EPA 747-R-96-006
May 1997
LEAD EXPOSURE ASSOCIATED WITH
RENOVATION AND REMODELING ACTIVITIES:
WORKER CHARACTERIZATION AND BLOOD-LEAD STUDY
Prepared By
Battelle
505 King Avenue
Columbus, OH 43201
for
Technical Programs Branch
Chemical Management Division
Office of Pollution Prevention and Toxics
Office of Prevention, Pesticides, and Toxic Substances
U.S. Environmental Protection Agency
Washington, DC 20460
-------
DISCLAIMER
Mention of trade names, products, or services does not
convey, and should not be interpreted as conveying official EPA
approval, endorsement, or recommendations.
This report is copied on recycled paper.
-------
Contributing Organizations
This study was funded and managed by the U.S. Environmental Protection Agency. The
study was conducted collaboratively by two organizations under contract to the Environmental
Protection Agency, Battelle, and Midwest Research Institute. Each organization's responsibilities
are listed below.
Battelle
Battelle was responsible for designing the study, recruiting participants, collecting worker
questionnaire data and blood samples, creating and maintaining data bases, conducting statistical
analysis, and producing the final report.
Midwest Research Institute (MRI)
MRI was responsible for chemical analysis and quality assurance for blood-lead chemical
analysis.
U.S. Environmental Protection Agency
The Environmental Protection Agency was responsible for oversight in developing the
study plan, managing and coordinating the study, and reviewing and editing this report. EPA
Project Managers included Dan Reinhart and Darlene Watford. Cindy Stroup was the Branch
Chief of the Technical Programs Branch under whose direction the study was conducted.
-------
Acknowledgements
United Brotherhood of Carpenters
The United Brotherhood of Carpenters and Joiners of America (UBC) provided a list of
current members that was used to develop the sampling frame for union carpenters. In addition,
UBC reviewed the questionnaire, assisted in pretesting the questionnaire, provided assistance in
recruiting their membership, and consulted in the design of the study. Special thanks go to John
Repko, Sigurd Lucassen, Edward Coryell, and Terry Nelson.
-------
TABLE OF CONTENTS
Page
EXECUTIVE SUMMARY viii
1.0 INTRODUCTION TO THE OVERALL RENOVATION AND REMODELING STUDY 1-1
1.1 Objectives of the R<£R Study 1-1
1.2 Scope of the R<£R Study 1-2
1.3 Overall Approach to the R<£R Study 1-3
1.4 Overview of the Worker Characterization and Blood-Lead Study 1-5
1.5 Peer Review 1-5
2.0 STUDY DESIGN AND METHODS 2-1
2.1 Study Objectives 2-1
2.2 Survey Design 2-1
2.2.1 Target Population 2-1
2.2.2 Sampling Frame 2-2
2.2.3 Sampling Methods 2-3
2.2.4 Recruitment 2-4
2.3 Sample Collection 2-5
2.3.1 Questionnaire Data 2-5
2.3.2 Blood Samples 2-6
2.4 Data Management 2-7
2.5 Statistical Methods 2-8
2.5.1 Construction of Exposure and Worker Group Variables 2-8
2.5.2 Descriptive Statistics 2-10
2.5.3 Exploratory Data Analysis for Ancillary Variables 2-10
2.5.4 Statistical Models 2-10
2.5.4.1 Blood-Lead Concentrations 2-10
2.5.4.2 Relationships Between Target Activities and
Blood-Lead Concentrations 2-11
3.0 RESULTS 3-1
3.1 Recruitment Results and Field Experiences 3-1
3.1.1 Recruitment of Respondents 3-1
3.1.2 Field Experiences 3-5
3.1.2.1 Field Experiences in St. Louis 3-5
3.1.2.2 Field Experiences in Philadelphia 3-6
3.1.2.3 Working with Difficult Respondents 3-6
3.2 Quality Control for Blood-lead Data 3-7
3.2.1 Laboratory Quality Control Data Summary 3-7
3.2.2 Field QC Data 3-8
3.2.2.1 CDC QC Reference Samples 3-8
3.2.2.2 Worker Blood-Lead Samples 3-9
-------
3.3 Telephone Screening Results 3-10
3.4 Worker Characterization Results 3-14
3.4.1 Demographics 3-14
3.4.2 Target Activities 3-16
3.4.3 Work Practices 3-16
TABLE OF CONTENTS
(Continued)
Page
3.5 Descriptive Statistics of Worker Blood-Lead Concentrations 3-21
3.6 Statistical Modeling Results 3-23
3.6.1 Statistical Model Building 3-23
3.6.1.1 Selection of Ancillary Covariates 3-23
3.6.1.2 Selection of the Form of the Model 3-25
3.6.2 Comparisons Between Blood-Lead Concentrations Among
Worker Groups 3-25
3.6.3 Relationships Between R<£R Target Activities and Worker
Blood-Lead Concentrations 3-27
3.6.4 Overall Statistical Model for Worker Blood-Lead Concentrations 3-28
4.0
SUMMARY AND OVERALL CONCLUSIONS 4-1
LIST OF APPENDICES
APPENDIX A: QUESTIONNAIRES
APPENDIX B: ASSESSMENT OF LABORATORY QUALITY CONTROL DATA
APPENDIX C: DESCRIPTIVE STATISTICS TABLES
APPENDIX D: ADDITIONAL QUESTIONNAIRE TABLES
APPENDIX E: STATISTICAL MODELS TABLE
APPENDIX F: ADDITIONAL FIGURES
APPENDIX G: HUMAN SUBJECTS APPROVAL AND INFORMED CONSENT
LIST OF TABLES
Table 1. Summary of Questionnaire and Rationale for Each Section 2-6
Table 2. Summary of Recruitment for Telephone Screening Interview 3-3
Table 3. Summary of Overall Recruitment and Participation for WCBS 3-3
Table 4. Quality Control and CDC Performance Samples 3-8
Table 5. 95% Confidence Intervals for Geometric Mean of Blood-Lead Concentrations for CDC
QC Reference Samples 3-9
Table 6. Variance Component Estimates for Blood-Lead Concentrations 3-10
Table 7. Summary of Telephone Screening Questionnaire Responses Relating to Type and Amount
of Specific Target Activities 3-11
-------
Table 8.
Table 9.
Table 10.
Table 11.
Table 12.
Table 13.
Table 14.
Table 15.
Demographic Comparison Between Participants and Nonparticipants 3-12
Summary of Demographic Data 3-15
Summary of Responses for Questions Pertaining to R<&R Target Activities 3-17
Summary of Responses for Questions Pertaining to Worker Practices 3-18
95% Confidence Intervals for Geometric Mean of Blood-Lead Concentrations for Each
Sampling Frame 3-22
95% Confidence Intervals for Geometric Mean of Blood-Lead Concentrations for Each
Worker Group 3-22
Geometric Mean Blood-lead Concentration and Log (Standard Deviation) for
Each Level of the Ancillary Covariates 3-25
95% Confidence Intervals for Geometric Mean of Blood-Lead Concentrations for Each
Worker Group Based on Covariate Adjusted Model 3-26
TABLE OF CONTENTS
(Continued)
Page
Table 16. Predicted Increase in Blood-Lead Concentration Associated with 10 days of
Work in Pre-1950 Buildings 3-28
Table 17. General F-Tests for the Combined Effects of All Target Activities on Worker
Blood Lead Concentrations, After Adjusting for the Effects of Covariates and
Worker Group 3-30
Table 18. Predicted Worker Blood-Lead Concentrations Associated with Low, Medium, and High
Exposure Indices for Each Worker Group 3-31
Table 19. Difference in Geometric Mean Blood-Lead Concentration Between WCBS and
NHANES III 4-4
LIST OF FIGURES
Figure 1. Overall Design Structure of the Renovation and Remodeling Study 1-4
Figure 2. Paradigm for Fitting Statistical Models to Blood-Lead Concentrations 2-12
Figure 3. Days in Last Month Spent Conducting General R<£R, Residential, Non-Residential and Pre-
1950 Work for Participants (shaded) versus Non-Participants 3-13
Figure 4. Summary of Worker Practices 3-20
Figure 5. Histogram of Blood-Lead Concentration (Semi-Logarithmic Scale) 3-21
Figure 6. 95% Confidence Intervals for Geometric Mean of Blood-Lead Concentrations for Each
Worker Group Based on Covariate Adjusted Model 3-26
VII
-------
EXECUTIVE SUMMARY
The Residential Lead-Based Paint Hazard Reduction Act (Title X) required the U.S.
Environmental Protection Agency (EPA) to conduct a study of lead exposure associated with
renovation and remodeling (R&R) activities (the R&R study). Information obtained from the
R&R study is to be used to help determine which groups of R&R workers require training,
certification, or educational materials because of the potential lead exposure resulting from the
R&R activities they perform. This report presents the results of one of the principal data
collection efforts in the R&R study: the Worker Characterization and Blood-lead Study (WCBS).
The primary goal of the WCBS was to collect data and information that would permit an
assessment of the relationship between R&R activities and lead exposure to the R&R workers
conducting the activities. The study surveyed two groups of workers (union carpenters and
employees of independent contractors) in two cities (Philadelphia, Pennsylvania, and St. Louis,
Missouri). The collected data included
1. Worker blood samples that were chemically analyzed for lead concentration
2. Questionnaire data on demographics, the extent to which specific R&R work
activities were conducted, work practices, previous training on or knowledge about
lead, and non-work activities and personal characteristics that are potentially related
to lead exposure.
Questionnaires were collected from a total of 585 workers. The questionnaire results
indicated that:
1. The R&R workers performed a wide variety of R&R activities, and spent
considerable time removing large structures and removing paint and preparing
surfaces, activities with potential for creating high dust-lead exposure.
2. 90% of the workers did not use a respirator.
3. 88% of the workers did not use cleanup methods recommended for use in a lead-
contaminated environment, and 99% used dry sweeping.
4. 97% of the workers used dry methods for paint removal.
5. 67% of the workers had not received any materials on lead hazards, and 87% has
received no lead exposure training.
Blood samples were collected from 581 of the 585 workers. Worker blood-lead
concentrations were generally low: 9.1% were above 10 |ig/dL, 1.2% were above 25 |ig/dL, and
only one worker had a blood-lead concentration greater than 40 |ig/dL. The geometric mean
blood-lead concentration for all workers was 4.5 |ig/dL.
VIII
-------
A statistical model was developed and fit to the data that included effects for variables
potentially related to lead exposure, such as education level, smoking status, and age of worker's
home; worker group; and the amount of R&R activity conducted during the past 30 days, last
year, and over the worker's career. Although blood-lead concentrations predicted by the model
for each worker group studied were low, there were significant differences among the worker
groups. Drywall workers and painters had the highest predicted blood-lead concentrations, and
floor layers had the lowest.
-------
1.0 INTRODUCTION TO THE OVERALL RENOVATION AND REMODELING
STUDY
Lead poisoning has long been recognized as one of this country's most important
environmental health problems. With the phase-out of lead in gasoline, lead-based paint is now
the primary source of lead exposure, particularly for children and construction workers. Federal
programs undertaken to understand and mitigate the lead exposure associated with lead-based
paint have focused on 1) deteriorated lead-based paint, and 2) methods of abatement. Therefore,
exposure data for both renovation and remodeling (R&R) workers and building occupants tend to
be related to either the presence of deteriorated paint or the occurrence of abatement.
However, disturbance of intact lead-based paint surfaces by R&R activities (conducted
with no abatement intent) may also result in lead exposure for both R&R workers and building
occupants. In fact, extensive R&R is often performed in older homes or public buildings with a
high probability of containing lead-based paint. Workers in many of these homes may not be
aware of a potential lead problem.
To address potential lead exposure associated with R&R, the United States Congress
directed the U.S. Environmental Protection Agency (EPA) Administrator to conduct a study of
lead exposure associated with R&R activities. The study is required by paragraph (2) of Section
402 (c) of Title IV of the Toxic Substances Control Act, contained in the Residential Lead-Based
Paint Hazard Reduction Act (Title X of HR 5334). The results of this study, hereafter referred to
as the R&R study, are documented in three reports:
p "Lead Exposure Associated With Renovation and Remodeling Activities: Summary
Report," containing overall study results
p "Lead Exposure Associated With Renovation and Remodeling Activities:
Environmental Field Sampling Study (EFSS)," a technical report on the results of one
component of the R&R study
p "Lead Exposure Associated With Renovation and Remodeling Activities: Worker
Characterization and Blood-Lead Study (WCBS)," a technical report on the results of
the WCBS, a second component of the R&R study.
Chapter 1 of this report includes a discussion of the overall design of the R&R study and
the complementary roles of its two principal data collection efforts: the WCBS and the EFSS.
Subsequent chapters deal only with the design, implementation, and results of the WCBS.
1.1 OBJECTIVES OF THE R&R STUbV
The primary purpose of the R&R study was to help determine which groups of people
require training, certification, or educational materials because of their potential lead exposure. In
particular the study was designed to satisfy two technical objectives:
l-l
-------
1. Determine the extent to which persons engaged in various types of R&R activities in
target housing, public buildings constructed before 1978, and commercial buildings
are exposed to lead.
2. Determine the extent to which persons engaged in various types of R&R activities
disturb lead and create a lead-based paint hazard, on a regular or occasional basis, to
building occupants or other exposed individuals.
1.2 SCOPE OF THE R&R STUE>V
The broad scope of the study mandated by Title X, along with time and budget
constraints, led to a design strategy that required multiple targeted field studies. Decisions had to
be made on priorities, focus, and representativeness. Details on the decisions related to
delineating the scope of this study, including key definitions and specification of populations,
environments and target activities, are given in Chapters 1 and 2 of the EFSS Technical Report.
A list of R&R activities associated with lead exposure was assembled by the EPA. As a
result of over 200 interviews with other government agencies, lead poisoning prevention experts,
industry representatives, labor unions, and other concerned groups. From input obtained in a
summary meeting with several of these contacted individuals, the EPA defined eleven categories
of R&R activity with potential for lead exposure that could be addressed by this study. These
activities, subsequently called target activities, were
1. Paint removal
2. Surface preparation
3. Removal of large structures
4. Window replacement
5. Enclosure of exterior painted surfaces (i.e., siding)
6. Carpet or other floor covering removal
7. Wallpaper removal
8. HVAC repair or replacement including duct work
9. Repairs or additions resulting in isolated small surface disruptions
10. Exterior soil disruption
11. Large renovation projects involving multiple target activities.
The data collection effort in the EFSS focused on the following six target activities:
removal of large structures (interior demolition), window replacement, carpet removal, HVAC
repair or replacement, surface preparation, and repairs resulting in limited surface disruption.
Target activities examined in the WCBS included window replacement, carpet removal, removal
of large structures (demolition), HVAC repair or replacement, and paint removal. Post-activity
cleanup was also observed.
1-2
-------
1.3 OVERALL APPROACH TO THE R&R STUbV
The R&R study consisted of three phases:
1. An information gathering and literature review phase to uncover the existing body of
information concerning lead exposure related to R&R activity. The major
conclusion of this phase was that, with the exception of paint removal, little data
was available.
2. The EFSS, which involved assessing the relative disturbance of and potential
exposure to lead associated with selected R&R activities by measuring lead in air
and dust.
3. The WCBS, which used questionnaire information and blood-lead measurements to
determine if specific worker groups or specific R&R activities are associated with
increases in blood-lead levels.
A follow-on study to assess the relationship between incidence of R&R activity and
elevated blood-lead concentration in children is currently in the design stages.
Discussion of the decision process related to environmental measurements versus blood-
lead measurements is given in the EFSS Technical Report. The EFSS and WCBS were part of an
overall design strategy to address the broad scope of the R&R study. This strategy included:
1. Use of the EFSS to relate environmental exposure estimates to specific R&R
activities. These activity-specific exposure estimates could then be used, along with
worker profile information, as building blocks for constructing exposure
assessments for a wide variety of R&R worker groups.
2. Use of the WCBS results to
a. Assess the health effect (internal dose) associated with exposure to R&R
activities for a subset of activities and worker groups
b. Validate the EFSS environmental exposure (potential dose) measurements
c. Provide worker profile information for a subset of R&R worker groups.
The overall design of the R&R study is presented in Figure 1. Information from the
WCBS was combined with the EFSS results in a summary report to help determine which groups
of people require training, certification or educational materials because of the potential lead
hazard associated with renovation and remodeling activities they perform. This report presents
the technical results of the WCBS.
1-3
-------
RENOVATION AND REMODELING STUDY
I
Delineation of Scope; Literature Review and Information Gathering
Worker Characterization and Blood-
Lead Study (Blood-Lead Study)
Identify worker groups
and recruit workers
Collect blood samples and
questionnaire information
Characterize workers.
Develop relationship between
blood lead levels and activities
WCBS
(Blood-Lead Study)
Technical Report
I
Environmental Field Sampling Study
(Environmental Study)
Large R&R
Projects with
Multiple Activities
Target
Activities
Decision on Assessment Method
Ji-Professional judgment
Jr- Literature
JT- New field studies
— Controlled designed field study
— Monitoring field study
Use of other extant data sources
Develop relationships between
activities and measurements of
lead in the environment
EFSS
(Environmental Study)
Technical Report
R&R Summary Report
Figure 1. Overall Design Structure of the Renovation and Remodeling Study
1-4
-------
1.4 OVERVIEW OF THE WORKER CHARACTERIZATION ANE> BLOOE>-LEAE> STUE>V
The WCBS involved a targeted survey of two groups of renovation and remodeling
workers (union carpenters and employees of independent contractors) in two cities (Philadelphia,
Pennsylvania, and St. Louis, Missouri). The data collected included:
1. Worker blood samples that were chemically analyzed for lead
2. Questionnaire data that were used to characterize the workers and understand
differences in blood-lead concentrations.
The questionnaire data included information relevant to lead exposures on (a) demographics (b)
work history (both current and long-term) (c) personal characteristics and habits (d) non-work
activities (e) medical history and (f) previous training or knowledge on lead.
Statistical models were used to determine if there were significant associations between
blood-lead levels and various target activities. The questionnaire data were used to characterize
the extent to which workers perform specific activities and use specific work practices. The
questionnaire data also were used to control for potential confounding factors when interpreting
blood-lead levels, such as smoking and the age of a worker's residence.
1.5 PEER REVIEW
This report on the Worker Characterization and Blood-Lead Study (WCBS) was reviewed
independently by members of a peer review panel. Comments which are important for
interpreting the study results or which resulted in important modifications to the report are
discussed below. All peer reviewers recommended publishing the report with minor revisions.
Concern was expressed over the inability to collect both blood-lead and environmental
lead measurements from the same group of workers and/or occupants. Human subjects review,
for both ethical and legal reasons, would not allow measuring blood-lead concentrations for
occupants (young children) before and after conduct of an activity that was suspected of causing a
hazard. For workers, the difficulty in this study was recruiting typical R&R workers operating in
an unregulated environment. For this group of workers, employers were very reluctant to
participate even as the study was conducted. Contractors were concerned over lawsuits by
workers in the event that the study revealed a worker's blood-lead increased as a result of a
specific job they were assigned to. We had very few contractors participating in either phase of
the study. Employees participated in the WCBS largely because of either their own interest or the
interest and encouragement of their national and local union. Gaining access to work sites for
environmental and biological sampling would have required participation of the contractors,
homeowners, and workers. If such sampling was conducted under forced cooperation, then the
results may have been biased. If the study had focused on lead abatement workers this may not
have been a problem, but with a focus on typical R&R workers who were not, at the time of this
study, using worker protection practices, there were many problems recruiting contractors to
1-5
-------
participate. In short, the difficulty in recruiting contractors was in getting at the population of
interest: unregulated R&R workers not specializing in lead abatement.
One reviewer requested more information to show that the QC data are consistent with
the statistical analysis applications and results. As a result of this comment, more documentation
was added to the report and inconsistencies in the presentation of QC results were resolved and
clarified.
Another reviewer questioned the basis for using 1950 as the dividing line between older
(meaning greater likelihood of lead exposure) and newer buildings. Prevalence statistics from the
HUD National Survey ("Report on the National Survey of Lead-Based Paint in Housing," U.S.
EPA, EPA 747-R-95-003, April, 1995) were used as the basis for selecting 1950 as the dividing
line between older (greater likelihood of lead exposure) and newer buildings. Although lead paint
was used well into the 1970s, the HUD National Survey indicates that homes built prior to 1950
contain significantly higher levels of lead in paint, dust, and soil, than homes built after 1950. In
general, the likelihood of lead-based paint, and other indicators of lead contamination such as
dust-lead levels, increase with the age of housing. For example, the HUD National Survey
estimates that 48% of occupied homes built prior to 1940 have dust-lead loadings above 100
jig/ft2, while only 3% of homes built between 1960-1979 have dust-lead loadings above 100
EPA has established a public record for the peer review under administrative record
AR152, "Lead Exposure Associated with Renovation and Remodeling Activities Peer Review."
The record is available in the TSCA Nonconfidential Information Center, which is open from
noon to 4 PM Monday through Friday, except legal holidays. The TSCA Nonconfidential
Information Center is located in Room NE-B607, Northeast Mall, 401 M Street SW, Washington,
D.C.
1-6
-------
2.0 STUDY DESIGN AND METHODS
The WCBS was conducted in two phases. In Phase I, workers were recruited from each
worker group in each city. A brief screening questionnaire was administered to the selected
workers over the telephone. The purpose of the telephone screening was to determine the
eligibility, as defined in Section 2.3.1, of the selected workers to recruit workers for Phase II, and
to collect preliminary information on targeted work activities. Phase II involved collecting
worker blood samples and questionnaire data from the workers recruited in Phase I.
This section presents the overall design of the study, including the study objectives,
method for chemical analysis of lead in blood, sampling plan, basis for sample size, and data
handling and analysis procedures. A detailed description of the field, laboratory, and data analysis
methods was provided in the document, "Quality Assurance Project Plan (QAPjP) for the R&R
Worker Characterization and Blood Lead Study" (July 8, 1994).
2.1 STUbV OBJECTIVES
The primary goal of the WCBS was to collect the data and information that would permit
an assessment of the relationship between renovation and remodeling activities and the actual
exposure to lead of the R&R workers conducting the activities. The objectives of the study were
to
1. Determine the relationship between blood-lead concentrations and work practices or
activities performed by R&R workers, after controlling for potential confounding
factors
2. Determine if the blood-lead concentrations of R&R workers in specific worker
groups differ after adjusting for potential confounding factors
3. Gather information on the types of work activities and work practices in which R&R
workers engage.
2.2 SURVEY EESI6N
Components of the WCBS survey design included defining the target population,
constructing a sampling frame, specifying sampling methods, and recruiting the targeted number
of workers in the survey.
2.2.1 Target Population
The target population for this study consisted of two groups of renovation and remodeling
workers:
p Union carpenters
2-1
-------
p Employees of independent, non-union contractors.
Carpenters were chosen because of the wide variety of R&R activities they perform. Carpenters
represent the generalists of the R&R industry, considered by some to comprise the backbone of
the industry. Employees of independent, non-union contractors were chosen because a large
portion of the remodeling and renovation business is conducted by these workers. Moreover,
work practices and activities conducted by this diverse group of workers is expected to vary
widely. Initially, laborers were targeted as a third group because they can be considered among
the most highly exposed groups of R&R workers. Demolition of a wall or ceiling (rip and strip),
generally performed by laborers, is often performed by pick and sledge hammer. Although
initially cooperative and interested in the study, the union representing laborers elected not to
participate in the study. As a result, it was not possible to construct a sampling frame of laborers.
However, some individuals participating in the study considered themselves laborers.
The WCBS targeted workers in two cities: St. Louis and Philadelphia. These two cities
were selected because of the support and cooperation of local union leadership, and because a
large number of children with elevated blood-lead levels have been found in both cities. Lead-
based paint exposure is considered a major factor in elevated blood-lead levels in children;
therefore, the presence of large numbers of children with elevated blood-lead levels is one
indicator of potential lead exposure in these two cities.
2.2.2 Sampling Frame
Sampling frames were defined separately for each group of workers. Union members
were identified using lists provided by union leadership. The population of potential respondents
in the independent, non-union group was much more diverse and not as well defined as the union
group. A list of independent workers was compiled by researching the local construction/
remodeling market in each city. An intense recruitment effort was conducted for potential
independent workers, which included obtaining information from the National Association of
Home Builders, advertisements in telephone books, newspaper advertisements, public service
announcements, and referrals by other workers. While the frame constructed for union members
was considered complete, including all eligible members, it was not possible to identify all eligible
non-union R&R workers.
Union Carpenters
The sampling frame for the union carpenters was based on a list of current union members
provided by the United Brotherhood of Carpenters and Joiners of America (UBC) leadership in
each city. In St. Louis the union provided four lists of workers defined by worker type:
1. Apprentice carpenters
2. Floor layers
3. Journeyman carpenters
4. Carpenters who worked specifically for the city of St. Louis.
2-2
-------
In Philadelphia the list included approximately 1,300 union members; but, unlike St. Louis, the list
was not subdivided by type of work.
Non-Union Workers
Three strategies were used to identify independent workers in St. Louis:
1. The local Home Builders Association (HBA) provided a list of 250 contractors who
are current members of the association.
2. One hundred forty-four (144) potential respondents were identified using
information found in local telephone books.
3. Advertisements in local newspapers, public service announcements on television,
and referrals from study participants yielded 194 potential respondents.
In Philadelphia, 202 potential respondents for the non-union group were identified using
advertisements in local newspapers.
2.2.3 Sampling Methods
Sampling methods differed for union and non-union workers. Workers from union
membership lists were randomly sampled for participation in the study. For the non-union groups,
an attempt was made to contact everyone identified in the sampling frame to maximize our ability
to find qualified, willing independent workers for the study.
Union Workers
In St. Louis, an attempt was made to contact 100% of the apprentice carpenters,
journeyman carpenters, and carpenters who were union members and worked specifically for the
city of St. Louis. However, only a random sample of approximately 60% of the floor layers was
contacted because it was believed that their potential lead exposure would be lower than in the
other three groups. In Philadelphia, a random sample of approximately 55% of the union
members working for the city was contacted.
Non-union Workers
An attempt was made to contact all non-union workers identified in the frame building
process. This included:
1. All contractors identified by the HB A and all employees identified by contacting
HBA contractors
2. Everyone identified through telephone book listings
2-3
-------
3. Everyone whose name was received as a referral from other workers
4. Everyone who responded to the newspaper advertisements and public service
announcements.
2.2.4 Recruitment
To meet the study objectives, R&R workers meeting specified criteria were needed to
participate in the study. Recruitment activities consisted of (1) gaining the support of national
leadership of the UBC and the National Association of Home Builders (NAHB), (2) obtaining the
support of local leadership, (3) compiling sample frames, and (4) contracting and recruiting
qualified workers from each sampling frame.
Support of the UBC national leadership was a prerequisite for obtaining cooperation at the
local level. The Director of the Health Effects Division of the UBC Health and Safely Fund
facilitated and coordinated contacts with national and local leadership in the UBC. The Director
arranged meetings between Battelle and the St. Louis District Council of the UBC, and between
Battelle and the Metropolitan District Council of the UBC in Philadelphia. These meetings were
instrumental in obtaining UBC membership lists in St. Louis and Philadelphia.
The Executive Director of the NAHB introduced Battelle to NAHB staff in St. Louis,
who provided Battelle with a membership list.
Workers in the sample frame were initially contacted by telephone. Recruitment during
telephone screening involved 1) determining if the worker was eligible, 2) convincing eligible
workers to participate, and 3) scheduling appointments for data collection. Recruitment did not
end with the telephone screening. It also involved getting a person to complete the questionnaire
and provide a blood sample during data collection and rescheduling appointments for individuals
that failed to show up at the scheduled appointment time. A $50 cash incentive was offered to the
workers who participated in the data collection procedure.
When it became clear that the telephone screening was not recruiting sufficient numbers of
non-union, independent R&R workers, other strategies were implemented. An advertisement was
placed in newspapers in both cities, a videotape was sent to the Public Broadcasting Service
(PBS) station in each city for broadcast, and a word-of-mouth referral approach was employed
with workers who did participate. The referral approach involved cash payments to each worker
that provided the names of at least three R&R workers that were eligible for the study. The
turnaround in the success of recruitment efforts with this group can be attributed primarily to the
newspaper advertisements. The referral approach was the next most helpful. Only the PBS
station in St. Louis broadcast the videotape; the broadcast was not a significant contributor in
bringing independent workers into the study.
2-4
-------
2.3 SAMPLE COLLECTION
2.3.1 Questionnaire Data
Two questionnaires were used in this study: a telephone screening questionnaire and the
main study questionnaire. Both questionnaires are provided in Appendix A. The telephone
screening questionnaire was administered by an interviewer, and the main study questionnaire was
self-administered. A complete pretest of both questionnaires was conducted to ensure that the
data would be reliable and useful.
For both questionnaires, questions with pre-coded responses were chosen to avoid
ambiguous answers to open-ended questions. This practice ensured consistency in the
respondents' answers and minimized the potential for information bias. It also facilitated data
editing, cleaning, coding, analysis, and interpretation. A few questions were not amenable to
closed-ended responses and were left open. For example, "What is your current job title and what
are your main responsibilities?"; responses were categorized retroactively.
Telephone Screening Questionnaire
The screening questionnaire included 11 questions related to current work activities and
demographic characteristics. These data were used to
p Determine whether a respondent was eligible for the study.
A worker was included if 1) the worker's primary source of income was derived
from R&R activities and 2) the worker actually did "hands-on" R&R work.
p Characterize the population.
Because a larger number of workers was contacted by phone than participated in the
main study, information from the phone screen was particularly useful for helping to
characterize the population of workers according to basic demographic
characteristics and current work activities such as carpet removal, paint removal,
and cleanup.
p Ensure variability in lead exposure.
The telephone screening questionnaire was used to ensure that workers with a range
of potential lead exposure were identified. In the union sample in St. Louis, for
every two respondents who indicated they had worked on buildings built before
1950, one worker was recruited who had been working in buildings built after 1950.
This maximized the chances of recruiting highly exposed workers. This practice
was initiated after determining that a large proportion of union workers initially
recruited were not working in older homes.
p Assess potential selection bias.
2-5
-------
Finally, the information from the phone screen allowed for comparison between basic
demographic characteristics and targeted work activities for workers who agreed to participate in
the main study and those who refused to participate.
Main Study Questionnaire
Table 1 summarizes the main study questionnaire and shows the links between the
questions and the study objectives. Although the questionnaire was self-administered, an
interviewer was always available to help the respondent with particular questions or to administer
the entire questionnaire, if necessary. The completed questionnaire was also reviewed for
consistency and completeness before the respondent left the data collection center.
2.3.2 Blood Samples
To minimize the potential for contamination and to insure comparability to data collected
in other studies, blood samples were collected by venipuncture, rather than by finger prick. Blood
draws were performed by trained and licensed phlebotomists. One to three phlebotomists
attended each data collection session, depending on the recruited number of workers.
The phlebotomy was performed immediately following administration of the questionnaire.
The protocol for collecting, storing, and shipping the blood samples appeared in the QAPjP for
the WCBS.
Table 1. Summary of Questionnaire and Rationale for Each Section
Section(s)
A. Demographics
B to H. Work History:
Performance of specific
Work Activities and
Practices
1. Work History: General
Work Practice Questions
J. Work History: Other
Occupational Exposures
K. Personal Characteristics
L. Non-work Activities
M. Medical History
N. Previous Training
Objective
1,2,3
1,2,3
1,2
1,2
1,2
1,2
1,2
1,2,3
Rationale
Includes questions on demographic factors — such as age, race,
education, and gender — which may be related to the worker exposure.
This information is also useful to characterize the population of workers
in terms of basic demographic factors.
Includes queries about each specific work activities, and specific work
practices associated with each activity. Targeted activities assessed are
carpet removal, window or door replacement, heating, ventilation and air
conditioning work, removal of large structures, paint removal/surface
preparation, and cleanup.
Includes questions about potential confounding factors such as smoking
and eating at the worksite.
Includes questions about potentially confounding non-R&R occupational
lead exposures.
Includes questions about the worker's home and dietary habits which
may be related to worker's non-occupational lead exposure.
Includes questions about hobbies and sporting activities which may be
related to a worker's non-occupational lead exposure.
