred sixteen months earlier.'91 In this latter case surface
dust on an area or loading basis (//g Pb/sq.m.) decreased with
vacuuming by between 64 and 94%. These experiences suggested that
cleaning "chronically-contaminated" carpets (those which received
-------
lead dust over an extended period of time, could actually increase
the likelihood of lead exposure, whereas cleaning of "acutely-
contaminated" (i.e., from a single episode) carpets may be
effective in reducing exposure. Based on preliminary work, it was
concluded that additional development of surface dust abatement
methods was needed, particularly for heavily-contaminated carpets.
There was also a need to determine the efficacy of the variety of
HEPA vacuum cleaners available, the length of cleaning time
necessary to significantly reduce lead loading, and the
variability in the effectiveness of cleaning among different
workers.
-i
METHODS
Two dust collection methods were used to measure the
effectiveness of cleaning. The first used the contents of the HEPA
vacuum cleaner paper bag as a measure of the total amount of
removed dust and lead. The second was a surface dust vacuum
method previously developed by Que Hee et al.(10) and which has,
been shown to produce results which correlate with exposure as
measured by blood lead.'1"61 The usefulness of the vacuum cleaner
as a sampler depends, in part, on whether the vacuum cleaner
removes a satisfactorily large proportion of the available dust.
-------
The method of Que Hee et al.1101 utilizes a light-weight, battery-
operated personal air sampling pump (2 liters per minute) to
collect dust in a filter holder while sampling over a known area.
This method, known as the "dust vacuum method", was not designed
to determine the total dust lead on a surface, or within a carpet
but it was intended to collect the dust most readily available to
a young child.
Two sources of contaminated carpets were used: (1)
contaminated carpets removed from homes of, children with high
blood lead; and (2) new carpets with artificial dust embedded in
them a laboratory setting. The variations in effectiveness of
different available vacuum cleaners and operators on different
carpet types were examined in the new carpets contaminated in the
laboratory. The effectiveness of methods for cleaning linoleum and
unpainted wood floors with different surface types was also
evaluated. Data were analyzed using the Statistical Analysis
System.1111
Laboratory Cleaning of Contaminated Carpets
Thirteen existing carpets from inner-city homes, several
where children had been lead poisoned, were removed and replaced
-------
with new carpets and brought to the laboratory for testing. Ten
cleaning "efforts", each defined as cleaning with a HEPA filter
equipped vacuum cleaner at a rate of one minute/square meter, were
performed on one square meter areas of each carpet. Thus, each
carpet was cleaned for a total of ten (10) minutes per square
meter. The extent of cleaning was chosen to be length of time of
cleaning and the unit rate of cleaning was chosen to be one minute
per square meter of carpet. This rate was also convenient for
manipulating the nozzle in a systematic way. The pattern used was
three up/back passes with the nozzle over the area. The size of
the "beater bar" nozzle was sufficient to just allow coverage, ;
within one minute, of a one .square meter area in these three
passes with very little overlap. This complete coverage of a
square meter in three up/back passes at a rate of 1 minute/square
meter is called "one cleaning effort". Preliminary studies
confirmed reports from industry sources'13', that a nozzle with a
rotary "beater bar" driven by an independent motor was more
effective than a nozzle without a "beater bar". A readily-
available commercial beater bar nozzle (Sears Kenmore) was used
for the cleaning. The weight and content of the vacuum cleaner
dust bag were obtained after successive cleaning efforts. Surface
dust measurements were collected using the method of Que Hee et
al.t101 at two pre-determined locations on each of the carpets
-------
before cleaning commenced, and after the first, third and tenth
cleaning efforts were completed.
, Equipment. Operator. Loading and Carpet Variability
In-laboratory cleaning of new carpets with artificially
embedded dust using a modified version of ASTM procedure F 608-
86(121 was used to determine the effects on cleaning performances
of equipment, loading, carpet type and operator. This ASTM
procedure specifies the method of imbedding a test dust mixture
into the new carpet prior to their being subjected to the
vacuuming protocol. The experimental design included three HEPA
vacuum cleaners, two carpet types (sculptured and short pile), two
dust loadings (100 and 400 g/sq.m.) and two operators.
The test dust mixture was made up of 100 g and 400 g
allotments and consisted of a sand and talc mixture in a 8:1
ratio. The talc was a USP Grade Supreme Talc purchased from a
local supplier. Ninety-five percent of the Keener sand used was
retained in the 210-299 //m size fraction. The two components of
the test dust mixture were weighed and mixed individually before
each test.
A one square meter surface area in the center of each carpet
served as the test area. The carpets were preconditioned as
10
-------
specified in the ASTM procedure by adding test dust to their
surface and removing it by vacuuming until <2 g of material was |
removed in a cleaning effort. Repeated trials of each carpet type
were conducted on the same square of carpet. For each test, the
carpet was affixed to a piece of plywood. The appropriate amount
of dust was spread onto the delineated area and was embedded into
it by dragging an embedment tool (as described in the ASTM
procedure)'121 over it for 30 strokes, alternating directions
forward and back.
i
Three commercially-available vacuum cleaners were used (see
Table 1 for details). Each of these was intended for industrial
use and had three levels of filtration. The initial filtration -
level was similar to a home vacuum cleaner in that a removable
paper bag catches most of the dust. The second level of
filtration consisted of a more permanent cloth filter. This
filtration level does accumulate dust over a period of time but
was not found to need changing over the period of these trials.
The third layer of filtration was a HEPA filter which removed the
remaining fine material and also did require changing during the
» ;
study.
A readily-available commercial (Sears Kenmore) "beater bar"
nozzle was substituted for the nozzles provided with the three
types of vacuum cleaners. Preliminary studies led to the
11
-------
Table 1
Vacuum Cleaners Tested
Cleaner
Model
Cubic Feet
Per Minute
Water Lift,
Inches
Nilfisk GS80
87
75
B
WAP 767
100
90
Euroclean UZ930
77
85
conclusion that the "long" pile setting for the nozzle always gave
the best cleaning regardless of the actual pile length. This
setting was the only one used in these experiments.
The Cleaning Effort
Ten cleaning efforts were performed in each trial. Two
experienced operators of the vacuum cleaners were used. The
number of different types of carpets, vacuum cleaners, dust
amounts, and operators made it impractical to perform cleaning
trials for each possible combination of these factors.
12
-------
Consequently, a randomized, incomplete block design was used.
Randomization of both treatment combinations to blocks and the
ordering of the treatment combinations was performed. The 24
treatment combinations were randomized to 3 blocks of 8 treatment
combinations each. Two of these three blocks were randomly
selected to be performed in the experiments.
Analysis
The dependent variable in these experiments was the amount
of the test dust collected in the vacuum cleaner bags after .1, .2,
3, 4, 5, 6, 7-8, and 9-10 cleaning efforts.
Statistical analysis was performed by the General Linear
Model (GLM) procedure of Statistical Analysis System.'111
Bare Floors
In order to test the effectiveness of proposed tare floor
cleaning procedures, homes with high levels of lead in the dust
were needed. Such homes were located through another research
project studying lead's neurobehavioral effects on young
children.'11 A list of homes meeting the following criteria was
compiled:
-------
(1) High lead concentrations in the interior dust;
(2) Either carpet, vinyl, or wood flooring or a
combination of those flooring types.
Permission from study participants was obtained to test cleaning
methods on different wood and vinyl floors.
The questions to be answered in these homes were:
(1) In order to reduce dust lead to an acceptable level
with a HEPA-equipped vacuum cleaner, what rate of
speed should the operator use and how many times
should the floor be vacuumed?
(2) Does wet washing after the final vacuum cleaning
remove additional lead?
*
The testing procedure was as follows:
(1) Three separate squares, each one meter by one meter,
were delineated on the surface of the floor type being
tested.
14
-------
(2) A bag for a HEPA-equipped vacuum cleaner '("Nilfi.sk)'
was tare-weighed and installed.
(3) The three squares were then vacuumed with the vacuum
cleaner at the rate of one minute for each square
meter.
(4) The vacuum bag was removed and weighed after each
cleaning.
I
(5) The process was repeated until the three squares were
'J ., •«• i '
cleaned a total of five times. • ,
(6) Each of the squares was then washed with 1,500 ml of
tap water from the residence. The washing was
performed with a new sponge by a researcher wearing
rubber gloves. An aliquot of 500 ml was taken from
each of the wash buckets. A 500 ml sample of clean
water from a wash bucket was also collected to provide
data on background lead levels in the tap water.
All the dust samples and water samples were analyzed for lead
concentration in a laboratory that was participating in a
proficiency program established for environmental lead samples by
15
-------
EPA through its Environmental Monitoring and Support Laboratory
(Las Vegas). The samples collected by the dust vacuum method were
digested using hot nitric acid and analyzed by a flame AA method
with a detection limit of 0.1 jjg lead. Dust samples collected in
vacuum cleaner bags were sieved at 250 jm and analyzed by XRF
(detection limit 20 ppm). Water samples were analyzed by graphite
furnace AA (detection limit 1 ppb).
