United States
Environmental Protection
Agency
Office of Pollution
Prevention and Toxics
Washington. DC 20460
EPA 747-R-99-001
March 1999
Lead Exposure Associated with
Renovation and Remodeling
Activities: Phase IV
Worker Characterization and Blood-Lead Study of
R&R Workers Who Specialize in Renovation of
Old or Historic Homes
-------
EPA 747-R-99-001
March 1999
LEAD EXPOSURE ASSOCIATED WITH
RENOVATION AND REMODELING ACTIVITIES: PHASE IV
WORKER CHARACTERIZATION AND BLOOD-LEAD STUDY
OF R&R WORKERS WHO SPECIALIZE IN
RENOVATION OF OLD OR HISTORIC HOMES
Prepared by
Battelle
505 King Avenue
Columbus, Ohio 43201-2693
for
Technical Branch
National Program Chemicals Division
Office of Pollution Prevention and Toxics
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
U.S. EPA DISCLAIMER
This report was prepared under contract to an agency of the United States
Government. Neither the United States Government nor any of its employees,
contractors, subcontractors, or their employees makes any warranty, expressed or
implied, or assumes any legal liability or responsibility for any third party use of or
the results of such use of any information, apparatus, product, or process disclosed
in this report, or represents that its use by such third party would not infringe on
privately owned rights.
Publication of the data in this document does not signify that the contents
necessarily reflect the joint or separate views and policies of each sponsoring agency.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
BATTELLE DISCLAIMER
This is a report of research performed for the United States Government by
Battelle. Because of the uncertainties inherent in experimental or research work,
Battelle assumes no responsibility or liability for any consequences of use, misuse,
inability to use, or reliance upon the information contained herein, beyond any
express obligations embodied in the governing written agreement between Battelle
and the United States Government.
-------
ACKNOWLEDGMENTS
This study was funded and managed by the U.S. Environmental Protection Agency. The
study was conducted by Battelle Memorial Institute under contract to the U. S. Environmental
Protection Agency. Each organization's responsibilities are listed below.
Battelle Memorial Institute (Battelle)
Battelle was responsible for designing the study, recruiting participants, collecting worker
and homeowner questionnaire data and blood samples, creating and maintaining data bases,
conducting statistical analysis, and producing the final report.
U.S. Environmental Protection Agency (EPA)
The U. S. Environmental Protection Agency was responsible for oversight in developing
the study plan, managing and coordinating the study, and reviewing and editing this report. The
EPA Work Assignment Manager was Dan Reinhart. The EPA Project Officers were Jill Hacker
and Sineta Wooten.
iii
-------
This page intentionally left blank.
IV
-------
TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS iii
EXECUTIVE SUMMARY ix
1.0 INTRODUCTION AND BACKGROUND 1
1.1 STUDY OBJECTIVES 2
1.2 LIMITATIONS OF THE STUDY 3
1.3 PEER REVIEW 3
2.0 STUDY DESIGN AND METHODS 5
2.1 SURVEY DESIGN 5
2.1.1 Target Population 5
2.1.2 Sample Size Determinations 6
2.1.3 Recruitment 7
2.1.3.1 Recruitment of Workers 7
2.1.3.2 Recruitment of Homeowners 8
2.1.4 Eligibility Criteria 9
2.2 SAMPLE COLLECTION 10
2.2.1 Questionnaires 10
2.2.1.1 Telephone Screener Questionnaires 10
2.2.1.2 Main Study Questionnaires 11
2.2.2 Blood Samples 11
2.3 DATA MANAGEMENT 12
2.4 STATISTICAL ANALYSIS 13
2.4.1 Construction of Variables 13
2.4.1.1 Construction of Exposure Variables 13
2.4.1.2 Construction of Other Variables 16
2.4.2 Descriptive Statistics 16
2.4.3 Exploratory Data Analysis for Ancillary Variables 16
2.4.4 Statistical Models 16
2.4.4.1 Blood-Lead Concentrations 16
2.4.4.2 Relationships Between Target Activities and Blood-Lead
Concentrations 17
3.0 RESULTS 20
3.1 RECRUITMENT RESULTS AND FIELD EXPERIENCES 20
3.1.1 Recruitment of Respondents 20
3.1.2 Field Experiences 22
3.1.2.1 Field Experiences in Charleston, South Carolina 22
3.1.2.2 Field Experiences in Savannah, Georgia 23*
3.1.2.3 Field Experiences in Baltimore, Maryland 24
3.2 QUALITY CONTROL FOR BLOOD-LEAD 25
3.3 TELEPHONE SCREENING RESULTS 27
3.4 PARTICIPANT CHARACTERIZATION 32
3.4.1 Demographics 32
3.4.2 R&R Work History 34
3.4.2.1 R&R Work History for Professional R&R Workers 34
-------
TABLE OF CONTENTS
(continued)
Page
3.4.2.2 R&R Work History for Homeowners 36
3.4.3 R&R Work Practices 36
3.5 OVERVIEW OF BLOOD-LEAD CONCENTRATIONS 40
3.5.1 Descriptive Statistics for the Blood-Lead Concentrations of Study
Participants 40
3.5.2 Comparison to the WCBS 43
3.5.2.1 Demographic Characteristics 43
3.5.2.2 Work History 45
3.5.2.3 Blood-Lead Concentrations 48
3.5.3 Comparison to NHANES III, PHASE 2 49
3.5.3.1 Blood-Lead Concentrations 50
3.6 STATISTICAL MODELING RESULTS 52
3.6.1 Statistical Model Building 52
3.6.1.1 Selection of Ancillary Covariates 52
3.6.2 Blood-Lead Concentration Comparisons Using Covariate-Adjusted
Model Results 54
3.6.3 Relationship Between R&R Target Activities and Participant Blood-Lead
Concentrations 56
3.6.3.1 Relationship Between R&R Target Activities and Worker Blood-Lead
Concentrations 56
3.6.3.2 Relationship Between R&R Target Activities and Homeowner
Blood-Lead Concentrations 58
3.6.4 Overall Statistical Models for Participant Blood-Lead Concentrations 60