Includes questions on worker's medical history, which may reflect his/her
potential for lead exposure (diagnosed with an elevated lead level).
Includes questions about training and information that the worker has
received about potential lead exposures in the workplace.
2.4 E>ATA MANAGEMENT
2-6
-------
Carefully designed data control procedures were employed to ensure that all data collected
were accurate, consistent, and complete. During all the steps of data management, measures were
taken to ensure confidentiality. A copy of the confidentiality pledge, which all interviewers and
field workers are required to sign, is provided in Appendix G. Locked file cabinets were assigned
in which all hard copies were kept. Access to these file cabinets was limited to those directly
involved in data collection, editing, and cleaning of data for this study.
There were four components to the data control procedures:
1. Data receipt and control system update
Data receipt and control procedures served as a link between data collection and
data preparation. The data receipt and control system ensured that all documents
required for each case were received and logged. Routine reports were produced
on the number of cases collected at each stage of processing. These reports allowed
for timely identification of documents not received from the field.
2. Data editing and coding
All data underwent a series of steps to ensure that they were maximally error-free
prior to electronic storage. When a data collection form was completed it was
edited for missed, inconsistent, or illegible responses. Any problems were checked
with the respondent while he/she was still present at the data collection site.
Completed data collection forms were logged in and sent to the data preparation
department to be thoroughly edited for completeness, accuracy, and consistency.
Editors conducted a question-by-question review of the data collection form.
During this step the data were checked again for inappropriate skips of questions,
double coding, inconsistencies, and illegible responses. Any inconsistencies or
unusual situations were referred to the Data Preparation Manager who was
responsible for handling all editing and coding decisions. Missed questions or
inconsistent responses were retrieved from the respondent whenever possible.
3. Data entry and verification
Once data passed the manual edit, they were transferred to data entry. Data sets
were keyed in-house using double entry to verify correct keying of the data. Any
discrepancies in keying were corrected before computer editing of data.
2-7
-------
4. Computer edits
Computer edits of the data took place after data were entered into the computer. A
set of edit specifications were created by the Data Preparation Manager to check
out-of-range values (e.g., more than 30 days worked in last month), inconsistencies
across variables and skip patterns. The data set was then checked against these
specifications, and a computer printout was produced to list all errors found in the
data. Errors identified by this procedure were corrected by the editing staff, and the
corrections were made to both the hard copy and the data disk. The data set was
run against these specifications a second time to ensure that all corrections were
made. This procedure was repeated until no errors were found in the data.
2.5 STATISTICAL METHODS
The statistical analysis included several preliminary steps, including constructing exposure
variables, calculating descriptive statistics, and exploring data analysis. Statistical models were
then fitted to the data to meet the study objectives listed in Section 2.1. The statistical models
were used to assess relationships between blood-lead concentrations and potential lead exposure
associated with the target activities. These relationships were investigated for three time periods:
exposure during the previous month, exposure during the past year, and historical exposure. All
statistical analyses were performed using the SAS® computing system (version 6.10).
2.5.1 Construction of Exposure and Worker Group Variables
Questionnaire responses and measured blood-lead concentrations were used to construct
variables for statistical analysis. The primary response variable for statistical analyses was blood-
lead concentration. Histograms, probability plots, and descriptive statistics were examined to
determine the distribution that best approximates the realized sample of blood-lead
concentrations.
Measures of potential lead exposure resulting from conducting R&R work were
constructed for three exposure periods: last 30 days, last year, and the entire career. These
exposure measures were constructed for each target activity (carpet removal, window
replacement, paint removal and surface preparation, work on HVAC systems, large structure
removal, and cleanup) and for conducting R&R work in general.
For each specific target activity, the potential lead exposure variables were constructed
from the following questions:
Short-term: In the last 30 days, how many days did you work on the target
(1 ast 3 0 day s) activity?
In the last 30 days, how many days did you work on the target
activity in homes or buildings built before 1950?
2-8
-------
Mid-term: Altogether in the past 12 months, how many weeks did you
(last year) work on the target activity?
(0) None
(1) < 1 Week
(2) 1-4 Weeks
(3) 4-8 Weeks
(4) 9-26 Weeks
(5) > 26 Weeks
Long-term: Think about all the years you've done renovation and
(entire career) remodeling. How many of these years did you work on the
target activity at least some of the time?
Variables constructed from the responses to these questions were used to assess the effects of
target activities on blood-lead concentration.
Since there were two questions related to short-term exposure, an effort was made to
determine which one was most strongly related to blood-lead concentration. The distinction
between a target activity in pre-1950 and post-1950 buildings is important because it is believed
that lead exposures resulting from R&R may be greater in older homes. For each target activity,
relationships were examined between worker blood-lead concentration and the number of days
the target activity was conducted, the number of days conducted in pre-1950 houses, and the
number of days conducted in post-1950 housing (calculated by computing the difference between
the number of days worked and the number of days worked in pre-1950 buildings). Based on
plots and univariate regressions, the number of days an activity was performed in homes built
before 1950 was selected as the measure of short-term exposure.
In addition to union status (union carpenter and non-union carpenter), each worker was
assigned to a specific worker group based on his response to the following question: What is
your current job title and what are your main activities at work? A listing of the main activities
was used to define the worker groups, independent of blood-lead concentrations and target
activities. When the subject's main activity response was insufficient for defining an appropriate
worker group, both job title and main activity were taken into consideration.
Demographic variables such as age, gender, race/ethnicity, and level of education were
constructed from the questionnaire responses. The questionnaire also provided information on
potential lead exposure that occurred outside of R&R work. An indicator (zero or one) variable
was constructed from the responses to questions on other activities (Appendix A). If a worker
responded positively to one or more of those questions then he was assigned a value of one for
the variable "Other Occupations," indicating potential occupational exposure outside of R&R. A
similar variable for potential lead exposure was defined based on the responses to questions on
non-work related activities. Variables were also constructed for the use of specific work practices
and work habits.
2-9
-------
2.5.2 Descriptive Statistics
Summary tables of demographic variables were prepared for each worker group and are
included in Appendices C and D. For specific worker groups, Table D-3 displays the amount of
time each target activity was performed and how often specific work practices were utilized.
Two-way frequency tables of the number of days the activity was performed versus the number of
days the activity was performed in pre-1950 homes were prepared for each target activity. Tables
in Appendix C summarize the distribution of blood-lead concentrations for each R&R worker
group and for each sampling frame. Tables (Section 3.5) were also prepared to assess the
variability in measured blood-lead concentrations between duplicate blood samples and among
duplicate chemical analyses.
2.5.3 Exploratory Data Analysis for Ancillary Variables
Exploratory data analyses were performed to assess the relationships between blood-lead
concentration and various ancillary variables describing worker demographics, worker practices,
and work site characteristics. The purpose of these analyses was to select ancillary covariates for
modeling the relationships between blood-lead concentration and target activities. For each
ancillary variable, the analyses included a plot against blood-lead concentration and a statistical
test to assess the significance of any functional relationship revealed in the plot. Analyses of
variance (ANOVA) were carried out for categorical variables and significance of slopes of linear
regressions were examined for continuous variables. These analyses were conducted over all
workers and for each sampling frame.
2.5.4 Statistical Models
2.5.4.1 Blood-Lead Concentrations
The QAPjP specified that at least one set of CDC blood-lead quality control reference
(CDC QC) samples be included in each shipment of blood samples. Nominal blood-lead
concentrations of the low, middle, and high CDC QC samples were 9, 23.3, and 41.1 |ig/dL,
respectively. An ANOVA model appropriate for random effects was fitted to the CDC QC
samples to assess the variability between replicate samples at same blood-lead concentration and
to estimate recovery rates at each concentration.
At least two chemical analyses were performed on each blood sample. In addition, as
specified in the QAPjP, approximately 10% of the workers was selected for duplicate blood
draws. An ANOVA model appropriate for random effects was fitted to the subset of workers
possessing two blood samples to assess the variability in blood-lead concentrations between
duplicate blood draws. The following random effects were included in this model: (1) worker,
(2) blood sample nested within worker, and (3) analysis nested within blood sample.
Analyses were conducted to determine if statistically significant differences exist between
the blood-lead concentrations of specific R&R worker groups and between the blood-lead
concentrations of different sampling frames. Side-by-side box plots of blood-lead concentrations
2-10
-------
for each worker group and for each sampling frame were prepared (Appendix F). ANOVA
models appropriate for random effects were then fitted to the blood-lead concentrations. The
following effects were included in these models: 1) fixed effects for either worker group or
sampling frame, and 2) random effects for worker and chemical analysis. Statistical contrasts
were conducted between the sampling frames. Because of the larger number of worker groups,
pairwise comparisons between worker groups were performed using Tukey's studentized range
test.
2.5.4.2 Relationships Between Target Activities and Blood-Lead Concentrations
A series of statistical models were fitted to the data to determine if there were any
significant associations between blood-lead concentration and various types of work or target
activities. The relationship between worker blood-lead concentrations and potential lead
exposure associated with R&R target activities was investigated for exposure during the previous
month, exposure during the past year, and historical exposure. Multiple regression models were
employed to examine these relationships. To simplify the regression models, results of multiple
chemical analyses and duplicate blood samples were averaged for each worker to provide a single
blood-lead concentration for each worker.
Figure 2 portrays the paradigm utilized for fitting the models. First, as shown at the top of
the figure, results of the preliminary analyses were employed to
1. Define the measures of exposure
2. Select the most appropriate distribution for blood-lead concentrations
3. Select covariates for the statistical models.
Second, presented on the left branch of Figure 2, separate models were fitted to the data
for each target activity. Initially, linear regression models were fitted to the log transformed
blood-lead concentrations using each of the exposure measures as the independent variable. Next,
the analyses were repeated incorporating the previously selected ancillary variables as covariates.
Finally, a linear regression model that incorporated the ancillary covariates and simultaneously
investigated the potential for lead exposures within the past month, the past year, and historically
was fitted to the data for each target activity.
The above analyses helped characterize the strength of the relationship between each
target activity and worker blood-lead concentrations. The final goal, however, was to develop a
model which explained how each of the target activities interacted in their association with blood-
lead concentrations while accounting for the effect of potentially confounding ancillary covariates.
Therefore, as illustrated by the right branch of Figure 2, regression models that examined all of
the target activities simultaneously, were fitted to the data. The initial models included effects for
all six target activities for each exposure period: short-term, mid-term, and long-term. For
instance, a model was fitted to the data that included the previously selected ancillary variables
and the number of days the activity was conducted in pre-1950 homes for each of the target
activities. Next, these models were repeated with worker group added to the model. Finally, an
attempt was made to construct a model that would assess the effects of short-term, mid-term, and
2-11
-------
long-term exposures simultaneously for all of the explanatory variables. Correlations among the
target activities and among the three exposure periods within each target activity were high.
Therefore, as shown in Figure 2, only a subset of the variables for the various exposure period
and target activity combinations were included in the final model.
PRELIMINARY ANALYSES
Analyses
For Each
Target Activity
Without Covariates
For Each
Exposure Measure
_L
With Covariates
For Each
Exposure Measure
I
With Covariates
Employing All
Exposure Measures
Analyses
Employing All
Target Activities
I
With Covariates and
Without Job Category
For Each Exposure
Measure
With Covariates and
With Job Category
For Each Exposure
Measure
I
With Covariates and
With Job Category
Employing Selected
Exposure Measures
Figure 2. Paradigm for Fitting Statistical Models to Blood-Lead Concentrations
2-12
-------
3.0 RESULTS
3.1 RECRUITMENT RESULTS AND FIELD EXPERIENCES
3.1.1 Recruitment of Respondents
Individual workers were recruited during the telephone screening process. Up to five
telephone calls were made to each worker. If a contacted worker was judged to meet study
eligibility criteria and scheduled for an appointment for data collection, then the worker was
classified as Recruited. Otherwise, the worker was assigned one of the following classifications:
Not Located: Potential respondents who could not be contacted (e.g., due to invalid
telephone number, or not at home).
Refused Screener: Potential respondents reached who refused to participate in the
telephone screening/recruiting interview.
Not Eligible: Respondents who participated in the telephone screening/recruiting
interview, but did not meet the criteria for the main study data
collection. A respondent failed to meet study eligibility criteria if 1) he
did not conduct "hands-on" R&R work for a living, or 2) he belonged to
a group (e.g. floor layers) whose quota for the study was already met.
Eligible Refused: Telephone screening/recruiting interview respondents who met the
participation criteria for the main study, but declined to participate.
Other: Telephone interview respondents who had dispositions other than those
listed above. Examples include potential respondents taken from the
union membership list who were no longer union members, potential
respondents who were deceased, and potential respondents whose
spouses did not cooperate in providing access to the respondent.
Table 2 provides a breakdown of recruitment results for the telephone screening for all
potential respondents. (A potential respondent is anyone the telephone screening process
attempted to reach. Total respondents are presented in the third column of Table 2.) A subset of
the recruited workers participated in the questionnaire and blood data collection. Workers who
completed both the questionnaire and the blood draw were classified as Complete To measure
the rate of screening recruitment, participation, and overall response, the following rates were
calculated:
o • T. -^ ^ T. ^ Recruited
Screening Recruitment Rate
Recruited Eligible Refused
3-1
-------
„,..,. D , Total Complete
Participation Rate
Recruited
D D , Total Complete
Response Rate
Recruited Eligible Refused Refused Screener
The screening recruitment rate measures the recruitment rate of the telephone screening
interview in Phase I. Given that a worker was successfully recruited in Phase I, the participation
rate measures the response rate of questionnaire and blood sample collection in Phase II. The
overall response rate of both study phases is measured by the response rate.
Summary of Overall Recruitment Results and Participation
Once workers were reached at home, the interviewers had little problem convincing them
to do the telephone screening interview. Of the 1,686 workers contacted, only 150 (9%) refused
the telephone interview. The interviewers had even less of a problem recruiting eligible
respondents to set an appointment date for the main data collection (self-administered
questionnaire and phlebotomy). Overall recruitment results and actual participation are
summarized in Table 3 for each sampling frame. "Total Contacted" in Table 3 represents the
number of potential respondents who were actually reached and represents the sum of Refused
Screener, Not Eligible, Eligible Refused, Recruited, and Other categories. Total self-administered
questionnaire (S AQ) represents the number of respondents who completed the questionnaire.
Blood samples were not collected from four of the 585 workers who completed the questionnaire
due to unsuccessful blood draws.
Recruitment Results for Each Sampling Frame
Recruitment of union workers in St. Louis began in mid-August, 1994. The screening
recruitment rate for St. Louis union workers was 94%. The participation rate was 60%, and the
response rate was 43%. This group had the lowest screening recruitment rate, participation rate,
and response rate.
Recruitment of union workers in Philadelphia began in mid-September, 1994. Unlike the
St. Louis sample, the Philadelphia workers were not divided into subgroups (i.e., apprentice,
journeyman, etc.). As shown in Table 3, recruitment of union members in Philadelphia went
better than in St. Louis. The screening recruitment rate for Philadelphia union workers was 96%.
The participation rate was 63%, and the response rate was 54%.
3-2
-------
Table 2. Summary of Recruitment for Telephone Screening Interview
City
ST. LOUIS
UNION
ST. LOUIS
NON-UNION
PHILADELPHIA
UNION
PHILADELPHIA
NON-UNION
Group
Apprentice
Floor Layer
Journeyman
Carpenter
Totals
Independents01
Local Home
Builders
Association
Referrals / Ads
Totals
Union
Non-Union
Total
Called
97
393
500
96
1086
103
32
189
324
674
189
Recruited
26
130
85
10
251
20
13
174
207
312
108
Not
Located
28
157
134
38
357
41
5
3
49
140
41
Refused
Screener
11
41
24
5
81
16
4
3
23
38
8
Not
Eligible
29
44
247
40
360
22
8
8
38
166
27
Eligible
Refused
3
7
7
0
17
2
2
0
4
12
1
Other
0
14
3
3
20
2
0
1
3
6
3
(1) Workers obtained through the telephone work.
Table 3. Summary of Overall Recruitment and Participation for WCBS
Group
St. Louis Union
St. Louis
Non-Union
Philadelphia
Union
Philadelphia
Non-Union
Totals
Total
Called
1086
324
674
189
2273
Total
Contacted
729
275
534
148
1686
Total
Recruited
251
207
312
108
878
Total
SAQ<"
151
161
197
76*
585
Total
Complete121
150
160
197
74*
581
Screening
Recruitment
Rate
94%
98%
96%
99%
96%
Participation
Rate
60%
77%
63%
69%
66%
Response
Rate
43%
68%
54%
63%
55%
(1) Number of respondents completing the self-administered questionnaire.
(2) Number of respondents completing both the questionnaire and the phlebotomy.
* From a total of 78 workers, 76 questionnaires were obtained, and 76 blood samples were collected. Both questionnaire and blood
data were collected from 74 workers.
3-3
-------
Recruitment of non-union carpenters in St. Louis began in September, 1994. After two
weeks only six workers had been recruited and therefore, the additional recruitment methods
presented below were employed:
1. R&R contractor names were abstracted from the telephone book and newspaper
advertisements. The intent was to reach small-time contractors who do remodeling
work themselves.
2. Classified advertisements were placed in five St. Louis newspapers: The Post,
Dispatch, The Suburban Journal, the Riverfront Times, and Today's Advantage.
3. A 30-second Public Service Announcement (PSA) promoting the WCBS was
videotaped and broadcast on public television. The St. Louis PBS affiliate, KETC
Channel 9, was contacted and asked if they would be willing to advertise the WCBS
on "This Old House," "Home Time," and other R&R instruction shows. The PSA was
shown multiple times following these programs.
4. A news release on the study was sent to Philadelphia and St. Louis newspapers on the
study. The press release was conducted on November 17, 1994.
5. Solicitation of references from workers already recruited was attempted. This
consisted of paying a recruited respondent $25 for the names of at least three people
who do remodeling work.
These various methods resulted in three general subgroups of non-union workers:
1. Independents — Workers drawn from the telephone book
2. HBA workers — R&R workers recruited through their employer, an HBA member
3. Referrals/ads — Workers referred by other respondents and those who called in
response to a newspaper/ television advertisement
Despite the early setbacks, recruitment of non-union workers in St. Louis ended
successfully. This success was especially due to the use of the advertisements and word-of-mouth
referrals from other workers. These two methods resulted in 84% (174) of the total St. Louis
non-union workers recruited. The screening recruitment rate for St. Louis non-union workers
was 98%. The participation rate was 77%, and the response rate was 68%.
Recruitment of non-union workers in Philadelphia did not begin until mid-October, 1994.
Newspaper advertisements were selected as the primary recruitment source based on the
recruitment experience in St. Louis. Advertisements were placed in the following Philadelphia
newspapers: The Philadelphia Inquirer, The Daily News, The Leader, The Breeze, The Review,
The Recorder, and the Germantown Paper. The screening recruitment rate for Philadelphia non-
union workers was 99%. The participation rate was 69%, and the response rate was 63%.
3-4
-------
3.1.2 Field Experiences
3.1.2.1 Field Experiences in St. Louis
All of the telephone interviewing and recruiting was performed in the St. Louis office of
Survey Research Associates (SRA). As a result of the high screening recruitment rate, no
substantial refusal conversion effort was necessary. However, no-shows were called to
reschedule data collection appointments.
Initially, Battelle anticipated scheduling half of the data collection sessions at central
locations (e.g., SRA's St. Louis office, union halls) and half in respondents' homes. In the St.
Louis area, however, only 12 in-home visits were required; six in Missouri and six in Illinois.
Respondents were very willing to travel to a central location for the data collection. The
weeknight dates worked best since the respondents could come in after work. The weekend
sessions had slightly poorer turnout and were stopped about halfway through the field period.
In St. Louis, two locations were used for the centralized data collection sessions: the
United Brotherhood of Carpenters Hall and SRA's office.
Respondents tended to arrive in groups of five to ten. A receptionist would check a
respondent in upon arrival and provide instructions on how to proceed. The original plan was to
have an interview specialist help respondents complete the questionnaire. It soon became clear,
however, that the respondents did not have much trouble with the questionnaire. The Study
Manager therefore handled most of the questions and the interview specialist assisted with editing
and data retrieval. (There were two respondents who required assistance reading the
questionnaire. One was a Russian immigrant who understood spoken English well, but had a hard
time reading the questions and response choices. The other was a respondent who asked that the
questionnaire be read to him after he had tried to go through it himself.)
The phlebotomy went smoothly with few problems being encountered. Although there
was some early concern with the need for multiple sticks for a successful draw, only two St. Louis
respondents were unable to complete the blood draw. The nurses were very good at explaining
the phlebotomy procedures and relaxing the respondents. Any respondent feeling unsteady after
the blood draw was not allowed to leave until he was feeling better.
As mentioned previously, only a small number of respondents required in-home visits. For
those respondents in Missouri, the primary reason for the in-home session was an inability to
come to a central location during the scheduled times; usually this was due to a conflicting work
schedule. For the in-home sessions in Illinois, the primary reason given by respondents was the
distance to the central locations (at least a half-hour drive through rush hour traffic). An
interview specialist and a phlebotomist traveled to the respondent's home and collected the data
and blood draw in a similar fashion to procedures used at the central location.
3-5
-------
3.1.2.2 Field Experiences in Philadelphia
Philadelphia union recruitment was even smoother than in St. Louis. One factor was that
recruitment in Philadelphia occurred after recruitment in St. Louis and, as a result, the
interviewers were more experienced with the process.
In Philadelphia, three locations were used for the centralized data collection sessions: the
UBC Carpenters Hall, the UBC training facility, and the Center City Hotel. The hotel facility was
secured to hold the data collection sessions for the non-union workers. Data collection sessions
were scheduled on Saturday during the day and on weeknights.
The procedures and data collection flow in Philadelphia were essentially the same as in St.
Louis. As in St. Louis, two respondents were unable to successfully provide a blood sample.
There were some problems with the Philadelphia data collection. The first concerned the
editing function. The purpose of the editing was to catch missing information and retrieve the
data before the respondent left. Some of the forms sent in for processing from Philadelphia did
not have proper editing. This caused the data preparation staff in St. Louis to telephone the
respondents to retrieve the missing data. In some cases it was not possible to reach the
respondent, and therefore the response(s) were coded as missing data. This problem was noticed
early in the data collection, allowing the Study Manager time to travel to Philadelphia and re-train
the entire staff.
The second problem was related to the payment of the $50 cash incentive. Cash payments
do yield more cooperation among prospective respondents compared to sending a check at a later
date. However, staff in Philadelphia were concerned with safety in handling the large amounts of
cash required for the study, sometimes as much as $2,500. The staff was especially concerned
with safety at the Center City Hotel site. To ameliorate that concern, the hotel provided a
security guard to provide a security presence in the room.
No in-home sessions were done in Philadelphia for two reasons. First, the telephone
interviewing staff were unfamiliar with the large geographic area around Philadelphia. Second,
with the high turnout in both cities, it was not worthwhile to work out the logistical problems
associated with in-home visits for a small number of respondents. No one refused to participate
because of the travel to a central location. In fact, some of the respondents indicated that they
had traveled from 45 minutes to 1-hour away.
3.1.2.3 Working with Difficult Respondents
A common problem in St. Louis and Philadelphia was dealing with difficult respondents.
The primary difficulty was respondents arriving for the data collection after drinking alcohol. This
was mostly a problem with the non-union workers in Philadelphia. While intoxicated respondents
were certainly not the rule, the questionnaire and blood data of two workers in Philadelphia were
not utilized because of the respondents' drunkenness.
3-6
-------
Without question, the predominant mood of the respondents was one of cooperation.
They seemed happy to participate and were genuinely concerned with the goals of the study as
well as their own blood-lead results.
3.2 QUALITY CONTROL FOR BLOOb-LEAb EATA
3.2.1 Laboratory Quality Control Data Summary
The assessment of the overall quality of the blood-lead data was performed by the Senior
Quality Assurance Officer at Midwest Research Institute using statistical quality control (SQC)
procedures. Data quality indicators, as defined in the QAPjP, were used to assess the data
obtained from quality control and performance evaluation specimens. This section summarizes
the laboratory quality control data evaluation; details are presented in Appendix B.
Data quality objectives were met with analytical data meeting all objectives of accuracy,
precision, and completeness of specimen collection. Analytical results are considered
representative of lead levels in the R&R population and comparable with results from other
studies.
Measurement quality objectives (MQO) were assessed using internal quality control and
CDC performance samples. The internal quality control samples consisted of QC blanks (matrix
modifier) and a series of standard reference materials that were used as calibration standards
(NIST 995a SRMs at Levels 1, 2, 3, and 4), calibration check standards (NIST SRM Level 2),
calibration verification standards (NY State SRM), and continuing calibration samples (BioRad-1,
BioRad-2, and BioRad-3). Performance criteria and results are summarized in Table 4. In
addition to the SRMs used for internal quality control, all of the samples were analyzed in
duplicate (same extract) and selected samples were analyzed in replicate (different sample extract)
to assess method precision. Precision criterion established for these samples was ±20%.
Results from internal quality control and external performance evaluation specimens
indicated that the blood-lead data were accurate and reliable. The instrument detection limit
(IDL), method detection level (MDL) and method quantitation level (MQL) were 1.3, 1.9, and
3.2 |ig/dL, respectively. Only 20 out of 581 worker blood lead concentrations were below the
IDL. Precision criteria were met for the majority of cases; those that did not meet established
precision criteria were reanalyzed. Details of all quality control data analysis, including Shewhart
plots and statistical evaluations of the data, are shown in Appendix B.
In summary, blood-lead data met the quality criteria with the exception of the positive bias
noted for the time period of December 15b30, 1994, for specimens in the range of 10p20 |ig/dL,
as assessed by the calibration verification quality control sample. As a result, reported lead levels
in these samples had the potential for being less than 1 |ig/dL above the actual lead level in the
sample. This bias has little clinical significance.
3-7
-------
Table 4. Quality Control and CDC Performance Samples
Sample
Identification
Blank
BioRad 1
BioRad 2
BioRad 3
New York State SRM
NIST SRM 995a
CDC RS 590
CDC RS 991
CDC RS 1394
MQO Accuracy
< 1 .0 |jg/dL
±25%
±25%
±25%
±10%
±10%
±20%
±20%
±20%
MQO
Precision
N/A
±10%
±10%
±10%
N/A
N/A
N/A
N/A
N/A
Type of Quality Control
Sample
Matrix modifier
Continuing calibration
Reference material
samples
Continuing calibration
Reference material
samples
Continuing calibration
Reference material
samples
Calibration verification
Calibration Check
Standard
Blind performance
Blind performance
Blind performance
Reference
Lead Cone.
ug/dL
0.0
5.0-8.3
24.4 - 24.5
55.4
16.0
13.53
9.0
23.3
41.1
Mean
Recovery
(%)
-0.08
99.74
95.24
95.98
98.25
101.0
93.35
98.72
102.6
Control
Limits
1.59
58-142
72-118
81-111
68-130
79-122
77-110
89-109
91 -114
3.2.2 Field QC Data
The quality of blood-lead results from the WCBS were characterized using several
different field samples. First, CDC Quality Control reference (CDC QC) samples were repeatedly
measured over the course of the study. Second, field duplicate samples (a second blood sample)
were analyzed for roughly 10% of the workers. Finally, at least two chemical analyses were
performed on each blood sample.
The quality of the blood-lead results will be discussed in terms of bias and precision. The
CDC QC samples were used to characterize any potential bias in the blood-lead data.
Duplicate analyses of the same sample were used to characterize the precision in the chemical
method of analysis. Blood-lead concentrations from the field duplicate samples were used to
characterize the variability in both sampling and chemical analysis.
3.2.2.1 CDC QC Reference Samples
Three CDC QC samples were analyzed 30 times each over the course of the study
(yielding a total of 90 reference samples). These reference standards consisted of bovine blood
which was spiked with known concentrations of lead (in units of 9.0, 23.3, and 41.1 |ig/dL). The
CDC QC samples were analyzed along with the regular worker blood samples to characterize any
potential bias in the blood-lead results. Each of the 90 CDC reference samples were shipped
3-8
-------
blindly to the laboratory in collection tubes similar to those that contained worker blood. The
lead concentration in each tube was then analyzed using duplicate injections (chemical
measurements) similar to the worker samples.
Table C-l in Appendix C presents descriptive statistics for the CDC QC samples. These
results are based on the geometric mean blood-lead concentration of the duplicate chemical
analyses made on each sample. The log standard deviation represents the variability between
different sample tubes of the same reference standard.
Table 5 presents 95% confidence intervals for the lead concentration in each CDC
reference standard. Only one of the three confidence intervals contained the nominal blood-lead
concentration. This, however, is not cause for immediate concern, due to the fact that the
estimated confidence intervals were very narrow and just barely excluded the nominal
concentrations. Additionally, differences between the observed geometric means and the nominal
concentrations were relatively small. All but one of the blood-lead concentrations, as discussed in
Appendix B, met the measurement quality objective for the CDC QC reference samples.
Table 5. 95% Confidence Intervals for Geometric Mean of Blood-Lead Concentrations
for COC QC Reference Samples
CDC
Reference
Number
590
991
1394
Nominal
Concentration
(uq/dL)
9.0
23.3
41.1
Measured Responses
N
30
30
30
Geometric
Mean
(uq/dL)
8.4
23.0
42.3
log
Std. Error
loq (uq/dL)
0.010
0.006
0.007
95% Confidence
Interval for
Geometric Mean
(uq/dL)
(8.2,8.6)
(22.8,23.3)
( 41 .8, 42.9 )
3.2.2.2 Worker Blood-Lead Samples
Duplicate blood samples were collected from 73 workers. Each of the duplicate samples
was chemically analyzed in the same manner as the regular worker samples and reference samples
(with at least two measurements of blood-lead level per sample collection tube). Table C-2
provides descriptive statistics on the log standard deviation of duplicate blood-lead
concentrations. The variability between duplicate samples was very small, and the variability
estimates were similar among the sampling frames. In fact, the variability in log blood-lead
concentrations between duplicate blood samples is comparable to that observed between duplicate
laboratory analyses made on the same sample.
Blood samples were obtained from 581 workers during the WCBS. Table C-3 provides
descriptive statistics on the distribution of the log standard deviation between duplicate chemical
measures of blood-lead among 580 workers (one worker had a missing value). These results do
not include the 73 duplicate blood-lead samples that were discussed above. The variability
3-9
-------
between duplicate chemical analyses within a sample was also very small and the variability
estimates were similar among the sampling frames.
Table 6 displays estimated variance components for the variability between workers,
between duplicate blood draws for a worker, and between duplicate chemical analyses on same
blood draw. These variance components are based on analyses of the worker blood samples, the
duplicate blood draw data, and the CDC QC reference samples. Based on the worker blood
samples, the estimated variability between workers [b=0.6158] was roughly an order of
magnitude greater than that attributed to chemical analysis [b=0.0677]. The estimated variance
components for duplicate blood samples also demonstrated that the variability between workers
[b=0.6800] was an order of magnitude greater than those for sampling variability [b=0.0677]
and chemical analysis [b=0.0602]. The estimated variance components for the CDC QC samples
showed that the variability within a sample [p=0.0315] (attributed to chemical analysis) and
between samples [p=0.0369] were similar.
Table 6. Variance Component Estimates for Blood-Lead Concentrations
Source of Variation
Between Workers
Between Duplicate Blood Draws
Between Duplicate Chemical Analyses
Pworker
(Field Samples)
log(ug/dl_)
0.6158
N/A
0.0677
Pworker
(Duplicates)
log(ug/dl_)
0.6800
0.0677
0.0602
PQuality Control
(Reference Samples)
log(ug/dl_)
0.0369
0.0315
Overall, the variability attributed to chemical analysis of worker blood-lead appears to be
twice that attributed to chemical analysis of CDC QC samples. This difference may be due to
differences between human and bovine blood samples. The variability in blood-lead
concentrations due to sampling and analyses was relatively small, exceeded expectations for this
study, indicated that measured blood-lead concentrations are of sufficient quality, and provided
evidence of an instrument detection limit of less than 1 |ig/dL.