RESULTS
Laboratory Cleaning of Contaminated Carpets Collected from
Homes of Inner-City Children
The amounts of lead removed (mg/sq.m.) by the HEPA vacuum
cleaners at each measurement interval are presented in Table 2.
The cumulative percentages removed, based on the total amount
removed after ten cleanings, which was assigned a value of 100%,
are also shown. The amount of lead removed by ten cleaning
efforts averaged.278.7 mg/sq.m. with a range of 8.6 to 1107.3.
Four cleaning efforts resulted in an average cumulative percentage
removal of 74% of the total amount removed with a range of 61-89%.
Lead concentrations of the material removed from the HEPA vacuum
cleaner bags were significantly different between carpets (p
<0.05) (Figure 2) ranging from 192 to 3226//g/g but did not vary
16
-------
significantly from first to last cleaning efforts for individual
carpets. Thus, the HEPA vacuum cleaner did not appear to
preferentially remove dust with either high or low lead
concentration.
A comparison of the initial surface dust lead loading by the
method of Que Hee et al.(10) with the total amount of lead removed
by 10 nvinutes/sq.m. cleaning with a HEPA vacuum cleaner revealed
that the surface dust lead loading was an average of 1.4% of the
total dust lead removed by extensive vacuum cleaning (range 0.12
to 3.3) (Table 3).
Surface dust lead levels (expressed as mg Pb/sq.m.), were
.-
also determined before cleaning and after the first, third and <
i
tenth cleaning efforts. The ratio of the surface dust lead levels
after the first, third or tenth cleaning efforts, to the surface
dust lead levels before the initial cleaning (Table 4) revealed
that surface loadings, on average, decreased with cleaning. j
However, in several instances, surface loading increased by up to
almost four-fold. These data reveal that vacuum cleaning if not
17
-------
O)
V)
W
CO
O
U
• =J
' 00
•o
(U
o
CO
01
CD
CD
at
CM •»->
,£ §
ja s-
nj CO
t s
fc S
a. s
CD
C
(U
co
c
o
O
(f.
o
(U
o
o
o-
00
CO
O)
en ^
•^ o*
5 o
^ i
n o:
u
I
S3
O
4J
O
<4-
«*-
LU
•o
co
I
«*-
LU
T3
C
CO
m ^
cu
CL
CD O O O O O CD
O O O O O O O
•—< i—< i—« i—i t—i i—i ,—1
oo o> r^. to co r*» i—i
• • • • . . .
CM oo CM co ot t^ co
CM ^S- t—I T—I CM
co •-*
5f CT> O* i-« CO «-i «ef-
OO CO U> VO t^ t£> VO
o» m **• 10 CM ,00 oo
• ••••••
C3 en o oo «a- oo cs
en CM
CT» in I-H •—< oo oo un
t^- oo to in 10 OO t—< to OO
•••••••
t—« 00 ^1- to to 1^ O
•—I CO CO
r—I
^*" oo en en en to in
r-* r-»- ^- co m co «*•
to i—i «a- CM oo oo r-.
• -.....
CM to «a- CM en t<~ i—i
co «*• «a-
CM
cr> co h«» >«*• in r~» to
in 10 CM CM «a- CM CM
in
o
to
cn
in
en
cr> CM
•!*• CM CO CM
cr> ca co •—i
CM I—
CO
LU
-------
^_,
•o
o>
c
•r"
C
O
U
CM
CU
n>
J-
3
S-
o
fc
LU
«C
4-»
0
1
in
E=
ii S*
*^
«i
El «
^Z
•
., 1
•fj
1
LU
_£.
4J
*3-
3 ^_^
ii S"
il N=
•
•r
••^
s-
o
^*~
LU
•a
CO
ii S"
1 ~E
•
3
3 — ^
4->
O
n~
LU
"0
c.
CM
if, 55
CT N
F— I ^^
*
B •-*
^ 3 ^
4->
I
LU
4J
00
«— I
0
^} ME
J3
4->
0)
CX
S-
n)
CJ
^^^^
o
o
I— 1
**"*
t— 1
in
cr>
CO
CO
in
CO
i
10
r— 1
in
CO*
to
" — '
CM
•
t— 1
in
^^
o
to
rt-
*
co
in
CO
••D
^^
O
i— <
**^^
CO
00
o
CO
CM
^.
o"
in
CM
to
* — '
CO
•
to
^_^
CO
CO
*
cn
r— <
^
^-^
o
0
1—4
^"^
CO
^
^
§
<•
o
CM
to
to
CM
CO
CM
in
* — '
r--.
•
5
^_^
CM
cn
•
UD
O
CO
0}
a>
OS
S-
co
>~
>
n>
•o
C
cn
oo
CO
(O
00
S-"
o
*i—
CU
cn
c
c
CU
7J
c
a>
"^
c
•o
>
o
cu
^
c
, 3
O
r—
rt
o
4->
Ci"
00
nj
CO
_J
TO
C
rt
(S
o
00
4^
d)
a.
s~
0
CU
•o
r^
£Z
•r-
0
4->
0)
r—
fO
•r™
rt
^
nj
T3
c:
O)
u
^
It-
~2
CO
C
1— I
J?
-------
Table 3
Comparison of Surface Dust Lead Loading with Total Lead Removed
by Vacuum Cleaning for Ten Minutes Per Sq Meter of Carpet3
(10 Cleaning Efforts)
Carpet
A
B
C
E
F
H
I
J
L
Average
Initial Surface
Oust Lead
Loading
mg/sq.m.
0.591
5.65
2.34
4.47
0.186
0.774
0.301
3.87
0.114
2.03
Total Lead Removed
by HEPA Vacuum
mg/sq.m.
17.8
252.7
t
1107.3
291.4
72.1
70.8
8.6
589.6
97.8
278.7
Initial Surface
Lead as a Percent to
Total Lead Removed
3.3
2.2
0.2
1.5
0.26
1.1
3.5
0.06
0.12
1.4
aCarpets tested were from homes of children with high blood lead.
20
-------
conducted for a sufficient time, has the potential for increasing
childhood exposure. In none of the carpets tested was the surface
dust lead loading after ten cleaning efforts as high as it was
prior to cleaning.
Table 4
Ratio of Surface Dust Lead Loading (mg Pb/sq.m.) After Specified
Number of Cleanings as a Fraction of the Initial Pre-Cleaning
Surface Dust Lead Loading"
Average
Ra^nge
% > 1.00
After
First
Cleaninq
0.55
(0.11-1.72)
1/f
After
Third
Cleaninq
0.47
(0.06-3.93)
5
After
Tenth
Cleaninq
0.20
(0.06-0.61)
0
"Carpets tested were from homes of children with high blood lead.
Variations in Cleaninq Performance of Vacuum Cleaners, Carpet
Types, Dust Loadings and Operators
A statistical analysis of the effects of vacuum cleaner
type, laboratory-embedded dust amount, and changes in amount of
dust collected during successive cleanup and interaction, is
presented in Table 5. Statistically significant differences were
21.
-------
Table 5
Final Model - ANOVA of Amounts of Synthetic Dust
Recovered in Repeated Vacuuming of Carpets
Source
Repeated Dust Weight
Repeated Dust Weight-
Vacuum Cleaner
Repeated Dust Weight-
Dust Amount
df
14
F-Value
PR>F
Between Sub.iect Effects
Vacuum Cleaner
Dust Amount
Within Sub.iect Effects
2
1
15.11*** 0.0005
20.32*** 0.0007
268.72*** 0.0001
2.30** 0.0101
2.50** 0.0222
**Significant at a = 0.05 level
***Siqnificant at a = 0.01 level
22
-------
observed between vacuum cleaners, amounts of embedded dust and in
amounts of dust collected during successive cleaning efforts.
The results of the cleaning of new carpets containing
laboratory-embedded dust are presented in Table 6 (a-c) as the j
mean proportion of the embedded dust removed by the ten cleaning
efforts for each of the three different vacuum cleaners, two
carpet types and two dust amounts used. Two of the three vacuum
cleaners (A and B) removed a statistically higher proportion of
the dust than the third (C). Contrasting the performance of the
vacuum cleaners (Table 6a), indicated no statistically significant
differences between the amounts of dust removed by vacuum cleaners
A and B. Results for vacuum cleaners A and B were different from
those for cleaner C for all cleaning effort (Figure 3). None of,
the cleaners removed an average of more than about one-fourth of
the embedded material with the first cleaning effort (one minute
per square meter of cleaning). Three cleaning efforts (three
minutes per sq.m.) were required to remove one-half of the dust
for two of the cleaners (A and B) while twice as long (6
min/sq.m.) was required for the third vacuum cleaner (C).