3.6.4.1 Overall Statistical Model for Worker Blood-Lead Concentrations .... 60
3.6.4.2 Overall Statistical Model for Homeowner Blood-Lead
Concentrations 65
4.0 SUMMARY AND CONCLUSIONS 68
APPENDIX A QUESTIONNAIRES A-1
APPENDIX B ADDITIONAL QUESTIONNAIRE TABLES B-1
APPENDIX C STATISTICAL MODEL TABLES C-1
APPENDIX D ADDITIONAL FIGURES D-1
APPENDIX E RECRUITMENT MATERIALS, HUMAN SUBJECTS APPROVAL, AND
INFORMED CONSENT E-1
LIST OF TABLES
Table 1. Summary of Recruitment Results by Study Group and City 21
Table 2. Summary of Laboratory Quality Control and CDC Performance Samples 26
Table 3. Variance Component Estimates for Blood-Lead Concentrations 27
vi
-------
TABLE OF CONTENTS
(continued)
Page
Table 4. Reasons for Worker Ineligibility 28
Table 5. Reasons for Homeowner Ineligibility 28
Table 6. Summary of Demographic Information for Participants and Non-Participants 29
Table 7. Summary of General R&R Work History of Participants and Non-Participants 30
Table 8. Summary of Demographic Data 33
Table 9. Summary of Worker Responses for Questions Pertaining to R&R Target Activities . . 35
Table 10. Summary of Homeowner Responses for Questions Pertaining to R&R Target
Activities 37
Table 11. Summary of Responses for Questions Pertaining to Work Practices 38
Table 12. 95 Percent Confidence Intervals for Geometric Mean of Blood-Lead Concentrations
by City and Study Group 42
Table 13. Selected Percentiles and Occurrence of Elevated Blood-Lead Levels by City and
Study Group 42
Table 14. Characteristics of Study Participants with Blood-Lead Levels Above 25 //g/dL 44
Table 15. Descriptive Statistics of Blood-Lead Concentrations by Worker Group 45
Table 16. Geometric Means and 95 Percent Confidence Intervals for Blood-Lead
Concentrations of Participants from the WCBS and the Current Study 48
Table 17. Occurrence of Elevated Blood-Lead Levels for Participants in the Current Study
(Phase IV) and Workers in the WCBS (Phase II) 49
Table 18. Geometric Means and 95 Percent Confidence Intervals for Blood-Lead Concentrations
of Participants in the Current Study and Adults in NHANES III, Phase 2 50
Table 19. Occurrence of Elevated Blood-Lead Levels for Participants in the Current Study
(Phase IV) and in the Adult U.S. Population 51
Table 20. Odds Ratios Comparing Odds of Blood-Lead Concentrations Greater Than 25 //g/dL
Among Study Participants to the General U.S. Population 51
Table 21. Geometric Mean Blood-Lead Concentration and Log (Standard Error) for Each
Level of the Ancillary Covariates 54
Table 22. Geometric Mean and 95 Percent Confidence Intervals for Each Study Group and
by Worker Group Based Upon Covariate-Adjusted Models 55
Table 23. Predicted Changes in Worker Blood-Lead Concentrations Associated with
10 Days of Work in Pre-1940 Homes 57
Table 24. Predicted Changes in Homeowner Blood-Lead Concentrations Associated With
Changes in the Number of Days Spent Performing a Target Activity in Their
Pre-1940 Home 59
Table 25. Predicted Changes in Homeowner Blood-Lead Concentrations Associated With
Changes in the Number of Hours Spent Performing a Target Activity in Their
Pre-1940 Home 60
Table 26. General F-Tests for Model Terms on Worker Blood-Lead Concentrations, After
Adjusting for the Effects of Covariates and Other Terms Included in the Model .... 62
Table 27. Predicted Worker Blood-Lead Concentrations Associated with Low, Medium, and
High Exposure Indices for Each Worker Group 64
Table 28. Predicted Homeowner Blood-Lead Concentrations Associated with Low, Medium,
and High Exposure Indices Using General R&R 67
VII
-------
TABLE OF CONTENTS
(continued)
Page
LIST OF FIGURES
Figure 1. Paradigm for Fining Statistical Models to Worker Blood-Lead Concentrations 19
Figure 2. Paradigm for Fitting Statistical Models to Homeowner Blood-Lead Concentrations . . 19
Figure 3. Percent of Participating and Non-Participating Workers Performing Target
Activities 31
Figure 4. Percent of Participating and Non-Participating Homeowners Performing Target
Activities 32
Figure 5. Histogram of Worker Blood-Lead Concentration (Semi-Logarithmic Scale) 41
Figure 6. Histogram of Homeowner Blood-Lead Concentration (Semi-Logarithmic Scale) .... 41
Figure 7. Comparison of Short-Term Work History for Workers in the Current Study
(Phase IV) to Those in the WCBS (Phase II) 46
Figure 8. Comparison of Mid- and Long-Term Work History for Workers in the Current Study
(Phase IV) to Those in the WCBS (Phase II) 47
Figure 9. 95 Percent Confidence Intervals for Geometric Mean of Blood-Lead Concentrations
for Each Worker Group and Study Group Based Upon Covariate-Adjusted Models . . 55
Figure 10. Predicted Increase in Homeowner Blood-Lead Concentrations as a Function of Hours
and Weeks Spent Performing Paint Removal/Surface Preparation 66
viii
-------
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 activities (R&R Study). Information obtained from the
R&R Study will be used to help determine which groups of workers require training,
certification, or educational materials because of the potential lead exposure to themselves or to
others resulting from the R&R activities that they perform. Three phases of the R&R Study,
each comprising a separate and distinct data collection effort, have been completed. They
include:
• Phase I. Environmental Field Sampling Study (EFSSX The EFSS was conducted to
measure the airborne (breathing zone) lead levels and lead levels in settled dust
resulting from several renovation and remodeling work activities.