3.3 TELEPHONE SCREENING RESULTS
The telephone screening phase had three main objectives. The first objective was to
recruit a sufficient number of workers for the main study and to ensure variability in lead
exposures among those workers. This first objective was successfully met, as discussed in
Section 3.1. The second objective was to provide preliminary information on work activities
practiced by R&R workers. Table 7 summarizes the responses to key questions on the
questionnaire. Conclusions about the R&R workers in the study that were reached from a
qualitative assessment of Table 7 include:
3-10
-------
Table 7. Summary of Telephone Screening Questionnaire Responses Relating to Type
and Amount of Specific Target Activities
Question Number: Description
3: Days(a) worked in last month (includes workers
responding "0')
4: Number of years of renovation and remodeling
work
5: Days worked in pre-1 950 buildings
6: Days worked in residential buildings
7: Days worked in nonresidential buildings
8A: Days spent removing carpets
8B: Days spent removing windows
8C: Days spent working on HVAC systems
8D: Days spent removing large structures
8E: Days spent removing paint
8F: Days spent performing cleanup
Group
Participant"1'
Nonparticipant(c)
Participant
Nonparticipant
Participant
Nonparticipant
Participant
Nonparticipant
Participant
Nonparticipant
Participant
Nonparticipant
Participant
Nonparticipant
Participant
Nonparticipant
Participant
Nonparticipant
Participant
Nonparticipant
Participant
Nonparticipant
0
4.6%
8.9%
1 .4%
1 .2%
29.9%
44.2%
32.6%
49.8%
47.3%
40.9%
64.3%
72.7%
47.9%
62.3%
76.8%
83.4%
30.2%
39.0%
46.3%
60.0%
19.9%
33.7%
1-5
5.5%
6.7%
17.3%
13.5%
15.8%
14.1%
1 1 .5%
8.9%
8.1%
5.8%
27.1%
21 .5%
35.7%
27.6%
17.8%
13.2%
29.8%
27.0%
26.1%
20.9%
22.8%
20.2%
6-10
9.3%
7.3%
22.1%
27.2%
13.2%
10.7%
10.3%
7.3%
6.7%
7.1%
4.3%
2.8%
8.4%
5.2%
2.4%
1 .8%
17.7%
14.7%
1 1 .3%
6.2%
1 1 .0%
8.6%
>10
80.6%
77.1%
59.2%
57.8%
41.1%
31 .0%
45.6%
33.9%
38.0%
46.2%
4.3%
3.1%
8.1%
4.9%
2.9%
1 .5%
22.3%
19.3%
16.3%
12.9%
46.3%
37.4%
(a) Number of days activity was performed during the last 30 days.
(b) Participants represented those workers actually participating in the WCBS.
(c) Nonparticipants encompass those individuals who were eligible but either refused participation or did not show.
1. R&R workers conducted a wide range of activities, as reflected by the percent of workers
who conducted the different R&R activities at least one day in the past month.
2. R&R workers spent considerable time doing large structure removal and paint removal or
surface preparation, activities with potential for creating high lead exposures. Therefore,
the generalists of the R&R industry — carpenters and small independent contractors — and
not just specialists such as painters, conduct a significant amount of these activities.
3. The R&R workers were divided evenly between those that work in residential and those
that work in non-residential buildings.
The third objective was to provide information on potential nonresponse bias. A
nonresponse bias arises whenever nonrespondents differ from respondents. The effect of
nonresponse was investigated by determining if participation was related to exposure.
Effects of nonresponse were characterized by comparing the participant group (eligible
workers who participated in the main study) to the nonparticipant group (eligible workers who
3-11
-------
did not participate in the main study). The nonparticipant group encompassed individuals who
were either eligible but did not show up for the WCBS (90%) or were eligible but refused
participation (10%). As described in Section 3.1.1, the total overall screening recruitment rate
was 96%; the participation rate was 66% and the response rate was 55%. There were 585
participants and 327 non-participants (293 were eligible but did not show and 34 eligible refused).
Demographic comparisons between participants and nonparticipants are summarized in
Table 8 (age levels were comparable between the two groups). Demographic characteristics were
comparable between participants and nonparticipants. As shown in Table 3, almost 60% of union
workers, and approximately 74% of independents participated in the study. In addition, as
displayed in Table 7, 81% of participants conducted more than 10 days of general R&R work in
the last 30 days compared to 77% of nonparticipants, and 41% of participants conducted more
than 10 days of work in pre-1950 housing compared to 31% of nonparticipants. These
percentages suggest that the study participants were more likely to perform large amounts of
R&R work in pre-1950 housing, and therefore more likely to be exposed.
Table 8. Demographic Comparison Between Participants and Nonparticipants
Category
Sender
City
St. Louis
Philadelphia
Race
Male
Female
Philadelphia
St. Louis
Union(a)
Independent
Union
Independent
White
Black
Other
Participant (%)
97
3
47
53
48
52
72
28
84
15
1
Nonparticipant (%)
98
2
49
51
70
30
79
21
89
10
1
(a) The large discrepancy between participants and nonparticipants in this category results from the higher proportion of individuals in St.
Louis who were independents. About 75% of Philadelphia workers were union versus 56% for St. Louis. Participation rates within
each city are consistent.
General R&R work comparisons are summarized in Figure 3. Nonparticipants performed
less R&R work, and worked fewer days in residential (more in nonresidential) and pre-1950
buildings. Means (and standard deviations) were comparable between participants and
nonparticipants for years of general activity (14.1 (9.00) and 14.0 (8.83) for participants and
nonparticipants, respectively).
3-12
-------
3-13
-------
100
80
60
40
20
77
81
9
0 1-5 6-10 >10
General R&R Activity
100
80
60
40
20
41
47
46
6^77
38
0 1-5 6-10 >10
Non-residential Activity
100
80
60
40
20
50
46
34
9 11 7 10
0 1-5 6-10 >10
Residential Activity
100
80
60
40
20
0
0 1-5 6-10 >10
Pre-1950 Activity
Figure 3. Days in Last Month Spent Conducting General R&R, Residential, Non-
residential and Pre-1950 Work for Participants (shaded) versus Non-
Participants
3-14
-------
Comparisons between the conduct of specific target activities for participants and
nonparticipants are presented in Figures F-l and F-2 in Appendix F. The mixture of target
activities performed by participants and nonparticipants is very similar, although nonparticipants
spent fewer days conducting each target activity compared to participants. Generally, the
percentage of workers performing an activity a large proportion of their time (>10 days) was
higher for participants. In fact, the workers (approximately 70%) conducting a large amount of
any specific target activity were participants.
In summary, the workers recruited and participating in the WCBS were more likely to be
exposed than nonparticipants, as determined by conduct of larger amounts of R&R work (both
general and specific). A higher percentage of participants than nonparticipants were independents
(non-union) and Black. The study was very successful in recruiting individuals who conducted a
wide variety of R&R activities, permitting the estimation of lead exposure associated with target
activities.
3.4 WORKER CHARACTERIZATION RESULTS
3.4.1 Demographics
Demographic information on WCBS participants is presented in Tables 3 and 9.
Noteworthy points include:
1. There is an almost even division between workers in St. Louis (53%) and workers in
Philadelphia (47%).
2. St. Louis workers were nearly evenly divided between union (48%) and non-union
(52%) workers while the Philadelphia workers were nearly 75% union.
3. Nearly half (45%) of the workers participating were carpenters, of which 60% were
union members.
4. The mean age of the workers was 38 years with 50% of the workers being between
the ages of 31 and 43 years. Supervisors had the highest mean age at 41 years.
5. For black workers, the greatest number were laborers (36%), painters (19%), and
non-union carpenters (14%). In fact, black workers made up approximately half of
the total number of laborers and painters.
6. Approximately 41% of workers lived in homes constructed prior to 1950.
7. Of those workers who had done home renovation in the last 12 months, 93%
performed the renovation themselves.
3-15
-------
Table 9. Summary of Demographic Data
Variable Description
Sample Size
Age
Gender
Race
Education
Age of Home
Home
Renovation in
the last 12
months
25th Percentile
Mean
75th Percentile
Male
Female
White
Black
Other
Didn't finish High
School
High School Graduate
Pre-1950
(1950-1978)
Post-1978
None
Contracted
Self-Performed
Worker Group
Union
Carpenter
159
32(a)
38.11
43
98.7% (b)
1.3%
93.7%
5.0%
1.3%
3.8%
96.2%
41 .5%
40.3%
18.2%
64.2%
2.5%
33.3%
Non-Union
Carpenter
105
32
38.01
43
99.0%
1.0%
85.7%
1 1 .4%
2.9%
7.6%
92.4%
48.6%
40.0%
1 1 .4%
61 .9%
2.9%
35.2%
Dry wall
Worker
64
29
37.16
45
100.0%
0%
93.8%
4.7%
1.6%
3.1%
96.9%
39.1%
35.9%
25.0%
54.7%
0%
45.3%
Floor
Layer
82
30
38.12
47
98.8%
1.2%
96.3%
3.7%
0%
6.1%
93.9%
24.4%
46.3%
29.3%
74.4%
2.4%
23.2%
Laborer
56
30
35.27
40.50
91.1%
8.9%
42.9%
55.4%
1.8%
14.3%
85.7%
58.9%
28.6%
12.5%
62.5%
5.4%
32.1%
Painter
34
31
39.03
46
94.1%
5.9%
50.0%
47.1%
2.9%
26.5%
73.5%
41 .2%
47.1%
1 1 .8%
64.7%
2.9%
32.4%
Supervisor
57
35
41.14
46
98.2%
1.8%
93.0%
3.5%
3.5%
0%
100.0%
31 .6%
38.6%
29.8%
54.4%
1.8%
43.9%
Window
14
29
34.86
45
100.0%
0%
71 .4%
28.6%
0%
0%
100.0%
21 .4%
50.0%
28.6%
71 .4%
14.3%
14.3%
Other
14
31
40.36
49
100.0%
0%
57.1%
42.9%
0%
0%
100.0%
50.0%
42.9%
7.1%
64.3%
0%
35.7%
Total
585
31
38.04
43
97.9%
2.1%
83.8%
14.5%
1.7%
6.5%
93.5%
40.5%
40.0%
19.5%
63.2%
2.7%
34.0%
(a) 25 percent of the 159 union carpenters were at or under the age of 32.
(b) 98.7 percent of all union carpenters were males.
-------
Generalizations based on worker groups in which only a small number were sampled should be
made with caution.
3.4.2 Target Activities
Tables 10 and D-2 summarize the extent to which each target activity was performed for
each worker group. The sampled workers spent an average of 17 days during the past month on
general R&R, of which 11 were spent in pre-1950 homes. The most frequent activity performed
was cleanup, which occurred an average of 11 days during the month. Over all workers, large
structure removal was conducted on an average of 7 days during the month, paint removal on 6
days, window or door replacement on 4 days, carpet removal on 2 days, and HVAC work on 1
day. As with all workers combined, for all of the worker groups except supervisor, the activity
conducted most frequently was cleanup.
As shown in Figure F-3, laborers, dry wall workers, non-union carpenters, painters, and
window installers spent more time performing the target activities in pre-1950 dwellings
compared to floor layers, union carpenters, and supervisors. As expected, paint removal was
conducted most often by painters, window replacement was performed most often by window
installers, and carpet removal was conducted most often by floor layers. Laborers, drywall
workers, non-union carpenters, and window installers conducted, on the average, more than 5
days of large structure removal in pre-1950 dwellings. Laborers and non-union carpenters
performed a wide mix of target activities, and floor layers and supervisors spent fewer days
conducting the selected target activities compared to the other worker groups.
Tables 10 and D-2 also present summary statistics on the number of days a respirator was
used while performing the activity, number of weeks spent performing activity in last year, and
number of years spent performing activity over career. Dust masks and/or respirators were used
less than half of the time the activities were performed. Generally, respirator usage was
proportionally greatest for paint removal and large structure removal.
The number of weeks and number of years spent conducting each target activity reflects
the number of days spent performing the activity in the last month. For example, large structure
removal comprises the second largest activity in terms of days performed, weeks performed, and
years performed. It is important to note that workers currently conducting target activities have
done so historically as well.
3.4.3 Work Practices
Summary statistics on work practices are presented for each worker group in Table 11,
and over all workers in Figure 4. Overall, 60% of the workers reported that they did not use a
dust mask or respirator, 30% reported using a dust mask, and 10% reported using a respirator.
Among the eight worker groups, respirator usage was highest for painters and laborers.
Relatively few of the workers had received formal training on the proper conduct for R&R in
lead-contaminated environments (13%) or had received any educational materials on lead
exposures (33%). Of the 293 workers that had performed paint removal during the past 30 days,
3-17
-------
Table 10. Summary of Responses for Questions Pertaining to R&R Target Activities
Variable
Description
Days performing the
activity in last month
Days performing the
activity in Pre-1 950
housing in last month
Days using a respirator
while performing activity
Number of Weeks
spent performing activity
in last year
Number of years spent
performing activity over
career
Statistic
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
Less than 1 Week
1 to 8 Weeks
More than 8 Weeks
25th Percentile
Mean
75th Percentile
Target Activities
General
R&R
10w
17.07(b)
24
0
10.74
20
(c)
(c)
(c)
4.1%(d)
20.9%
75.0%
6
13.28
20
Carpet
Removal
0
2.31
2
0
1.40
1
0
0.39
0
64.4%
28.7%
6.8%
1
5.15
8
Window
Replacement
0
3.84
5
0
2.98
3
0
0.77
0
44.6%
44.8%
10.6%
1
5.94
10
HVAC
Work
0
1.24
0
0
0.93
0
0
0.44
0
82.6%
14.4%
3.1%
0
2.49
2
Large
Structure
0
6.78
10
0
4.38
5
0
1.41
0
30.4%
44.4%
25.1%
2
7.10
11
Paint
Removal
0
5.54
10
0
3.51
5
0
1.58
0
47.0%
35.4%
17.6%
0
6.28
10
Cleanup
2
10.92
20
0
6.63
10
0
1.70
0
22.6%
41 .4%
36.1%
3
10.51
15
co
i—k
00
(a) 25% of all workers responded that they did R&R 10 days or less over the past 30 days.
(b) Mean number of days spent doing R&R in the past 30 days was 17.07 over all workers.
(c) Did not ask this question for general R&R activity.
(d) 4.1 % of workers spent less than one week doing general R&R in the last year.
-------
Table 11. Summary of Responses for Questions Pertaining to Worker Practices
Variable
Description
Sample Size
Smoke
Respirator
Other
Occupational
Exposures
Hobbies with Pb
Exposure
Received Pb
Training
Received
Educational
Material
Hours of Cleanup
Variable
Category
Don't
Smoke, not on
job
Smoke on job
Don't Use
Use Dust-mask
Use Respirator
Yes
Yes
Yes
Yes
< 1/2 Hrs/Day
1/2-1 Hrs/Day
1 -4 Hrs/Day
>4 Hrs/Day
Worker Group
Union
Carpenter
159
64.8%(a)
5.0%
30.2%
57.2%
38.4%
4.4%
30.2%
45.9%
16.2%
38.4%
30.8%
34.6%
26.9%
7.7%
Non-Union
Carpenter
105
37.1%
7.6%
55.2%
65.7%
21 .9%
12.4%
36.2%
69.5%
16.2%
27.6%
4.4%
40.6%
49.3%
5.8%
Dry wall
Worker
64
54.7%
6.3%
39.1%
67.2%
31 .3%
1.6%
31 .3%
59.4%
6.3%
26.6%
3.8%
28.3%
37.7%
30.2%
Floor
Layer
82
48.8%
9.8%
41 .5%
70.4%(b)
22.2%
7.4%
8.5%
52.4%
8.5%
26.8%
16.7%
50.0%
30.4%
2.9%
Laborer
56
37.5%
8.9%
53.6%
39.3%
37.5%
23.2%
35.7%
75.0%
16.1%
41.1%
18.2%
54.6%
18.2%
9.1%
Painter
34
35.3%
8.8%
55.9%
32.4%
47.1%
20.6%
29.4%
70.6%
20.6%
18.2%
12.1%
33.3%
45.5%
9.1%
Supervisor
57
66.7%
0%
33.3%
68.4%
26.3%
5.3%
19.3%
50.9%
22.8%
50.9%
22.5%
57.5%
20.0%
0.0%
Window
Installer
14
57.1%
14.3%
28.6%
78.6%
7.1%
14.3%
21 .4%
71 .4%
14.3%
21 .4%
24.6%
51 .8%
19.3%
4.4%
Other
14
57.1%
7.1%
35.7%
42.9%
21 .4%
35.7%
35.7%
64.3%
21 .4%
35.7
38.5%
30.8%
23.1%
7.7%
Total
585
52.0%
6.7%
41 .4%
59.8%
30.5%
9.8%
27.7%
58.3%
13.2%
33.4%
17.7%
44.2%
30.6%
7.6%
co
H^
\o
(a) 64.8% of the 159 union carpenters responded that they do not smoke.
(b) Response was missing for one of the 82 workers in the floor layer category, and therefore, total is only 81.
-------
Table 11. Summary of Responses for Questions Pertaining to Worker Practices (continued)
Variable
Description
Sample Size
Paint
Removal*0'
Cleanup10'
Variable
Category
Dry power-sanding
Dry hand-sanding
Dry scraping
Burning, torching, or a
heat gun
Wet-scraping
Wet-sanding
Chemical stripping
Use dust collector
when sanding
Total number of
people who performed
paint removal
Broom
Vacuum
HEPA vacuum
Wet mop with TSP
Clean power tools
using any method
Clean tools using
Compressed Air
Total number of
people who performed
cleanup
Job Category
Union
Carpenter
159
35.6%
66.7%
66.2%
15.6%
2.2%
4.4%
13.3%
11.1%
45
100%
46.8%
0.9%
1.8%
26.1%
15.3%
111
Non-Union
Carpenter
105
53.1%
72.8%
80.2%
19.8%
8.6%
8.6%
25.9%
16.0%
81
98.0%
69.6%
2.0%
1 1 .8%
42.2%
26.5%
102
Dry wall
Worker
64
59.1%
86.4%
63.6%
22.7%
22.7%
13.6%
40.9%
22.7%
22
98.1%
38.5%
3.9%
7.7%
32.7%
7.7%
52
Floor
Layer
82
44.7%
42.1%
89.5%
26.3%
2.6%
0%
10.5%
21.1%
38
100.0%
43.5%
1.4%
4.3%
40.6%
14.5%
69
Laborer
56
54.1%
75.7%
86.5%
37.8%
43.2%
32.4%
29.7%
40.5%
37
98.1%
55.8%
19.2%
25.0%
55.8%
40.4%
52
Painter
34
51 .6%
87.1%
74.2%
41 .9%
6.52%
12.9%
29.0%
22.6%
31
97.0%
48.5%
6.1%
18.2%
30.3%
15.2%
33
Supervisor
57
52.6%
89.5%
47.4%
5.3%
0%
0%
5.3%
21.1%
19
100.0%
41 .0%
0%
2.6%
30.8%
12.8%
39
Window
14
22.2%
77.8%
88.9%
22.2%
11.1%
0%
22.2%
22.2%
9
92.3%
69.2%
7.7%
15.4%
69.2%
538.5%
13
Other
14
45.5%
72.7%
81 .8%
27.3%
27.3%
18.2%
27.3%
27.3%
11
100.0%
81 .8%
45.5%
18.2%
72.7%
63.6%
11
Total
585
48.5%
72.0%
75.8%
24.2%
12.3%
10.2%
22.5%
21 .2%
293
98.8%
52.3%
5.0%
9.3%
38.4%
21 .0%
482
ro
O
Workers may have checked one or more responses for this question.
-------
100
80 --
o>
a 60 --
CD
ffi 40 +•
Q_
20 --
0
100
30.5
9.8
None Dustmask Respirator
Protection Used
Dry Wet Chemical Heat
Paint Removal Method Used
-------
3.5 DESCRIPTIVE STATISTICS OF WORKER BLOOD-LEAD CONCENTRATIONS
Blood samples were obtained from 581 workers. The lead concentration of each sample
was measured in at least two separate chemical analyses. The geometric mean blood-lead
concentration (in units of |ig/dL) among the multiple chemical injection measurements was used
to represent the blood-lead concentration of each worker. Table C-4 provides descriptive
statistics for the worker blood-lead concentrations.
The distribution of blood-lead concentrations for this sample was skewed; and therefore, a
natural log transformation was employed to improve the assumption of normality. Normal and
log-normal probability plots are displayed in Figures F-4 and F-5, respectively. Figure 5 shows a
histogram of the blood-lead concentrations for the sampled workers. Worker blood-lead
concentrations ranged from (below the detection limit of) 1 to 55 |ig/dL, with a geometric mean
of 4.5 |ig/dL. Less than 10% of the workers (51 of 581) had blood-lead concentrations greater
than 10 |ig/dL, less than 1.5% had blood-lead concentrations greater than 25 |ig/dL, and only one
worker had a blood-lead concentration greater than 40 |ig/dL.
18
15
e
0)
D_
0
X
\
\
\
\
2 4 8 16
Blood — lead Concentration
32
64
Figure 5. Histogram of Blood-Lead Concentration (Semi-Logarithmic Scale).
3-22
-------
Table C-4 also provides descriptive statistics for the blood-lead concentrations for each
sampling frame. Estimated 95% confidence intervals for the geometric mean of blood-lead
concentrations for each sampling frame are provided in Table 12. Worker blood-lead
concentrations were lower in St Louis (in comparison to Philadelphia, p-value <0.0001), and were
less for union workers (in comparison to non-union workers, p-value <0.0001). The variability
between worker blood-lead concentrations (measured by the log standard deviation) was similar
among the sampling frames. Figure F-6 displays side-by-side boxplots of blood-lead
concentrations for each sampling frame.
Each worker was assigned to a worker group based on his current job title and job
activities. Table C-5 provides descriptive statistics and Figure F-7 displays side-by-side boxplots
of blood-lead concentrations for each worker group. Estimated 95% confidence intervals for the
geometric mean blood-lead concentration for each worker group are presented in Table 13.
Table 12. 95% Confidence Intervals for Geometric Mean of Blood-Lead
Concentrations for Each Sampling Frame
Sampling Frame
Philadelphia Union
Philadelphia Non-Union
St Louis Union
St Louis Non-Union
All Four Groups Combined
N
197
74
150
160
581
Geometric Mean
(MQ/dL)
5.1
5.6
3.0
4.9
4.5
log Std. Error
log (|jg/dL)
0.044
0.072
0.050
0.049
0.027
95% Confidence
Interval (jjg/dL)
(4.6, 5.6)
(4.9, 6.5)
(2.7, 3.3)
(4.5, 5.4)
(4.2, 4.7)
Table 13. 95% Confidence Intervals for Geometric Mean of Blood-Lead
Concentrations for Each Worker Group
Worker Group
Union Carpenter
Non-Union Carpenter
Drywall Worker
Floor Layer
Laborer
Painter
Supervisor
Window Installer
Other
N
159
104
64
81
54
34
57
14
14
Geometric
Mean
4.4
5.0
5.8
2.6
4.9
7.2
3.8
5.4
5.3
95 % Confidence
Interval
(4.0 , 4.8)
(4.5 , 5.7)
(5.0 , 6.8)
(2.3 , 3.0)
(4.1 , 5.7)
(5.8 , 8.8)
(3.2 , 4.4)
(3.9 , 7.4)
(3.9 , 7.3)
3-23
-------
3.6 STATISTICAL MOEELIN6 RESULTS
3.6.1 Statistical Model Building
Statistical models were used to investigate the relationship between log-transformed
worker blood-lead concentrations and potential lead exposure associated with specific R&R
activities. Questionnaire responses were used to construct variables that represent potential lead
exposure resulting from the R&R target activities. The lead exposure variables were constructed
for carpet removal, window replacement, HVAC work, large structure removal, paint removal,
cleanup, and general R&R work. There were three different periods of exposure captured by the
questionnaire for each target activity: short-term exposure, mid-term exposure, and long-term
exposure.
The number of days an activity was performed in pre-1950 buildings was selected as the
measure of short-term exposure in the statistical models. Mid-term exposure for each target
activity was characterized by the ordinal response number for the number of weeks a worker
performed the activity over the past year. Long-term exposure for each target activity was
characterized by the number of years a worker has performed the activity over his career. Thus, a
total of 21 variables, one for each combination of target activity and exposure period, were used
to characterize the potential lead exposure resulting from R&R.
3.6.1.1 Selection of Ancillary Covariates
Questionnaire information not related to worker group or the conduct of R&R may also
have a significant effect on worker blood-lead concentrations. These ancillary variables were
divided into two categories:
b Variables related to the conduct of R&R
— Respirator use
— General work practices
— R&R activity in own home
— Previous lead training and education
p Variables not related to the conduct of R&R
— Other occupations with potential lead exposure
— Age of home
— Hobbies with potential lead exposure
— Race and ethnicity of worker
— Age
— Smoking status
— Education level
The large number of candidate variables required an initial screening to select ancillary covariates
for use in the statistical models.
3-24
-------
Each potential ancillary covariate was classified as being either discrete or continuous.
Scatterplots and regression lines were used to characterize the relationships between continuous
potential covariates and worker blood-lead concentrations. For each discrete covariate, boxplots
of worker blood-lead concentrations (Figures F-8 to F-13) were generated for each level of the
response. These boxplots helped identify discrete variables that had a significant relationship with
blood-lead, and also provided insight for collapsing the levels of some of the discrete variables.
Collapsing the response levels of the discrete variables may provide more power for statistical
tests. For example, Question A4 related to worker education level had the following response
levels:
(a) Grades 1 through 8
(b) Some High School
(c) High School Graduate
(d) Apprenticeship Training
(e) Some College/Tech School Graduate
(f) College/Tech School
Preliminary boxplots suggested a fundamental division between high school graduates and
those who did not finish high school. Therefore, response levels (a) and (b) were collapsed into
one category, and response levels (c) through (f) were collapsed into another category (See
Figure F-12).
In addition, an effort was made to reduce the number of potential covariates, when
possible, by creating a single quantal variable from the responses of several questions. For
example, two different variables were constructed to reflect the number of other occupations or
hobbies a worker had with high potential for lead exposure.
The statistical significance of each candidate covariate on blood-lead concentration is
displayed in Table E-l of Appendix E. Covariates with significant relationships with blood lead
were then simultaneously placed into a single model with blood lead. Based on this analysis, the
following variables were selected for inclusion in covariate adjusted models:
1. Race
2. Education
3. R&R in Own Home
4. Age of Home
5. Smoking Status
6. Respirator Use
Table 14 provides the geometric mean blood-lead concentration and log(standard deviation) for
each level of the selected ancillary covariates.
3-25
-------
Table 14.
Geometric Mean Blood-lead Concentration and Log (Standard
Deviation) for Each Level of the Ancillary Covariates
Variable
Race
Education
Smoking Status
Age of Home
Performed R&R at Home
Used Respirator or
Dustmask
Level
White
Black
Other
Less than HS
Finished HS
Non Smoker
Smokes, not on Job
Smokes on Job
Post 1978
1978 - 1950
Pre-1950
No
Yes
No
Yes
Sample
Size
488
83
10
36
545
303
39
239
113
232
236
366
215
347
233
Geometric Mean
Blood-lead (j/g/dL)
Mean
4.26
5.81
4.73
6.23
4.37
3.97
4.64
5.14
3.48
4.38
5.15
4.27
4.81
4.12
5.04
Std Dev
0.029
0.071
0.206
0.109
0.028
0.037
0.103
0.042
0.060
0.042
0.042
0.034
0.045
0.035
0.043
3.6.1.2 Selection of the Form of the Model
The paradigm for fitting statistical models presented in Figure 2 was used to investigate
the relationship between R&R target activities and worker blood-lead concentrations. Prior to
investigating the data, there was no compelling basis for selecting the form of the functional
relationship between blood-lead concentrations and measures of potential lead exposure
associated with target activities. Therefore, exploratory methods were used to assess these
relationships. This involved plotting the data and using robust locally weighted regression. If a
functional relationship exists between blood lead and a particular exposure measure, be it linear,
quadratic or exponential, the robust locally weighted regression procedure will yield a curve
which approximates the functional relationship. Results of the exploratory analyses demonstrated
that the relationship between log of blood lead and measures of potential lead exposure associated
with each combination of target activity and exposure period could be adequately described by a
linear model.
3.6.2 Comparisons Between Blood-Lead Concentrations Among Worker Groups
Geometric mean blood-lead concentration was also estimated for each worker group after
adjusting for ancillary covariates. The geometric means are presented in Table 15. The covariate
adjusted geometric means utilized a linear combination of the levels of each covariate with
weights equal to the percent of workers at each level. The covariate adjusted geometric means
3-26
-------
are similar to those unadjusted for covariates for all worker groups except painters. Figure
graphs the 95% confidence intervals for the covariate adjusted geometric means.
Table 15.
95% Confidence Intervals for Geometric Mean of Blood-Lead
Concentrations for Each Worker Group Based on Covariate
Adjusted Model
Worker Group
Union Carpenter
Non-Union Carpenter
Drywall Workers
Floor Layer
Laborer
Painter
Supervisor
Window Installer
Other
Geometric Mean
4.5
4.8
6.1
2.8
4.1
5.9
4.1
5.8
4.9
95% Confidence Interval
( 4.1 , 5.0)
(4.3 , 5.4)
(5.3 , 7.1)
(2.5 , 3.2)
(3.5 , 4.9)
(4.8 , 7.3)
(3.5,4.8)
(4.3 , 7.9)
(3.6 , 6.7)
3-27
-------
nniiuall UUmirar
ufywaii wofKer
Painter
rtuiliei
VVHHJUW HlSliHJKM
INLHI unimi oai|jBiiiBi
UIIHJII OOI|JBillBf
1 ahrmar
LjeUjufoT
oupennsor
riuof Ljayof
6.1
n_ei r A 1
n-04 [ fl J
53
n_ii r A i
n-34 1 f J
5*
n 11 T Ji 1
n 14[ fl J
44
n_inr. T A 1
4,5
n im r A i
n U»[ R J
4,1
„_« r A i
n-5B i ^ j
4,1
n-rrr T A 1
n— a/ L ^f j
as
n_B-I T A 1
n-82 I ^ J
Blood-Lead Concentration (pg/dL)
Figure 6. 95% Confidence Intervals for Geometric Mean of Blood-Lead Concentrations
for Each Worker Group Based on Covariate Adjusted Model
3.6.3 Relationships Between R&R Target Activities and Worker Blood-Lead
Concentrations
The effect of each of the 21 target activity variables, unadjusted for covariates or the
conduct of other target activities, was estimated by fitting separate linear regression models.
Estimated slopes, standard errors, and p-values for each exposure measure are displayed in Table
E-2. With the exception of carpet removal, there was a statistically significant positive
relationship between worker blood-lead concentration and short-term conduct in pre-1950 houses
for each target activity. The relationships between worker blood-lead concentration and mid-term
and long-term exposure associated with target activities were also generally positive.
Many of the slope estimates presented in Table E-2 were positive and statistically
significant (p-value < 0.05). For each target activity, Table 16 presents the predicted increase in
blood-lead concentration for 10 days of work activity in pre-1950 buildings. (Mean number of
days of work activity in pre-1950 buildings, as shown in Table 10, was approximately 10 days.)
The second column displays predicted blood-lead concentration for each target activity, for
workers who performed the activity for zero and ten days in pre-1950 houses during the previous
3-28
-------
month. The predicted increases in blood-lead concentration ranged from 0.5 |ig/dL for carpet
removal to 1.2 |ig/dL for HVAC work. The maximum predicted blood-lead concentration
associated with 10 days of work in pre-1950 buildings, 5.6 |ig/dL, was estimated for HVAC
work.
The covariate adjusted effects for each combination of target activity and exposure period
were estimated using separate linear regression models. Estimated slopes, standard errors, and p-
values are shown in Table E-3. The estimated slopes for target activities were generally positive
after adjusting for the effects of selected ancillary covariates. As expected, the magnitudes of the
covariate adjusted slopes estimates were generally less than the unadjusted slope estimates.
The predicted increase in blood-lead concentrations, based on the covariate adjusted
models, associated with 10 days of work in pre-1950 buildings are presented in the third column
of Table 16. The smaller slope estimates for the covariate adjusted models compared to the
unadjusted models generally resulted in smaller predicted increases in worker blood-lead
concentrations. For example, the predicted increase in blood-lead concentrations associated with
10 days of paint removal was 0.5 |ig/dL for the covariate adjusted model compared to 1.1 |ig/dL
for the unadjusted model. For all activities, the estimated increase in blood-lead concentration
from the base level was very small (less than 1 |ig/dL) for the covariate adjusted model.