Vacuum cleaner C did not do as well as either A or B
beginning with the first cleaning efforts (Table 6a). Although
somewhat more dust was removed during the first five cleaning
efforts for short pile carpets than for sculptured carpets,
slightly more was removed during later cleaning efforts from
sculptured carpets (Table 6b). These small differences in dust
removal between the carpet types were statistically significant
only for the first cleaning effort. A larger proportion of dust
23
-------
Table 6
Proportion of Total Test Dust Removed After Specified Number
of Cleaning Efforts by Vacuum Cleaner (Table 6a),
Carpet Type (Table 6b) and Dust Amount (Table 6c)
(a) Vacuum Cleaner Comparisons
.Cleaning
Efforts
1
2
,3
4
5
6
8
10
Least Square Means
Vac A
0.260
0.455
0.553
0.620
0.671
0.709
0.769
0.811
Vac B
0.268
0.439
0.557
0,642
0.701
0.755
0.825
0.875
Vac C
0.142
0.232
0.309
0.371
0.429
0.481
0.573
0.645
Probability of Difference
Among Vacuum Cleaners
AC BC AB
0.0078 0.0072 0.834
0.0074 0.0150 0.8209
0.0024 0.0031 0.9583
0.0007 0.0006 0.6947
0.0004 0.0002 0.5674
0.0004 0.0001 0.3454
0.0006 0.0001 0.2116
0.0012 0.0001 0.1304
24
-------
Table 6b
Carpet Type Comparisons
Cleaning
Efforts
1
2
3
4
5
6
7-8
9-10
Least Square Means
Sculptured
0.192
0.329
0.443
0.529
0.598
0.657
0.740
0.801
Short-pile
0.256
0.422
0.503
0.560
0.602
0.640
0.704
0.753
Probability of
Difference Between
Carpet Types
0.0306
0.1085
0.2822
0.5316
0.9304
0.6876
0.3348
0.1440
25
-------
Table 6c
Dust Loading Comparisons
Cleaning
Efforts
1
2
3
4
5
6
7-8
9-10
Least Square Means
100 G/m2
0.294
0.478
0.578
0.644
0.690
0.731
0.791
0.834
400 G/m2
0.153
0.273
0.368
0.445
0.511
0.565
0.653
0.720
Probability of
Difference Between
Dust Amounts
0.0006
0.0039
0.0019
0.0010
0.0010
0.0012
0.0022
0.0047
26
-------
was removed when the 100 gram amounts were applied (Table 6c) and
the differences between cleaning efficiency for the two amounts
were statistically significant. (The dust amounts applied were in
the range of those removed from contaminated carpets as shown
elsewhere in this paper.)
There was no significant difference in dust removal between
the two operators tested.
The fraction of dust removed was lower when the quantity
embedded in the laboratory was higher (Table 6c). This difference
was statistically significant. ,,
Bare Floor Cleaning |
Two rooms with unpainted wood floors and two with linoleum-
covered floors were test cleaned with the HEPA vacuum cleaner and
wet washing procedure described earlier. Results of the cleaning
efforts show that for wood floors, a total of about 95% of the|
dust removed was removed by five vacuum cleaning efforts and the
follow-up washing yielded only 2-5% of the total dust collected by
both methods (Table 7). Results for the linoleum floors were
markedly different with the vacuuming effects. After the second
minute, these vacuum cleaning efforts were relatively
unproductive, with 19-22% of the dust remaining and being removed
by the wet washing. About one-fifth of the total dust collected
on linoleum was not amenable to removal by the vacuum cleaner |
method.
27
-------
DISCUSSION
There are few reports in the literature on the efficiencies
of methods of cleaning carpets and bare floors. In a pilot study
of methods to remove lead dust from homes near a secondary lead
smelter in Toronto, Canada, a pilot testing of vacuuming of carpet
and floors followed by wet shampooing was carried out in eight
houses.'141 An average of 5.2 times as much lead was removed by
the vacuum procedures (about 1 min/sq.m) than by the wet method.
The average quantity of lead removed from all areas of the houses,
including heating ducts and basement ceilings, was 8.1 grams of
lead per house.
Kominsky et al.1151 found that for new carpets, embedded in
the laboratory with asbestos fibers, use of HEPA filtered dry
vacuum cleaners at 1.4 min/sq.m was ineffective in removing the
asbestos fibers. HEPA-filtered hot water extractions operated for
about 1.4 min/sq.m. reduced the level of asbestos contamination in
the carpet by approximately 70%.
A pilot study of the efficiency of HEPA vacuuming and
shampooing of carpets and upholstered furniture from homes near a
lead smelter in Bunker Hill, Idaho was conducted.1161 The cleaning
protocol involved an initial vacuuming
28
-------
Table 7
Percentage of Dust Removed From Floors by HEPA Vacuum
Cleaning and Follow-up Washing With Tap-Water8
(Vacuum Cleaning
Type Surface
Bare Wood
Bare Wood
Bare Wood
Bare Wood
Linoleum
Linoleum
Linoleum
1st
62
60
53
14
69
69
67
2nd
13
12
18
35
12
7
11
Efforts)
3rd
5
9 ...
14
16
0
5
0
4th
7
9
4
21
0
0
0
5th Follow-
up
Washing
[
9 3
5 5 ;
6 4 . i
12 2
19
19
22
"One cleaning effort « one minute per square meter of floor. •
i
—Fifth cleaning effort not completed because prior cleaning did
not yield a detectable amount of dust.
29
-------
followed by five consecutive shampooings and a final vacuuming.
The total amount of lead in the carpet and furniture was
determined following laboratory digestion and analysis of measured
portions of those items. Samples were analyzed by graphite
furnace atomic absorption. A summary of the results of this study
(Table 8) indicates that the cleaning efforts for the carpets were
able to remove an average of less than 8% of the lead in the
carpet. For furniture with only three shampooings and no final
vacuuming, an average of 18% of the lead was removed.
An extrapolation for the first carpet in Table 8 revealed
that 74 additional shampooings would be needed to remove all of
the lead. The total lead loading (mg/m2) in the carpet in the
homes near the smelter (average 1068, range 185-3044) was somewhat
higher than the amount removed during the current study (average
418, range 8.6-1107) with some overlap.
CONCLUSIONS
Carpets removed from poorly-maintained houses containing
lead-based paint sources were not able to be cleaned effectively
in the laboratory by repetitive cleaning with HEPA-filtered vacuum
cleaners. The lead concentration in the removed dust did not
change significantly from the initial cleaning (1 min/sq.m.) to
the final cleaning (total cleaning time of 10 min/sq.m.).
30
-------
CO
0)
s
CO
t—
"S
o>
0)
.JJ
CO
{«•••
o
Q.
•o
CO
t/1
CU o
C> 5^
(O O
O 4->
Cfw CU
C3 ^~
CO
en
C "O
t— CO
O flj
0 —1
I"-* flJ
JC S-
CO CO
cu
"C3 r^
CO CO
& 3T
•t—
^^ E
3 ®
u cil
CO
> OJ
"*» 3
•Q 4J
"O *C
> *~
g u.
cu
OS
•o
ro
CU
•
••••1
<*-
o
c
=3
C£
r— jl
E c c
cS ul *=
•O C 3
CO *r« rj
Q^ m
__i -.;
•CJ jr
a> w
> o
S > -r-
O) .J- O
er: u_ o
•o c E
CO .£ CO
a) .c
— I CO
•a
cu
5- -M „
o co o
cu .i T-
o: u. E
3
T3 c: =3
CO JH (J
CU CO
"-' ^"
CM
"O "^.
CO O1
CU E
_J
r—
CO
| ^
O
h-
E
^
D>
£
^5
co
E
ss
g
55
Cn
E
(U
1—4
vtT
*
CO
H
^j.
O
CO
in
CO
•
o
en
H
en
CM
. CO
in
CO
vH
in
*3*
v
•*
O
0
ft
(N»
CM
H
VO
^4*
CM
CM
4J
0)
Ou
^
rt
O
H"
CM
H
CO
C\
m
N
«.
o
£«v
H
H
Q
!>•
H
^
en
J*s.
M1
H
•i
O
[^
VO
in
in
CM
CO
VD
t^
VO
•*
4J
(U
CU
M ;
(0
CJ
o"
f~
H
CO
CO
rH
••*
•
0
CM
i^g
o
H
CO
CO
vo
H
CM
in
,
CO
H
OY
in
CO
H
en
rH
i1
in
4J
0)
Pj
^i
m
o
en
o
CO
CO
•<*
CO
o
•
a
^*
H
CM
VO
O
O
CM
CO
O
CM
t-
0
^
O
H
^<
*?