• Phase n. Worker Characterization and Blood-Lead Study (WCBSX TheWCBSwas
designed to assess the relationship between the conduct of R&R activities and lead
exposure to the R&R workers who engaged in those activities. The WCBS collected
questionnaire information and a blood sample for analysis from each of 581 R&R
professionals.
• Phase HI. Wisconsin Childhood Blood-Lead Study. The Wisconsin Childhood
Blood-Lead Study was a retrospective case-control study designed to systematically
examine the association between R&R activities and elevated blood-lead levels
0> 10 |ig/dL) among children.
This report presents the fourth Phase in this series of data collection activities. Phase IV closely
resembles Phase n (WCBS) in design and functions as an extension of the earlier study.
Whereas Phase n explored lead exposure among general R&R workers, Phase IV focused on
individuals who worked extensively in older (pre-1940) or historic homes where the risk of lead
exposure associated with R&R work is considered to be especially high. In addition to R&R
professionals, homeowners of older or historic homes who reside in their homes while
performing R&R activities themselves were included in Phase IV.
Study participants were recruited using several approaches. Newspaper advertising
produced the greatest number of study subjects. In-person recruitment and personal referrals
were also effective methods for identifying participants. In total, questionnaire information and
blood samples were collected from 243 participants (161 workers and 82 homeowners) in three
cities: Charleston, SC; Savannah, GA; and Baltimore, MD.
As was the case with Phase n, environmental samples were not collected as part of
Phase IV. There were several reasons for this decision. Both phases were designed to be
minimally intrusive to potential participants. Collection of environmental samples would have
reduced the participation rates for these two studies, and clearly would have substantially
increased the time and cost of conducting them.
ix
-------
The questionnaire results indicated that
1. Workers spent more time performing R&R than homeowners (on average 24 days and
14 days in the past 30 days respectively).
2. Both workers and homeowners spent time performing a variety of R&R activities.
For both groups, a large amount of time was spent performing large structure removal
and paint removal/surface preparation.
3. Despite the OSHA Lead in Construction Standard, only 23.7 percent of workers
reported using a respirator in the last 30 days. Similarly, only 16 percent of
homeowners used a respirator in the last 30 days. Homeowners and workers in every
work group used a dust mask more frequently than any type of respirator
(44.9 percent of workers and 29.6 percent of homeowners reported using dust masks).
4. Seventy-six percent of workers had not received any lead exposure training, and
67 percent of workers and 62 percent of homeowners had not received any
educational materials on lead hazards (homeowners were not asked about training).
5. Over 75 percent of workers and homeowners reported using dry sanding/scraping to
remove paint. Roughly 41 percent of workers reported using chemical stripping
(37 percent of homeowners) and 32 percent of workers (31 percent of homeowners)
reported using burning/torching/heat gun to remove paint.
The geometric mean blood-lead concentrations of study participants were 5.73 ug/dL for
workers and 4.45 ug/dL for homeowners. Forty-nine out of the 243 study participants
(20.2 percent) had blood-lead concentrations greater than 10 ug/dL. Several participants
(2.9 percent) had blood-lead concentrations greater than 25 ug/dL, and three participants (all
workers) had blood-lead concentrations above 40 u£/dL.
Overall, the blood-lead data indicate that Phase IV study participants were more highly
exposed to lead than the general R&R workers from Phase n (i.e., workers not necessarily
working in historic or older homes). The geometric mean blood-lead concentration for workers
in this study (Phase IV) was significantly higher than the geometric mean blood-lead
concentration of workers in the WCBS (Phase H). On the other hand, the geometric mean
blood-lead level for homeowners in Phase IV was not significantly different from that of workers
in the WCBS. Perhaps most important, a significantly larger percentage of all participants
(workers and homeowners) in Phase IV had elevated blood-lead levels than did workers in the
WCBS. In general, the percentage of Phase IV participants with blood-lead levels above
10 ng/dL, 15 ug/dL, and 25 ug/dL was about twice as high as the corresponding percentages for
workers in the WCBS.
Statistical models were developed and fitted to the worker and homeowner data. These
models were used to investigate the relationship between blood-lead concentrations and potential
lead exposure associated with specific R&R activities. These models indicate that there is a
significant relationship between the conduct of certain R&R activities and blood-lead
-------
concentrations. For homeowners, paint removal/surface preparation was the single target activity
that explained the most variability in blood-lead concentrations. For workers, the models
indicated that worker blood-lead concentrations were most strongly associated with the
combined effects of the following: the number of days spent performing cleanup, the number of
weeks spent performing paint removal/surface preparation, the number of weeks spent
performing carpet removal, and the number of years in the R&R career where some time was
spent replacing window or door casements. The total number of hours in the last 30 days and the
number of weeks in the last 12 months that the homeowner spent performing paint
removal/surface preparation were related to increased blood-lead concentrations. The number of
hours and weeks spent performing general R&R were also related to increased blood-lead levels
in homeowners.
xi
-------
This page intentionally left blank.