Covariate adjusted models that included all three exposure periods (short-, mid-, and long-
term) were also fitted for each target activity. This series of models was used to assess which
exposure period, if any, was the best predictor of blood-lead concentrations. Estimated slopes,
standard errors, and p-values are displayed in Table E-4. Parameter estimates in Table E-4 are
not consistent across the seven target activities. Short-term exposure (days pre-1950) appeared
to dominate for paint removal and cleanup, mid-term exposure (weeks in last year) for window
replacement and carpet removal, and long-term exposure (years over career) for large structure
removal. For the conduct of R&R work in general, all three exposure periods were significantly
associated with blood-lead concentrations. There were no significant relationships
Table 16. Predicted Increase in Blood-Lead Concentration Associated with 10
days of Work in Pre-1950 Buildings
Target Activity
Carpet Removal
Window
Replacement
Paint Removal
HVAC Work
Large Structure
Removal
Based on Model Unadjusted for
Covariates
Level When Worker Conducts
Base . an Additional 10 Days per
Level r Month of Activity
4.4 \> 4.9
4.2 \> 5.1*
4.1 \> 5.2*
4.4 \> 5.6*
4.2 \> 4.8*
Based on Covariate Adjusted Model
Level When Worker Conducts
Base . an Additional 10 Days per
Level P Month of Activity
4.5 p 4.1
4.4 \> 4.8
4.3 p4.8*
4.4 \> 4.7
4.3 \> 4.7
3-29
-------
Cleanup 4.1 p 4.7* 4.3 p 4.6*
| general R&R || 3.7 p 4.4* |[ 3.9 p 4.4*
' Slope parameter estimate for days per month of activity was significant at the alpha = 0.05 level.
found at all for HVAC work. The inconsistent results of these models may be attributed to the
high degree of correlation among the three exposure period measures for each target activity.
3.6.4 Overall Statistical Model for Worker Blood-Lead Concentrations
The right branch of Figure 2 presented the paradigm for developing an overall statistical
model. An initial series of models were fitted to the data to assess which of the specific target
activities, if any, had the largest impact on blood-lead concentration during each exposure period.
For each period of exposure, the effects of all of the target activities (with the exception of
general R&R work) were estimated simultaneously using covariate adjusted models. Estimated
slopes, standard errors, and p-values are presented in Table E-5.
When all of the target activities were considered simultaneously, only one or two of them,
within each exposure period, had a significant impact on worker blood-lead concentrations. This
is a result of the strong inter-correlations among the target activity variables.
In Section 3.3, it was observed that target activities varied across the worker groups and
in Section 3.6.2 that there were differences in the geometric mean blood-lead concentrations
among the nine worker groups. The next series of models assessed which of the specific target
activities, if any, have a statistically significant impact on worker blood-lead concentrations after
adjusting for the effects of ancillary covariates and worker group. Estimated slopes, standard
errors, and p-values are displayed in Table E-6. Although mid-term window replacement had a
significant impact on worker blood-lead concentrations, the major result was that virtually no
statistically significant relationships were found between target activities and blood lead after
adjusting for the effects of the ancillary covariates and worker group.
This result is illustrated in Table 17, which provides general F-tests for the effects of all
target activities combined on worker blood-lead concentrations after adjusting for the effects of
selected covariates and worker group. The first row shows the R2 value (12.4%), F-test (8.97)
and corresponding p-value (< 0.001) for a model which included only the selected ancillary
covariates. The next three rows correspond to models which, for each exposure period, included
the ancillary covariates and exposure measures for the six target activities. The combined effect
of all six target activities was statistically significant for each exposure period, and accounted for
between 1.9 and 5.5% of the variability in worker blood-lead concentrations after adjusting for
the covariates. The next row displays the results for a model that included the effects of the
selected ancillary covariates and worker group. The effect of worker group was statistically
significant and explained 12.4% of the variability in blood lead after adjusting for the covariates.
The last three rows describe, for each exposure period, the combined effect of all six target
activities after adjusting for the effects of ancillary covariates and worker group. The F-test for
3-30
-------
the combined exposure measures showed that the combined effects of target activities did not
have a significant impact on worker blood-lead concentrations after adjusting for covariates and
worker group.
Based on the previously fitted models, the following information was used to develop a
final predictive model:
1. Worker group appears to be the most predictive measure of worker blood-lead
concentration. Table 17 showed that the partial R2 for the effect of worker group
was 0.124 after adjusting for the effects of the ancillary covariates.
2. The conduct of general R&R work for all three exposure periods was statistically
significant when fitted simultaneously in a covariate adjusted model.
3. The combined effect of the six target activities (carpet removal, window
replacement, HVAC work, large structure removal, paint removal, and cleanup)
did not have a significant impact on worker blood-lead concentration after
adjusting for the effects of the ancillary covariates and worker group.
Final Model
The final predictive model fitted to worker blood-lead concentrations was
log(blood lead) = Worker Group + Covariates + (General R&R * Exposure Period)
Estimated parameters, standard errors, and p-values are presented in Table E-7.
This model included an intercept for each worker group and a slope for the effect of
performing general R&R in each exposure period. The estimated worker-group intercepts
represent the baseline blood-lead concentration for each worker group. Baseline blood-lead
concentrations were highest for painters, drywall workers, and window installers, and were lowest
for supervisors and floor layers.
3-31
-------
Table 17. General F-Tests for the Combined Effects of All Target Activities on
Worker Blood-Lead Concentrations, After Adjusting for the Effects
of Covariates and Worker Group
Base Model
Covariates
Covariates
Covariates +
Worker Group
Variables Under Investigation
Covariates
Target Activities (Pre-1 950)
Target Activities (Weeks)
Target Activities (Years)
Worker Group
Target Activities (Pre-1 950)
Target Activities (Weeks)
Target Activities (Years)
R2
0.124
0.145
0.172
0.141
0.233
0.235
0.246
0.235
p2
RP
0.024
0.055
0.019
0.124
0.003
0.017
0.003
F
8.97
2.48
5.42
2.67
10.00
0.94
1.56
1.19
P-Value
<0.001
0.022
<0.001
0.015
<0.001
0.466
0.157
0.310
R2 = The coefficient of determination for the base model and variables under investigation.
R2P = The partial coefficient of determination for the variables under investigation, after adjusting for the effects of variables indicated in
the base model.
F-test pertains only to the variables under investigation, after adjusting for variables included in the base model.
The estimated slopes for the conduct of general R&R were positive and statistically
significant for each exposure period. Since estimated slopes are the same for all worker groups,
differences between worker groups are captured by the estimated intercepts. Thus, the
interpretation of this final model is that there are differences in the blood-lead concentrations of
different worker groups, and that the amount of general R&R work conducted has the same effect
for all workers, regardless of worker group.
The linear relationships between worker blood-lead concentration and conduct of general
R&R within each worker group are displayed graphically for short-term exposure, mid-term
exposure, and long term exposure in Figures F-14 through F-16, respectively. The fitted line
displayed in each graph is based on the common slope estimated for the exposure period. Based
on these graphs, it appeared that the assumption of a common slope across the worker groups for
the conduct of general R&R in each exposure period was justified.
To illustrate how blood-lead concentration is a function of the amount of general R&R
work performed, Table 18 displays the estimated blood-lead concentration (and 95% confidence
interval) associated with low, medium, and high exposure indices based on the final model. The
low, medium, and high exposure indices were based on the 25th, 50th, and 75th percentiles of
questionnaire responses for short-, mid-, and long-term conduct of general R&R. Although the
model predicts a 60% increase in blood-lead concentration between the low and high exposure
indices, the actual differences in predicted blood-lead concentrations ranged from 1.5 |ig/dL for
floor layers to 3.0 |ig/dL for painters.
3-32
-------
Table 18. Predicted Worker Blood-Lead Concentrations Associated with Low, Medium,
and High Exposure Indices for Each Worker Group
Worker Group
Union Carpenter
Non Union
Carpenter
Dry wall Worker
Floor Layer
Laborer
Other
Painter
Supervisor
Window Installer
Low Exposure Index
0 Days Pre-1950
1-4 Weeks in Last Year
5 Years over Career
Geometric
Mean
4.0
3.9
5.0
2.5
3.5
4.2
5.1
3.4
4.4
95% C.I.
(3.5 , 4.5)
(3.3 , 4.5)
(4.3 , 5.9)
(2.2 , 2.9)
(2.9 , 4.2)
(3.1 , 5.8)
(4.1 ,6.3)
(2.9,4.1)
(3.2,6.1)
Medium Exposure Index
10 Days Pre-1950
9-26 Weeks in Last Year
10 Years over Career
Geometric
Mean
4.7
4.6
6.0
3.0
4.2
5.0
6.0
4.1
5.2
95% C.I.
(4.3 , 5.2)
(4.1 , 5.2)
(5.2 , 6.9)
(2.6 , 3.4)
(3.5 , 4.9)
(3.7 , 6.8)
(4.9 , 7.4)
(3.5 , 4.8)
(3.8,7.1)
High Exposure Index
25 Days Pre-1 950 >26
Weeks in Last Year
25 Years over Career
Geometric
Mean
6.4
6.2
8.1
4.0
5.6
6.8
8.1
5.5
7.0
95% C.I.
(5.5 , 7.3)
(5.3,7.1)
(6.8 , 9.6)
(3.4 , 4.8)
(4.6 , 6.8)
(4.9 , 9.3)
(6.4, 10.2)
(4.6 , 6.6)
(5.1 , 9.6)
Tukey's multiple comparison procedure for unbalanced data was employed to conduct
pairwise comparisons between geometric mean blood-lead concentrations predicted for each
worker group, based on the covariate-adjusted model. The nine worker groups were separated
into three groups, as shown below.
Non-Union Union Window Drywall
Floor Layer Supervisor Laborer Carpenter Carpenter Installer Painter Worker
3.0 ug/dL 4.1 ug/dL 4.2 ug/dL 4.6 ug/dL 4.7 ug/dL 5.2 ug/dL 6.0 ug/dL 6.0 ug/dL
ppppppppppppppppppppppppppppp
ppppppppppppppppppppppppppppp
pppppppppppppppppppppppppppp
Mean blood-lead concentrations for floor layers were statistically less than those for the other
worker groups. Mean blood-lead concentrations for painters and drywall workers were greater
than those for supervisors.
3-33
-------
4.0 SUMMARY AND OVERALL CONCLUSIONS
The WCBS surveyed two groups of R&R workers (union carpenters and employees of
independent contractors) in two cities (Philadelphia, Pennsylvania, and St. Louis, Missouri). A
total of 585 questionnaires and 581 blood samples were collected from R&R workers. The
results were utilized to address three objectives:
1. Determine the relationship between blood-lead concentrations and work practices
or target activities performed by R&R workers after controlling for other factors
that may affect worker blood-lead concentrations.
2. Determine if blood-lead concentrations of R&R workers in specific worker groups
differ after adjusting for other factors that may affect worker blood-lead
concentrations.
3. Gather information on the types of work activities and work practices in which
R&R workers engage.
Overall results for each of these objectives, presented in reverse order, are summarized below.
Questionnaires were collected from a total of 585 workers in the WCBS. The sample of
R&R workers consisted primarily of white males between the ages of 30 and 45. The
questionnaire captured data on how often each worker conducted specific target activities in any
home, as well as in pre-1950 homes, during the past 30 days. The questionnaire results indicated
that the sampled workers spent an average of 17 days during the past month on general
renovation and remodeling. The workers spent on average of 11 of these 17 days in pre-1950
homes. The questionnaire results indicated that:
1. The workers performed a wide variety of R&R activities, and spent considerable
time doing large structure removal and paint removal or surface preparation,
activities with potential for creating high dust-lead exposures.
2. The R&R workers were evenly divided between those that worked in residential
and nonresidential buildings.
3. 90% of the workers did not use a respirator.
4. 88% of the workers did not use cleanup methods recommended for use in a lead
contaminated environment and 98% reported using dry sweeping methods.
5. 66% of the workers had not received any materials on lead hazards and 86% had
received no lead exposure training.
4-1
-------
Blood samples were successfully collected from 581 of the 585 workers. Worker blood-
lead concentrations were generally low: 9.1% were above 10 |ig/dL, 3.8% were above 15 |ig/dL,
1.2% were above 25|ig/dL, and only one worker had a blood-lead concentration greater than 40
|ig/dL. The blood-lead results indicated that
6. The distribution of blood-lead concentrations was approximately log-normal with a
geometric mean of 4.5 |ig/dL and a log standard deviation of 0.659 log(|ig/dL).
7. Blood-lead concentrations were significantly different between the sampling
frames. This was mostly attributed to lower blood-lead concentrations for the
union carpenters in St. Louis.
8. Although geometric mean blood-lead concentrations were low for all of the
worker groups, there were significant differences among the worker groups that
differentiate the groups with the high mean blood-lead concentrations from those
with the low mean blood-lead concentrations. Drywall workers (6.1 |ig/dL),
painters (5.9 |ig/dL), and window installers (5.8 |ig/dL) had the highest blood-lead
concentrations and floor layers (2.8 |ig/dL) had the lowest.
9. Several of the ancillary variables were significantly related to worker blood-lead
concentrations. Of all the factors investigated, race, education level, smoking
status, age of worker's home, recent R&R work in worker's home, and respirator
usage were determined to be significantly related to worker blood-lead
concentrations. In general, the estimated effect of each factor was anticipated:
smokers, Blacks, non-high school graduates, workers residing in older homes, and
workers residing in homes that recently underwent R&R had higher blood-lead
concentrations. However, the observed affect associated with respirator usage
was opposite of what was anticipated with an increase in mean blood-lead
concentration associated with use of a respirator. One possible explanation for this
may be that workers who report they wear respirators are much more likely to be
exposed to lead-based paint than those who report otherwise.
A series of statistical models were used to investigate the relationship between blood-lead
concentration and the conduct of specific R&R activities. Initially, separate models were fitted to
the data for each target activity. Based on the separate models fitted to each target activity that
adjusted for the ancillary covariates:
10. The number of days worked in pre-1950 buildings in the past month was
significantly related to increases in blood-lead concentrations for general R&R
work, paint removal, and cleanup. However, the estimated increase in predicted
blood-lead concentration associated with performing any of these activities in pre-
1950 buildings for 10 days per month was very small (less than l|ig/dL) for all
activities.
4-2
-------
11. The number of weeks worked in the past year was significantly related to increases
in blood-lead concentrations for general R&R work, window replacement, and
large structure removal.
12. The number of years worked was significantly related to increases in blood-lead
concentrations for general R&R work, window replacement, and large structure
removal.
13. Predicted increases in blood-lead concentrations, although sometimes statistically
significant, were less than 1 |ig/dL for each additional 10 days of work conducted.
A final model was developed that included effects for ancillary covariates, worker group,
and conduct of general R&R activity for each of three exposure periods: short-term, mid-term,
and long-term. Results of this model indicated that:
14. Much of the statistical association between specific target activities and blood-lead
concentrations was also captured by the effects of worker group. In fact, worker
group was the most predictive measure of blood-lead concentration.
15. Worker blood-lead concentrations were predicted to be highest for painters,
drywall workers, and window installers, and were predicted to be the lowest for
floor layers and supervisors.
16. Conduct of general R&R work was significantly related to increases in worker
blood-lead concentrations for all three exposure periods.
17. An empirically based index of low potential for lead exposure resulting from
conduct of R&R work was estimated to be 0 days worked in pre-1950 buildings in
the past month, 1 to 4 weeks worked in the past year, and 5 years of R&R
experience. Similarly, an empirically based index of high potential for lead
exposure was estimated to be 25 days worked in pre-1950 buildings in the past
month, more than 26 weeks worked in the past year, and over 25 years of R&R
experience. The increase in worker blood-lead concentrations between a low and
high index of potential lead exposure was predicted to be 60%. Although
statistically significant, the maximum predicted blood-lead concentration for a high
potential for exposure was only 8.1 |ig/dL.
To place the results of the WCBS into perspective, the geometric mean blood-lead
concentrations of non-Hispanic White and Black workers aged 20-49 were compared to national
averages reported in NHANES III. The geometric mean of blood-lead concentrations for the
R&R workers were only slightly higher than those reported in NHANES III. Table 19 indicates
that the difference in geometric mean blood-lead concentration between R&R workers and the
general population was 0.4 |ig/dL for non-Hispanic Whites and 1.1 |ig/dL for non-Hispanic
Blacks.
4-3
-------
The results of this study indicated that although R&R workers may be exposed to high
levels of lead during the conduct of their work, there was little evidence of elevated blood-lead
concentrations among the workers. It is possible that there exist specialized groups of R&R
workers who may have higher lead exposures, for example, workers specializing in historic
renovations. However, the WCBS study included workers in cities with a documented lead
problem who were conducting a significant amount of work in older buildings. In fact, the study
was weighted toward highly exposed general R&R workers. Nevertheless, only seven workers
out of 581 had a blood-lead concentration greater than 25 |ig/dL, and only one worker out of 581
had a blood-lead concentration greater than 40 |ig/dL; workers with blood-lead concentrations
greater than 40 |ig/dL require increased medical surveillance under the interim OSHA rule for
lead in construction.
Table 19. Difference in Geometric Mean Blood-Lead Concentration Between WCBS
and NHANES III
Control
Non-Hispanic Whites
Non-Hispanic Blacks
WCBS
Geometric
Mean
0/9/dL)
4.2
5.6
95%
Confidence
Interval
(3.9 , 4.4)
(4.8 , 6.4)
NHANES III
Geometric
Mean
0/9/dL)
3.8
4.5
95%
Confidence
Interval
(3.6 , 4.1)
(4.2 , 4.8)
4-4
-------
APPENDIX A:
QUESTIONNAIRES
-------
RECORD 01
ID*-I I I I |
RECORD #: I ° 11J
01-04
05-06
Hello, my name is
WORKER EXPOSURE STUDY
TELEPHONE INTERVIEW SCREENER
and I'm calling for the Environmental Protection Agency
about a study of renovation and remodeling workers. You may have received a letter from (your
union/the National Association of Home Builders) in the past couple weeks explaining about the study.
I'd like a little of your time to ask a few questions about your work. This should take about five
minutes, and everything you say will be kept confidential. [IF R ASKS ABOUT CONFIDENTIALITY,
REFER TO ATTACHED SHEET DESCRIBING CONFIDENTIALITY PROCEDURES.]
1. Do you work in home or building construction or
renovation and remodeling for a living? That is, do you
earn money to support yourself doing this?
2. Do you do hands-on work? For example, do you do
painting; surface preparation such as sanding, scraping,
torching or floor refinishing; carpentry; demolition;
cleanup or installation? (This does not include
administrative or sales work.)
3. In the last 30 days, how many days did you spend
doing any kind of renovation and remodeling work?
Please include any home improvement or building
construction work.
4. How many years altogether have you earned your living
by working in the renovation and remodeling industry?
5. In the last 30 days, how many days did you spend
renovating or remodeling homes or buildings built
before 1950?
6. In the last 30 days, how many days have you worked in
residential buildings such as homes or apartments?
7. In the last 30 days, how many days have you worked in
non-residential buildings such as offices, schools or
government buildings?
YES (CONTINUE)
NO (THANK R AND STOP)
YES (CONTINUE)
NO (THANK R AND STOP)
# DAYS WORKED
07
08
09-10
NUMBER OF YEARS
DAYS IN BLDGS < 1950 . .
DAYS IN HOMES
DAYS IN NON-RES.
TJ-*cr*«n.v5
in. 06/17/94
11-12
13-14
15-16
17-18
-------
RECOR& 01
8A. In the last 30 days, how many days did you spend
removing carpets?
88. In the last 30 days, how many days did you spend
removing windows or door casements?
8C. In the last 30 days, how many days did you work to
maintain, repair or clean heating, ventilation or air
conditioning systems?
8D. In the last 30 days, how many days did you work
removing large structures such as making openings for
large windows or doorways, tearing down ceilings,
putting up walls or removing kitchen cabinets?
8E. In the last 30 days, how many days did you spend
removing paint or preparing surfaces?
8F. In the last 30 days, how many days did you spend
doing the hands-on dirty cleanup, where you cleaned up
the dirt, dust and debris caused by the renovation and
remodeling activities?
8G. In the last 30 days, about how many hours in a typical
day did you spend doing dirty cleanup work?
9. What is your current job title and what are your main activities at work?
• JOB TITLE:
MAIN ACTIVITIES:
'I—JLJ 21-22
LI I 23-24
DEDNT DO CLEANUP ........ 1
4 HOURS/DAY ............. 5
10.
11.
Now I'd like to ask you just a few more questions about
yourself. How old are you?
What is your race and ethnic group?
(PROBE FOR HISPANIC ORIGIN)
AGE
SPECIFY:
12.
RECORD GENDER WITHOUT ASKING, IF POSSIBLE.
IF NOT, ASK "Are you male or female?"
WfflTE, NOT HISPANIC 01
WHITE, HISPANIC 02
BLACK, NOT HISPANIC 03
BLACK, HISPANIC 04
AMERICAN INDIAN 05
ASIAN/PACIFIC ISLANDER ... 06
OTHER .. . (SPECIFY) 07
MALE 1
FEMALE 2
36-37
33-39
Thank you very much for taking the time to speak with me about your work. If you have any questions
or concerns, the study manager, Beth Moore, would be happy to speak with you. You can reach her
at (614) 424-4560.
40
END 01
/•v 06/17/9-i
-------
RECORD 01
J
ID#: U - U -1 I I I I 01-06
RECORD #: I ° 11 I 07-°8
DATE: LU - U_l - L_U 09-14
MM DD YY
WORKER EXPOSURE STUDY
-------
This questionnaire collects basic information about you, your work, and some
of your hobbies and activities. All information you give will be kept confidential. No
names will be used and no one outside the study will be able to tell which person
gave which answers. If you have any questions about this study, the Study Manager
will be happy to talk with you.
Please answer the questions as completely as you can. In some cases, you
will be asked to write in an answer. Other times, you will be asked to check a box
next to the answer that best applies to you.
If you have any questions about what a question means or how to answer a
question or section, the study staff will be happy to help you. Study staff on duty can
provide you with a copy of the letter you were mailed earlier, explaining about the
study and how we will use the information you provide.
Thank you for your help in this important research project.
-------
-1 -
SECTION A: DEMOGRAPHICS
A1. How old are you?
A2. What is your race and ethnic group?
SPECIFY:
A3. Are you male or female?
A4. How much schooling have you had? Check the
highest level completed.
RECORD 01
Vi
j
AGE
WHITE, NOT HISPANIC D 17-18
WHITE, HISPANIC D
BLACK, NOT HISPANIC D
BLACK. HISPANIC D
AMERICAN INDIAN D
ASIAN/PACIFIC ISLANDER D
OTHER ... (SPECIFY) D
MALE D
FEMALE D
GRADES 1 THROUGH 8 D 20
SOME HIGH SCHOOL D
HIGH SCHOOL GRADUATE/GED . D
APPRENTICESHIP TRAINING . .. D
SOME COLLEGE/TECH SCHOOL D
COLLEGE/TECH SCHOOL GRAD D
A5. How many children under age 6 currently live with you
in your home?
A6. Do you belong to the following trade organizations (not including
unions)?
the National Association of the Remodeling Industry (NARI)? ----
the National Association of Homebuilders? .................
any other trade organizations?
# CHILDREN UNDER 6
Yes No
D D
n D
(Please write name of other trade organization.)
OFFICE USE
(Please write name of other trade organization.)
OFFICE USE
TJ-SAQ.V7
rev. 09/07/94
21'22
23
24
25-26
27-23
-------
-2-
A7. Do you belong to any of the following unions?
the United Brotherhood of Carpenters?
the International Brotherhood of Electrical Workers?
the Laborers International Union of North America?
the International Brotherhood of Painters and Allied Trades?
any other union? (SPECIFY)
Yes
D
D
D
D
No
D
D
D
D
RECORD 01
j
29
30
31
32
33-34
OFFICE USE
The following sections ask about your work. By work, we mean any work you've done on the
job or on the side. That is, we are interested in work you've done for an employer and work
you do in your spare time for yourself or for family and friends.
SECTIONS: WORK HISTORY
B1. What is your current job title and what are your main activities at work?
JOB TITLE:
MAIN ACTIVITIES:
OFFICE USE
B2.
B3.
B4.
B5.
During the last 30 days, how many days did you spend
doing any kind of renovation and remodeling work?
(Please include any home improvement or building
construction.)
During the last 30 days, how many days did you spend
renovating or remodeling homes or buildings built
before 1950?
During the last 30 days, how many days did you work in
residential buildings (homes, apartments)?
During the last 30 days, how many days did you work in
non-residential buildings (offices, schools, government
buildings)?
days
days
days
days
Tj-SAQ.v7
t«v. 0*07/94
35-36
37-38
39-40
41-42
43-44
45-46
-------
B6.
-3-
During the last 12 months, how many weeks did you
spend doing renovation and remodeling?
B7.
How many years altogether have you done renovation
and remodeling work?
NONE D
LESS THAN 1 WEEK D
1 - 4 WEEKS D
5 - 8 WEEKS D
9 - 26 WEEKS
(ABOUT 3 - 6 MONTHS) D
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) U
years
RECORD 01
v •
J
47
43-49
TJ-SAQ.V7
rev. 09/07/94
-------
-4 -
SECTION C: CARPET REMOVAL
C1. In the last 30 days, how many days did you remove days
carpet?
If you have not removed any carpet in the last 30 days, go to question C4.
C2. In the last 30 days, how many days did you use a days
respirator while you removed carpet?
RECORD 01
C3. In the last 30 days, how many days did you remove
carpet from homes or buildings that were built before
1950?
C4. Altogether in the last 12 months, how many weeks did
you remove carpet?
C5. Think about all the years you've done renovation or
remodeling. How many of these years did you remove
carpet at least some of the time?
If you have never removed carpet, go to SECTION D.
C6. Think about all the years you've removed carpet. In an
average year, how many weeks did you spend
removing carpet?
days
NONE D 56
LESS THAN 1 WEEK D
1 - 4 WEEKS D
5 - 8 WEEKS D
9 - 26 WEEKS
(ABOUT 3 - 6 MONTHS) Q
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) Q
years 57-58
NONE D 59
LESS THAN 1 WEEK D
1 - 4 WEEKS O
5 - 8 WEEKS Q
9 - 26 WEEKS
(ABOUT 3 - 6 MONTHS) LJ
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) U
TJ-SAQ.v?
rev. 09/07/94
50-51
52-53
54-55
-------
-5-
SECTION D: WINDOW OR DOOR CASEMENT REPLACEMENT
D1. In the last 30 days, how many days did you remove
windows or door casements?
If you have not removed any windows or door casements in the last 30 days, go to question
D4.
RECORD 01
days
D2.
D3.
D4.
In the last 30 days, how many days did you use a
respirator while you removed windows or door
casements?
In the last 30 days, how many days did you remove
windows or door casements from homes or buildings
built before 1950?
Altogether in the past 12 months, how many weeks did
you remove windows or door casements?
days
days
NONE
LESS THAN 1 WEEK
1 - 4 WEEKS
5 - 8 WEEKS
D5.
Think about all the years you've done renovation or
remodeling. How many of these years did you remove
windows or door casements at least some of the time?
9 - 26 WEEKS
(ABOUT 3 - 6 MONTHS) .
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
years
.D
.0
.D
.D
.D
,D
If you have never removed window or door casements, go to SECTION E.
D6.
Think about all the years you've removed windows or
door casements. In an average year, how many weeks
did you spend removing window or door casements?
NONE
LESS THAN 1 WEEK
1 - 4 WEEKS
5 - 8 WEEKS
9 - 26 WEEKS
(ABOUT 3 - 6 MONTHS) .
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
.D
.D
.0
.D
.n
.D
TJ-SAO.V7
rev. 09/07/94
60-6;
62-63
64-65
66
67-63
69
END 01
-------
-6-
SECTION E: HEATING, VENTILATION AND AIR CONDITIONING — When you answer these
questions, include any work you have done to maintain, repair or clean heating, ventilation or
air conditioning systems.
days
E1. In the last 30 days, how many days did you work on
heating, ventilation or air conditioning systems?
If you have not done any work on heating, ventilation or air conditioning systems in the last 30
days, go to question E4.
E2.
E3.
E4.
In the last 30 days, how many days did you use a
respirator while you worked on heating, ventilation or air
conditioning systems?
In the last 30 days, how many days did you work on
heating, ventilation or air conditioning systems in homes
or buildings built before 1950?
Altogether in the past 12 months, how many weeks did
you work on heating, ventilation or air conditioning
systems?
days
days
NONE D
LESS THAN 1 WEEK D
1 - 4 WEEKS D
5 - 8 WEEKS D
9 - 26 WEEKS
(ABOUT 3 - 6 MONTHS) U
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) D
E5.
years
Think about all the years you've done renovation or
remodeling. How many of these years did you work on
heating, ventilation or air conditioning systems at least
some of the time?
If you have never worked on heating, ventilation or air conditioning systems, go to SECTION F.
E6.
Think about all the years you've worked on heating,
ventilation or air conditioning systems. In an average
year, how many weeks did you spend working on these
systems?
NONE D
LESS THAN 1 WEEK D
1 - 4 WEEKS D
S - 8 WEEKS D
9 - 26 WEEKS
(ABOUT 3 - 6 MONTHS) U
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) LJ
RECORD 02
*a *
J
TJ-SAQ.v?
rev. 09/07/94
09-10
11-12
13-14
15
16-17
18
-------
-7-
RECORD I
SECTION F: LARGE STRUCTURE REMOVAL — These questions are about removing any kind
of large structures. Include things like:
• making openings for large windows or doorways
• tearing down ceilings
• putting up walls
• removing kitchen cabinets
F1. In the last 30 days, how many days did you remove
large structures?
_days
If you have not removed any large structures in the last 30 days, go to question F4.
F2. In the last 30 days, how many days did you use a
respirator while you removed large structures?
F3. In the last 30 days, how many days did you remove
large structures from homes or buildings built before
1950?
F4. Altogether in the past 12 months, how many weeks did
you remove large structures?
F5. Think about all the years you've done renovation or
remodeling. How many of these years did you remove
large structures at least some of the time?
If you have never removed large structures, go to SECTION G.
F6.
Think about all the years you've removed large
structures. In an average year, how many weeks did
you spend removing large structures?
.days
.days
NONE
LESS THAN 1 WEEK D
1 - 4 WEEKS D
5 - 8 WEEKS D
9 - 26 WEEKS
(ABOUT 3 - 6 MONTHS) U
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) U
years
NONE
LESS THAN 1 WEEK
1 - 4 WEEKS
5 - 8 WEEKS
9 - 26 WEEKS
(ABOUT 3 - 6 MONTHS) ...... LJ
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) ..... LJ
TJ-SAQ.v?
lev. 09/07/9*
19-20
21-22
23-24
25
26-27
23
-------
-8 -
RECORD£Z
SECTION G: PAINT REMOVAL AND SURFACE PREPARATION — By surface preparation, we
mean activities such as sanding, scraping, torching, or floor refinishing.
G1. During the last 30 days, how many days did you
remove paint or prepare surfaces?
days
29-30
If you didn't remove paint or prepare surfaces at all in the last 30 days, please go to question
G6 and continue from there.
G2. During the last 30 days, how many days did you
remove paint or prepare surfaces in homes or buildings
built before 1950?
days
31-32
G3. During the last 30 days, how many days did you use a
respirator when you removed paint or prepared
surfaces?
days
33-34
G4. During the last 30 days when you removed paint or
prepared surfaces, how many days did you do this
inside?
days
35-36
G5. When you removed paint in the last 30 days, how many days
did you use the following?
A. dry power sanding?
days
37-38
B. dry hand sanding?
days
39-40
C. dry scraping?
.days
41-42
D. burning, torching, or a heat gun?
.days
43-44
E. wet scraping?
days
45-46
F. wet sanding?
days
47-48
G. chemical stripping?
days
49-50
H. a dust collector when you sanded?
days
51-52
TJ-SAQ.v/
lev. 09/07/9-1
-------
G6.