O
co
co
vo
**
H
in
-P
0}
cu
S-l
(0
u
CO
CO
1-1
in
"i*
in
M
CM
•
o
en
CO
o\
CO
CO
vo
rH
H
H
9
a\
CO
CO
H
CO
CM
co
en
o
^t
-P
Q)
ft
(0
U
^
o
H
in
CM
«cj*
CM
CM
•
O
cn
en
CO
H
^
t^
0
CO
9
CO
en
CO
CO
vo
vo
CO
CO
CO
o
^
-P
Q)
ft
tO
U
en
c-
CM
H
in
CO
CO
•
o
CM
CO
in
CM
CO
in
CO
CM
en
CM
,
•>*
vo
en
H
CO
vo
0
H
H
O
en
H
0)
en
-------
&5
f s
CO
=3
C
C
O
QJ
"
CO
cu
r»—
_Q
CO
CO
.3 -g
> 01
"cO §
4-> §5
O rv
r—
T3 S3
0)
> j.
0 -"Q.
£ 3 O1
"cO U" ^ E
_OJ C CO
•O J» S5
O) v>
> a
0 £=
E -r-
^3 ' « E
CO CO
NE
•0 E
CO
co
CO
0 01
r— E
CU
i— i
OJ
•
01
t-l
^4*
M
H
•
O
a\
0
^•*
a\
a\
CO
H
H
•*'
01
H
<*
VO
•*
^*
o\
o
H"
•H
(0
X!
U
<*T
•
VO
OJ
0
*
in
•
r-
in
CO
^•^
vo
r!
P
OJ
OJ
*
CO
't
in
OJ
H
n
tH
OJ
-H
(0
in*
•
CO
c*>
H
*
0
CO
1
fT
o
r>
^^
r^
^*
H
n
_•*
in
vo
a\
OJ
CO
o
OJ
X!
O
0
o
U
^^
CO
•
CO
•*
•
O]
1
CO*
n
•
OJ
• **•• '
•^
t^
~
^
vo
o
o
OJ
in
o\
H
rH
r>
•H
(0
U
OJ
-------
•5"
0)
c
•H
•P
O
Q
*M»
CO
0)
•H
<0
EH
#>
•d
<0
G) U
Hi g)
rH O
(tf g
JJ flj t7>
O « g
EH
0)
> H
o rt -^
g G g <«>
Q) -H .3
K » g
ffl 'H 3
3 | I1
13 .a
a) a) w > G
O -H C
Q} O
« c o
•H O
•0 g
(fl tO
Q) J2 &
A W g
13
Q)
> -P
O W O)
i *| ^> c
13
m
•^
r*
rt
4J
o tr*
t* f*
g
0)
-p
H
~- •, vo ^"
en co
t, i ^i
^* t~1 f^
n 03 co
• • •
in H CM
H in
i i
i i
co H en
r- in vo
• • •
^1* co ro
n o co
• • •
in eg r*
*-^ ^.
^7 vo
• •
'-x VO *3<
0 H H
o in
o en *#
**
CM en H
in rH CO
*"<
CM r- ro
rH in in
H ro
0)
&
H J-i to
•H -H ^
10 10 0)
g g 5
Q)
W
3
4^
"c
rH
(2
m
p^
-------
For wooden floors, over 95% of the total dust removed by the combination
of dry vacuuming (5 min/sq.m.) followed by wet washing was removed by the dry
vacuuming. 49-75% was removed after vacuuming for 2 min/sq.m. and 65%-85% was
removed after 3 min/sq.m. For linoleum, more than 75% was removed by vacuuming
for 2 min/sq.m. However, little was removed in subsequent vacuuming and 20%
was removed in the final wet washing step.
HEPA vacuuming of new carpets embedded with a test dust in the laboratory
according to an ASTM procedure was used to determine the differences in
cleaning efficiency between three commercial vacuum cleaners and two carpet
types. Two of the vacuum cleaners were essentially equivalent and each removed
significantly more dust than a third vacuum cleaner when used for between 4
min/sq.m. and 10 min/sq.m. Cleaning for 6 min/sq.m. was necessary to remove
more than 70% of the embedded dust by the two more efficient vacuum cleaners;
the third vacuum cleaning effort only removed 50% after this amount of
cleaning. Cleaning efficiencies were about the same for short pile and
s-
sculptured carpets.
Cleaning by HEPA vacuuming and/or shampooing does not appear to be
effective for carpets "chronically contaminated" in homes containing lead dust.
For carpets with recent contamination, HEPA vacuuming for at least 6 min/sq.m.
was necessary to remove over 70% of the laboratory-embedded dust. Not all
commercially-avail able HEPA vacuum cleaners tested with laboratory-embedded
dust appeared to remove equivalent amounts of the dust.
34
-------
ACKNOWLEDGMENTS
Financial support for the work presented in this paper came in part from
U.S. Environmental Protection Agency Grant No. VOO 5035-01. The efforts of
Chris Kuettner in the carpet cleaning and that of Sandy Roda and her\staff in
chemical analysis of the samples are appreciated. Portions of this material
were initially presented at the 1989 American Industrial Hygiene Conference.
35
-------
REFERENCES
1. Bornschein, R.L., Succop, P., Dietrich, K., Clark, C.S., Que Hee, S.S. and
Hammond, P.B.: "The Influence of Social and Environmental Factors on Dust
Lead, Hand Lead and Blood Lead Levels in Young Children", Environ. Res.
38:108-118 (1985).
2. Bornschein, R.L., Succop, P.A., Krafft, K.M., Clark, C.S., Peace, B. and
Hammond, P.B.: "Exterior Surface Dust Lead, Interior House Dust Lead and
Childhood Lead Exposure in an Urban Environment", Symposium on Trace
Substances in Environmental Health, II, 1986, D.D. Hemphill (Ed.), University
of Missouri, Columbia, pp. 322-332, 1987.
3. Bornschein, R.L., Clark, C.S., Grote, J., Peace, B., Roda, S. and Succop, P.:
"Soil Lead-Blood Lead Relationship in a Former Lead Mining Town", In: Lead in
Soil: Issues and Guidelines, B.E. Davies and B.G. Wixson (Eds.), Science
Reviews Ltd., pp. 149-160, 1988.
4. Bornschein, R., Clark, S., Pan, W. and Succop, P.: "Midvale Community Lead
Study". Chemical Soeciation and Bioavailabilitv 3(3/4):149-162 (1991).
36
-------
5. Clark, C.S., Bornschein, R.L., Succop, P.A., Hammond, P.B., Peace, B.,
Krafft, K. and Dietrich, K.: "Pathways to Elevated Blood Lead and Their
Importance in Control Strategy Development", International Conference on
Heavy Metals, Vol. 1, S.E. Lindberg and T.C. Hutchinson (Eds.), CEP
Consultants Ltd., pp. 159-161, 1987.
i
6. Clark, S., Bornschein, R., Succop, P., Roda, S. and Peace, B.: "Urban Lead
Exposures of Children in Cincinnati, Ohio", Chemical Speciation and
Bioavailabilitv 3(3/4):163-171 f!991).
7. U.S. Department of Housing and Urban Development: "Lead-Based Paint;: Interim
Guidelines for Hazard Identification and Abatement in Public and Indian
Housing", Washington, D.C., September 1990 (Revised May 1991).
8. Milar, C.R. and Mushak, P.: "Lead Contaminated Housedust: Hazard Measurement
«. and Decontamination", In Chilsolm, J. and O'Hara, D.. (Eds.) Lead Absorption
in Children, Urban & Schwarzenberg, Baltimore-Munich, 1982.
37
-------
9. Clark, S., Bornschein, R., Ryan, J., Kochanowski, A., Succop, P. and Peace,
B.: "The Cincinnati Soil-Lead Abatement Demonstration Project", In: Lead in
Soil: Issues and Guidelines, B.E. Davies and B.C. Wixson (Eds.)» Science
Reviews Ltd., pp. 287-300, 1988.
10. Que Hee, S.S., Peace, B., Clark, C.S., Boyle, J.D., Bornschein, R.L. and
Hammond, P.B.: "Evolution of Efficient Methods to Sample Lead Sources Such as
Housedust and Handdust in the Homes of Children", Environ. Res.. 8:77-95
(1985).
11. Statistical Analysis System (SAS) (1988).
12. American Society for Testing and Materials, Standard Laboratory Method for
Evaluation of Carpet-Embedded Dirt Removal Effectiveness of Household Vacuum
Cleaners, Philadelphia, PA, 1986.
>
13. Personal Communication, John N. Balough, The Hoover Company, September 1988.
38
-------
14. City of Toronto Department of Public Health: "South Riverdale Lead Reduction
Program Housedust Cleaning Demonstration", Final Report Prepared by Concord
i
Scientific Corporation and Gore & Storrie Limited in Association with South
i
t
Riverdale Community Health Center, Toronto, Ontario, May 1989.
15. Kominsky, J.R.: "Evaluation of Two Cleaning Methods for the Removal of
Asbestos Fibers from Carpet", Am. Ind. Hyg. Assoc. J. 51(9):500-504 (1990).