XII
-------
1.0 INTRODUCTION AND BACKGROUND
To address potential lead exposure associated with R&R, the U.S. Congress directed the
U.S. Environmental Protection Agency (EPA) Administrator to conduct a study of lead exposure
associated with R&R activities (R&R Study). The R&R 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). Paragraph (2) of Section 402 (c)
states:
"The Administrator shall conduct a study of the extent to which persons engaged
in various types of renovation and remodeling activities in target housing, public
buildings constructed before 1978, and commercial buildings are exposed to lead
in the conduct of such activities or disturb lead and create a lead-based paint
hazard on a regular or occasional basis."
The overall objectives of the R&R study were to
• Determine the extent to which persons engaged in various types of R&R activities are
exposed to lead; and
• 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.
Previous results of the R&R Study were derived from three principal data collection
efforts: Phase I: Environmental Field Sampling Study1 (EFSS); Phase H: Worker
Characterization and Blood-Lead Study2 (WCBS); and Phase ffl: Wisconsin Childhood
Blood-Lead Study3. The EFSS was conducted to measure the airborne lead levels and lead levels
in settled dust resulting from various renovation and remodeling work. The WCBS was designed
to collect data and information that could be used to assess the relationship between R&R
activities (termed 'Target Activities") and lead exposure to the R&R workers conducting these
activities. Questionnaire information on work history, work habits, specific work activities, etc.,
was obtained for 585 workers in the WCBS, and blood samples were collected from 581 of the
585 workers. Because children may be the population most sensitive to lead exposures from
R&R activities, additional data were needed to assess the impact of the conduct of R&R
activities on household occupants. Phase HI was added to the overall R&R study with the
U.S. Environmental Protection Agency, "Lead Exposure Associated with Renovation and Remodeling
Activities: Environmental Field Sampling Study, Volume I: Technical Report; Volume n
Appendices," EPA Report No. EPA 747-R-96-007, May 1997.
U.S. Environmental Protection Agency, "Lead Exposure Associated with Renovation and Remodeling
Activities: Worker Characterization and Blood-Lead Study," EPA Report No. EPA 747-R-96-006,
May 1997.
3 U.S. Environmental Protection Agency, "Lead Exposure Associated with Renovation and Remodeling
Activities: Wisconsin Childhood Blood-Lead Study," EPA Report No. EPA 747-R-99-002,1999.
1
-------
objective of determining the impact of residential R&R work on child occupant's blood-lead
levels.
Although there were significant differences among worker groups identified in the
WCBS, the average blood-lead concentrations for the sampled workers were well below
10 ng/dL (the geometric mean blood-lead concentration among all workers was 4.5 fig/dL).
However, because such a diverse group of R&R workers was sampled, it is possible that one or
more subgroups of high-risk R&R workers were not well represented in the WCBS. This study,
Worker Characterization and Blood-Lead Study of R&R Workers Who Specialize in Renovation
of Historic Homes (WCBS-HH), is a follow-up to the WCBS. It is a focused examination of
those workers who routinely perform R&R activities in homes where there is likely to be a high
risk of exposure to lead. In particular, this study focused on workers who conduct R&R
activities in historic/older4 homes likely to contain lead-based paint. In addition, homeowners of
older or historic homes that reside in their homes while performing R&R activities themselves
may also be at high risk for lead exposure. Information on the renovation activities conducted by
homeowners was also collected.
1.1 STUDY OBJECTIVES
The overall objective of this study was to conduct a focused study of workers and
homeowners who are believed to have the greatest potential for exposure to lead-based paint
because they routinely perform R&R activities in older homes. These homes are considered
high-risk because they are likely to contain large amounts of lead-based paint which is often
disturbed by R&R. Additionally, because these are older homes, fine dust and debris containing
lead may have accumulated behind woodwork, plaster, and other architectural components and
may be released as a result of the R&R activities.
The specific objectives of this study were to
1. Determine the relationship between blood-lead concentrations and work practices or
activities performed by workers or homeowners in old or historic homes, after
controlling for potential confounding factors.
2. Determine if the blood-lead concentrations of workers in specific worker groups
differ after adjusting for potential confounding factors.
3. Determine if the blood-lead concentrations of homeowners differ from those of
workers.
4. Gather additional information on the types of work activities and work practices that
workers and homeowners performing R&R conduct.
Throughout the remainder of this report, the term "old" or "older" will be used to indicate homes built
prior to 1940.
-------
5. Compare the blood-lead results of workers and homeowners specializing in R&R in
high-risk homes with those observed among general R&R workers (WCBS).
6. Determine if the blood-lead concentrations of persons routinely performing R&R
activities in high-risk homes differ from those of the general U.S. population.
1.2 LIMITATIONS OF THE STUDY
There are several limitations to this study, all of which may have an impact on the
analyses and conclusions presented in this report. First, though attempts were made to recruit
specific workers, this study is predominantly a 'Volunteer" study. That is, workers and
homeowners were not selected with fixed probabilities from a pre-determined sampling frame.
Thus, the respondents participating in this study may have characteristics that differ from those
that are not apt to volunteer. For example, if respondents that volunteer are more conscientious
workers or have an increased awareness of the dangers of lead paint, they may take precautions
that non-respondents do not.
Another limitation of this study is that the information collected is based upon recall of
selected activities performed at various times throughout the last month, year, or throughout the
lifetime of the respondent. Therefore, there is some inherent degree of recall error. However,
every attempt was made to design questionnaires that would facilitate accurate recall and
minimize the associated recall errors.