- 9 -
During the past 12 months, how many weeks did you
spend doing paint removal or surface preparation?
RECOR$t)2
NONE '.....'. D
LESS THAN 1 WEEK D
1 - 4 WEEKS D
5 - 8 WEEKS D
9 - 26 WEEKS
(ABOUT 3 - 6 MONTHS) D
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) U
If you have never done paint removal or surface preparation, please go to SECTION H.
G7.
Think about all the years you've done renovation and
remodeling. How many of these years did you do paint
removal and surface preparation at least some of the
time? (Please include in your answer any time you •
spent doing this while in the military.)
A.
When was the last time you did any paint
removal or surface preparation?
B.
Think about all the years you've done paint
removal and surface preparation. In an
average year, how many weeks did you spend
doing paint removal and surface preparation?
TJ-SAQ.V7
years
WITHIN THE LAST MONTH .
MORE THAN A MONTH AGO
MORE THAN A YEAR AGO ..
MORE THAN 5 YEARS AGO ,
...D
...D
...D
NONE U
LESS THAN 1 WEEK D
1 - 4 WEEKS D
5 - 8 WEEKS D
9 - 26 WEEKS
(ABOUT 3 - 6 MONTHS) U
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) U
rev. 09/07/94
S3
54-55
56
57
-------
- 10-
SECTION H: CLEANUP — By cleanup work, we mean the hands-on (dirty) cleanup, where you
cleaned up the dirt, dust and debris caused by the renovation and remodeling activities. We
want to know about the time you spent doing this kind of work. Don't include times when you
were around others doing cleanup, but you weren't.
RECORD
days
H1. During the last 30 days, how many days have you
spent doing dirty cleanup?
If you didn't do any dirty cleanup activity at all in the last 30 days, please go to question H6
and continue from there.
H2. In the last 30 days, about how many hours in a typical
day did you spend doing dirty cleanup?
H3. In the last 30 days, how many days did you spend
doing dirty cleanup in homes or buildings'built before
1950?
H4. In the last 30 days, how many days did you use a
respirator while doing dirty cleanup work?
H5. In the last 30 days while you did dirty cleanup, how many
days^did you...
A.
use a broom?
B. use any kind of vacuum?
C. use a HEPA vacuum?
D. use a wet mop with TSP, trisodium phosphate
or ledisolve?
E. clean power tools using any method?
F. clean power tools using compressed air?
LESS THAN V4 HOUR/DAY D
V4 TO LESS THAN 1 HOUR/DAY . D
1 TO 4 HOURS/DAY D
MORE THAN 4 HOURS/DAY D
days
.days
_days
_days
_days
.days
days
days
TJ.SAd.v7
r»v. 09/07/94
55-59
60
61-62
63-64
65-66
67-68
69-70
71-72
73-74
75-76
END 02
-------
H6.
- 11 -
During the past 12 months, how many weeks did you
spend doing dirty cleanup?
NONE n
LESS THAN 1 WEEK Q
1 - 4 WEEKS D
5 - 8 WEEKS D
9 - 26 WEEKS
(ABOUT 3 - 6 MONTHS) LJ
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) D
RECORD 03
v **
j
09
If you have never done dirty cleanup of any kind, please go to SECTION I.
H7.
How many years altogether have you spent doing dirty
cleanup work at least some of the time?
A.
When was the last time you did any dirty
cleanup work?
B.
Think about all the years you've done dirty
cleanup. In an average year, how many weeks
did you spend doing dirty cleanup?
TJ-SAQ.V7
years
WITHIN THE LAST MONTH .
MORE THAN A MONTH AGO -
MORE THAN A YEAR AGO . ,
MORE THAN 5 YEARS AGO .
...D
...D
...a
NONE D
LESS THAN 1 WEEK D
1 - 4 WEEKS D
5 - 8 WEEKS D
9 - 26 WEEKS
(ABOUT 3 - 6 MONTHS) U
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) D
(8V. 09/07/94
10-11
12
13
-------
- 12 -
RECORD 03 „,
CT1
11.
12.
13.
14.
3.V7
ON I: GENERAL WORK PRACTICE QUESTIONS
Please answer the following questions about your job in the last 30 days.
A. How many days did you have water available at the
worksite?
B. How many days did you work at sites that were
dusty and dirty?
C. How many days did you eat at the worksite?
D. How many days did you wash your hands before you
ate?
E. How many days did you wash your hands before you
went home?
F. How many days did you change your clothes before
you went home?
G. How many days did you change your shoes before
you went home?
H. How many days did you take a radio to the worksite
with you?
Do you currently use snuff or chewing tobacco?
Do you currently smoke cigarettes?
If you do not smoke cigarettes, go to question 14. If you d
answer questions A through C below.
A. Do you smoke while you work?
B. Do you smoke while on break?
C. Do you carry your cigarettes in your shirt or
pants pocket at work?
When you used a respirator in the last 30 days, what
type did you use? Check all that apply.
days
days
days
days
days
days
days
days
YES
NO
YES
NO
o smoke cigarettes, please
YES
NO
YES
NO
YES
NO .
DIDNT USE RESPIRATOR
DUST MASK
HALF MASK
FULL FACE MASK ....
TYPE C SUPPLIED
AIR MASK • ....
PAPR
SCBA
n
n
n
n
n
n
n
n
n
n
D
n
n
D
n
n
n
19V. 09/07/94
14-15
16-17
18-19
20-21
22-23
24-25
26-27
28-29
30
31
32
33
34
35-36
37-38
39-40
41-42
43-44
45-46
47-48
-------
SECTION J: OTHER OCCUPATIONS
Have you ever worked full-time in the
following industries?
- 13
1 . For a lead abatement contractor?
2. In a firing range?
3. In a smelter?
4. In a battery plant?
5. Doing radiator repair?
67 As a bridge painter or in bridge demolition?
7. In the shipbuilding industry?
8. In an oil refinery?
9. In paint manufacturing?
10. As a welder?
11. As a solderer?
12. As a lead miner?
13. In a lead mill?
14. In a scrap and waste material company?
YES
D
D
D
D
D
n
D
D
D -
D
D
D
D
D
NO
D
D
D
D
D
D
D
D
D
D
D
D
D
D
For each job you answered "YES" to above,
please fill out a line on the following table.
Line # of What was your job and most
job important duties?
(from table
above)
l_LJ
LJ_J
L_LJ
When was the last time you worked in this industry?
Within the More than a
last month month ago
D D
D D
D D
More than More than 5
a year ago years ago
D D
D D
D D
RECORD 03
W
j
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63-67
68-72
73-77
TJ-SAQ.V7
lev. 09/07/94
-------
- 14 -
RECORD CM
SECTION K: RESIDENTIAL INFORMATION — In this section, we want to find out about your
home. By home, we mean the house, apartment building, or any other building where you live,
whether or not you own it
K1. How old is your home? If you've lived in more than one NEW CONSTRUCTION Q 09
home during the last 12 months, how old is the home
you lived in the longest? NOT NEW BUT BUILT ,_,
1 a AFTER 1978 D
BUILT BETWEEN
1950 AND 1978 D
BUILT BEFORE 1950 Q
K2. Have any room additions or any major remodeling YES D 10
changes been made to your home in the last 12
months? NO LJ
If no additions or changes have been made to your home in the last 12 months, leave
questions A through F blank and continue with SECTION L.
A. Was the inside of the house remodeled in the last YES D n
12 months? _
NO D
B. Was the exterior of the house remodeled in the YES D
last 12 months? _
NO D
C. Was paint stripped or sanded during the YES D 13
remodeling?
NO U
D. Did the remodeling include major work on the YES D 14
kitchen or the bathroom? _
NO U
E. Did you do any of the work yourself? YES D 1S
NO D
If you did any of the work yourself, please describe what you did on the SPECIFY line
below.
16-17
SPECIFY:
F. Did you live in the house while the work was YES D 1S
being done? ._.
NO U
TJ-SAQ.V7
rev. 09/OWW
-------
- 15-
SECTION L: NON-WORK ACTIVITIES
L1. Do you do any of the following?
a. Shoot guns at an indoor firing range?
b. Cast lead into bullets or fishing sinkers?
c. If you go fishing more than 10 times a year, do
you crimp your sinkers onto your line with your
teeth?
d. Refinish, restore or repaint old cars or bicycles?
e. Refinish furniture?
f. Work on old metal radiators?
g. Dismantle car or truck batteries?
h. Paint with artist's paints?
i. Work with stained glass?
j. Work with ceramics?
k. Eat out of imported ceramic dishes or decorated
pottery?
SECTION M: MEDICAL HISTORY
M1. Have you ever been diagnosed by a health care
professional as having an elevated lead level in your
blood?
M2. In the last 12 months, have you been diagnosed by a
health care professional as having anemia?
M3. Has anyone else living in your household ever had their
blood tested for lead? Check all that apply.
M4. Has anyone else living in your household ever been
identified by a health care professional as having an
elevated blood lead level? Check all that apply.
M5. Do you take calcium supplements?
Yes
D
D
D
D
D
D
D
D
D
No
n
n
n
n
n
a
a
a
a
a
YES
NO
YES .
NO .
YES, OTHER ADULT
YES. CHILD
NO
YES, OTHER ADULT
YES. CHILD
NO
YES
NO
RECORD Q.4.
j
19
20
21
22
23
24
25
26
27
28
29
.a
.n
.a
.a
.a
.a
.a
.D
.a
.a
.a
TJ-SAQ.V7
fgv 09/07/94
30
31
32
33
34
35
36
37
38
-------
- 16-
RECORD 0-1
M6. Do you use any of these medicines: Azarcon, greta, YES Q 39
liga, maria luisa, alarcon, coral, rueda or pay-loo-ah?
NO n
M7. Have you ever been shot by a gun? YES Q 40
NO D
SECTION N: PREVIOUS TRAINING
N1. Have you ever received any training about how to YES D 41
reduce potential lead exposures at work?
NO D
If you have never had any lead training, please
continue with question N2.
A. Who provided this training? Check all that apply. EMPLOYER D 42~43
UNION D 44~45
LOCAL OR STATE „ AR ,_.
HEALTH DEPT D 46'4?
PRIVATE COMPANY D 48~49
FEDERAL AGENCY D 50~51
COLLEGE OR UNIVERSITY D 52~53
- SPECIFY: SOMEONE ELSE (SPECIFY) D 54'55
N2. Have you ever received any pamphlets or other information about
how to reduce potential lead exposures in the workplace from any
of the following groups? Yes No
a. A trade organization, such as NARI? D D 56
b. Union newsletters? D D 57
c. Trade magazines? D D 58
d. Supplier pamphlets? ' I—' "-J 59
e. Federal government, such as EPA '—' '-' 60
f. Non-profit or advocacy organization '—' '—' 61
g. Any other source? If yes, please specify on line below. . . .
SPECIFY:
TJ-SAQ.V7
62
-------
- 17-
N3. Have you ever had the water in your home tested for
lead?
N4. Do you run your tap water for a minute before taking a
drink?
N5. Has your home ever been tested for lead paint?
N6. Do you think that lead poses a potential problem for you
at work?
N7. Are there other ways you think you may be exposed to
lead that we haven't asked about?
RECORD Q4--
YES
NO
YES
NO .
YES .
NO .
YES .
NO .
N8. Is there anything else you'd like to tell us, or do you
have any additional comments you would like to make?
.D
.n
.D
.D
.D
.D
.D
.D
63
64
65
66
67
68
END 04
Thank you very much for taking the time to help us understand how people may or may not be
exposed to lead. If you have any questions about the study, you can call John Egel at
1-800-444-5234.
PLEASE RETURN THIS QUESTIONNAIRE TO THE STUDY MANAGER
If you would like the results of this study when it is complete, please call John Egel at 1-800-444-5234
to request them.
TJ-SAQ.V7
(8v. 09/07/9-t
-------
APPENDIX B:
ASSESSMENT OF LABORATORY QUALITY CONTROL DATA
-------
Appendix B
Assessment of Laboratory Quality Control Data
Quality control activities were conducted on this project in order to assure the accuracy
and reliability of the data. The assessment of the overall quality of the data was performed by the
Senior Quality Assurance Officer of Midwest Research Institute using statistical quality control
(SQC) procedures. The quality control data from the blood analyses for lead were evaluated
using one or more statistical techniques (e.g., Gaussian distribution, ANOVA statistics, or
Shewhart performance charts).
This section evaluates the data in terms of the data quality indicators as defined in the
QAPjP. These indicators are the data quality objectives (DQOs) for the program and
measurement quality objectives (MQOs) for the analytical procedures. Finally, the analytical
results are assessed using the data obtained from the various quality control and performance
evaluation samples.
1. DATA QUALITY OBJECTIVES
The analytical objective of the study was to determine the levels of lead in blood samples
obtained from R&R workers. Using these data, the primary objective was to determine the
relationship between blood-lead levels and work practices performed by R&R workers in specific
work groups or work activities. To achieve the analytical objective of assuring that the data will
permit an assessment of the correlation between R&R activities and actual exposure, the
analytical data must meet the quantitative QA objectives of precision, accuracy, and
completeness, and the qualitative QA objectives of representativeness and comparability.
With the selection of standard analytical methodology having known and acceptable
criteria of performance for method detection limit, precision, and accuracy, the basic analytical
objectives were met for this project. The quantitative objective for completeness required to
achieve a specific statistical level of confidence also was achieved with 100% of the samples
collected and shipped to the laboratory where they were to be analyzed. By using an acceptable
and validated standard analytical method for blood-lead analysis, the analytical results are
considered to be representative of lead levels in this population and comparable to the results of
other similar studies.
2. MEASUREMENT QUALITY OBJECTIVES
The analytical methodology selected for the analysis of blood samples for lead was
expected to provide quantifiable lead levels based on the expected amount of sample to be
collected. To ensure the accuracy and reliability of the data generated during the analytical
process, a series of internal quality control samples and CDC performance samples were included
as part of the analytical design. These quality control and performance samples (Table B-l) with
B - 1
-------
proposed criteria were designed to allow not only a controlling mechanism for the procedure
during the analytical process but also to provide the means to evaluate the results from each set of
analyses and to assess the quality of the data.
Table B-l. Quality Control and CDC Performance Samples
Sample
Identification
Blank
BioRad 1
BioRad 2
BioRad 3
New York State
SRM
NIST SRM 995a
CDC RS 590
CDC RS 991
CDC RS 1394
MQO
Accuracy
< 1.0 pg/dL
±25%
±25%
±25%
±10%
±10%
±20%
±20%
±20%
MQO
Precision
±10%
±10%
±10%
Type of Quality
Control Sample
Matrix modifier
Continuing calibration
Reference material
samples
Continuing calibration
Reference material
samples
Continuing calibration
Reference material
samples
Calibration verification
Calibration Check
Standard
Blind performance
Blind performance
Blind performance
Lead Cone.
j/g/dL
0.0
5.0 - 8.3
24.4 - 24.5
55.4
16.0
13.53
9.0
23.3
41.1
Mean
Recovery
-0.08
99.74
95.24
95.98
98.25
101.0
93.35
98.72
102.6
Control
Limits
1.59
58 - 142
72 - 118
81 - 111
68 - 130
79 - 122
77 - 110
89 - 109
91 - 114
The results from the internal quality control samples and the external performance samples
show that the selected method provided quantifiable lead levels that were accurate and reliable.
The internal quality control samples were used during the analytical process as a control
mechanism. The precision criteria were met for the majority of the internal quality control
samples analyzed. Of those samples that did not meet the precision criteria, all were reanalyzed.
The accuracy criteria were met for the internal quality control samples with the exception of six
occasions or control situations. In two of the six control situations occurring during the analytical
process, the process was terminated and restarted because the quality controls sample results
were outside the acceptance criteria. In three of the six situations, the preceding and subsequent
quality control samples were acceptable; therefore, the analytical process was continued. The
samples analyzed between the two acceptable quality control samples were individually reviewed
and either repeated in another batch or accepted with explanation and corrective action. The last
control situation resulted from a shift of the control posture of the calibration verification
reference (New York State SRM). These control situations and the control posture for all of the
quality control and performance samples are discussed under data assessment (3.0).
B - 2
-------
3. DATA ASSESSMENT
Six quality control samples and three CDC performance evaluation samples were evaluated
using a statistical evaluation program based on Gaussian distribution (Tables B-2 through B-10)
and Shewhart control charts (Figures B-l through B-10). Of the nine performance charts, five
charts (BioRad-2, BioRad-3, CDC RS 590, CDC RS 991, and CDC RS 1394) show that the
analytical process was in control for the specific level of lead. The other four charts indicated
different control situations during the analytical phase of the project. These situations are
discussed below.
The low level BioRad-1 (5.0 to 8.3 |ig/dL) reference material quality control sample shows
a broader accuracy range (58% to 142%) than had been expected (±25%) due largely to an
observation (216%) on 22 November 1994. However, on that day, the failure of that quality
control sample resulted in stopping the analytical sequence and restarting the analysis. When this
sample is deleted, the variability is reduced from a standard deviation of 14.6% to a standard
deviation of 10.4 and the accuracy range is reduced to ±30%, close to laboratory expectations and
in compliance with QAPjP data quality objectives of ±30% for spiked samples. Moreover, over
95% of the BioRad-1 samples fell within ±25% of the certified concentration of the reference
material.
The NIST SRM (13.53 |ig/dL) used as the continuing calibration check was determined to
have a broader accuracy range (79% to 122%) than originally considered (±10%). The accuracy
range for the NIST SRM is consistent with the CDC RS 590 (9.0 (ig/dL) and the BioRad-2 (24.4
to 24.5 |ig/dL) ranges. The accuracy is consistent with the fact that it is a matrix SRM rather
than a calibration standard that is used in similar analytical methods. In view of these facts, the
accuracy range should be considered consistent with other procedures using matrix SRM material
with an accuracy acceptance range of ±25%. In addition the accuracy range was calculated
including a 22 November 1994 sample (62.31%) that resulted in stopping the analytical sequence
and restarting the analysis, as well as a 15 November 1994 sample (62.31%) that was determined
to be caused by a bad burn as discussed below. Deletion of these two samples results in an
accuracy range of ±12%.
The normal instrument detection limit for the GFAA analysis is approximately 1 |ig/dL.
However, the results for the matrix modifier (blank) indicate a standard deviation of 1.59. A
review of the data for the matrix modifier blanks indicated that four of the 288 samples were
unusually large. Two of the four samples resulted in the analytical sequence being stopped and
reanalysis of the samples. The other two samples were rejected because of their large difference
from a replicate analysis of the same sample. When these four samples are deleted, the standard
deviation is reduced to 0.64 resulting in an DDL of 1.27. Since all of the calibration standards and
quality control samples were blood matrices, the MDL (3 x noise level) and method quantitation
level (MQL = 5 x noise level) were also evaluated. The resultant MDL and MQL were 1.9 |ig/dL
and 3.2 |ig/dL, respectively, for this procedure. These results are consistent with the results of
the duplicate analyses (presented in Section 3.2.2.2). The log standard deviation of the duplicate
blood draws was 0.0677. Assuming a lognormal distribution and a geometric mean of 4.5 |ig/dL
B - 3
-------
(the mean blood-lead concentration in the study), the standard deviation of the sampling and
analysis method is 0.31 which agrees with the expected precision of the method.
The last performance chart control situation occurred on the calibration verification quality
control sample (NY State SRM). The control limits for the NY State SRM
(16 |ig/dL) show an accuracy range of ±31%. However, due to a shift in the control posture of
the quality control sample, there are actually two ranges. The first range is for the analytical
period from 23 September 1994 to 15 December 1994, where the mean recovery was 93.4%, and
the accuracy range was 83.4% to 103.4%. The second range is for the analytical period from 15
December through 30 December 1994, where the mean recovery was 116.8% and the analytical
range 106.8% to 126.8%. Each range by itself was within the acceptable criterion; however, the
shift in control posture is unacceptable. The control posture of the NIST SRM (13.35|ig/dL) was
also reviewed and compared with the NY State SRM results. Although there was no shift in
control posture for the NIST SRM, the data show a mean that has increased consistently over the
analytical sequence. The NIST data are supportive of the shift in control posture for the NY
State SRM. The analytical results, based on the calibration curves from 15 December through 30
December 1994, should be flagged as having potentially high bias.
As discussed earlier under 2.0, there are six individual control situations. There also are
four blanks that are above the IDL. Each of these situations were reviewed by the laboratory and
corrective actions taken. For two of these situations on 22 November 1994, the failure of the
BioRad-1 (216%) and NIST (62.31%) quality control samples resulted in stopping the analytical
sequence and restarting the analysis.
On 23 September and 11 November 1994, one of the BioRad-2 quality control samples for
each analytical run failed (125.6% and 70.35%, respectively). The quality control samples, both
prior and subsequent, were within their respective control criteria. After investigation by the
laboratory, the results for these standards were flagged as being outside of the normal data set,
and the samples analyzed between the acceptable standards were accepted as valid.
The fifth situation occurred on 15 November 1994 for one of the NIST quality control
samples. The result (62.31%) for this sample was determined to be caused by a bad second
analytical burn (analysis). The result was flagged as such, and no further action was taken.
The last situation was uncorrected by the laboratory because it occurred on one of the
blind performance evaluation samples (CDC RS 590). The result for the CDC RS 590 was
76.1%, which was outside of the statistical control limits (77 to 110%) but within the preset
control criterion of ±20% (73% to 113%). No further action is required on these data.
For the matrix modifier (method blank), there were four analyses that had results greater
than the IDL. Two of these results were from a single analysis with the second analytical result
falling below the IDL; therefore, these are considered to be acceptable (showing no possible
interference). The other two results were from a replicate analyses where both were above the
DDL. These occurred on 22 November 1994 and were part of a normal analytical system
shutdown when it is shown not to meet the QC criteria. The analytical sequence was restarted
B - 4
-------
with calibration standards followed by matrix modifier, and the samples analyzed after the
instrument failed QC criteria were then reanalyzed.
4. SUMMARY
The overall assessment of the data indicates that the data are accurate and reliable. The
data meet the DQOs and MQOs as defined in the QAPjP with only the exceptions flagged in this
section.
B - 5
-------
B - 6
-------
Table B-2. ESA Internal Quality Control Sample Statistics Report for the Matrix
Modifier Blank
Column Name: Matrix Modifier
Number of column points: 288
Number of valid values: 288
Number of missing value: 0
Number of negative value: 166
Number of positive value: 119
Number of zero: 3
Maximum:
Minimum:
Sum of raw value:
Sum of absolute value:
Arithmetic mean:
Geometric mean:
Quadratic mean:
Harmonic mean:
Absolute mean:
Median:
Sum of squares:
Variance:
Standard deviation:
Absolute deviation:
Standard error:
95 % confidence interval:
[-0.265625218 , 0.102708551]
99 % confidence interval:
[-0.324086927 , 0.161170260]
Coeff. of variance: -1949.34582
Skewness: 10.0738714
Coeff. of skewness: 5.03693571
Kurtosis: 132.385783
Coeff. of kurtosis: 135.385783
22.0900000
-1.97000000
-23.4600000
170.040000
-0.0814583333
0.000000000
1.58723707
0.000000000
0.590416667
-0.0800000000
725.564600
2.52144107
1.58790462
0.583909144
0.0935681769
Percentiles:
10 percentile;
25 percentile:
50 percentile;
75 percentile;
90 percentile:
Quartiles:
First quartile:
Second quartile:
Third quartile:
-1.31000000
-0.410000000
-0.0800000000
0.180000000
0.430000000
-0.410000000
-0.0800000000
0.180000000
B-6
-------
«p
bO
1
OS
25
20
15
"
0
-5
g
(Triton X-100 [1%] in a buffer solution)
1 1 1
20
o
o
•a B-
B o
1 1 1 1 1 1
40
60
Analytical Date Sequence
80
100
Figure B-1. ESA Matrix Modifier (Blank).
-------
Table B-3. ESA Internal Quality Control Sample Statistics Report for BioRad-1
0
131
0
Colximn Name : RECOVERY
Number of column points: 132
Number of valid values: 131
Number of missing value: 1
Number of negative value
Number of positive value
Number of zero
Maximum:
Minimum:
Sum of raw value:
Sum of absolute value:
Arithmetic mean:
Geometric mean:
Quadratic mean:
Harmonic mean:
Absolute mean:
Median:
Sum of squares:
Variance:
Standard deviation:
Absolute deviation:
Standard error:
95 % confidence interval:
[97.2197214 , 102.267929]
99 % confidence interval:
[96.4085450 , 103.079105]
216.111111
73.8202247
13066.4411
13066.4411
99.7438252
98.9031773
100.799016
98.1887767
99.7438252
97.1910112
1331017.8553
213.238749
14.6026966
8.67560181
1.27584353
Coeff. of variance:
Skewness:
Coeff. of skewness:
Kurtosis:
Coeff. of kurtosis:
Percentiles:
10 percentile:
25 percentile:
50 percentile:
75 percentile:
90 percentile:
Quartiles:
First quartile:
Second quartile:
Third quartile:
14.6402011
3.93842938
1.96921469
28.6586439
31.6586439
86.6292135
93.9887640
97.1910112
102.777778
115.900000
93.9887640
97.1910112
102.777778
B-8
-------
BR1 Blood Lead Reference 5.0 - 8.3 /xg/dL
CD
CO
22b
200
175
£* 150
8
« 125
c
8
«D
100
75
50
*
0
- .
1 , . 1* i
• •.•-•".«• X«!
- M ? ".:? ;^ !<»«•..•
L § ' '* . . '' '•
-
-
0
20
40
60
80
100
Analytical Date Sequence
Figure B-2. ESA Internal Quality Control Sample (5.0 to 8.3 //g/dL).
-------
Table B-4. ESA Internal Quality Control Sample Statistics Report for BioRad-2
Column Name : RECOVERY
Number of column points: 72
Number of valid values: 72
Number of missing value: 0
Number of negative value: 0
Number of positive value: 72
Number of zero: 0
Maximum:
Minimum:
Sum of raw value:
Sum of absolute value:
Arithmetic mean:
Geometric mean:
Quadratic mean:
Harmonic mean:
Absolute mean:
Median:
Sum of squares:
Variance:
Standard deviation:
Absolute deviation:
Standard error:
95 % confidence interval:
[93.4595462 , 97.0195159]
99 % confidence interval:
[92.8766869 , 97.6023752]
Coeff. of variance: 7.95339386
Skewness: 0.731306870
Coeff. of skewness: 0.365653435
Kurtosis: 4.41858805
Coeff. of kurtosis: 7.41858805
125.571429
70.3469388
6857.24624
6857.24624
95.2395311
94.9479326
95.5361114
94.6574073
95.2395311
94.3979592
657154.6983
57.3772166
7.57477502
4.79680046
0.892695797
Percentiles:
10 percentile:
25 percentile:
50 percentile:
75 percentile:
90 percentile:
Quartiles:
First quartile:
Second quartile:
Third quartile:
88.4221311
91.8852459
94.3877551
97.3469388
102.356557
91.8852459
94.3877551
97.3469388
B-10
-------
BR2 Blood Lead Reference 24.4 -24.5
130
120
110
4J
1
8
£
100
90
80
70
60
0
20
40
60
80
100
Analytical Date Sequence
Figure B-3. ESA Internal Quality Control Sample (24.4 to 24.5 //g/dL).
-------
Table B-5. ESA Internal Quality Control Sample Statistics Report for BioRad-3
Column Name : RECOVERY
Number of column points: 71
Number of valid values: 71
Number of missing value: 0
Number of negative value: 0
Number of positive value: 71
Number of zero: 0
Maximum:
Minimum:
Sum of raw value:
Sum of absolute value:
Arithmetic mean:
Geometric mean:
Quadratic mean:
Harmonic mean:
Absolute mean:
Median:
Sum of squares:
Variance:
Standard deviation:
Absolute deviation:
Standard error:
95 % confidence interval:
[94.7932469 , 97.1738521]
99 % confidence interval:
[94.4032499 , 97.5638491]
Coeff. of variance: 5.23925334
Skewness: -0.728428036
Coeff. of skewness: -0.364214018
Kurtosis: 2.18955629
Coeff. of kurtosis: 5.18955629
106.317690
76.0649819
6814.83201
6814.83201
95.9835495
95.8493468
96.1133426
95.7101037
95.9835495
96.0486891
655881.9990
25.2890439
5.02882132
3.65195459
0.596811291
Percentiles:
10 percentile:
25 percentile:
50 percentile:
75 percentile:
90 percentile:
Quartiles:
First quartile:
Second quartile:
Third quartile:
90.8052434
93.2116105
96.0486891
98.3801498
102.481949
93.2116105
96.0486891
98.3801498
B-12
-------
BR3 Blood Lead Reference 55.4 /-ig/dL
130
120
110
CD
CJ
4>
100
90
•••
5
80
70
60
0
20
40
60
80
100
Analytical Date Sequence
Figure B-4. ESA Internal Quality Control Sample (55.4 //g/dl).
-------
Table B-6. ESA Internal Quality Control Sample Statistics Report for New York
State SRM
Column Name : RECOVERY
Number of column points: 43
Number of valid values: 43
Number of missing value: 0
Number of negative value: 0
Number of positive value: 43
Number of zero: 0
Maximum:
Minimum:
Sum of raw value:
Sum of absolute value:
Arithmetic mean:
Geometric mean:
Quadratic mean:
Harmonic mean:
Absolute mean:
Median:
Sum of squares:
Variance:
Standard deviation:
Absolute deviation:
Standard error:
95 % confidence interval:
[95.0659486 , 101.451978]
99 % confidence interval:
[93.9900751 , 102.527851]
Coeff. of variance: 10.5590345
Skewness: 1.06031767
Coeff. of skewness: 0.530158836
Kurtosis: -0.317295166
Coeff. of kurtosis: 2.68270483
121.281250
86.7812500
4225.13542
4225.13542
98.2589632
97.7589271
98.7925362
97.2933080
98.2589632
94.2500000
419678.5040
107.644731
10.3751979
8.25766180
1.58220284
Percentiles:
10 percentile:
25 percentile:
50 percentile:
75 percentile:
90 percentile:
Quartiles:
First quartile:
Second quartile:
Third quartile:
88.4375000
91.2812500
94.2500000
100.437500
114.562500
91.2812500
94.2500000
100.437500
B-14
-------
CO
_^
ui
SJ
8
o
8
140
130
120
110
100
90
80
70
60
0
NY State Lead Reference 16.0 jig/dL
20
40
60
80
100
Analytical Date Sequence
Figure B-5. ESA Initial Calibration Verification (16.0//g/dL).
-------
Table B-7. ESA Internal Quality Control Sample Statistics Report for NIST SRM
189
0
0
189
0
Column Name : RECOVERY
Number of column points: 189
Number of valid values:
Number of missing value:
Number of negative value:
Number of positive value:
Number of zero:
Maximum:
Minimum:
Sum of raw value:
Sum of absolute value:
Arithmetic mean:
Geometric mean:
Quadratic mean:
Harmonic mean:
Absolute mean:
Median:
Sum of squares:
Variance:
Standard deviation:
Absolute deviation:
Standard error:
95 % confidence interval:
[99.9664885 , 102.031487]
99 % confidence interval:
[99.6369693 , 102.361006]
Coeff. of variance: 7.12444179
Skewness: -1.70068007
Coeff. of skewness: -0.850340037
Kurtosis: 9.12367233
Coeff. of kurtosis: 12.1236723
119.586105
59.0785908
19088.8087
19088.8087
100.998988
NAN
101.253634
100.380741
100.998988
100.406504
1937684.4096
51.7768622
7.19561410
4.92037472
0.523403909
Percentiles:
10 percentile:
25 percentile:
50 percentile:
75 percentile:
90 percentile:
Quartiles:
First guartile:
Second guartile:
Third guartile:
94.9741316
97.8196600
100.406504
105.210643
109.312639
97.8196600
100.406504
105.210643
B-16
-------
CD
_4
•vl
o
o
o
o
c
o
NIST Blood Lead Reference 13.53 jtig/dL
130
120
110
100
90
80
70
60
50
u
0
20
f
•
T
40
60
80
100
Analytical Date Sequence
Figure B-6. ESA Internal Quality Control Sample (13.53 //g/dL).