16. CH2M Hill: "Final House Dust Remediation Report for The Bunker Hill CERCLA
Site Population Areas RI/FS", Document No.: BHPA-HDR-F-RO-052091, prepared
for Idaho Department of Health and Welfare, Boise, Idaha, May 1991:
39
-------
LIST OF FIGURES
Figure I—Child lead exposure model.
Figure 2--Lead content of vacuum cleaner dust after successive cleaning efforts.
Figure 3--Mean fraction of laboratory - embedded dust removed/cleaning efforts by
HEPA vacuum cleaner.
40
-------
-3
&a
Q
o
C/D
O
X
-------
CD O)
**- CO
O CO
4- ^-
co
E
Q.
a.
mo
ro
o
c
o
o
•o
(0
0
\ - . • *.
^v »
1
CO
CM
O O O O O O O
o o o o o o o
10 o LO o m o
CO CO CM CM T- T-
O
-^
CD
Q.
u. „
0)
0) g
Q. o
*— 3L- tS
(0
CO
O
:::::
CO -
o I
_»
u
ro
r CQ LU ^
CO O)
O =
^J
o
II
•c
LU
O
^C
'E
ra
j>
O
o
O
-------
CO
•C
C
CD
O
i
in
CO
CM
O O O O O O O
in o u> o LO o
CO CO CM CM T- i-
CO
O
fi X
OJ
.E
"E
03
O
O
o
Q.
co
O
~ c
O ra
Q. jl>
co
O
E
3
3
<3
^f UJ
O "T"
• ' " " ^~
CO >
0 E
(0
g- a
CO g>
o |
o
5=
LU
D)
.E
'E
(0
o
0)
c
o
-------
-------
-------
APPENDIX G
Training Manual for Paint and Water Sampling
-------
ENVIRONMENTAL MONITORING
TRAINING MANUAL FOR PAINT AND WATER SAMPLING
Section 1 ;
What is paint and water sampling?
Water sampling is the collection and analysis of water samples from the subject residences in
the Soil Project study areas. This sampling and analysis is done to determine the ampum of lead
in the water.
Paint sampling is the non-destructive analysis of painted surfaces on the interior and exterior
of the subject residences in the Soil Project study areas. This, analysis will give us an
indication of the lead content of the paint on those surfaces.
Section 2
Why do we care about lead in paint and lead in water?
The major purpose of the research study is to determine the effect of the abatement of interior
dust, exterior dust and lead contaminated soil on the blood leads of children under five years of
age. In order to make an assessment of the actual impact of the various abatement strategies it
Is necessary to know the total exposure to other sources of lead in the subject child's
environment. Lead in drinking water and .lead injDaint are potential immediate sources of
exposure to the subject children.
Lead can contaminate drinking water in several ways. The water source can be contaminated. In
this case the potential exposure would be uniform across all of the study areas because the water
originates from the same pumping station and distribution system. Drinking and cooking water
can also be contaminated with lead as a result of the presence of either lead pipe, copper pipe
with soldered joints containing lead, or excessive lead in the faucet structures. The amount of
lead dissolved from the sources in the plumbing system increases as the acidity of the water
increases.
Lead content in paint within a residence has been identified as an immediate
source of exposure to children, especially very young children.
Paint may either flake or chip off of wall surfaces and woodwork surfaces or it may come off in
the form of fine dust from the operation of windows and any abrasion of painted surfaces.
Knowledge of the lead concentration in these interior and exterior painted surfaces may assist in
the explanation of blood lead concentrations in children living in the environment plus it may
assist in the explanation of the rates of re-contamination of abated apartments and residential
areas. !
-------
Section 3
When will paint and water sampling occur?
Paint and water sampling will occur two times during the course of the Soil Lead Abatement
Demonstration Project. The first such sampling will occur in the early spring of 1990. This
will be Phase 04 of the sampling phases. The second paint and water sampling will occur
approximately one year after the initial sampling. That will be the spring of 1991; Phase 08 of
the environmental monitoring.
Section 4
Who will perform the environmental paint and water sampling?
It is expected that the environmental monitors, who have collected the soil exterior dust,
interior dust, dust fall, and handwipes will perform the paint and water sample collection. As
in other interior environmental sampling, teams consisting of two individuals will visit the
subject residences. Present plans call for two teams of two monitors to complete the work.
Section 5
Scheduling of paint and water monitoring.
Scheduling of the initial paint and water monitoring visit will be performed by the Subject
Matters staff of the Cincinnati Soil Project. Initially the Data Management section will provide
a current up to date list of subject families along with addresses and telephone numbers.
Subject Matters staff will then start the scheduling of appointments probably using those
families with telephones for the initial appointments. Once the families with phones or phone
contacts are exhausted then the Subject Matters staff will be make home visits to do the
scheduling.
After the scheduling of the initial visit, a work sheet for lead paint screening and water
sampling will be filled out and forwarded to the abatement staff. The abatement staff will be
responsible for owner contact. As with other environmental sampling, the resident's
permission is adequate to perform the interior monitoring of the residences. Sampling in
common areas and the exterior of buildings requires permission of the owner of the building. As
long as the subject families agree, we can perform the paint and water sampling on the interior
of the units. The exterior paint sampling will require permission from the owner.
Once the abatement staff receives the work sheets they will make an assessment of the presence
or absence of paint on the exterior of the building. If there is paint present then the abatement
staff person will contact the owner and ask permission to do the required sampling. The
abatement staff person will complete two questions on the lead paint screening work sheet. Those
two questions are: owner permission; yes or no and owner contacted by. Owner permission
question is answered by circling either the "n" or the "y"- The "owner contacted by section" is
completed by entering the initials of the abatement staff person making the contact. When the
abatement staff completes this information, the work sheets for lead paint screening and water
sample collection are forwarded to the environmental monitoring teams who will then use these
-------
sheets for the collection of data in the field.
In the case of "no shows", the environmental monitoring team or teams will have the
responsibility of doing their own rescheduling. This will be accomplished by either using the
telephone to contact the family or by stopping in occasionally to see if the family is at home. It
is expected that rescheduling will be less of a problem in Phase 04 because the children do not
necessarily have to be home in order to accomplish this phase of sample collection. In fact, it
would be easier if the children were not at home.
Two water samples are required in this phase of environmental monitoring. One water sample
is collected at the time of the environmental visit by the environmental monitoring team. The
second water sample is to be collected by the family at the beginning of the following day. One
scheduling slot will be alloted to revisiting families to collect the second water sample. The
same environmental monitoring team does not have to collect the second sample. It will be
likely be more efficiently accomplished by whichever team happens to be in a particular
neighborhood on a given day.
Section 6
The Environmental Visit
The Environmental Visit will begin with the environmental monitoring team greeting the family
and explaining briefly the purpose of collecting the paint and water samples. An explanation of
what will occur is also appropriate at this time. The key points to be covered in this initial
explanation are : why we want to collect the sample. How the samples are collected. And most
importantly, what the potential benefits are to the family. Potential benefits include the-
knowledge of both paint and water exposure to the family. If there is high lead content in the
paint, parents who are aware of this might be more cautious about children playing on the floor
or allowing paint chips or other dust to accumulate in the residence. Families with high lead
content in the water could be instructed to reduce exposure by allowing the water to run a
significant amount of time before using that water for drinking or cooking purposes.
At the conclusion of the explanation one team member will start the water collection process
while the other team member begins the calibration of the XRF instrument. After the water
collection process is initiated, and the XRF instrument is calibrated then both team members
can proceed with the paint screening. At the conclusion of the visit the monitoring teams will
assure that all of the equipment has been collected, the toys are collected, and balloons are
distributed to the children if present. Finally a reminder that a team will return the following
day to collect the second water sample is given to the family.
Section 7
i
Water Collection
Two water samples will be collected. Those are W-1 which is a 30 minute stagnation sample and
W-2 which is an overnight stagnation sample. Stagnation samples are collected in order to
provide some uniformity of samples and to determine the amount of lead which will dissolve into
the water over a fixed time period. Significant quantities of lead will not dissolve in water as it
-------
runs through pipes which contain a lead source. One can only measure, the amount of lea
-------
subject residence. This calibration must occur at the beginning of each environmental monitor
visit. The first step in calibration is to turn on the instrument and in the case of the XK3 press
the ress* button and he'd it in for a period of 6-8 seconds. The work sheet should be available
for the calibration procedure. The number of the instrument is entered in the space provided on
the work sheet. Three readings are taken for each standard and recorded on the work Sheet. No
standards are the "no lead standard", the "mid-range standard", and the "high-range standard".
The high-range standard is used only with the XK2, it is not necessary with the XK3.
During the initial introduction to the family several questions should have been answered. The
answers to those questions will determine the location of painted surfaces to be sampled.
Paint testing will occur on surfaces in the three rooms most frequently used by subject
children. The three rooms will most likely be the living room, kitchen, and bedroom: If there
are no painted surfaces in the most commonly used rooms then alternative roms will be used.