Finally, this study was conducted in cities where the majority of the housing stock is
known to contain lead-based paint. However, since environmental samples were not collected in
this study, there is no guarantee that workers or homeowners were actually performing tasks on
surfaces that contained lead-based paint. This lack of knowledge is a potential confounding
component to the analysis in that the inclusion of information from respondents that are not
exposed to lead-paint could result in the distortion of the relationships between blood-lead levels
and R&R activities. For example, consider a homeowner who reported performing extensive
R&R in a historic home that, unbeknownst to the homeowner, has been previously abated. In
this case, the blood-lead levels of this homeowner may not be related in any way to the amount
or nature of the R&R activities performed. Thus, including this homeowner's blood and
questionnaire information would cause the positive relationship between R&R and blood-lead
levels to be underestimated.
1.3 PEER REVIEW
This report 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.
Several comments were made on the sampling methodology and statistical analysis of the
data. No changes were made to the methodology or analysis as a result of these comments but in
many instances additional text or tables were added to clarify the interpretation of the results.
For example, one reviewer expressed concern with the eligibility criteria for workers and
-------
homeowners. As a result of comments from the reviewers, Section 2.1.4 was added to discuss
the rationale for requiring study participants to have a history of routinely performing R&R
activities in older homes over an extended period so as to focus the study on chronically exposed
workers rather than workers engaged in intense activities of short duration. Additional text was
added to Section 2.1.4 in response to a comment to clarify differences in selection criteria for
professional workers and homeowners. Another reviewer commented on the lack of discussion
on specific methods of clean-up and the ramifications that this has on the analysis. As a result,
additional text was added to discuss these limitations in the study design and the ramifications on
the study results. In response to one reviewer's comment, text was added to discuss the impact
of not sampling the worker's environment to confirm that participants were actually working
with lead-based paint.
In a few cases, suggested changes or recommendations by reviewers did not ultimately
result in substantial changes to the interpretation or the discussion contained in this report.
However, these recommendations and suggestions were carefully investigated, and served to
improve the report through innumerable small changes in either the methodology or in the
phrasing of the discussion. One reviewer suggested changes to the methods used to fit the
statistical models. However, it was determined that the suggested changes would result in
comparisons that were less statistically powerful and therefore, the suggestion was not
implemented. Other concerns included the appropriateness of modeling job/task predictors of
historical exposure based on blood-lead concentrations, the lack of specific questions related to
time indices for performing the R&R activities, and the classification of workers into groups.
Several responses were received that just commented upon the findings, compared them
to previous studies, or suggested areas where language or tables could be clarified or emphasis
added. Two reviewers commented upon the negative association between carpet removal and
blood-lead concentrations. Additional text was added in Section 3.6.3.1 discussing the various
interpretations of this result. Two reviewers also felt that more detailed data should be given
concerning the type of respirator used by the participants; additional discussion was included and
the number of categories in Table 11 was expanded to provide more information relative to this
topic. Discussion of the relationship between clean-up method and R&R activities and its
implications for the results presented in this report was also included in Section 3.6.3.1.
Additional text was added to further highlight the differences in paint removal work practices
between homeowners and workers at the suggestion of a reviewer. Table 14 was expanded to
provide more information on respirator use, work practices and education. At the suggestion of a
reviewer, Table 20, and corresponding text, was changed to include odds ratios comparing the
odds of a blood-lead concentration greater than 25 ug/dL for study participants to the general
population.
EPA has established a public record for the peer review under administrative record
AR-210. 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.
-------
2.0 STUDY DESIGN AND METHODS
This study (WCBS-HH) was a focused follow-up to the WCBS. As expected in a follow-
up study, the study design was based upon the experiences acquired in the WCBS, and the design
for this follow-up study is similar to that used in the WCBS. However, several modifications
were made to the design of the WCBS to reflect field experiences and so that resource and
schedule limitations could be met.
This study involved a targeted survey of two groups of persons engaged in renovation and
remodeling activities (professional workers and homeowners who were renovating their own
home) in three cities (Charleston, South Carolina; Savannah, Georgia; and Baltimore, Maryland).
The data collected included:
1. Questionnaire data that were used to characterize the workers and homeowners and to
understand differences in blood-lead concentrations
2. Blood samples that were analyzed for lead.
Information for this study was collected in two stages. In the first stage, workers and
homeowners from each city were recruited into the study. Brief screening questionnaires
(separate questionnaires for workers and homeowners) were administered to potential
participants over the telephone to determine eligibility and collect preliminary information on
targeted work activities. The second stage involved collecting blood samples and questionnaire
data from the eligible participants recruited in the first stage.
This section presents the overall design of the study including the survey design, sample
collection, data management, and statistical methods.
2.1 SURVEY DESIGN
Components of the WCBS-HH study design included defining the target population,
selecting the sampling locations, sample size determinations, and recruiting the targeted number
of workers and homeowners in the survey.
2.1.1 Target Population
The target population for this study consisted of two different groups of persons engaged
in renovation and remodeling in historic homes or homes built before 1940:
• Professional R&R workers
• Homeowners that perform a large amount of R&R in their home.
Homeowners were included in this study because it was believed that they perform a
wide variety of renovation and remodeling activities in then" historic or older homes and are
potentially exposed to lead as a result of these activities.
-------
Similarly, professional R&R workers, both union and non-union, who perform R&R in
historic or older homes were included in the study. However, initial investigations indicated that
a large portion of commercial R&R work performed in historic homes is being conducted by
employees of independent, non-union contractors. Therefore, unlike Phase n of the R&R Study
(WCBS), this study did not specifically target workers based upon union membership.