-------
Table B-8. ESA Performance Evaluation Sample Statistics Report for CDC RS 590
(9.0 //g/dL)
Column Name : RECOVERY
Number of column points: 35
Number of valid values: 35
Number of missing value: 0
Number of negative value: 0
Number of positive value: 35
Number of zero: 0
Maximum:
Minimum:
Sum of raw value:
Sum of absolute value:
Arithmetic mean:
Geometric mean:
Quadratic mean:
Harmonic mean:
Absolute mean:
Median:
Sum of squares:
Variance:
Standard deviation:
Absolute deviation:
Standard error:
95 % confidence interval:
[91.4720453 , 95.2374785]
99 % confidence interval:
[90.8271166 , 95.8824072]
Coeff. of variance: 5.87092503
Skewness: -0.996913625
Coeff. of skewness: -0.498456813
Kurtosis: 1.48435599
Coeff. of kurtosis: 4.48435599
104.944444
76.0833333
3267.41667
3267.41667
93.3547619
93.1909478
93.5109210
93.0188233
93.3547619
94.8333333
306050.2323
30.0390380
5.48078808
4.04340136
0.926422274
Percentiles:
10 percentile:
25 percentile:
50 percentile:
75 percentile:
90 percentile:
Quartiles:
First quartile:
Second quartile:
Third quartile:
86.1666667
91.0833333
94.8333333
97.1388889
97.7777778
91.0833333
94.8333333
97.1388889
B-18
-------
CDC Blood Lead Reference 9.0
120
110 -
u
o
o
-------
Table B-9. ESA Performance Evaluation Sample Statistics Report for CDC RS 991
(23.3//g/dL)
98
98,
Column Name : RECOVERY
Number of column points: 35
Number of valid values: 35
Number of missing value: 0
Number of negative value: 0
Number of positive value: 35
Number of zero: 0
Maximum:
Minimum:
Sum of raw value:
Sum of absolute value:
Arithmetic mean:
Geometric mean:
Quadratic mean:
Harmonic mean:
Absolute mean:
Median:
Sum of squares:
Variance:
Standard deviation:
Absolute deviation:
Standard error:
95 % confidence interval:
[97.5778438 , 99.8690045]
99 % confidence interval:
[97.1854227 , 100.261426]
Coeff. of variance: 3.37802936
Skewness: 0.138362618
Coeff. of skewness: 0.0691813088
Kurtosis: 0.492721042
Coeff. of kurtosis: 3.49272104
107.263948
90.5793991
3455.31984
3455.31984
98.7234241
6687853
7781267
98.6141893
98.7234241
98.4334764
341499.1409
11.1215997
3.33490625
2.47177350
0.563702042
Percentiles:
10 percentile:
25 percentile:
50 percentile:
75 percentile:
90 percentile:
Quartiles:
First quartile:
Second quartile:
Third quartile:
94.4527897
97.1351931
98.4334764
100.407725
102.167382
97.1351931
98.4334764
100.407725
B-20
-------
CDC Blood Lead Reference 23.3
120
110 -
O
100
I •
90
80
70
0
20
40
60
80
100
Analytical Date Sequence
Figure B-8. ESA Performance Evaluation Sample (23.3 //g/dL).
-------
Table B-10. ESA Performance Evaluation Sample Statistics Report for CDC RS 1394
Column Name : RECOVERY
Number of column points: 35
Number of valid values: 35
Number of missing value: 0
Number of negative value: 0
Number of positive value: 35
Number of zero: 0
Maximum:
Minimum:
Sum of raw value:
Sum of absolute value:
Arithmetic mean:
Geometric mean:
Quadratic mean:
* Harmonic mean:
-^ Absolute mean:
Median:
Sum of squares:
Variance:
Standard deviation:
Absolute deviation:
Standard error:
95 % confidence interval:
[101.351694 , 103.908299]
99 % confidence interval:
[100.913808 , 104.346185]
109.482968
93.3333333
3592.04988
3592.04988
102.629997
102.563833
102.695513
102.497027
102.629997
103.406326
369122.8947
13.8478871
3.72127492
3.04425245
0.629010267
Coeff. of variance:
, Skewness:
Coefif. of skewness:
Kurtosis:
Coeff. of kurtosis:
Percentiles:
10 percentile:
25 percentile:
50 percentile:
75 percentile:
90 percentile:
Quartiles:
First guartile:
Second quartile:
Third quartile:
3.62591352
-0.323345875
-0.161672938
-0.366972074
2.63302793
97.1715328
100.206813
103.406326
105.182482
106.429440
100.206813
103.406326
105.182482
B-22.
-------
CDC Blood Lead Reference 41.1 j*g/dL
120
110
CD
M
GO
1
€
8
100
90
80
70
I 1 I I I I I I I I
20
40
60
80
100
Analytical Date Sequence
Figure B-9. ESA Performance Evaluation Sample (41.1 //g/dL).
-------
CDC Blood Lead Reference
GO
NJ
120
110
b
> 100
8
o
1
£ 90
80
70
00
oo ° o Q o
° D 3 o ^ a <> °
D ^r^ O O f-t O.
* n 9-1 ° 5 o
89 B °OD otB
° D0 ° D000
D oo J3 a
0 o * o °
a u ° oo
o
0 0
0
O Lcfai
O- 9.0/.t/dL
0 0-23.3^
1 1 1 1 1 1 1 1 1 1 t 1 | 1 1 1 1 1 1
0 20 40 60 80 10
Analytical Date Sequence
Figure B-10. ESA Performance Evaluation Sample.
-------
APPENDIX C:
DESCRIPTIVE STATISTICS TABLES
-------
Table C-l. Descriptive Statistics of Lead Concentrations for CDC Quality
Control Reference Samples
CDC
Referenc
e
Number
590
991
1394
Known
Value
9.0
23.3
41.1
N
30
30
30
Min
(ug/dl)
7.4
21.8
38.4
25th
Percentile
(ug/dl)
8.3
22.7
41.4
Median
(ug/dl)
8.6
23.0
42.6
75th
Percentile
(ug/dl)
8.8
23.4
43.5
Max
(ug/dl)
9.5
25.0
45.1
Geometric
Mean
(ug/dl)
8.4
23.0
42.3
log
Std. Dev.
log(ug/dl)
0.056
0.031
0.037
Table C-2. Descriptive Statistics of Log Standard Deviations Between Duplicate
Blood Draws for Each Sampling Frame
Sampling Frame
Philadelphia Union
Philadelphia Non-Union
St Louis Union
St Louis Non-Union
All Four Groups Combined
N
25
8
19
21
73
Min
log(ug/dl)
0.006
0.003
0.000
0.000
0.000
25th
Percentile
log(ug/dl)
0.010
0.008
0.000
0.017
0.009
Median
log(ug/dl)
0.020
0.016
0.025
0.028
0.020
75th
Percentile
log(ug/dl)
0.035
0.031
0.080
0.042
0.042
Max
log(ug/dl)
0.113
0.097
0.504
0.109
0.504
Mean
log(ug/dl)
0.031
0.026
0.076
0.037
0.044
Table C-3. Descriptive Statistics of Log Standard Deviations Among Multiple
Chemical Analyses for Each Sampling Frame
Sampling Frame
Philadelphia Union
Philadelphia Non-Union
St Louis Union
St Louis Non-Union
All Four Groups Combined
N
197
74
149
160
580
Min
log(ug/dl)
0.00
0.00
0.00
0.00
0.00
25th
Percentile
log(ug/dl)
0.006
0.00
0.010
0.006
0.005
Median
log(ug/dl)
0.016
0.015
0.031
0.020
0.019
75th
Percentile
log(ug/dl)
0.035
0.031
0.062
0.036
0.040
Max
log(ug/dl)
0.267
0.243
0.485
0.325
0.485
Mean
log(ug/dl)
0.030
0.028
0.051
0.031
0.036
C - 1
-------
Table C-4. Descriptive Statistics of Blood-Lead Concentrations for Each
Sampling Frame
Sampling Frame
Philadelphia Union
Philadelphia Non-Union
St Louis Union
St Louis Non-Union
All Four Groups
N
197
74
150
160
581
Min
(ug/di)
1.55
1.00
1.00
1.00
1.00
25th
Percentile
(ug/di)
3.80
3.80
1.95
2.95
2.90
Median
(ug/di)
5
5.08
3.03
4.73
4.50
75th
Percentile
(ug/di)
6.60
8.45
4.50
7.95
6.60
Max
(ug/di)
55.25
26.2
13.80
36.25
55.25
Geometric
Mean
(ug/di)
5.10
5.61
3.02
4.90
4.46
log
Std. Dev.
log(|jg/dl)
0.517
0.657
0.592
0.727
0.659
n>10
14
12
3
22
52
Table C-5. Descriptive Statistics of Blood-Lead Concentrations for Each Worker
Group
Worker Group
Union Carpenter
Non-Union Carpenter
Drywall Worker
Floor Layer
Laborer
Other
Painter
Supervisor
Window Installer
N
159
104
64
81
54
14
34
57
14
Min
(ug/di)
1.00
1.05
1.80
1.00
1.00
1.00
1.00
1.10
1.65
25th
Percentile
(ug/di)
3.10
3.28
4.50
1.95
3.55
3.50
4.75
2.55
2.55
Median
(ug/di)
4.50
5.00
5.65
2.75
4.80
4.68
7.15
4.05
5.55
75th
Percentile
(ug/di)
6.25
8.03
7.23
3.85
7.15
6.45
8.80
4.95
10.40
Max
(ug/di)
32.60
26.20
24.45
6.65
55.25
21.40
36.25
13.60
15.65
Geometric
Mean
(ug/di)
4.37
5.03
5.85
2.63
4.86
5.30
7.17
3.75
5.37
log
Std. Dev.
log(ug/dl)
0.579
0.665
0.484
0.497
0.716
0.829
0.781
0.524
0.778
n>10
9
15
5
0
5
3
7
3
4
C - 2
-------
APPENDIX D:
ADDITIONAL QUESTIONNAIRE TABLES
-------
Table D-l. Main Activities and Job Titles for Each Worker Group
Job Title
Carpenter
Carpenter
Carpenter
Carpenter
Carpenter
Maintenance
Man
Carpenter
Owner
Contractor
Carpenter
Owner
Main Activity
Main Activities for Carpenter Classification
GENERAL CARPENTRY-(CONSTRUCTION REMODELING, ALL PHASES,
FRAMING, HANDYMAN)
GENERAL CARPENTRY-(CONSTRUCTION REMODELING, ALL PHASES,
FRAMING, HANDYMAN)
CUSTOM WOOD WORKING
PRE-HUNG DOOR ASSEMBLER
ROOF SPRAYING, SIDING
INSULATION OF THE CEILING AND SIDING AND CONCRETE
PUNCH LIST WORK ON NEW CONSTRUCTION
RENOVATION OF WATER PLANT BUILDINGS
TENANT REMODEL DEMOLITION AND RECONSTRUCTION
REMODEL NEW GROCERY STORE AND INSTALL NEW COMM FREEZER
WORKING AT HOME
BUILDING- SCAFFOLDING/HOUSING WORK
METAL ROCK, MILLWORK, CABINETRY
WORKED ON FURNITURE
ROOF AND WALL FRAMING
BUILDING NEW HOMES
BUILDING MAINTENANCE COMP. WIRING
BUILDING MAINTENANCE COMP. WIRING
ROOFS TO RATHSKELLERS, ALL PHASES
FINISH WORK
OUT HURT
CONCRETE FORM WORK/RENOVATION
NEW RENOVATIONS
NEW RENOVATIONS
NEW RENOVATIONS
REMODELING APARTMENTS/HOMES
INTERIOR RENOVATIONS
INTERIOR RENOVATIONS
INSTALL- CORIAN COUNTERTOPS, KITCHEN, AND BATHS
WEATHERIZATION
REPAIR WATER AND ELECTRIC
FIT UP FOR STORES AND OFFICES
SUB-CONTRACTOR
CARPENTER/ELECTRICIAN
HOME MAINTENANCE
FURNITURE AND REMODELING
HANGING METAL PANELS
REMODELED KITCHEN AND BATHS
Main Activities for Drywall Worker Classification
HANGS SHEET ROCK- DRYWALL
DRYWALL, METAL STUDS, DOOR FRAMES
DRYWALL AND INTERIOR FINISH
ACOUSTIC CEILINGS & DRYWALL/INSULATION
INTERIOR WALLS AND CEILINGS
DRYWALL, CONCRETE, RENOVATIONS
DRYWALL- SETTING DOORBUCKS
D-1
-------
Table D-1. Main Activities and Job Titles for Each Worker Group (continued)
Job Title
Main Activity
Main Activities for Floor Layer Classification
INSTALL FLOORS (CARPET, TILE, VINYL, HARDWOOD, VICTORIAN, PARQUET, MARBLE)
FLOOR SURFACE PREPARATION (SAND, SCRAPE, TEAR UP, FINISH, PREP
UNDERLAYMENT, REMOVAL)
TEACHES FLOOR LAYING
INSTALLING & PAINTING ASTROTURF
CUTTING CARPET AND VINYL
Main Activities for Laborer Classification
Laborer
Laborer
Laborer
Laborer
Laborer
PRE-HUNG DOOR ASSEMBLER
ROOF SPRAYING, SIDING
GUTTING WINDOWS
CLEANS UP
REMODEL NEW GROCERY STORE AND INSTALL NEW COMM FREEZER
DEMOLITION/INSTALLATION
GUTTING INTERIORS, DEMOLITION
NEW RENOVATIONS
INTERIOR RENOVATIONS
LABOR
PLUMBING, ELECTRICAL, DRYWALL, INSULATION
BREAKING DOWN FLOORS AND WALLS
REMOVING AND PLACING SIDING
CLEAN UP-ELECTRICAL WIRING, SHEETROCKING, FIRE/WATER DAMAGE
BRING IN AND TAKE OUT MATERIALS
Main Activities for Other Classification
Mill Worker
CUSTOM WOOD WORKING
COOLING TOWERS
CLEAN RESTROOM AND OTHER CLEANING DUTIES
SHOPWORK, ASSEMBLY , SANDING, GLUING, ETC.
BUILDING STEEL CONVEYOR/BUILDING MACHINES
UNEMPLOYED
PLUMBING AND HEATING
CUTTING STONE TO BE PUT IN PLACE
LINING AND SETTING (BRICK LAYER)
PRINT
DETOX LEAD FROM HOME
TEACHING STUDENTS
Main Activities for Painter Classification
Painter
ROOF SPRAYING, SIDING
CUTTING GRASS, PAINTING
PAINT STRUCTURES, HOUSES
REMODELING/PAINTING
PAINTING, REPAIRING, CLEANING
PAINTING AND PULLING UP CARPET
D-2
-------
Table D-1. Main Activities and Job Titles for Each Worker Group (continued)
Job Title
Main Activity
Main Activities for Supervisor Classification
SUPERVISOR (JOB MANAGEMENT, ESTIMATING), RUN WORK AND WORKERS
INSPECTS CARPENTRY WORK (REMODELING AND NEW WORK)
OFFICE DUTIES
JOB LAYOUT
Main Activities for Window Installer Classification
INSTALLATION OF WINDOWS, DOORS, WALLS, ETC.
WINDOWS- TEAR DOWN AND REPAIR
D-3
-------
Table t>-2. Summary of Responses for Questions Pertaining to R&R Target Activities for Each Worker Group
Variable
Description
Statistic
Target Activities
General
R&R
Carpet
Removal
Window
Replacement
HVAC
Work
Large
Structure
Paint
Removal
Cleanup
UNION CARPENTER
Days performing the activity in
last month
Days performing the activity in
Pre- 1950 housing in last month
Days using a respirator while
performing activity
Number of Weeks spent
performing activity in last year
Number of years spent
performing activity over career
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
Less than 1 Week
1 to 8 Weeks
More than 8 Weeks
25th Percentile
Mean
75th Percentile
5
-1 490(")
22
0
8.65
15
(<0
(c)
(=)
5.0%(d)
25.2%
69.8%
5
12.74
18
0
0.84
1
0
0.36
0
0
0.09
0
87.4%
12.6%
0%
0
3.21
3
0
3.04
5
0
2.35
3
0
0.27
0
45.9%
47.2%
6.9%
1
6.56
10
0
0.38
0
0
0.22
0
0
0.04
0
91 .8%
7.5%
0.6%
0
1.71
1
0
6.42
10
0
3.57
5
0
1.01
0
25.8%
51 .6%
22.6%
2
7.77
13
0
1.84
2
0
0.68
0
0
0.43
0
72.3%
24.5%
3.1%
0
4.13
5
0
6.88
12
0
3.96
5
0
0.79
0
38.4%
49.7%
1 1 .9%
2
8.75
15
(a) 25 percent of all workers responded that they did R&R 5 days or less over the past 30 days.
(b) Mean number of days spent doing R&R in the past 30 days was 14.90 over all workers.
(c) Did not ask this question for general R&R activity.
(d) 5 percent of workers spent less than one week doing general R&R in the last year.
-------
Table D-2. (continued)
Variable
Description
Statistic
Target Activities
General
R&R
Carpet
Removal
Window
Replacement
HVAC
Work
Large
Structure
Paint
Removal
Cleanup
NON-UNION CARPENTER
Days performing the activity
in last month
Days performing the activity
in Pre- 1950 housing in last
month
Days using a respirator while
performing activity
Number of Weeks spent
performing activity in last year
Number of years spent
performing activity over
career
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
Less than 1 Week
1 to 8 Weeks
More than 8 Weeks
25th Percentile
Mean
75th Percentile
17(a)
20.38(b)
24
7
14.91
20
(c)
(<0
(c)
0%(d)
12.4%
87.6%
6
13.26
18
0
1.82
3
0
1.29
2
0
0.25
0
60.0%
36.2%
3.8%
1
5.91
10
2
5.58
9
0
4.19
5
0
0.67
0
15.2%
66.7%
18.1%
3
8.85
14
0
1.68
1
0
1.26
1
0
0.24
0
63.8%
31 .4%
4.8%
1
5.22
7
3
8.70
13
0.5
6.56
10
0
1.16
0
1 1 .4%
50.5%
38.1%
5
9.70
14
1
7.19
12
0
5.51
10
0
0.99
0
19.0%
56.2%
24.8%
3
8.93
15
5
13.68
20
1
9.67
20
0
1.74
0
8.6%
40%
51 .4%
6
12.08
17
(a) 25 percent of all workers responded that they did R&R 17 days or less over the past 30 days.
(b) Mean number of days spent doing R&R in the past 30 days was 20.38 over all workers.
(c) Did not ask this question for general R&R activity.
(d) 0 percent of workers spent less than one week doing general R&R in the last year.
-------
Table D-2. (continued)
Variable
Description
Statistic
Target Activities
General
R&R
Carpet
Removal
Window
Replacement
HVAC
Work
Large
Structure
Paint
Removal
Cleanup
DRYWALL WORKER
Days performing the activity
in last month
Days performing the activity
in Pre- 1950 housing in last
month
Days using a respirator
while performing activity
Number of Weeks spent
performing activity in last
year
Number of years spent
performing activity over
career
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
Less than 1 Week
1 to 8 Weeks
More than 8 Weeks
25th Percentile
Mean
75th Percentile
10(a)
17.84(b)
25
0
11.33
20
(=)
(c)
(<0
0%(d)
14.1%
85.9%
7
13.59
19
0
1
1
0
1.20
0
0
0.06
0
78.1%
21 .9%
0%
0
2.78
2.5
0
2.05
3
0
1.52
1.5
0
0.22
0
57.8%
39.1%
3.1%
1
4.56
5.5
0
0.75
0
0
0.52
0
0
0.45
0
90.6%
9.4%
0%
0
1.61
1
1
8.28
14.5
0
5.53
9
0
1.05
0
26.6%
45.3%
28.1%
2
8
13
0
3.58
3.5
0
2.61
0
0
0.63
0
60.9%
32.8%
6.3%
0
4.34
5.5
2
8.31
15
0
5.83
7.5
0
0.88
0
32.8%
51 .6%
15.6%
1
8.19
12
(a) 25 percent of all workers responded that they did R&R 10 days or less over the past 30 days.
(b) Mean number of days spent doing R&R in the past 30 days was 17.84 over all workers.
(c) Did not ask this question for general R&R activity.
(d) 0 percent of workers spent less than one week doing general R&R in the last year.
-------
Table D-2. (continued)
Variable
Description
Statistic
Target Activities
General
R&R
Carpet
Removal
Window
Replacement
HVAC
Work
Large
Structure
Paint
Removal
Cleanup
FLOOR LAYER
Days performing the activity
in last month
Days performing the activity
in Pre- 1950 housing in last
month
Days using a respirator while
performing activity
Number of Weeks spent
performing activity in last year
Number of years spent
performing activity over
career
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
Less than 1 Week
1 to 8 Weeks
More than 8 Weeks
25th Percentile
Mean
75th Percentile
7
14.79(b)
23
0
5.21
6
(<0
(=)
(=)
13.4%(d)
25.6%
61 .0%
8
15.32
21
0
5.79
10
0
1.93
3
0
0.07
0
18.3%
50.0%
31 .7%
4
11.67
17
0
0.57
0
0
0.45
0
0
0.18
0
91 .5%
7.3%
1.2%
0
1.67
1
0
0.61
0
0
0.32
0
0
0
0
96.3%
1.2%
2.4%
0
1.12
0
0
0.74
0
0
0.24
0
0
0
0
84.1%
14.6%
1.2%
0
1.38
1
0
8.28
20
0
2.0
2
0
1.17
0
42.7%
24.4%
32.9%
0
6.90
12
5
14.53
24
0
3.25
5
0
0.91
0
13.4%
24.4%
62.2%
6
13.01
20
(a) 25 percent of all workers responded that they did R&R 7 days or less over the past 30 days.
(b) Mean number of days spent doing R&R in the past 30 days was 14.79 over all workers.
(c) Did not ask this question for general R&R activity.
(d) 13.4 percent of workers spent less than one week doing general R&R in the last year.
-------
Table D-2. (continued)
Variable
Description
Statistic
Target Activities
General
R&R
Carpet
Removal
Window
Replacement
HVAC
Work
Large
Structure
Paint
Removal
Cleanup
LABORER
Days performing the activity
in last month
Days performing the activity
in Pre- 1950 housing in last
month
Days using a respirator while
performing activity
Number of Weeks spent
performing activity in last year
Number of years spent
performing activity over
career
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
Less than 1 Week
1 to 8 Weeks
More than 8 Weeks
25th Percentile
Mean
75th Percentile
12w
18.02(b)
24.5
5
13.73
20
(=)
(c)
(<0
0%(d)
26.8%
73.2%
5.5
10.38
15
0
4.82
7
0
3.68
4.5
0
2.13
0
46.4%
44.6%
8.9%
1
4.63
6
1.50
6.95
10
0
6.38
10
0
3.71
3
25.0%
53.6%
21 .4%
1
4.93
7
0
3.95
7
0
3.64
5
0
2.45
1.5
62.5%
28.6%
8.9%
0
2.11
2
3
10.04
15
0.50
7.20
13
0
4.50
6.5
17.9%
51 .8%
30.4%
2
5.38
8
0
8.57
14.5
0
7.54
12.5
0
4.30
5
23.2%
55.4%
21 .4%
1
5.92
10
10
16.64
25
3
12.46
20
0
5.27
8
3.6%
39.3%
57.1%
4
8.92
13
C7
00
(a) 25 percent of all workers responded that they did R&R 12 days or less over the past 30 days.
(b) Mean number of days spent doing R&R in the past 30 days was 18.02 over all workers.
(c) Did not ask this question for general R&R activity.
(d) 0 percent of workers spent less than one week doing general R&R in the last year.
-------
Table D-2. (continued)
Variable
Description
Statistic
Target Activities
General
R&R
Carpet
Removal
Window
Replacement
HVAC
Work
Large
Structure
Paint
Removal
Cleanup
PAINTER
Days performing the activity
in last month
Days performing the activity
in Pre- 1950 housing in last
month
Days using a respirator while
performing activity
Number of Weeks spent
performing activity in last year
Number of years spent
performing activity over
career
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
Less than 1 Week
1 to 8 Weeks
More than 8 Weeks
25th Percentile
Mean
75th Percentile
13(a)
18.38(b)
22
10
14.18
20
(<0
(=)
(=)
2.9%(d)
26.5%
70.6%
5
11.12
15
0
2.88
3
0
3.06
2
0
0.97
1
58.8%
32.4%
8.8%
1
3.76
5
0
4.76
5
0
3.97
4
0
1.15
0
41 .2%
44.1%
14.7%
1
4
5
0
0.82
0
0
0.82
0
0
0.50
0
82.4%
14.7%
2.9%
0
1.97
1
0
5.91
10
0
3.65
5
0
2.24
3
35.3%
44.1%
20.6%
1
4.91
8
7
14.09
20
3
10.91
20
0
7.09
15
1 1 .8%
32.4%
55.9%
4
9.44
15
5
13.18
20
3
11.09
20
0
4.82
5
5.9%
38.2%
55.9%
2
8.94
15
C7
VO
(a) 25 percent of all workers responded that they did R&R 13 days or less over the past 30 days.
(b) Mean number of days spent doing R&R in the past 30 days was 18.38 over all workers.
(c) Did not ask this question for general R&R activity.
(d) 2.9 percent of workers spent less than one week doing general R&R in the last year.
-------
Table D-2. (continued)
Variable
Description
Statistic
Target Activities
General
R&R
Carpet
Removal
Window
Replacement
HVAC
Work
Large
Structure
Paint
Removal
Cleanup
SUPERVISOR
Days performing the activity
in last month
Days performing the activity
in Pre- 1950 housing in last
month
Days using a respirator while
performing activity
Number of Weeks spent
performing activity in last year
Number of years spent
performing activity over
career
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
Less than 1 Week
1 to 8 Weeks
More than 8 Weeks
25th Percentile
Mean
75th Percentile
go
16.21(b)
22
0
8.19
14
(<0
(c)
(=)
1 .8%(d)
19.3%
78.9%
9
15.21
20
0
0.71
1
0
0.16
0
0
0
0
75.4%
24.6%
0%
1
4.68
9
0
2.56
3
0
1.07
1
0
0.23
0
45.6%
50.9%
3.5%
1
7.72
14
0
0.46
0
0
0.05
0
0
0.07
0
87.7%
12.3%
0%
0
2.60
3
0
6.84
10
0
3.09
5
0
0.98
0
17.5%
50.9%
31 .6%
3
9.37
15
0
1.60
2
0
0.56
0
0
0.21
0
70.2%
26.3%
3.5%
0
6.70
10
0
5.28
8
0
2.21
2
0
0.35
0
35.1%
45.6%
19.3%
7
12.70
18
C7
H^
o
(a) 25 percent of all workers responded that they did R&R 9 days or less over the past 30 days.
(b) Mean number of days spent doing R&R in the past 30 days was 16.21 over all workers.
(c) Did not ask this question for general R&R activity.
(d) 1.7 percent of workers spent less than one week doing general R&R in the last year.
-------
Table D-2. (continued)
Variable
Description
Statistic
Target Activities
General
R&R
Carpet
Removal
Window
Replacement
HVAC
Work
Large
Structure
Paint
Removal
Cleanup
WINDOW INSTALLER
Days performing the activity in
last month
Days performing the activity in
Pre- 1 950 housing in last
month
Days using a respirator while
performing activity
Number of Weeks spent
performing activity in last year
Number of years spent
performing activity over career
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
Less than 1 Week
1 to 8 Weeks
More than 8 Weeks
25th Percentile
Mean
75th Percentile
21M
22.43(b)
22
20
19.50
22
(=)
(c)
(<0
7.1%(d)
7.1%
85.7%
7
15.14
20
0
2.07
2
0
1.86
2
0
1
0
85.7%
7.1%
7.1%
0
4.43
6
7
14.14
21
5
11.14
16
0
2.29
0
7.1%
35.7%
57.1%
4
9.79
17
0
2.14
0
0
1.93
0
0
1.57
0
85.7%
7.1%
7.1%
0
3
0
2
10.86
21
2
9.57
20
0
1.57
0
28.6%
35.7%
35.7%
5
10.93
17
0
6.71
10
0
6.07
10
0
2.36
0
35.7%
28.6%
35.7%
0
6.29
14
20
19.14
24
8
16.86
22
0
1.64
0
14.3%
21 .4%
64.3%
6
13.14
20
(a) 25 percent of all workers responded that they did R&R 21 days or less over the past 30 days.
(b) Mean number of days spent doing R&R in the past 30 days was 22.43 over all workers.
(c) Did not ask this question for general R&R activity.
(d) 7.1 percent of workers spent less than one week doing general R&R in the last year.
-------
Table D-2. (continued)
Variable
Description
Statistic
Target Activities
General
R&R
Carpet
Removal
Window
Replacement
HVAC
Work
Large
Structure
Paint
Removal
Cleanup
OTHER
Days performing the activity in
last month
Days performing the activity in
Pre- 1 950 housing in last
month
Days using a respirator while
performing activity
Number of Weeks spent
performing activity in last year
Number of years spent
performing activity over career
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
Less than 1 Week
1 to 8 Weeks
More than 8 Weeks
25th Percentile
Mean
75th Percentile
10(a)
17.71*)
25
3
13.57
25
(<0
(c)
(=)
14.3%(d)
21 .4%
64.3%
10
13.50
20
0
3.43
4
0
3.14
2
0
0.64
0
64.3%
28.6%
7.1%
0
2.29
3
0
7.29
15
0
6.14
5
0
1.07
0
35.7%
50.0%
14.3%
1
6.29
10
0
6.14
10
0
4
5
0
1.14
0
57.1%
21 .4%
21 .4%
1
4.64
6
1
9.79
15
0
6.86
10
0
4.93
10
21 .4%
42.9%
35.7%
2
8.21
14
1
9.93
20
0
8.50
15
0
6.14
5
28.6%
50.0%
21 .4%
2
7.92
10
2
13.43
25
0
10.71
20
0
3.64
5
28.6%
28.6%
42.9%
4
12.29
20
C7
I—*
ro
(a) 25 percent of all workers responded that they did R&R 10 days or less over the past 30 days.
(b) Mean number of days spent doing R&R in the past 30 days was 17.71 over all workers.
(c) Did not ask this question for general R&R activity.
(d) 14.3 percent of workers spent less than one week doing general R&R in the last year.
-------
Table D-3. Frequency Tables of Days Activity Performed Versus Days Performed in
pre-1950 Buildings for Each Target Activity
RENOVATION AND REMODELING
Number
of Days
Activity
Pre-1950 Activity
Freq.
Missing
0
1-5
6-10
>10
TOTAL
Missing
0
0
0
0
2
2
0
0
35
26
13
76
150
1-5
0
0
27
17
58
102
6-10
0
0
3
35
38
76
>10
0
1
0
4
250
255
TOTAL
0
36
56
69
424
585
CARPET REMOVAL
Number
of Days
Activity
Pre-1950 Activity
Freq.
Missing
0
1-5
6-10
>10
TOTAL
Missing
1
0
1
0
0
2
0
0
315
82
8
4
409
1-5
0
0
114
17
4
135
6-10
0
0
1
7
10
18
>10
0
0
2
4
15
21
TOTAL
1
315
200
36
33
585
WINDOW REPLACEMENT
Number
of Days
Activity
Pre-1950 Activity
Freq.
Missing
0
1-5
6-10
>10
TOTAL
Missing
0
0
1
1
0
2
0
0
247
60
11
4
322
1-5
0
0
155
11
4
170
6-10
0
0
1
33
9
43
>10
0
0
2
1
45
48
TOTAL
0
247
219
57
62
585
D-13
-------
Table D-3. Frequency Tables of Days Activity Performed Versus Days Performed in pre-
1950 Buildings for Each Target Activity (continued)
HVAC
Number
of Days
Activity
Pre-1950 Activity
Freq.
Missing
0
1-5
6-10
>10
TOTAL
Missing
0
0
1
0
0
1
0
0
457
31
1
0
489
1-5
0
0
59
3
3
65
6-10
0
0
1
12
2
15
>10
0
0
0
0
15
15
TOTAL
0
457
92
16
20
585
LARGE STRUCTURE REMOVAL
Number
of Days
Activity
Pre-1950 Activity
Freq.