After the instrument is calibrated and the location of the painted surfaces to be screened has
been determined, it will be time to do the actual paint screening with the XRF instrument.
For taking readings on trimmed surfaces, it is better to select flat surfaces. Very often there
are fiat surfaces on base boards, window sills, and some window trim. If none of these surfaces
Is flat then an alternative surface might be a door. In most cases there are flat surfaces on
doors. The disadvantage of doors is very often doors were initially varnished and stained when
older buildings were constructed. Doors may have been the last wood surfaces in dwejling units
to be painted. Therefore painted trim may provide the best opportunity to determine the
maximum exposure from lead based paint. Three separate readings will be taken from the trim
!n the selected rooms. Once a piece of trim has been selected as a typical piece of trim in a given
room three readings will be made on that piece of trim. Three separate readings may be
obtained by taking one reading and then moving the instrument 4-6 inches then taking the
second reading, moving the instrument again taking the final reading. ',
Walls are sampled in the same manner. Three separate readings are taken from a typical wall
In the selected room. When moving the instrument to obtain the second and third readings it is
better to not to move the instrument horizontal or vertical direction, but in some other axis.
The reason for this is heights hidden in the wall behind the plaster or drywall, which give us
various reading typically run in vertical or horizontal directions. Therefore if the instrument
is moved in either of those directions and there happens to be a hype behind the plastered
surface, then one could obtain subsequent readings which are potentially influenced by the
presence of a iron or lead pipe.
After the readings are obtained with the XRF instrument, there are several other questions to be
answered about the painted surfaces. We would like some assessment of the condition of the
paint on the painted surfaces. By condition is meant by the physical state of the painted surface.
Paint can be either very tight, adhering securely to the wall; or it could be flaking off in
varying degrees. The condition code which will be entered on the work sheet is as follows:
"1" indicates a very tight secure paint. In this condition there is no loose paint on the
wall.
"2" indicates that there is some loose paint; it may be bubbling or generally coming
loose from the surface in some manner. Condition "2" is considered to be
intermediate between Condition "1" and Condition "3".
-------
"3" is paint which is actually flaking off of the wall and landing on the floor. It is a
very loose peeling condition for the paint.
These condition codes should be entered for both the wall and trim surfaces in the rooms selected
for paint monitoring.
The work sheet for paint screening also contains a place for indicating the use of the room.
Room use should be written in this section. Examples of room use are, kitchen, bedroom, living
room, hallway, bathroom, etc.
The final information needed for paint and wall screening is the sub-straight over which the
paint is applied. The sub-straight of a painted surface is the kind of material over which the
paint is applied. For trims or doors it could be and would very likely be wood for trim. Doors
could be either wood or metal. Walls will very likely consist of paint over drywall or paint
over plaster. Other painted surfaces could include painted brick, painted block or painted
plaster over brick or block.
If the residents ask questions about the XRF readings the monitors should provide the requested
information. It may not be necessary to give the actual readings to the residents if requested for
example, it may not make a difference to a resident if the reading is 6.5 or 7 but the difference
between the reading of 1 and the reading of 9 may make a difference to the resident. As a general
guideline a reading of 2 or less would indicate there is little or no exposure to lead in paint for
the residents. Readings between 2 and 6 might indicate moderate exposure. Readings above 6
could be defined as high exposure to lead in paint.
Two instruments will be used for paint screening, the XK2 and the XK3. The XK3 is the later
model and some advantages in terms of calibration and ease of use. The XK2 has the capability of
reading above 10 mg. per square centimeter. The XK3 has the capability of reading up to 10 mg
per square centimeter. In those cases where the environmental monitoring team obtains
readings with the XK2 of 10, then another screening of the paint with the XK2 instrument will
be necessary. This could very well be accomplished during the return when the water sample W-
2 is collected.
Section 9
Completion of the Environmental Monitoring Team's Day.
When the Environmental Monitoring Team returns to the Main Street facility, transfer of
custody of the water samples will occur. The water samples and the accompanying data sheets
will be given to the Environmental Coordinator, who will accept custody of those samples at that
time. At the end of the day, water samples will be checked for pH and following that the samples
will be acidified with concentrated nitric acid.
-------
-------
APPENDIX H
Worker Safety and Health Plan Publication
A publication:"Development and Implementation of a Safety and
Health Program for Employees Involved with Residential Soil and
Dust Lead Abatement and Monitoring" APPL. OCCUP. ENVIRON. HYG.
7(6), pp 398-404, June 1992.
-------
Development and Implementation of a Safety
and Health Program for Employees Involved
with Residential Soil and Dust Lead
Abatement and Monitoring
C.F. Thomson*, B. Poppe", C.S. Clark0, C.H. Rice0, and D. Linz°
AEastman Kodak Company, Health and Environment Laboratories, 343 State Street, Rochester, New York 14650;
8Friertds of the Homeless, 924 East Main Street, Columbus, Ohio 43205; °University of Cincinnati, Department of'
Environmental Health, 3223 Eden Avenue, Cincinnati, Ohio 45267-0056 I
The safety and health plan that was developed for the •
Cincinnati Soli lead Abatement Demonstration Project
Is described from Its development through its Initial im-
plementation, review, and revisions. This demonstra-
tion project, funded under the Superfund Amendments
and Rcauthorlzatlon Act, was developed to demonstrate
the effectiveness of soil and dust lead abatement in
inner-city neighborhoods hi reducing the blood lead of
young children. The project involved extensive sam-
pling of soil, exterior and interior dust, water, paint
lead, and blood lead. Abatement activities included soil
excavation and replacement, debris removal, pavement
cleaning, and housedust abatement by cleaning and fur-
niture/carpet removal. Air lead concentrations were
minimal and no significant blood leads •were observed.
Experiences described should prove to be useful to
others engaged In similar lead abatement activities; the
overall approach used may also prove useful in the de-
velopment of safety and health plans for lead paint
abatement projects. Thomson, C.F.; Poppe, B.; Clark, C.S.; Rice,
C.H.; Llnz, D.: Development and Implementation of a Safety and
Health Program for Employees Involved with Residential Soil and
Dust Lead Abatement and Monitoring. Appl. Occup. Environ. Hyg.
7(6}:398-4Q4; 1992.
Background
Residential lead contamination and its abatement are
receiving increased public and private attention. The
development and implementation of effective safety
and health programs for workers involved in these activ-
ities is therefore of considerable interest/" The lead
found in the soil and dust in urban areas is primarily
from paint and atmospheric fallout, although mining
and industrial sources(2) are also a problem in some
areas. Lead-based paint was banned for use in housing in
1972GX however, many older homes still contain lead-
based paint, and soils and dust are contaminated in
many locations from residences previously painted with
lead-based products. The primary source of lead in at-
mospheric fallout in most urban areasw was leaded gas-
oline, which was phased out beginning in the 1970s.
Exposure to lead for children living in urban areas
may result from a number of sources, including the
lead-based paint used during the last 100 years. During
this time, natural weathering and other processes have
eroded away portions of the paint, depositing it in the
soil. Sanding, abrasive blasting, and chemical removal
of the lead-based paint during urban housing renovation
and rehabilitation produce small particles. When not
properly contained, these particulates may contaminate
the soil and contribute to high lead concentrations in
urban dust. Once the soil becomes contaminated,
weathering and the additional mechanical action from
the movement of adults, children, and pets further break
down the lead-containing dust. Wind, vehicular traffic,
and pedestrians transport the dust throughout the
.neighborhoods and into residences. Children playing
outdoors or indoors may then ingest the soil and dust
through normal hand-to-mouth behavior; ingestion is
greatest among children who exhibit pica behavior/55
Both the U.S. Department of Housing and Urban De-
velopment (HUD) and the U.S. Environmental Protec-
tion Agency (EPA) are in the process of investigating
effective ways to abate lead from housing and soil. On
April 1,1990, HUD released interim guidelines (revised
September 1990) for lead-based paint abatement proce-
dures for public and Indian housing units around the
country.(6) The plans include measuring the levels of
lead on painted surfaces and determining the types of
abatement that could be used. Decisions will then be
398
1(M7-322X/92/0706-398$5.00/7 © 1992 AIM
APPL OCCUP. ENVIRON. HYG. 7(6) • JUNE 1992
-------
made as to what type of abatement will be most success-
ful and which housing units require abatement.
The EPA is engaged in a soil and dust lead abatement
demonstration program.™ The Cincinnati Soil/Lead
Abatement Demonstration Project is one of three re-
search projects funded by the EPA through the Super-
fund Amendments and Reauthorization Act to deter-
mine the impact of soil and dust lead abatement.® The
other two projects are located in Baltimore and Bos-
ton.®-10'
The primary objective of the Cincinnati Soil/Lead
Abatement Demonstration Project is to determine
whether procedures to abate high lead concentrations
in soil, exterior dust, and house dust, applied separately
and in combination, are effective in reducing the blood
lead levels in children up to 5 years of age. Secondary
objectives of the project are to determine the effective-
ness of the abatement procedures in reducing the quan-
tity of lead on the hands of the children and in the dust at
residences, and to determine the rate of recontamina-
tion of household dust and soil and the factors asso-
ciated with such recontamination.