Workers and homeowners were targeted in three cities: Charleston, SC; Savannah, GA;
and Baltimore, MD. These cities were selected because they fulfilled several key criteria
including:
• Potential Respondent Base: In order to fulfill the sample size requirements, each city
area needed to have a sufficient number of workers and homeowners that perform
R&R activities in historic or older homes on a routine basis. Therefore, each city
needed to have a large number of historic or older homes. Local historic preservation
societies and other local organizations in several cities were contacted in an effort to
determine cities that have a large population of historic homes.
• Presence of Local Organizations: The presence of local organizations, such as a local
historic preservation society, was considered to be a key component in identifying
potential respondents.
• Geographic Location: In order for the study to be cost effective, it was desirable to
have on-site data collection support in or near the sampled city.
Several other cities including: Boston, Chicago, New Orleans, San Francisco, and
Cleveland were considered for this study, but were eliminated because they did not fulfill all
three of the above criteria.
2.1.2 Sample Size Determinations
The targeted sample sizes were developed to allow an assessment of whether workers (or
homeowners) mat perform R&R activities in historic homes have elevated blood-lead levels as a
result of their R&R activities. Moreover, the resulting sample size represents the minimum
sample size necessary to detect a 2.5-fold difference between the geometric mean blood-lead
levels observed in the general U.S. population and blood-lead levels of workers that routinely
perform R&R activities in historic homes or homeowners that perform R&R activities in their
historic home while retaining residency.
The following assumptions were made to derive the minimum sample size required:
• Statistical hypothesis tests were assumed to be conducted at the 5 percent significance
level to examine whether the blood-lead levels of workers and homeowners were
significantly higher than the national average. Also, these tests were assumed to be
conducted with at least 90 percent power. That is, the Type I and n error rates were
assumed fixed at 0.05 and sO.10, respectively.
-------
• The analysis was assumed to consist of a general linear model with the
log-transformed blood-lead as the response variable since blood-lead data are
generally assumed to follow a log-normal distribution.
• The log standard deviation of the blood-lead levels was assumed to be twice that
observed in the WCBS. These respondents were expected to have higher blood-lead
levels than workers in the WCBS. Since blood-lead is assumed to follow a
log-normal distribution, higher expected mean levels would also result in more
variability.
Based upon these assumptions, the minimum sample size was calculated to be 40 professional
R&R workers, and 40 R&R homeowners in each city area. For the purposes of determining
sample size estimates, Charleston, SC, and Savannah, GA, were considered as one city area
while Baltimore, MD, was considered to be a second city area. Therefore, the total target sample
size of 160 individuals was set for the study (80 professional R&R workers and 80 R&R
homeowners).
2.1.3 Recruitment
With professional workers, many contractors who have listings or advertisements
indicating that they specialize in R&R of historic homes spend only a small fraction of their time
actually working in historic homes. Similarly, much of the hands-on work may be completed by
subcontractors who spend a large portion of time working in newer construction. Also, as was
experienced in the WCBS, large contracting companies are reluctant to allow their workers to
participate in a study of this type. Homeowners often do not want to participate in this type of
study because they are performing R&R work without a permit.
A similar, multi-pronged approach was used to recruit both workers and homeowners.
Section 2.1.3.1 presents the recruitment approach used for workers and Section 2.1.3.2 presents
the recruitment approach for homeowners.
2.1.3.1 Recruitment of Workers. Initial investigations indicated that a large portion
of the work performed in historic homes is done by employees of independent, non-union
contractors. Therefore, membership lists in unions such as the United Brotherhood of Carpenters
that were employed in the WCBS, were not used as a sampling frame. Instead, a three-pronged
approach that consisted of (1) newspaper, magazine, and radio advertisements; (2) development
of contractor lists; and (3) referrals and other networking activities, was used to identify potential
respondents. The following activities were conducted to recruit workers:
• Advertisements. Newspaper advertisements were the most successful method for
recruitment of eligible workers; both as a primary and as a secondary recruitment
method (i.e., a worker was referred to the study by someone who saw an
advertisement). Advertisements were placed in wide circulation daily newspapers
and smaller community-based newspapers in each city. Examples of the newspaper
advertisements are presented in Appendix E.
-------
Approximately 60 15-second radio advertisements were also run on local stations in
Charleston, SC, and Savannah, GA, to recruit potential study subjects. These
advertisements identified the project as a study on renovation and remodeling and
provided a phone number that interested persons could call.
• Media Coverage. A written press release was prepared and sent to all media in each
city (see Appendix E). A video press release was prepared and sent along with the
written press release to all television stations in each city's broadcast area. Two
television stations in Savannah, GA, covered the study.
• Networking and Outreach. Posters were put in local hardware stores and lumberyards
(see Appendix E). Local neighborhood associations were contacted, and posters
about the study were distributed to their members. Other local associations such as
historical societies were contacted and asked to recruit workers into the study.
In each city, several in-person visits were made to neighborhoods where there was
observable renovation and remodeling being performed. During these visits,
preprinted fliers were distributed to the workers and they were encouraged to
participate in the study.
• Referrals. Referrals were also highly effective in recruiting additional participants.
To encourage eligible participants to provide referrals, a $10 cash incentive was given
to participants for each R&R worker (or homeowner) they referred to the study. This
incentive was only given if the person referred was subsequently determined to be
eligible and participated in the study. Eligible participants were asked to provide
referrals during a telephone interview (screener) and were again reminded while
participating in their data collection session.
• Contact List. Lists of non-union contractors were developed through listings in local
phone books, advertisements, and by contacting local organizations, such as historic
preservation societies, landmark commissions, and architectural firms. An attempt
was then made to contact and enroll all workers on the list. Although this method
was more time consuming and only a handful of participants were identified through
this approach, it allowed for the inclusion of participants that might otherwise have
been omitted from the study (e.g., persons who work by themselves and do not read
the paper).