Missing
0
1-5
6-10
>10
TOTAL
Missing
0
0
0
1
1
2
0
0
196
51
23
29
299
1-5
0
0
116
14
16
146
6-10
0
0
2
38
8
48
>10
0
0
0
1
89
90
TOTAL
0
196
169
77
143
585
PAINT REMOVAL AND SURFACE PREPARATION
Number
of Days
Activity
Pre-1950 Activity
Freq.
Missing
0
1-5
6-10
>10
TOTAL
Missing
0
0
0
0
2
2
0
0
275
43
16
19
353
1-5
0
0
91
7
17
115
6-10
0
0
1
27
12
40
>10
0
0
0
2
73
75
TOTAL
0
275
135
52
123
585
D-14
-------
Table D-3. Frequency Tables of Days Activity Performed Versus Days Performed in pre-
1950 Buildings for Each Target Activity (continued)
CLEANUP
Number
of Days
Activity
Pre-1950 Activity
Freq.
Missing
0
1-5
6-10
>10
TOTAL
Missing
1
0
0
0
2
3
0
0
97
57
20
36
210
1-5
0
0
111
15
49
175
6-10
0
0
3
27
23
53
>10
0
0
1
1
142
144
TOTAL
1
97
172
63
252
585
D-15
-------
APPENDIX E:
STATISTICAL MODELS TABLE
-------
Table E-l. Summary of Relationships Between Discrete Covariates and Blood-Lead Concentrations
Categorical Covariate
Description (Levels)
Smoking status (No/Yes)
Smoking status (No/Yes, but not
on job/Yes, on job)
Previous training
Room additions (No/Yes)
Room additions
(No/Contracted/Self-performed)
Age of home (Pre-/Post-1950)
Age of home (Pre-1 950/1 950-
1978/Post-1978)
Race (White/Black/Other)
Race (all levels)
Educational level (At most High
School/More than High School)
Educational level (all levels)
Respirator use
(None/Dustmask/Respirator)
Respirator use (None/Respirator)
Non-work activities (hobbies)
Combined
F(a)
20.59
10.71
0.34
4.40
2.35
19.30
14.37
8.16
3.09
10.01
2.50
6.90
4.43
0.26
P-value(b)
<0.001
<0.001
0.562
0.036
0.097
<0.001
<0.001
<0.001
0.005
0.002
0.030
0.001
0.036
0.610
Philadelphia Union
F
11.34
6.70
0.74
0.47
1.32
0.086
0.391
0.124
0.114
0.923
1.88
4.62
2.36
1.07
P-value
<0.001
0.002
0.390
0.494
0.269
0.770
0.677
0.883
0.952
0.338
0.115
0.033
0.097
0.303
Philadelphia
Non-union
F
17.68
8.81
2.32
0.39
1.34
0.25
0.22
0.14
0.28
6.8
4.33
1.05
0.82
0.20
P-value
<0.001
<0.001
0.132
0.532
0.267
0.620
0.801
0.867
0.921
0.011
0.002
0.357
0.367
0.652
St. Louis
Union
F
15.65
8.32
1.84
0.57
0.57
11.58
7.56
11.29
7.48
6.91
1.76
1.82
1.38
3.35
P-value
<0.001
<0.001
0.177
0.452
0.564
<0.001
<0.001
<0.001
<0.001
0.009
0.140
0.165
0.241
0.069
St. Louis
Non-union
F
0.35
1.49
0.72
0.19
0.29
6.79
5.59
1.95
1.38
0.51
0.25
1.93
2.89
0.02
P-value
0.554
0.230
0.400
0.665
0.749
0.010
0.005
0.146
0.250
0.477
0.942
0.149
0.091
0.892
m
(a) F-values represent test statistics for analysis-of-variance of covariate on log-blood concentrations.
(b) P-values measure influence of covariate. P-values < 0.05 indicate significant covariate effect.
-------
Table E-1. (continued)
Categorical Covariate
Description (Levels)
Other occupations
Runs water before drinking
Takes radio to worksite
Changes shoes before going
home
Changes clothes before going
home
Washes before going home
Washes before eating
Eats at worksites
Dirty worksites
Uses water at worksite
Combined
pfa)
6.71
3.78
0.98
0.18
0.22
1.24
0.38
1.78
0.60
0.24
P-value(b)
0.010
0.052
0.403
0.911
0.877
0.293
0.769
0.149
0.616
0.870
Philadelphia
Union
F
0.18
0.64
1.54
1.28
0.39
1.18
0.81
0.23
0.89
0.32
P-value
0.675
0.427
0.204
0.283
0.760
0.319
0.488
0.874
0.449
0.810
Philadelphia
Non-union
F
3.63
0.93
0.17
0.25
1.19
0.94
0.02
0.53
0.10
0.17
P-
value
0.061
0.338
0.919
0.862
0.321
0.426
0.996
0.662
0.959
0.917
St. Louis
Union
F
7.27
3.73
0.81
0.78
0.07
2.05
0.16
0.42
4.54
0.64
P-
value
0.008
0.055
0.490
0.505
0.975
0.110
0.923
0.736
0.005
0.590
St. Louis
Non-union
F
0.69
0.24
0.58
1.49
0.34
0.37
1.50
2.22
0.04
0.78
P-
value
0.409
0.622
0.631
0.221
0.798
0.772
0.216
0.088
0.987
0.505
m
(a) F-values represents test statistics for analysis-of-variance of covariate on log-blood concentrations.
(b) P-values measure influence of covariate. P-values < 0.05 indicate significant covariate effect.
-------
Table E-1. (Continued)
Continuous Covariate
Description
Worker's age
Number of non-work activities
(hobbies)
Number of other occupational
hazards
Days radio taken to worksite
Days change shoes before going
home
Days change clothes before going
home
Days wash before going home
Days wash before eating
Days eat at worksite
Days worksite dirty
Days water available at worksite
Combined
F(a)
29.11
0.32
3.37
1.96
0.46
2.34
0.09
0.10
1.70
1.65
1.65
P-value(b)
<0.001
0.570
0.067
0.162
0.500
0.127
0.761
0.751
0.192
0.200
0.200
Philadelphia Union
F
12.81
0.71
0.21
3.00
3.53
0.002
2.01
2.71
0.59
1.00
1.76
P-value
<0.001
0.400
0.645
0.084
0.062
0.970
0.157
0.101
0.445
0.319
0.186
Philadelphia
Non-union
F
12.68
0.03
0.07
0.37
0.18
3.26
0.06
1.13
0.47
0.54
0.06
P-value
<0.001
0.866
0.799
0.543
0.675
0.075
0.802
0.291
0.500
0.463
0.814
St. Louis
Union
F
12.17
6.75
6.25
1.76
0.0001
0.73
0.07
0.21
0.40
4.63
1.31
P-value
<0.001
0.010
0.013
0.186
0.992
0.394
0.787
0.649
0.531
0.033
0.254
St. Louis
Non-union
F
6.23
0.35
0.001
0.30
3.37
0.52
1.36
1.79
1.15
0.03
0.16
P-value
0.014
0.553
0.971
0.583
0.068
0.474
0.246
0.183
0.286
0.870
0.685
m
(a) F-values represent test statistics for regression F-test of covariate on log blood concentrations.
(b) P-values measure influence of covariate. P-values < 0.05 indicate significant covariate effect.
-------
Table E-2. Summary of Univariate Relationships Between Conduct of Target Activity and Blood-
Lead Concentrations for each Target Activity (Unadjusted)
Target Activity
Renovation and
Remodeling
Carpet Removal
Window Replacement
HVAC Work
Large Structure
Paint Removal
Cleanup
Days of Pre-1950 Activity
€
0.0187
0.0100
0.0177
0.0243
0.0136
0.0219
0.0138
se(€)
0.0027
0.0077
0.0051
0.0083
0.0040
0.0043
0.0031
P-value
<0.001
0.190
<0.001
0.003
<0.001
<0.001
<0.001
Weeks in Last Year
€
0.1155
-0.0388
0.1125
0.0666
0.0630
0.0249
-0.0044
se(€)
0.0215
0.0217
0.0198
0.0259
0.0174
0.0169
0.0175
P-value
<0.001
0.074
<0.001
0.010
<0.001
0.143
0.802
Years of Activity
€
0.0078
-0.0019
0.0124
0.0104
0.0135
0.0063
0.0008
se(€)
0.0032
0.0041
0.0041
0.0054
0.0041
0.0036
0.0033
P-value
0.017
0.646
0.003
0.056
<0.001
0.083
0.803
m
Table E-3. Summary of Univariate Relationships Between Conduct of Target Activity and Blood-
Lead Concentrations for each Target Activity (Adjusted for Ancillary Covariates)
Target Activity
Renovation and
Remodeling
Carpet Removal
Wndow Replacement
HVAC Work
Large Structure
Paint Removal
Cleanup
Days of Pre-1950 Activity
€
0.0135
-0.0096
0.0088
0.0055
0.0070
0.0108
0.0067
se(€)
0.0027
0.0079
0.0051
0.0084
0.0039
0.0044
0.0032
P-value
<0.001
0.227
0.084
0.515
0.073
0.015
0.039
Weeks in Last Year
€
0.1124
-0.0654
0.0854
0.0167
0.0412
-0.0037
-0.0278
se(€)
0.0205
0.0210
0.0194
0.0256
0.0169
0.0165
0.0168
P-value
<0.001
0.002
<0.001
0.514
0.015
0.824
0.099
Years of Activity
€
0.0090
-0.0041
0.0112
0.0066
0.0130
0.0028
0.0009
se(€)
0.0031
0.0040
0.0039
0.0052
0.0039
0.0034
0.0031
P-value
0.004
0.298
0.004
0.204
0.001
0.411
0.768
-------
Table E-4. Summary for Each Target Activity of the Covariate Adjusted Relationship Between
Blood-Lead Concentration and the Combined Effect of All Three Exposure Period
Measures.
Target Activity
Renovation and
Remodeling
Carpet Removal
Window Replacement
HVAC Work
Large Structure
Paint Removal
Cleanup
Days of Pre-1950 Activity
€
0.0093
0.0055
-0.0110
0.0087
0.0027
0.0137
0.0125
se(€)
0.0029
0.0094
0.0064
0.0107
0.0047
0.0054
0.0036
P-value
0.001
0.557
0.086
0.419
0.568
0.011
<0.001
Weeks in Last Year
€
0.0810
-0.0754
0.0979
-0.0177
0.0155
-0.0376
-0.0622
se(€)
0.0219
0.0278
0.0273
0.0369
0.0217
0.0218
0.0196
P-value
<0.001
0.007
<0.001
0.631
0.475
0.085
0.002
Years of Activity
€
0.0068
0.0021
0.0036
0.0071
0.0109
0.0037
0.0016
se(€)
0.0030
0.0045
0.0045
0.0062
0.0043
0.0039
0.0032
P-value
0.026
0.636
0.427
0.254
0.012
0.338
0.620
R2
0.190
0.136
0.156
0.129
0.145
0.135
0.142
m
Note: There are a total of seven models being fitted in this Phase. Each model represents a single target activity and contains
all three measures of potential lead exposure (short term, intermediate, and long term). Thus the Table reads across the
rows.
-------
Table E-5. Summary for Exposure Period of the Covariate Adjusted Relationship Between Blood-Lead
Concentration and the Combined Effect of All Six Target Activities Exposure Variables.
Target Activity
Carpet Removal
Window Replacement
HVAC Work
Large Structure
Paint Removal
Cleanup
R2
Days of Pre-1950 Activity
€
-0.0227
0.0037
0.0045
0.0040
0.0116
0.0023
se(€)
0.0088
0.0062
0.0089
0.0048
0.0051
0.0040
0.145
P-value
0.010
0.553
0.613
0.407
0.024
0.557
Weeks in Last Year
€
-0.0538
0.0876
0.0058
0.0068
0.0021
-0.0293
se(€)
0.0238
0.0232
0.0279
0.0197
0.0183
0.0197
0.172
P-value
0.024
<0.001
0.836
0.732
0.910
0.136
Years of Activity
€
-0.0059
0.0079
-0.0008
0.0110
0.0025
-0.0056
se(€)
0.0050
0.0063
0.0066
0.0060
0.0046
0.0045
0.141
P-value
0.233
0.210
0.898
0.068
0.590
0.217
Note: There are a total of four models being fitted in this Phase. Each model represents a single period of exposure (Short-Term, Mid-Term
and Long-Term) and contains all six target activities. Thus the Table reads down the three columns.
-------
m
Table E-6. Summary for Exposure Period of the Relationship Between Blood-Lead Concentration and the
Combined Effect of All Six Target Activities Exposure Variables, After Adjusting for the
Effects of Covariates and Worker Group
Target Activity
Carpet Removal
Window Replacement
HVAC Work
Large Structure
Paint Removal
Cleanup
R2
Days of Pre-1950 Activity
€
-0.0130
0.0031
0.0080
-0.0018
0.0086
0.0015
se(€)
0.0085
0.0061
0.0087
0.0047
0.0052
0.0039
P-value
0.130
0.608
0.361
0.700
0.099
0.707
0.235
Weeks in Last Year
€
0.0205
0.0683
-0.0059
-0.0335
-0.0040
-0.0080
se(€)
0.0267
0.0245
0.0281
0.0208
0.0188
0.0200
P-value
0.442
0.005
0.834
0.108
0.830
0.691
0.246
Years of Activity
€
0.0107
0.0032
-0.0038
-0.0012
-0.0002
0.0007
se(€)
0.0054
0.0062
0.0064
0.0060
0.0045
0.0044
P-value
0.048
0.605
0.555
0.846
0.959
0.870
0.235
Note: There are a total of four models being fitted in this Stage. Each model represents a single period of exposure (Short-Term, Mid-Term
and Long-Term) and contains all six target activities. Thus the Table reads down the three columns.
-------
Table E-7. Parameter Estimates for Final Model
Variable
Category
Worker Group
Covariates
General R&R
Variable
Union Carpenter
Non-Union Carpenter
Drywall Worker
Floor Layer
Laborer
Other
Painter
Supervisor
Window Installer
Race: Black
Race: Other
Education: Finished High School
Smokes
Smokes While Working
Age of Home: Pre 1950
Age of Home: Pre 1978 (Post 1950)
Performed R&R on Own House
Use of Respirator or a Dustmask
Days of R&R in Pre 1950 Buildings
Weeks of R&R in Last Year
Years of R&R over Career
Effect on
log (Blood Pb)
0.961
0.929
1.198
0.499
0.832
1.024
1.202
0.811
1.058
0.190
-0.052
-0.223
0.162
0.223
0.212
0.134
0.054
0.123
0.0059
0.0685
0.0093
Standard
Error
0.145
0.156
0.159
0.150
0.163
0.209
0.164
0.163
0.213
0.080
0.185
0.104
0.101
0.052
0.069
0.067
0.050
0.051
0.0029
0.0213
0.0029
P-Value
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
0.017
0.780
0.032
0.108
<0.001
0.002
0.045
0.283
0.016
0.041
0.001
0.002
E-8
-------
APPENDIX F:
ADDITIONAL FIGURES
-------
HVAC
Window
Replacement
Carpet
Removal
Large Structure
Removal
Cleanup
Figure F-1. Average Number of Days Spent Conducting Each Target Activity for
Participants Versus Nonparticipants.*
* The maximum possible length of each axis is 20 days. The outer polygon represents the average number of days activity was performed
during the past 30 work days for participants. The inner polygon presents the average number of days activity was performed for non-
participants.
F-1
-------
100
so
60
40
2O
O
O 1—5 6—1O >10
Large Structure Removal
39
27
30
15
18
nl
100
80
60
2O
73
64
&l
n
3434
100
SO
6O
4O
2O
O
1OO
SO
60
40
2O
46
34
37
20 20
23
n
O 1—5 6—1O :»1O
Cleanup
60
46
21
n
26
6
11
n
O 1—5 6—1O >1O
Carpet Removal
O 1-5 6-10
Paint Removal
100
so
60
4O
20
83
I ni ^± z
3
O 1—5 6—1O
HVAC Work
100
so
60
4O
20
O
62
48
36
28
S * 5 8
O 1—5 6—1O >1
Window Replacement
Figure F-2. Days in Last Month Spent Conducting Specific Target Activities for Participants
(shaded) Versus Nonparticipants.
F-2
-------
HVAC
Window
Replacement
HVAC.
Window
.Replacement
Carpet
/Removal
PaintN
Removal\
Large Structure
Removal
Cleanup
Large Structure
Removal
Cleanup
Laborer
Floor Layer
HVAC,
Window
.Replacement
HVAC
Window
Carpet
/Removal
PaintN
RemovaA
Large Structure
Removal
Cleanup
Large Structure
Removal
Cleanup
Union Carpenter
Drywall Worker
Figure F-3. Average Number of Days Spent Conducting Each Target Activity for Each Worker
Group.*
* There are two polygons shown in each graph, and each polygon has six vertices, one for each target activity. Each vertex of the
outer polygon represents the average number of days the activity was performed during the past 30 days, and each vertex of the
inner polygon denotes the average number of days the activity was performed in pre-1950 buildings. The border of the graphs
represent 20 days of work activity. The closer a vertex is to the border, the more that particular target activity was performed.
F-3
-------
HVAC
Window
Replacement
HVAC,
Window
^Replacement
Carpet
^Removal
Paint ^
RemovaA
Large Structure
Removal
Cleanup
Large Structure
Removal
Cleanup
Non-union Carpenter
Supervisor
HVAC
Window
Replacement
HVAC
Window
Replacement
Large Structure
Removal
Cleanup
Large Structure
Removal
Cleanup
Painter
Window Installer
Figure F-3. (continued).*
* There are two polygons shown in each graph, and each polygon has six vertices, one for each target activity. Each vertex of the
outer polygon represents the average number of days the activity was performed during the past 30 days, and each vertex of the
inner polygon denotes the average number of days the activity was performed in pre-1950 buildings. The border of the graphs
represent 20 days of work activity. The closer a vertex is to the border, the more that particular target activity was performed.
F-4
-------
60
50
30
20
10
0.1 2.3 15.9 50.0 84.1 97.7
Cumulative Normal Probability
99.9
Figure F-4. Normal Probability Plot of Blood-Lead Concentrations.*
1 Straight line indicates normality.
100
-------
64
32
18
8
+ +-H-f
0.1 2.3 15.9 50.0 84.1 97.7
Cumulative Normal Probability
99.9
Figure F-5. Semi-log Probability Plot of Blood-Lead Concentrations.*
1 Straight line indicates log-normality.
100
-------
64
32
16-
8
Philadelphia
Union
Philadelphia
Non-union
St. Louis
Union
St. Louis
Non-union
Sampling Frame
Figure F-6. Boxplot of Blood-Lead Concentrations for Each Sampling Frame.
-------
64
32
16
5 8
O 4
I 2
c5 .
Worker Group
Figure F-7. Boxplot of Blood-Lead Concentration for Each Worker Group.
-------
64
32
16
8
<
<
S
<
Philadelphia St. Louis Philadelphia St. Louis
Union Union Non-union Non-union
Figure F-8. Boxplot of Blood-Lead Concentration Within Each Sampling Frame by Race.
-------
64
32
fie
8
Philadelphia
Union
St. Louis
Union
Philadelphia
Non-union
St. Louis
Non-union
Figure F-9. Boxplot of Blood-Lead Concentration Within Each Sampling Frame by Smoking Status.
-------
64
32
^ 16
8
Philadelphia St. Louis Philadelphia St. Louis
Union Union Non-union Non-union
Figure F-10. Boxplot of Blood-Lead Concentration Within Each Sampling Frame by Age of Home.
-------
64
32
16
8
2
NO
YES
NO
YES
NO
YES
NO
YES
Philadelphia St. Louis Philadelphia St. Louis
Union Union Non-union Non-union
Figure F-11. Boxplot of Blood-Lead Concentration Within Each Sampling Frame by Respirator Use.
-------
64
32
8
Philadelphia St. Louis Philadelphia St. Louis
Union Union Non-union Non-union
Figure F-12. Boxplot of Blood-Lead Concentration Within Each Sampling Frame by Educational Level (HS Indicates
High School).
-------
64
32
16
8
2
NO
YES
NO
YES
NO
YES
NO
YES
Philadelphia St. Louis Philadelphia St. Louis
Union Union Non-union Non-union
Figure F-13. Boxplot of Blood-Lead Concentration Within Each Sampling Frame by Incidence of Home
Renovation/Remodeling in Last Year.
-------
UNION CARPENTER
o
o° 8
0
° ° ° o 8
2 I 800 o ° 8 °
1 ° o i n o °
So ° i i o o
i M§° p ° -o
a °o8 ° 0 g | °
I o0° °
O O o O o
o
o
o ° °
o
-s , , , , , -
64 64
32 32
^
\
16 -5 16
o
s
8 1 8
Q
o
4 •§ 4
&
1
2 I 2-
OQ
1 1
NON-UNION CARPENTER
o
0
o
o 8
0 0 g
0 o° go80 °
0 8 0 00 0°
o o° „
O 0 0 °
O O "O^ ©
Q — Q O S ~ Q^
0 ° 0 ° 0°
o 0 oo ° @ Oo
0 ,,0 0 o 0 0
° o
° o o Oo o 0
o o o o ° o
o
8 ° 8
o
0 5 10 15 20 25 30
Days of Pre-1950 R&R Activity
0 5 10 15 20 25 30
Days of Pre-1950 R&R Activity
PAINTER
64
32
16
64
32
16
0 5 10 15 20 25 30
Days of Pre-1950 R&R Activity
WINDOW INSTALLER
0 5 10 15 20 25 30
Days of Pre-1950 R&R Activity
Figure F-14. Modeled Relationship Between Blood-Lead Concentration and Days in Last
Month for Each Job Category of Pre-1950 R&R Activity.
F-15
-------
FLOOR LAYER
g o o
0°e° oo o
§ 8
64 64
32 32
1
16 -5 16
CD
SUPERVISOR
0 5 10 15 20 25 30
Days of Pre-1950 R&R Activity
0 5 10 15 20 25 30
Days of Pre-1950 R&R Activity
DRYWALL WORKER
64 64
32 32
16 -3 16
0 5 10 15 20 25 30
Days of Pre-1950 R&R Activity
LABORER
o o
0 5 10 15 20 25 30
Days of Pre-1950 R&R Activity
Figure F-14. (continued).
F-16
-------
UNION CARPENTER
o
o o
o
8 °
8 | §
0 a
° „ 8 1
8 g § §
S g
OB B
Q " 5
§ o 8
0 ° §
00° 8
0 °
0
00 0
64 64
32 32
^
16 3 16
I
8 I 8
i
O
4 -o 4
03
*P
i
2 "8 2
m
-i ^
NON-UNION CARPENTER
o
g
O a
0 8 1
oo 1
o o o
o 0 - 1
o @
o
0 0 5
o
0 0 8
o
0-8 9-26 >26 0-8 9-26 >26
Weeks of R&R Activity Weeks of R&R Activity
PAINTER
o
° o
o
o
o
o
0 g
° 8 8
g
8 §
o
0 0
o
o
o
64 64
32 32
^
16 3 16
y—
Q
^Q
S
8 | 8
|
Q
4 -g 4
q)
1
2 "8 2
m
1 1
WINDOW INSTALLER
8
o
0
o
o
o
0
o
o
o
8
o
0-8 9-26 >26
Weeks of R&R Activity
0-8 9-26 >26
Weeks of R&R Activity
Figure F-15. Modeled Relationship Between Blood-Lead Concentration and Weeks in Last
Year of General R&R Activity for Each Job Category.
F-17
-------
FLOOR LAYER
o
a
<9 . o
08 8
00° 0
o g
o
§ °
0 °
0 0
0
o
o
1
0
e
i
o
8
s
64 64
32 32
1ft
ID
16
0-8 9-26 >26
Weeks of R&R Activity
SUPERVISOR
0-8 9-26 >26
Weeks of R&R Activity
DRYWALL WORKER
o
o
o
° 8
0 ° g
O D
o 8
o e
o
0
e
8
R
1
@
Q
g
g
0
O
o
0-8 9-26 >26
Weeks of R&R Activity
64 64
32 32
16 3 16
i
8
o
o
8
o
o
o
o
o
0 0
o LABORER
o
o
o
g 0
o
8 9
O 5
1 8
i §
0-8 9-26 >26
Weeks of R&R Activity
Figure F-15. (continued).
F-18
-------
UNION CARPENTER
10 20 30 40
Years of R&R Activity
64 64
32 32
a-
16 -5 16
50
NON-UNION CARPENTER
° ^o
10 20 30 40
Years of R&R Activity
50
RAINIER
10 20 30 40
Years of R&R Activity
64 64
32 32
16 -5 16
WINDOW INSTALLER
50
10 20 30 40 50
Years of R&R Activity
Figure F-16. Modeled Relationship Between Blood-Lead Concentration and Years of
General R&R Activity for Each Job Category.
F-19
-------
FLOOR LAYER
0 Cfa
GO O
10 20 30 40
Years of R&R Activity
64 64
32 32
1
16 -5 16
50
SUPERVISOR
10 20 30 40
Years of R&R Activity
50
DRYWALL WORKER
10 20 30 40
Years of R&R Activity
64 64
32 32
16 -5 16
50
LABORER
10 20 30 40
Years of R&R Activity
50
Figure F-16. (continued).
F-20
-------
APPENDIX G:
HUMAN SUBJECTS APPROVAL AND INFORMED CONSENT
-------
ASSURANCE OF CONFIDENTIALITY OF SURVEY DATA
Statement of Policy
SRA is firmly committed to the principle that the confidentiality of individual data obtained
through SRA's surveys must and shall be protected. The principle holds true whether or not any
specific guarantee of confidentiality was given at time of interview, or whether or not there are
specific contractual obligations to the client. When guarantees have been given or contractual
obligations regarding confidentiality have been entered into, they may impose additional
requirements which are to be strictly adhered to by you.
Procedures for Maintaining Confidentiality
A. You shall sign this assurance of confidentiality. This assurance may be supplemented
by another comparable assurance for a particular project.
B. You shall keep completely confidential the names of respondents and/or study
subjects, all information or opinions collected in the course of conducting work, and
any information about respondents and/or study subjects otherwise learned, directly
or indirectly, during work. You shall exercise reasonable precaution to prevent
access by others to all survey data not in their possession or under their control and
responsibility.
C. Unless specifically instructed otherwise for a particular project, you, upon
encountering a respondent or information pertaining to a respondent whom you
know personally, shall immediately cease the activity and contact your supervisor for
further instructions.
Pledge of Confidentiality
A. I hereby certify that I have carefully read and will cooperate fully with the above
procedures on confidentiality. I will keep confidential all information arising from
surveys concerning individual respondents and/or study subjects to which I gain
access. I will not discuss, disclose, disseminate or provide access to survey data and
identifiers except as specifically authorized by SRA for a particular contract. I will
devote my best efforts to ensure that there is compliance with the required
procedures by any personnel whom I supervise. I understand that violation of this
pledge is sufficient grounds for disciplinary action, including immediate dismissal. I
also understand that violation of the privacy rights of individuals through such
unauthorized discussion, disclosure, dissemination, or access may make me subject to
criminal or civil penalties. I give my personal pledge that I shall abide by this
assurance of confidentiality.
G - 1
-------
B. I shall not, during or after my employment with SRA, for any reason whatsoever,
unless I receive express written permission from an SRA officer, reproduce, copy,
disclose or divulge to anyone, directly or indirectly, any information or knowledge
relating to the past, present or future business operations or internal structure of any
project conducted by SRA.
C. I acknowledge and agree that all files, records, reports, manuals, memoranda,
notebooks, documents, correspondence, and all other information or records and
similar items relating to the business or SRA, whether prepared by me or otherwise
coming into my possession, are, and shall remain, the exclusive property of SRA, and
shall be promptly delivered to SRA upon demand by an SRA Officer.
7. SIGNATURE. I have read and understand and agree to abide by the provisions contained
in this memorandum, and have received a copy of this memorandum which is hereby
acknowledged, I understand that a copy, signed by me, will be placed in my employment
file.
Accepted:
Survey Research Associates, Inc.
By: (SEAL) Name:
Duly Authorized Agent (Please print)
Date: Signature:
SS#:
Address:
G - 2
-------
\
MIDWEST RESEARCH INSTITU1
425 Volker Bouleva
Kansas Chy, Missouri 641
Telephone (816) 753-76i
Telefax (816) 753-84:
June 30, 1994
Mr. Daniel Reinhart
U.S. Environmental Protection Agency
Technical Programs Branch, 7404
East Tower
401 M Street, SW
Washington, DC 20460
Subject: Contract No. 68-DO-01 37, MR! Project No. 9803-A, Work Assignment 4-1 4,
MRI Project No. 9803-A 14.
Dear Mr. Reinhart:
I have reviewed the work to be carried out under the above contract and am in
agreement with the action of the Battelle Human Subjects Review Committee and the
SRA Internal Review Board. These actions grant authorization for Battelie to proceed on
EPAJs Worker Characterizations and Blood-lead Study Under Work Assignment No. 4-14
of MRI contract No. 68-DO-0137 with the EPA.
Please call me at (816) 753-7600, Ext. 1184 if you have any questions.
MRI HSC Assurance Identification No.: M1051
IRB Identification No.: 01
32,
Eugene G. Podrebarac, Ph.D. / Date
Chairman, Human Subjects Committee
/hs
c:' Paul Constant
9803-A 4-1 4
-------
' Project Number B-5 12208-07
Putting Technology To Work
Date June 23, 1994
TO John Menkedick
From David Snediker/HSC
subject EPA/Worker Characterization
and Blood Lead Study
D Snediker/J Greenway
P Brusky
J Manuel
HSC No. = 0010.2
Project No. = G301106-1414
On June 23, 1994, the HSC Committee received and reviewed the pre-screen letters
submitted in response to our follow-up memo dated May 9, 1994. Your compliance
adequately addressed our concerns. Therefore, final authorization has been granted for this
study to proceed. .^
No changes, amendments, or addenda may be made to this protocol or to the informed
consent form without Committee re-review and approval. You, as project manager, or your
designate, are responsible for promptly informing Phyllis Brusky, HSC Secretary,
Ext. 6536, if:
• Adverse consequences are encountered in the course of the project.
• New information becomes available which could change the actual or perceived
risk associated with the project.
• A change in the scope of work alters the nature of human subject participation
in the project.
The Committee must review all changes and new information to determine if the program,
protocol, or informed consent form should be modified, discontinued, or should continue as
approved.
If I can be of further assistance, please contact me at Ext. 4-4633.
DSrjlg
: final.app
-------
Protection of Human Subjects
Assurance Iderrtlflcation/CartJfication/Declaration
(Common F»4«f«i Rui«)
fUqueet Type
ORIGINAL
FOLLOWUP.
EXEMPTION
2. Type of
GRANT CONTRACT
COOPERATIVE AGREEMENT
OTHER-
FEU-OWSHJP
X Appttcacon or r*rqnoM< Mw«Ae«ton No.
,J0.
4. TiO« o< Applicxaen or Acavtty
. Prognun Director. F»ltew. or Oc
cr
t Auuranc* Sttfu* of ffu« Piujtct (Respond to on* of ffM /o*o»ing>
1^^ This A*»ur«no, on fit* with Vr» D*p«rtm«fK oY H*«Jtn «nd Human S*ryio««. eovw« th» actvlty:
A*«wr«nc» id»«Jfic«oon no. U-
IRS i
no.
01
Thk Aasuranea. on ffic w«h
id*ntifie>bon no.
IRB i
I No ssourBnea has b««n fil»d for tfu« project. Thi» institution d*d>rM Bi* it win provide «n Aourmne* snd Ccrtifiotlon of IRB
upon r»qu««t.
_J EnmpO'on Tl^afr: Human njbj^rai ara invo»v»d. but thim activity quaiifia* for aKampbon undar Section 101 (b). paragraph
and apprevi
'. Carcrftcation of IRS fU»i«<» |7taxpond to oo« o/ffM toUa**rrg IF you nrrt *n Assurmncm on tit*)
_J Thi» aeovtty haa bun r»viao«d and approvad by Bv« IRS HI aeeordano wittl tfv» common rui» and any offxr governing r»flui«l>on» or •ubpart* o
June 23, 1994
w
by:
Pud IRS
Q Exp*dit»d FUwww.