The primary housing type in the Cincinnati abatement
areas is either rehabilitated or nonrehabilitated housing
in satisfactory condition. In the area of the Cincinnati
project, approximately one third of the children under
the age of 5 are likely to have blood lead levels exceed-
ing 25 fig/dl (the former Center for Disease Control
(CDC) limit)<"> at least once per lifetime. The recently
established current CDC goal calls for using primary
prevention efforts such as community-wide environ-
mental interventions and nutritional and educational
campaigns to reduce levels below 10 #g/dl.<12> In-
terruption of the exposure pathways is thought to be the
most successful means of reducing lead absorption. If
the intervention is at the soil lead stage, there should be
a consequent reduction of lead in urban dust.
Rationale and Methods
• The health and safety program needed for the workers
involved in the soil and dust lead abatement demonstra-
tion project activities and associated environmental and
biological monitoring will be described in this article
from its inception, through its development, imple-
mentation, and midcourse review. These workers were
involved in lead-contaminated soil removal and re-
placement, interior and exterior dust removal, environ-
mental sample collection and analysis, and blood col-
lection and analysis. The abatement activities occurred
at numerous sites in six inner-city neighborhoods. Soil
abatement activities included some handwork involving
picks, shovels, and wheelbarrows, and the use of me-
chanical equipment such as front-end loaders, bobcats,
and dump trucks. Exterior dust abatement involved use
of vacuum-assisted street and sidewalk cleaning equip-
ment; interior dust abatement involved the use of vac-
uum cleaners with high efficiency paniculate air
(HEPA) filters, wet-mopping, and carpet and furniture
removal and replacement.
The University of Cincinnati is a state university in
Ohio, where occupational safety and health enforce-
ment activities are conducted by the federal govern-
ment. Consequently, its employees, including the in-
house staff of the abatement project, are exempt from
Occupational Safety and Health Administration (OSHA)
regulations/135 The soil abatement workers employed
by outside contractors are covered under the OSHA
construction standard/145 Since lead is considered a
hazardous waste, University employees are covered by
the EPA Hazardous Waste Operations and Emergency
Response standard, which was adopted as a companion
regulation to OSHA's for the protection of public-sector
employees. This standard requires a safety and health
plan and an officer to monitor the plan.(ls> Regulation
notwithstanding, the EPA mandated the project staff to
have a safety and health plan and a safety and health
officer for the duration of the demonstration. Good
public health policy and the University's commitment to
comply with OSHA regulations also supported develop-
ment of a comprehensive safety and health plan.
To prepare a safety and health plan, a committee of
knowledgeable persons was gathered to review and as-
sess the potential hazards that must be addressed. A
committee was assembled because few persons alone
have the knowledge that was required for writing a
comprehensive safety and health plan.<16> The commit-
tee formed to write the Cincinnati Soil/Lead Safety and
Health Plan encompassed the following fields of exper-
tise: lead toxicity, safety, industrial hygiene, medicine,
and construction hazards. Evaluation of existing and
perceived hazards was conducted as a group process.
Comprehensive evaluation of the hazards was neces-
sarily a lengthy process. It was recognized early that a
flexible plan would be required so that newly identified
hazards could be incorporated. The elements of the
plan were lead exposure hazards, engineering controls,
work practices, personal protective equipment, radia-
tion, construction and safety hazards, medical surveil-
lance, worker training, air monitoring, and laboratory
audits.
Potential lead exposure for both project and contrac-
tor personnel was of concern. University of Cincinnati
personnel would be collecting, processing, and testing
contaminated soil and dust samples for the duration of
the 3-year study. Contractor personnel would perform
the abatement work on sites for limited time periods
where lead levels in the soil were found to be elevated.
Both groups would also be exposed to physical hazards.
During collection of soil and dust samples or abatement
of contaminated soil, sharp objects such as broken glass
and metal may be present, posing a risk of puncture or
cut for the environmental monitors and the contractors.
Contractors have an additional physical hazard that
APPL. OCCUR. ENVIRON. HYG. 7(6) « JUNE 1992
399
-------
exists when using heavy power tools. Because the
abatement sites are located in heavily traveled urban
areas, which are typically characterized with above-
average crime rates and drug use, personal safety was
another concern that required evaluation. Collection of
the soil and dust would be performed outdoors
throughout the year, raising concerns for both heat and
cold stress. Finally, radiation sources exist in two analyt-
ical instruments: one was a portable X-ray fluorescence
(XRF) in situ paint lead analyzer and the other a
laboratory-based XRF instrument used for soil and dust
lead analysis. All of the potential hazards were evalu-
ated, analyzed, and addressed by the committee prior to
writing and implementing the safety and health plan.
Development of the Safety and Health Plan
Development of the safety and health plan began with
a brainstorming session in which potential exposures,
hazards, and safety concerns were listed. From these
Items an initial outline of the safety and health plan was
developed. Particular concerns were noted in the out-
line in addition to applicable OSHA standards. After the
Initial safety and health plan was written, internal review
was performed before sending it to the EPA. The in-
house reviews were performed by a Certified Industrial
HygienSst, a physician, and the Project Directors. The
EPA reviewed the document at two separate times dur-
ing the initial writing. Comments and recommenda-
tions were specific and included items such as the use of
cartridge-specific respirators as opposed to dust masks,
discussions of sampling methods and instruments, and
field testing procedures. The EPA took an active support
role in reviewing and critiquing the document. Had this
not been the case, an external review procedure would
have been necessary to prevent an in-house bias and to
identify omissions. The development of the plan took
about 9 months and required about three person-
months of full-time effort.
Content of the Safety and Health Plan
Lead Exposure Hazards
Lead was the initial hazard to be addressed. The most
common routes of exposure for lead are inhalation and
ingestSon. Inhalation, being the primary concern for
adults, was evaluated thoroughly. Monitoring, abate-
ment activities, and sample sieving were judged to have
the greatest potential to create elevated airborne lead
concentrations. Ingestion is a less common exposure
route in the workplace but may result from eating con-
taminated food, smoking contaminated cigarettes,
chewing fingernails, and applying contaminated cos-
metics.
The OSHA Lead Standard07' was referenced thor-
oughly for compliance and control methods. Engineer-
ing controls were specified as the preferred way of re-
ducing the potential lead dust hazards. To minimize and
reduce dust inhalation, hoods were required for labora-
tory processing of the soil and exterior dust samples.
These hoods are surveyed quarterly during the struc-
tured laboratory audit to ensure proper ventilation rates.
The use of HEPA filters on the vacuum cleaners used for
cleaning laboratory and field office space was another
specified engineering control. Contractors were re-
quired to use HEPA-equipped vacuum cleaners during
abatement.
Work practices and personal hygiene are additional
methods of reducing lead exposure and other hazards.
University employees working out of the field office
were provided with shower and locker facilities.
Restrictions were implemented to prevent eating,
drinking, smoking, applying cosmetics, chewing to-
bacco, or chewing gum at work sites. Environmental
monitors were required to change into street clothes
before entering the lunchroom. i
A commercial cleaning service was utilized for the
uniforms to prevent, worker take-home of soil and dust.
All streets and sidewalks within the abatement neigh-
borhoods and all participating apartments were dust
abated regardless of the lead levels in the dust. The
lead-contaminated soil to be abated was well below the
level that would classify it as hazardous waste. There-
fore, the contamination of workers' clothing was ex-
pected to be similar to that associated with inner-city
street and housing maintenance, gardening, and apart-
ment cleaning activities, and no special procedures
were thought to be necessary for the commercial clean-
ing service.
Personal protective equipment was used when engi-
neering controls were deemed infeasible or the worker
required protection from a physical hazard. Uniforms
were worn by the site inspectors and environmental
monitors to prevent the transport of lead-contaminated
dust into the field office, personal automobiles, and
homes. Soil abatement tools and equipment were kept
separate from those used for replacement soil unless
they were thoroughly cleaned between uses. Steel-toed
safety shoes were the required footwear for the environ-
mental monitors when collecting soil and exterior dust.
The monitors were working with steel soil core collec-
tion devices and were also exposed to broken glass, etc.,
on the sites. Additionally, rubber gloves are required to
be worn when handling the soil and exterior dust sam-
ples. This applied to both field and laboratory person-
nel.
Radiation
Two radiation sources exist in the instruments used
for lead analysis. These instruments and the laboratories
where they were located were tagged with radioactive
warning signs. All laboratory personnel and other per-
sons associated with these instruments attended
400
APPL OCCUR. ENVIRON. HYG. 7(6) • JUNE 1992
-------
courses on safety hazards conducted by both the manu-
facturer and the University.