Of these activities, newspaper advertisements, word-of-mouth referrals, and in-person
recruitment were the three most successful methods for obtaining eligible worker participants.
2.1.3.2 Recruitment of Homeowners. Recruitment of homeowners that perform
R&R work in their own historic/older home while retaining residency was conducted in a fashion
similar to that described in Section 2.1.3.1. Again, a multi-pronged approach was used.
Newspaper advertisements were the primary method for recruitment of eligible participants and
separate, targeted advertisements were developed to recruit this group of participants (see
Appendix E).
-------
Although news advertisements and identification of R&R homeowners through local
organizations (e.g., neighborhood associations, historical societies, etc.) were successful in
recruiting a number of homeowners, other recruitment activities were conducted to ensure that
the target sample size was reached. For example, fliers and posters were prepared and distributed
to local hardware stores and supermarkets and radio advertisements were developed and aired on
various media. However, these recruitment activities were labor intensive and were used only
after it appeared that newspaper advertisements alone would not be successful in reaching the
target sample size.
2.1.4 Eligibility Criteria
As in any study, the eligibility criteria are a function of both the study objectives and
restraints on resources available to conduct the study. Further, the eligibility criteria needed to
account for differences in the nature of exposure from R&R activities. For example, workers can
perform activities for a brief (less than 10 days) but intense period, which may result in an
elevated blood-lead level due to the intense exposure. This situation, however, could be
investigated using data from the WCBS data (where days spent performing R&R were not used
as an eligibility criterion). What could not be investigated using the WCBS data was the
relationship between routinely (over many days) performing R&R activities in high-risk
residences and blood-lead levels (only 41 percent of the respondents performed R&R work in
pre-1950 homes and on-average WCBS respondents only spent approximately 10 days working
in homes built before 19505). Therefore, this study was focused on examining the relationship
between performing R&R activities in "worst-case" exposure scenarios where workers routinely
day-after-day, spend the majority of their time performing R&R activities in high-risk, older
homes. Given mis focus, the information that was already collected in the WCBS, and the
resource constraints for the study, the eligibility criteria were established so that we focused on
the chronically exposed workers.
The eligibility criteria were different for homeowners and workers because the type of
exposure for homeowners due to conducting R&R differs from that of professional R&R
workers. First, R&R workers tend to work for 8-hour periods during the business week while
homeowners tend to perform R&R as a "second job" on nights and weekends. Second,
homeowners have the potential for additional exposure because they are living in the home while
they are performing R&R. Thus, we utilized different eligibility criteria for homeowners
because we wanted to capture both types of exposure. A criterion based solely on the number of
days in the last 30 days that the homeowner performed the activity may account for secondary
exposure due to living in the home, but may not account for the intensity of the activities being
performed. Conversely, using criteria based solely on the intensity of the activity (hours in the
last 30 days) may not account for exposure due to living in the home. We established the
specific eligibility criteria for homeowners based upon the results of a pilot test of the
questionnaire.
The WCBS only collected detailed information on (he number of days worked in all residences and in
pre-1950 residences.
-------
Ideally, we hoped to measure worker's exposure in the same fashion as homeowners
(i.e., in terms of both days and hours). However, it proved unrealistic to have R&R workers
determine, even in a general sense, how many hours they spent performing a particular R&R
activity.
2.2 SAMPLE COLLECTION
2.2.1 Questionnaires
One set of questionnaires was used to obtain information such as detailed work history,
personal characteristics, and general work practices from professional R&R workers and another
set was used to obtain similar information from homeowners that conduct R&R activities in their
own historic/older home. Each set of questionnaires consisted of a telephone administered
screener and a main self-administered questionnaire (S AQ). All four questionnaires are
presented in Appendix A.
The worker questionnaires were modifications of those used in the WCBS and had
essentially been pre-tested. However, the homeowner questionnaires were necessarily more
complex. A pre-test of the homeowner questionnaires was conducted with homeowners to
determine whether the flow of the questionnaire caused confusion or item non-response and to
estimate the amount of time needed to complete the screener and main questionnaire.
In all four questionnaires, most questions were worded with pre-coded responses to avoid
ambiguous answers to open-ended questions. This also minimizes the potential for information
bias, ensures consistency in the respondents* answers, and facilitates data editing, cleaning,
coding, and analysis. The few questions that were not amendable to closed-ended responses
were left open and categorized retroactively.
2.2.1.1 Telephone Screener Questionnaires. Questionnaire information was
collected on all potential participants using a telephone interview (screener) as part of the
recruitment process. Specifically, the telephone interview was used to:
1. Determine Eligibility. To be eligible for the study, a R&R worker had to meet four
criteria: (a) perform R&R for a living, (b) conduct "hands-on" R&R work, (c) have
worked more than 10 days (in the last 30 days) in historic homes or homes built
before 1940, and (d) have worked more than 9 weeks (in the last 12 months) in
historic homes or homes built before 1940.
Homeowners were considered to be eligible for the study if they met all of the
following seven criteria: (a) did not perform R&R for a living, (b) had made room
additions or renovations to their home in the last 12 months, (c) lived in the house
while the R&R was being performed, (d) lived in a home built before 1940, (e) did a
large portion of the "hands-on" R&R work themselves, (f) did more that 20 hours of
R&R work in their home in the last 30 days, and (g) did some R&R work in their
home in at least 9 weeks (in the last 12 months).
10
-------
2. Assess Potential Selection Bias. Information from the screener allowed for a
comparison between basic demographic characteristics and targeted activities for
those workers and homeowners that participated in the study and those that refused to
participate.