_] ThU activny contain* muttipl* profvca. com* of which hava not b*Mi nmi«<»»d. Th« IRS ha* grantad approval on condition ffiat an pm't»cti covw
by ffM common ml* wiS b« favian'id and appravW baton they am initiaiad and that appropriate tumar eaftifieation wU b«
. The official signing b«k» oertfiea that 9te information provided above
i» correct and that, as required, future, reviews wffl be performed and
cercrficalion wfll be provided.
I. Phone No. (Wrffi anM csx>*)
614) 424-4633
12. Fax No. (Wff» area cod*;
(614) 424-7274
L Nan>e of Official
Dr. David K. Snediker
VLx'^AiO
Subjects
1993
.Uxw-c»d for local reproducec
by HHS/PHS.
-------
SRA Survey Research Associates, Inc.
6115 Falls Road. Baltimore. Maiyiand 21209 (410) 377-5660
May 18, 1994
Ron Menton, Ph.D.
Battelle Memorial Institute
505 King Avenue
Columbus, OH 43201-2693
Dear Dr. Menton:
As IRB Chairperson I have reviewed the modified consent form and protocol
that you provided for the study entitled, "Worker Characterization and Blood-
Lead Study" and have determined that this study has now received final
approval.
As with all SRA studies, this study will be subject to an annual IRB review
at the end of next year. We will send you the necessary form for annual
review at the appropriate time. In the meantime, should any changes occur in
your protocol or questionnaire, please inform the IRB. Similarly, the IRB
needs to be notified in the event of any injury or unexpected outcome arising
from this study.
I wish you the best in your study.
Sincerely,
Elizabeth E. Hogue
IRB Chairperson
-------
STUDY ID:
ENVIRONMENTAL PROTECTION AGENCY WORKER CHARACTERIZATION AND
BLOOD-LEAD STUDY OF RENOVATION AND REMODELING WORKERS
INFORMED CONSENT DOCUMENT
The U.S. Environmental Protection Agency (EPA) has contracted with Battelle, Survey
Research Associates (SRA) and Midwest Research Institute (MRI) to study the lead exposure hazard
associated with renovation and remodeling. The objectives of this study are to (1) characterize the
relationship (if any) between renovation and remodeling activities and worker blood-lead levels, and
(2) gather information on the types of work activities and work practices engaged in by renovation
and remodeling workers.
As a participant in this study you will be asked to complete a questionnaire which includes
information relevant to lead exposures on a) work history (both current and long term); b) personal
characteristics and habits related to lead exposures; c) non-work activities; d) medical history related
to lead exposures; and e) previous training or knowledge on lead. After completion of the
questionnaire, a trained and licensed phlebotomist will collect a 1.5 ml blood sample which will later
be analyzed for blood-lead content only. We estimate that it will take approximately one hour for
completion of the questionnaire and collection of the blood sample. What we learn about the
relationship of blood-lead levels to renovation and remodeling activities will help EPA determine what
if any guidance is needed for renovation and remodeling workers.
If you would like a summary of study results and the
result of your blood-level measurement, please check here
There is currently no established minimum acceptable level of lead in the bloodstream. OSHA
has established 50 jig/dl as the blood-lead level at which workers must be removed from jobs having
significant lead exposures. Some states require that a blood-lead measurement in adults greater than
or equal to 25 /ig/dl be reported to the state registry. If your blood-lead measure is greater than or
equal to 25 /xg/dl, we will attempt to notify you of this regardless of whether you requested your
results. We will also pay you $25.00 as compensation for completion of the questionnaire and $25.00
for collection of the blood sample.
Risks: The risk incurred by participation in this study is the risk associated with having a
venous blood sample drawn. The sample will be drawn by a licensed and trained phlebotomist using
standard procedures and precautions including the use of a new sterile syringe and needle for every
blood draw. However, there is a slight risk of local infection and you may experience discomfort,
bruising, and/or bleeding at the site of the needle insertion, or feel dizzy, faint or upset to your
stomach.
Confidentiality; All reasonable efforts will be made to protect the confidentiality of
information obtained from this study in keeping with legal requirements. A participant identification
number will be assigned to your questionnaire and blood sample and will be the only identifier
associated with that information. The file listing participants' names and their participant
identification number will not be released outside of SRA.
If you have any questions or comments regarding this study, or if you experience any
difficulties as a result of participation in this study, please contact: Ms. Beth Moore, Battelle
Memorial Institute, 505 King Avenue, Columbus Ohio 43201. Her phone number is (614) 424-4560.
YOU HAVE THE RIGHT TO WITHDRAW FROM PARTICIPATION IN THIS STUDY
AT ANY TIME WITHOUT PENALTY TO YOUR COMPENSATION.
Battelle/SRA will retain a copy of this Informed Consent Document. A copy of this form will
also be provided to you upon completion of the study.
I consent to participate in this study by completing the associated questionnaire and allowing a
venous bjood sample to be collected.
I, _, UNDERSTAND THE NATURE OF THIS
STUDY AND AGREE TO PARTICIPATE.
SIGNATURE DATE
-------
May 11, 1994
Mr. John Moran
Laborers' International Union of North America
905 16th Street, N.W.
Washington, D.C. 20006-1765
Dear Mr. Moran:
Thank you for returning my call. I appreciate your interest in EPA's Renovation and
Remodeling (R&R) Study even though you are very busy with other matters. Enclosed is a
document that covers in detail the Worker Characterization and Blood-lead Study (WCBS);
one of two field studies that comprise the EPA R&R Study. We will call you again in the
next week to continue our conversation on the WCBS, discuss our needs, and to answer any
questions you may have.
The continued success of EPA's WCBS depends on the input and cooperation of individuals
such as yourself and the Laborer's International Union of North America. Your earlier
comments had a direct impact on revisions made to the questionnaire which was pretested in
April 1994, with the help of the United Brotherhood of Carpenters and Joiners of America.
We are planning to pretest the telephone screening questionnaire and begin constructing our
sampling frames in the next few weeks. In addition, we are anticipating receipt of OMB
approval by the end of this month which means we could be out in the field as early as
June 1, 1994. As we are moving toward implementation of the field study, we are looking
forward to an increased role and participation of the Laborer's International Union of North
America. Please feel free to call me at (614)424-4560 or Ron Menton at (614)424-7165 if
you have any questions about the study.
Sincerely,
Beth E. ,Moore
Statistics and Data
Analysis Systems
BEMimk
Enclosure
-------
June 13, 1994
John Moran, Executive Director
National Health and Safety Fund
Laborer's Union International
Dear John,
We were disappointed not to have been able to meet with you while we were in Washington on
June 6. We do appreciate the time Bill Kijola spent with us to discuss the Worker Characterization
and Blood-lead Study (WCBS).
This study is not looking for any renovation and remodeling jobs. The WCBS is looking to
recruit field workers belonging to the laborer's union to: 1) complete a questionnaire and 2) have a
blood sample taken. Completion of the questionnaire and having a blood sample taken will take
approximately 1 hour of a person's time. Each participant will be compensated $50.00 for
participating.
In our last phone conversation, you had indicated that the local unions would not be
cooperative if we were to call them and ask for a copy of their membership list. Bill also confirmed
this with us. I thought I had understood you to say that you could provide us with the membership
lists in the cities of interest if I sent you a memo staring that as a request. Could you please let me
know as soon as possible if that is true or whether I have misunderstood you. If this is not feasible, I
would ask to get a random sample from the membership lists or work out another compromise.
We have not pursued going through the AGC or other avenues to locate labor union members.
If we need to be go through the AGC or another means to locate labor union members, we would
appreciate any help you can provide. Our time frame for getting started on the study is getting very
close and we need to decide how we are going to recruit workers from your union.
We believe our recruitment success depends on your cooperation and support of the study. If
you do not feel you can help us, we may forego including the laborer's union hi our study.
I look forward to hearing back from you sometime this week to discuss our needs and yours
with respect to the WCBS. You may call me at (614) 424-4560.
Sincerely,
/
Beth E. Moore
Statistics and Data Analysis Systems
cc: Dan Reinhart, EPA
-------
June 24, 1994
Mr. John Moran, Executive Director
laborers' International Union of North America
National Health and Safety Fund
905 16th Street, N.W.
Washington, D.C. 20006-1765
Dear Mr. Moran:
The United States Congress passed the Residential Lead-Based Paint Hazard Reduction Act
in 1992. Part of this act requires the Environmental Protection Agency (EPA) to conduct a
study of lead exposures among renovation and remodeling (R&R) workers, such as those
belonging to the Laborer's Union. Two research firms, Battelle and Midwest Research
Institute, are assisting EPA with the study, referred to as the Worker Characterization and
Blood-lead Study (WCBS). The objectives of the study are to:
1.- Determine the relationship between blood-lead levels and work practices or
activities performed by R&R workers.
2. Determine if blood-lead levels of R&R workers in specific worker groups differ.
3. Gather information on the types of work activities that various categories of
R&R workers engage in, and the personal background and behavioral
information relevant to their potential lead exposures.
EPA will use the results of this study to prepare training and certification guidelines that may
affect R&R workers who are performing activities involving lead-based paint. These
guidelines cannot be properly constructed if this study does not succeed.
A random sample of workers belonging to the Laborers' Union from St. Louis and, most
likely, Philadelphia are needed to participate in this study. Your cooperation in providing to
us the names, addresses and phone numbers of a: least 450 members in St. Louis and 450
members in Philadelphia will assist us in locating participants for the study. We anticipate
needing to contact 450 members from each city in an effort to get 100 to participate.
-------
Mr. John Moran
Laborers' International Union of North America
National Health and Safety Fund
Page 2
A phone call will be made to determine a worker's eligibility for the study. If a worker
meets the criteria of the study (i.e. currently does or has done hands-on renovation work in
homes or buildings built prior to 1950), he will be asked to come to a central location to
complete a questionnaire and have a blood sample taken. This will take approximately 1
hour of his time. He will be compensated $50.00 for participating.
We believe our recruitment success and results of the study depend on your cooperation and
support. I look forward to hearing back from you this week to discuss our needs and yours
with respect to the WCBS. You may call me at (614)424-4560.
Sincerely,
^2®^^J
Beth E< Moore fl
Statistics and Data
Analysis Systems
-------
.
AIRCSTEES
rAHlE BOOKER
Administrator
LABORERS' HEALTH & SAFETY FUND OF NORTH AMERICA
October 17, 1994
Ronald G. Menton, Ph.D.
Statistics and Data Analysis Systems
Battelle
505 King Avenue
Columbus, OH 43201-2693
Dear Mr. Menton:
Your letter of September 13, 1994 to John Moran with a
copy to me is acknowledged. John and I have discussed
the matter of the EPA sponsored R&R Worker
characterization and Blood-lead Study which Battelle is
conducting and your requested participation of members
of the Laborers' International Union of North America
in that study.
The attachment to your letter, the Informed Consent
Document, is one which we have been seeking for some
months. John informs me that both your office and EPA
committed to providing this form to him months ago.
The form does not bear an OMB approval number, a matter
which John discussed with Mr. Reinhart.
We remain concerned about the study protocol and the
fact that you will only "attempt" to get the results of
the individual blood-lead analysis to the participating
worker according to the Consent Document. Further,
there is no indication that you will bear
responsibility for reporting elevated blood-leads to
the state registry. Indeed, earlier discussions on
that matter indicated the intent to select the cities
for the study in states which did not have reporting
requirements.
Accordingly, we do not feel that it is the best
interest of LIUNA members to participate in the study
under the current study protocol. As a result, we will
not recommend participation to the International Union.
Sincerely,
OVATION!
Brian M. McQuade
Executive Director
BMcrSZ
cc: John Moran
HEADQUARTERS. 905-16th Street. NW • Washington, D.C 20006-1765 • (202)628-5465 • Fax:(202)623-2613
-------
Carpenters' District Council of Greater Saint Louis, AFL-CIO
Affiliated with the United
Brotherhood of Qflrpentei
and Joiners of America,
AFL-CIO':.
Carpenters' Building
1401 Hampton Avenue
St. Louis, Mo. 63139-3199
314-644-4800
Terry Nelson
Executive Secretary-Treasurer
Dear Member:
Two research companies, Battelle and MidWest Research Institute, are assisting EPA with a
study, the Worker Characterization and Blood-lead Study (WCBS), that looks at lead exposures
among remodeling field workers. The study will examine the possible relationship between lead
levels in the blood and work practices performed by renovation and remodeling workers. EPA
will use the results of this study to prepare training and certification guidelines that may directly
affect you and your fellow workers.
The United Brotherhood of Carpenters and Joiners of America supports this study. As Executive
Secretary-Treasurer of the St. Louis District Council, I also support this study and strongly
encourage you to participate in the study.
A Battelle representative may call you in the next week to ask you a few brief questions. The
representative may also ask you to schedule a time to come to the Carpenters' Hall to complete
a questionnaire and have a blood sample taken. You will be paid $50.00 for completing the
questionnaire and having a blood sample taken.
We are counting on you to help us collect this important information for EPA. Your participation
will make a difference to union carpenters everywhere. If you have any questions about the
study, call John Egel at Battelle: (314) 993-5234 or toll free (800) 444-5234. Thank you in
advance for your cooperation.
Sincerely and fraternally,
TN/lp
TERRY NELSON
Executive Secretary-Treasurer
-------
; 'SEP is '34 i2-.a?p« I-ETRO DISTRICT COUNCIL
PHONEt
METROPOLITAN DISTRICT COUNCIL
CORrat J^^SS^BJJSiiip^nt FRAMCB J. UVFEY
OfTCS
PHILADELPHIA, PENNSYLVANIA
1809 SPRING GARDEN 8TWEST. WIUO&FH1X r*
22, 1994
Dear Keaber:
Two research companies, Battalia and HidWest
Irastitura^ ar* asfeis-cing ctu Enviranaerrtai »rot»crtion Agaacy with
a study/ vtiieh is! called th» "Vortour gharaot:«rl»a'tien ant^jBloqA**
» to inveatlgatQ Jxow ocposure to
__ _
load aifacts vorJcera in th« Hone 'Renovation field.
study vill stamina the possible relationship batvcion lead
levels in the bl«od and tha vork practicas of our Coitnoil'a
rtnovation and remodeling Carpenters. SPA will use the results of
ttiis study to prepare Training and certification guidelines that
Bay directly affect you and your fellow vorXers in the future.
A. list of our carpenters vho have done this type of vorX Ms
been supplied to the Organisation conducting thia etudy. certain
ocabers vill be selected at random from this list to receive thia
letter.
The General .Office of our United Brother-hood is sponsoring
this Program through our National Health & Safety fund. Our
Jfetropolitan District council also supports it/ eftd we strongly
encouragm you to participate in ths study if you are contacted.
A Battelle repreie&tative may oall you some
ovening; ia, the aeact veek ta »ak yoti a few
ehcurt q^tatioua. Tha representative may also
».ai yon, to case "to tit* oajrpeater*' Snildlztg-^ at
1803 JpriiKj Garden street , ^hiLadeipnia, PX, to
ooaplece :a
-------
:S£P IS '34 12--27PM
DISTRICT COWC1L
P.3
-2-
Xf you *r* selected, v« are counting on you to help tut collect
fchia laapcortant. -indorsation for tha SPA.. Your participation if ill
r«al
-------
10104 OLD OLIVE STREET ROAD • ST. LOUIS, MO 63141-1509 • (314) 994-7700 • FAX (314) 432-7j
home builders association
of greater saint louis
August 22, 1994
Dear HBA Member:
Two research companies, Battelle and MidWest Research Institute, are assisting EPA with a
study, the Worker Characterization and Blood-lead Study (WCBS), that will look at lead
exposures among remodeling field workers. The study will examine the possible relationship
between the possible relationship between blood-lead levels and work practices performed by
renovation and remodeling workers. EPA will use the results of this study to prepare training
and certification guidelines that may affect you and your workers.
The NAHB Remodelors Council supports this study and encourages your cooperation so that
you can have an impact on the EPA's upcoming policy. You can help by taking just a few
minutes to provide the names, addresses and phone numbers of your field workers. Use the
enclosed FIELD WORKER INFORMATION sheet and FAX or mail in the next three days to:
Mr. John Egel
Battelle/Survey Research Associates
401 N. Lindbergh, Ste. 330
St. Louis, MO 63141
FAX: 314-993-5163
The enclosed page, Summary of the WCBS. briefly explains how your remodeling workers
would be involved in the study if they choose to participate. If you are hi a one-person shop,
and would like to participate yourself hi the study, or feel your subcontractors would be
interested hi participating, please call Mr. Egel. If you have any questions or cannot return
the FIELD WORKER INFORMATION sheet in the next three days, please call John Egel at
(314) 993-5234. Thank you for your cooperation.
Sincerely,
'' /(.•' '•"••^ / I \- ^-"/ /'?
Jzmes P. Quiniey^CGR / c
'Chairman, NAHB Remodelors Council
l
-
co yagers o
ARCH • PATRON
~hbo
-------
Summary of the Worker Characterization and Blood-lead Study
Battelle and MidWest Research Institute (MSI) are assisting the United States Environmental
Protection Agency (U.S. EPA) in planning and implementing the Worker Characterization and Blood-
lead Study (WCBS) which is a component of EPA's research on lead exposures associated with
renovation and remodeling (R&R). The WCBS is a targeted field study of lead exposures for three
workers groups: union carpenters, union laborers, and non-union, independent carpenters. The data
collected include: 1) telephone screening questionnaire, 2) self-administered questionnaire, and 3)
blood samples to be analyzed for lead concentration. The results of this study will help EPA to
prepare guidelines for training and certification of R&R workers.
Rather than comprising a nationally representative sample, this study will target these three
worker groups in just two cities: St. Louis and Milwaukee. The two cities were selected because of
support and cooperation of local union leadership, and because a large number of children with
elevated blood-lead levels have been found hi both cities.
The first step in recruiting workers to participate in the study is to contact their employer.
You were randomly selected from a membership list of employers provided to us by your local
NAHB Remodelors Council Chairman. By completing and returning the enclosed EMPLOYEE
INFORMATION sheet, we can begin to randomly select workers to call. Other employers will be
submitting EMPLOYEE INFORMATION sheets to us, so not all employees will be selected and
called.
Employees that are selected will receive a letter in the mail that gives them a brief summary of
the study, informs them that the NAHB and their employer have endorsed the study, and lets them
know that they will be receiving a phone call from a Battelle representative. When the Battelle
representative calls, the employee will be asked to answer a few questions and may also be asked to
participate in the main study. Participation in the main study requires that the employee come to a
central location to complete a questionnaire and have a blood sample taken. Employees are
compensated 550.00 for completing the questionnaire and giving a blood sample.
-------
ROUTE TO;-'
Volume 10, No. 8 April 13, 1994 FAX# 432-7185 HBA Phone 994-7700
St. Louis Selected for Lead Study .... EPA
is working with NAHB in a blood lead level
study to provide statistical data in lead levels
in blood, specifically of those who work in the
remodeling industry .... Data could have a
tremendous impact on rules and regs coming
from EPA .... 100 volunteers from this area
needed for study .... Will involve a
confidential blood test and completion of a
survey .... Volunteers will receive $50 for
participating .... Only requirement is the
participant must have worked on a remodeling
project within 30 days of test.... Preliminary
data show that blood lead levels are lower
than anticipated .... Lack of participation will
mean no data, and tougher regs for
employers and employees in lead abatement
.... Call Jim Kuhn at Consolidated
Construction at 647-9077 if you and/or your
employees can help. HBA staff contact:
Roxanne Radunzel.
-------
BATTELLE
NEWS
RELEASE
Contact:
Robin Yocum
Telephone 614-424-5544
Will Kopp
Telephone 614-424-7984
Celebrating
No. 82-94
November 17, 1994
Years
For Immediate Release
BATTFTT.F. CONDUCTING STUDY IN PHILADELPHIA
TO EXAMINE LEVELS OF LEAD IN REMODELERS
Battelle is conducting a study of home improvement workers and
carpenters in the Philadelphia area to determine if home remodelers have been
exposed to lead.
The study was mandated by Congress in October 1992. Battelle is
conducting the study for the U.S. Environmental Protection Agency.
Battelle is attempting to survey between 225-250 union and non-union
carpenters by Dec. 15. Participants are asked to fill out a questionnaire and
allow a blood sample to be taken. Participants receive $50.
"What we're trying to determine is whether any particular work
activities or practices are associated with elevated blood-lead levels," said
Battelle researcher Ronald Menton, who is heading the study.
Lead poisoning can cause neurological and cognitive development
problems in children. In adults, lead exposure has been linked to high blood
pressure, kidney problems, headaches, fatigue, and stomach problems. Extreme
exposure can result in seizures.
Menton said the survey is being conducted in Philadelphia and St. Louis
because they are older cities with higher documented instances of elevated
blood-lead levels in children. Older urban homes tend to have more lead in
paints, dust, and plumbing.
Carpenters ingest lead through airborne particles in the work place.
Carpenters can reduce the likelihood of suffering from lead exposure, Menton
said, by taking simple precautions, such as:
(MORE)
505 King Avenue
Columbus, Ohio 43201-2693
Fax: (61 4| 424-3889
(614] 424-5544
-------
-2-
• Wearing a respirator
• Wetting down surfaces before working
• Avoiding dry power sanding
• Sealing off work areas
• Washing hands before eating
Although the study involves union and non-union workers, Menton said
the study needs to recruit more non-union workers.
"We've worked with the United Brotherhood of Carpenters and Joiners
and received tremendous cooperation," Menton said. "But, we need to look at
both union and non-union workers.
"If work activities differ between union and non-union workers, then
lead exposure also may differ. Union carpenters tend to be more specialized
than a non-union, residential remodelers. The independent handyman often
performs a variety of jobs in a diverse range of buildings."
Anyone interested in participating in the study should call 800-444-5234.
Battelle serves industry and government by developing, commercializing,
and managing technology. With a wide range of scientific and technical
capabilities, Battelle puts technology to work for clients in 30 countries.
###
-------
2B"
-ST.LDUIS PCS'
BRIEFS
REGION
MISSOURI
LAMBERT RELD INQUIRY ' ;
Tapes Show Cessna Pilot Was Told To Stop
• .The pilot of the Cessna that crossed into the path of
a Trans World Airlines jet atlambert Field last month
acknowledged a flight controller's order to stop but:
kept moving, federal investigators have found. :
' -Investigators for the National.Transportation Safe-
ty:Board discovered the new information, in listening to
taped conversations between the control tower and ;
thfe Cessna pilot, said Alan Pollock, a spokesman for
the board.
. Jn addition, investigators found that the TWA plane,
an' MD-82/ was traveling down Runway 30 Right
around 125 mph, or 25mph faster ffa»T previously
believed, Pollock said.
The Cessna, a charter plane, rolled onto Runway 30
Right about 10:03 pjn. oaNov. 22. The. TWA pilot,'
Rick Carf, swerved to avoid a head-on collision, and
the TWA's wing sliced open the Cessna. The Cessna's
pQot and a passenger were killed. ••
ST. LOUIS COUNTY
Port Authority Makes' Plans For Rfverboat
The SL Louis County Port Authority says it hopes
by mid-April to recommend to the County Council a '
developer and operator of a riverboat gambling facility^j
on the authority's property in Lemay. ([
. The County Council will select the developer-opera^
tor. Council members- got the responsibility when :
voters in November agreed to allow riverboat gam- .
bung in the unincorporated area.
.Vernon "Bud" Schertel, chairman of the Port Au- ,
thority Board of Commissioners, said he hopes a ded-.
sion can be made quickly so the developer could have
niost of the construction season to build a facility. •
The authority plans to send out requests, for qualifi-
cations of.developer-operators this week with replies ;
due by Jan. 6.
ST. PETERS ,:
Hate Crime May Be Civil Rights Issue
The FBI has launched a preliminary investigation to
determine whether any civil rights violations occurred
in St. Peters over the weekend. A pickup was torched
in front of a house in Millwood subdivision, and the
'owner's house was painted with a swastika and the
words "White Power."
• The- victim, who is white, said he believed the
vandalism occurred because his children play with the
children of a black family down the street.
FBI spokesman Michael Roman said Tuesday that
the agency's report would be forwarded to the Justice
Department, which would decide whether to prose-
cute any suspects. •
WELLSTON
Officials Approve Private Detention Facility
Weflston officials, meeting Tuesday night, approved
the SL Louis area's first privately owned and operated
detention center on the site of an abandoned foundry.
It is expected to open in about a year.
Former Missouri LL Gov. Ken Rothman is a partner
in Secured Housing Inc., which will begin construction
of the $6 million facility this spring in the 1400 block of
Ogden Avenue, south of the Mbog Industries plant
The facility win house people awaiting trial on '
misdemeanor federal and state charges or on munici-
pal charges, and others serving sentences of up to one
year.
The Wellstqn City Council voted 5-0 at the end of a
long public hearing that drew about 100 participants.
Four people who live near the site objected.
But council members said the estimated 70 new jobs
and $200,000 in taxes to be paid each year.by the
facflity justified the approval.
^-^===*-
LEAD-EXPOSURE STUDY
fat. Louis Carpenters Sought For Survey
Scientists conducting a leaerai smay 01 ieaa expo-
sure say they need help from carpenters and home-
improvement workers.
Scientists with Battelle, a research company based
in Columbus, Ohio, are surveying union and non-union
carpenters and remodelers in SL Louis and Philadel-
phia as part of a study for the federal Environmental
Protection Agency.
The two cities were chosen because they have a
•history of high blood-lead levels in children.
Battelle researchers want to survey about 250
union and non-union carpenters by Dec. 15, but they
are having difficulty recruiting non-union workers,
Menton said. Participants wfll be asked to fill out a
questionnaire and allow a blood sample to be taken.
Each will receive $50.
The SL Louis portion of the study is run by Survey
Research Associates, a Battelle subsidiary, at 401N.
Lindbergh Boulevard. Anyone interested in participat-
ing can call 1-800-444-5234.
v•••••-: '- -"•ocv°".
•• r-i -> *."~ *•-~ '-'•?•. -'•;
-------
Yes! You Can Make a Difference
St. Louis Selected for Lead Level Study
Soon you will receive a letter from the National Association of Home
Builders and the HBA of Greater St. Louis regarding member
participation in a lead blood level study. The results of this study will
have a great impact in the rules and regulations developed by EPA for
lead abatement. Other studies have indicated that remodeling and
abatement work do not affect a field worker's lead level.
j Your help is needed to make the study a success.
i
Watch your mail for a letter from Jim Quinley, Chairman, NAHB
Remodelors Council.
8/16/94
-------
bcr-
rr: nur'ic BL)lL.utr,3 i-
REMODELORS COUNCIL
Sf. LOLHS Selected to
Take Part in Lead Study
by Roxanne Radunrel
Staff V.P7A330c!atIon Services
St Louis has been selected
for lead based paint testing
by the EPA, to be conducted
by Batteile, a national research
firm. The goal of the study is to
show the correlation between ren-
ovation and remodeling field work
and actual human exposure to
lead. The NAHB Remodelors
Council Is fully supporting the
testing.
In October 1992, the U.S.
Congress passed the Residential
Lead-Based Paint Hazard
Reduction Act This act falls
under the jurisdiction of the
Environmental Protection Agency
(EPA) to set the guidelines and
any requirements for working with
lead based paint in remodeling
and renovation projects.
As part of the legislation, the EPA
is required to conduct a study of
lead exposures among remodeling
field workers. It is hoped that this
research will indicate the lead
based paint risk to remodeling/
rehab field workers. The EPA
attempted to locate other existing
sources of data for the study, but
none addressed the lead hazard
(in terms of a health effect) asso-
ciated with remodeling and reno-
vation activity. A painting trade
association has conducted a
smaller sampling which did not
indicate a lead problem with field
workers. However, the format and
siza of the sampling Is not ade-
quate for the EPA's needs.
VbbntMiv far CMstouu In April w**f m***»
*m they tcnfil p«*A>9 hudJuMttl ptlnt tnm
Vttir ttmflgniUd /MAM.
The objectives of the St Louis
study are to:
^Determine the relationship
between blood-lead levels and
work practices or activities per-
formed by remodeling and
renovation field workers (after
including potential confounding
factors such as personal hob-
bies with lead exposure, etc.).
^Determine tf blood-lead levels of
remodeling and renovation
workers in specific worker
groups differ.
^ Gather information on the types
of work activities and work
practices In which various cate-
gories cf remodeling and reno-
vation workers engage.
Participants will be asked to com-
plete a questionnaire and submit
to a blood test The individual
results of the testing will be kept
confidential. Ail participants will
receive $50 for the testing. All
types of field workers, illustrating a
range of lead exposure, are need-
ed for the testing. The results of
the test will determine the extent
to which field workers are
exposed to lead. The EPA will
then use the results to set forth
guidelines or determine guidelines
that are not needed.
. Member participation Is needed
to make this research success-
ful. Without definite data, the
EPA will develop worker guide-
lines without regard to the true
Impact of lead In the remodeling
and renovation industry. Your level
of participation is limited to giving
the Batteile research company a
list of your field workers' names,
addresses, and phone numbers
for follow-up. The testing is done
on the employees' time.
You should have already received
in the mail a letter briefly outlining
the program and asking for a list
of employees. If you have not yet
responded, please do so by con-
tacting John Egel at Batteile,
Survey Research Associates at
(3U) 993-5234. Egel will be able
to answer any questions you may
have about the survey process, or
assist you in getting a worker
Information sheet prepared.
The final results of the survey will
be known within 6 months of sur-
vey completion, with the EPA
using the results In the final
report It is up to St Louis to
make a difference In the out*
come of this report by partici-
pating In the survey research I
D
Sept '94 Builder News 17
-------
50272-101
REPORT DOCUMENTATION
PASE
1. REPORT NO.
EPA 747-R-96-006
3. Recipient's Accession No.
4. Title and Subtitle
Lead Exposure Associated with Renovation and Remodeling Activities: Worker
Characterization and Blood-Lead Study
5. Report Date
May 1997
6.
7. Author(s)
Menton, R.G., Menkedick, J.R., Schulman, J., Strauss, W.J., Egel, J.N., Lowry, L.
8. Performing Organization Rept. No.
9. Performing Organization Name and Address
Battelle Memorial Institute
505 King Avenue
Columbus, Ohio 43201-2693
10. Project/Task/Work Unit No.
and Midwest Research Institute
425 Volker Boulevard
Kansas City, Missouri 64110
11. Contract(C) or Srant(S) No.
(C) 68-D5-0008
(S)
12. Sponsoring Organization Name and Address
U.S. Environmental Protection Agency
Office of Pollution Prevention and Toxics
401 M Street, S.W.
Washington, D.C. 20460
13. Type of Report & Period Covered
Technical Report
14.
15. Supplementary Notes
In addition to the authors listed above, the following staff were major contributors to the study: Patsy Henderson, Beth Burkhart,
Halsey Boyd, and Karen Christie of Battelle; and Paul Constant and Jack Balsinger of MRI.
16. Abstract (Limit 200 words)
The U.S. Environmental Protection Agency, in response to the Residential Lead Based Paint Hazard Reduction Act of 1992 (Title X)
conducted a study of lead exposure associated with renovation and remodeling (RAR) activities. This report presents the results of one of
the principle data collection efforts of the study: the Worker Characterization and Blood-Lead Study (WCBS). The WCBS collected blood-
lead measurements and questionnaire information from 585 RAR workers in two cities. Questionnaire results include the type and frequency
of renovation work conducted, worker protection and clean-up methods, and lead training. Blood-lead concentrations were generally low, with
only 7 out of 581 workers above 25 ^/g/dL and only one out of 581 above 40 ^/g/dL. The geometric mean blood-lead concentration for all
workers was 4.5 ^/g/dL. Statistical models were fit to the data to assess the effect of worker groups, the amount of RAR activity
conducted, and significant covariates.
17. Document Analysis
a. Descriptors
Lead-based paint, lead hazards, renovation and remodeling, blood-lead measurements, work habits, worker certification, worker
survey
b. Identifiers/open-ended Terms
Lead, renovation and remodeling, worker exposure, blood-lead, Title X
c. COSATI Field/Sroup
18. Availability Statement
Release Unlimited
19. Security Class (This Report)
Unclassified
20. Security Class (This Page)
Unclassified
21. No. of Pages
187
22. Price
(See ANSI-239.18)
OPTIONAL FORM 272 (4-77)
(Formerly NTIS-35)
Department of Commerce
------- |