Construction Safety Standards
Construction safety hazards were primarily a concern
during the abatement phases of the project. Contractors
used heavy equipment (front-end loaders, bobcats,
dump trucks, etc.) as well as hand tools in highly popu-
lated areas, which posed a risk not only to the operator
but to project personnel and neighborhood persons.
Specific safeguards were required to address these
needs* The work site was fenced to restrict adults, chil-
dren, and animals from the abatement area. The fencing
remained around the site on a 24-hour basis until activi-
ties were completed. This prevented contact with the
equipment by the public and avoided disruption of the
soil and dust, as well as providing additional security for
the equipment left on the site overnight. Specifics con-
cerning vehicles* hand power tools, concrete activities,
etc., were addressed in the contractor safety and health
plan according to OSHA Construction Standard 1926.
Medical Surveillance
A medical surveillance program included preplace-
ment examinations, annual reexaminations, and moni-
toring for the effects of specific potential hazards.(17)
The preplacement portion of the medical surveillance
program established baseline data on each individual to
better safeguard the health of the employee. An occupa-
tional/meclical history was taken. Past occupational ex-
posures were noted along with any pertinent medical
history. A physical examination was performed to assess
the employee's general health and fitness to perform
required duties and to use air-purifying respirators.
Baseline data were collected for blood lead levels. Be-
cause collection and analysis of numerous blood sam-
ples from inner-city children and adults was an integral
part of the demonstration project, it was necessary to
include prevention of blood-borne diseases and injuries
in the training program and prudent to offer vaccination
against hepatitis B to employees involved with blood
collection and analysis. Finally, tetanus shots were
given to any employee during the preplacement exam
whose previous immunization was more than 10 years
old or of unknown date. Employees covered by the
medical monitoring program were the environmental
monitors, site inspectors, and laboratory personnel be-
cause of their frequent exposure to the lead-containing
soil. Specific medical monitoring requirements for each
group are shown in Table I.
At 6-month intervals another blood lead determina-
tion was made; results were compared with previous
data to detect whether changes in the levels had oc-
curred. The annual medical examination also included
an assessment to verify if the employee was still medi-
cally capable to carry out the tasks required in his/her
position. Blood lead level was also evaluated at termina-
tion of employment.
There were 37 employees in the safety and health
program. This included office staff, field monitors, and
laboratory technicians. Twenty-seven of those employ-
ees were involved in the abatement and monitoring ac-
tivities. Mean blood lead levels for these 27 employees
were 2.4, 5.8, and 5.0/zg/dl for the initial, 6-month, and
annual determinations, respectively, with standard de-
viations of 0.55, 2.39, and 1.00/ig/dl. The change from
the initial value to 6-month value had a p-value of less
than .0005 and the initial to 1-year value had a p-value of
less than .005. These changes were not considered clin-
ically significant. The highest blood lead value was 12
fig/dL It was the only value of 10 or above. Blood sam-
ples were analyzed for lead using an ESA Model 3010A
anodic stripping voltammeter. Details of the laboratory
TABLE I. Medical Surveillance Program by Job Type
HBV
Job
Type
Users of vacuum dust
collection methods
Clinic or blood laboratory
Site inspectors
Environmental monitors
and assistants
Safety inspector/
abatement
' coordinator
HBV = hepatitis B virus;
N = not applicable;
0 = optional;
r = recommended;
R = required.
History/
Physical
R
r
R
R
0
Baseline
Laboratory
R
r
R
R
0
Blood
Lead
R
N
R
R
0
Respiratory
Fitness
R
N
N
N
N
Physical
Capacity
N
N
N
0
N
Antibody
Status
N
r
N
N
N
HBV
Vaccine
N
r
N
N
N
APPL. OCCUP. ENVIRON. HYG. 7(6) • JUNE 1992
401
-------
methods and quality control and quality assurance pro-
gram have been provided elsewhere/185 The detection
limit of this method is 1.4 ± 0.7 /zg/dl.
Physical Hazards
Year-long environmental monitoring exposed the
field personnel to various weather conditions that
needed to be closely watched. When a heat alert was
declared by the City Health Department, a heat stress
monitoring protocol went into effect. Included in the
protocol were the following elements: worker informa-
tion and training, work practices, heat alert program,
environmental surveillance, medical surveillance, and
record keeping. Results were obtained by a series of
calculations, and effective measures were taken appro-
priate to the findings, i.e., reduce the work day, increase
consumption of fluids, stop all outdoor activities, etc.
Injury occurrence was monitored by means of an In-.
cident Report. The report was filed immediately after an
incident and then reviewed by those persons involved.
Appropriate action and follow-up was taken relevant to
the incident.
Worker Training
Training was needed for hazard recognition and to
assure proper use of protective equipment and other
control measures. Several training programs were de-
veloped and implemented during the Lead Demonstra-
tion Project. A general health and safety training session
was given for laboratory and field personnel. This cov-
ered general laboratory practices, chemical and physical
hazards, legal rights and responsibilities, personal pro-
tective equipment, emergency response, hazard con-
trol, work practices and procedures, and medical sur-
veillance. A lead hazard session was provided for all
employees. This session encompassed information on
how lead enters the body, health effects, and ways to
reduce exposure. A first aid program given by the Red
Cross was required for field monitors. Each field crew
has at least one person trained in first aid. Heat and cold
stress training was given to all field monitors. The heat
stress protocol was also reviewed at this session. The
Red Cross also provided CPR training and a lifting tech-
niques and back injury prevention program for those
employees required to lift heavy objects. Workers in-
volved in blood collection also received training in
AIDS awareness, phlebotomy work practices, and medi-
cal surveillance. Supervisors were required to attend all
of the training sessions. Training sessions were per-
formed for short durations of time (1 to 5 hours) and in
small groups with interactive discussions. This ap-
proach to training, which actively involves the trainees,
was used to foster learning through participation/19)
Training occurred annually or more often if the work
acil%»ity repeated at more frequent intervals. For exam-
ple, training of blood collection clinic staff occurred five
times over about a 26-month period. The total amount of
training ranged from about 8 to 16 hours for each worker
on an annual basis.
Air Monitoring
Air sampling strategies were developed to obtain an
initial determination of levels of airborne lead at each
work site. Both personal and area air samples were col-
lected. Analyses were performed in an American Indus-
trial Hygiene Association accredited laboratory. Levels
must be maintained below the OSHA permissible expo-
sure limit (PEL) of 50 ^g/m3 and preferably below the
action level of 30 ^g/m3.(17) Provisions were also made
to include any new or changed activities fof additional
monitoring if airborne lead concentration might ap-
proach the action level.
Area and personal air monitoring was performed at a
temporary laboratory used for a field office and during
soil sieving and preparation at the beginning of the proj-
ect. The mean concentration of lead found during activ-
ities exposed employees to less than 1 /ig/m3. When the
soil processing laboratory was then moved to its perma-
nent location, sampling was performed again, as re-
quired in the safety and health plan. The mean concen-
tration found at this facility was less than 3 /*g/m3.
During the initial vacuum cleaner performance, air
levels were below 0.25 /tg/m3. The lead levels found in
the laboratory facilities were well below the OSHA PEL
and action level and therefore are in compliance.
At the time of the development of this safety and
health program, little information could be located in
the literature on air lead levels during similar activities
elsewhere. Municipal street cleaners in Cincinnati,
using the dustier mechanical broom-type sweepers,
were previously found to be exposed to a geometric
meanair lead concentration of 3.2y«g/m3 (maximum 6.6
/fg/m3) .(20) More recently, extensive monitoring of air
lead exposure during interior dust abatement at an inte-
rior residential dust abatement pilot program in a
neighborhood near a secondary lead smelter in Tor-
onto, Canada, was reported by others.00 Air lead con-
centrations were determined from both personal and
area air samples collected before, during, and after
abatement activities in eight houses. Airborne dust con-
centrations were also determined from the area sam-
ples. Results of the area monitoring revealed that air-
borne lead concentrations were all belowthe OSHA PEL
of 50jUg/m3, with the highest value being 9.2/zg/m3. The
average airborne lead concentration was somewhat
higher during abatement (0.76 //g/m3) than before
(0.31 /zg/m3) or after (0.14 /ig/m3). Area dust concen-
trations followed a similar pattern, with the highest level
(1.3 $g/m3) occurring during abatement. This value was
less than 10 percent of the OSHA nuisance dust limit of
15 £?g/rn3.' .Personal air samples revealed somewhat
higher levels than area samples with a.mean concentra-
tion during duct cleaning of 2.9 fig/m* (range: 0.76 to
4.39 /zg/m3; standard deviation: 1.38/zg/m3), which was
402
APPL. OCCUP, ENVIRON. HYG. 7(6) • JUNE 1992
------- |