3. Schedule Appointments and Collect Referral Information. Once the eligibility of a
participant had been determined, an appointment for further data collection was
scheduled for eligible participants. Both eligible and ineligible participants were
asked at the end of the telephone screener to provide referrals for the study.
Each telephone interview took approximately five minutes to complete.
In some cases, the respondent did not complete a telephone screener prior to arrival at the
data collection facility. In these cases, a screener, identical to the telephone screener, was
administered in-person to determine eligibility of these "walk-in" participants.
2.2.1.2 Main Study Questionnaires. Among participants that were eligible for the
study, detailed information on work history, personal characteristics, and general work practices
was obtained during a data collection session at a collection facility. A central location (either a
hospital or health department building) was used in each city as the data collection facility.
Three, one-night data collection sessions were held in Charleston, SC, at Roper Hospital on
4/21/97,5/12/97, and 5/28/97. Two one-night data collection sessions were held in
Savannah, GA, at the Public Health Department on 5/14/97 and 5/29/97. Three data collection
sessions were held in Baltimore, MD, at Johns Hopkins University (Bayview Campus) on
6/26/97 and 7/15/97-7/16/97. During the data collection sessions, detailed work history, personal
characteristics, and information on general work practices were collected using a
self-administered questionnaire (SAQ), which took approximately 30 minutes to complete. A
trained interviewer was present during all of the data collection sessions to answer questions
concerning specific questions in the questionnaire or to administer the entire questionnaire, if
necessary.
A field review of this questionnaire was performed before the respondent left the facility
in an effort to minimize item non-response. Respondents that satisfactorily completed the SAQ
received $25.
2.2.2 Blood Samples
Two trained and licensed phlebotomists collected a venous blood sample from eligible
study participants immediately following the completion of the SAQ. A duplicate blood draw
was collected from approximately 15 percent of the respondents for quality assurance purposes.
Respondents for whom a successful blood sample was collected received an additional $25.
The protocol for collecting, storing, and shipping the blood samples is described in
"Quality Assurance Project Plan (QAPjP) for the R&R Worker Characterization and Blood-Lead
Study," July 8,1994.
11
-------
2.3 DATA MANAGEMENT
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 of personal information obtained from study
participants. 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 any documents not
received from the field.
2. Data Editing and Coding. All data were subject to 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.
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.
12
-------
2.4 STATISTICAL ANALYSIS
The statistical analysis included several preliminary steps, including constructing
variables, calculating descriptive statistics, and exploratory data analysis. Statistical models were
then fit to the data to meet the study objectives listed in Section 1.1. 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
for workers: exposure during the previous 30 days, exposure during the past 12 months, and
historical exposure. For homeowners, only two time periods were investigated: exposure during
the previous 30 days and exposure during the past 12 months. All statistical analyses were
performed using the SAS® computing system (Version 6.12).
2.4.1 Construction of 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.
The following two sections discuss the two types of variables constructed prior to
performing the statistical analyses: variables related to potential lead exposure from conducting
R&R work (Section 2.4.1.1), and variables related to demographic characteristics and other
sources of exposure (Section 2.4.1.2). The constructed variables were used to assess the effects
of target activities on blood-lead concentration.
2.4.1.1 Construction of Exposure Variables. Measures of potential lead exposure to
professional workers resulting from conducting R&R work were constructed for three exposure
periods: last 30 days, last 12 months, and the entire career. The exposure measures were
constructed for each target activity (large structure removal, paint removal and surface
preparation, window replacement, carpet removal, and cleanup) and for conducting R&R work in
general.
For each specific target activity, the potential lead exposure variables for workers were
constructed from the following questions:
Short-term: In the last 30 days, how many days did you work on the target
(last 30 days) activity?
In the last 30 days, how many days did you work on the target
activity in homes or buildings built before 1940?
13
-------
Mid-term: Altogether in the past 12 months, how many weeks did you
(last 12 months) work on the target activity?
(0) None
(1) < 1 Week
(2) 1-4 Weeks
(3) 5-8 Weeks
(4) 9-12 Weeks
(5) 13-26 Weeks
(6) > 26 Weeks
Long-term: Think about all the years you have done renovation and
(entire career) remodeling. In how many of these years did you work on the
target activity at least some of the time?
In addition, one objective of this study was to determine if specific groups of workers are
more exposed than others due to the nature of their work. Therefore, each worker was assigned
to a specific worker group based on their response to the following question: What is your
current job title and what are your main activities at work? The main activities were 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. In all, four worker groups were
defined: Carpenter, Laborer, Painter, and Other. Table B-l in Appendix B contains a listing of
the main activities and job titles for each worker group.
There were two questions related to short-term exposure of workers: (1) In the
last 30 days, how many days did you spend doing any kind of R&R work? and (2) In the
last 30 days, how many days did you spend doing any kind of R&R work in historic homes or
homes built before 1940? Therefore, an effort was made to determine which was most strongly
related to. worker blood-lead concentration. For each target activity, relationships were examined
between worker blood-lead concentration and the number of days the target activity was
conducted, and the number of days conducted in pre-1940 houses. Based on plots and univariate
regressions, the number of days an activity was performed in homes built before 1940 was
selected as the measure of short-term exposure.
Measures of potential lead exposure to homeowners resulting from conducting R&R were
constructed for two exposure periods: last 30 days and the last 12 months6. As with the workers,
exposure measures were constructed for each target activity and for conducting R&R work in
general.
Information corresponding to the third exposure period (entire career) was not collected for
homeowners because of the intermittent nature of renovation performed by homeowners and
difficulties of recall.
14
-------
For each specific target activity, four potential lead exposure variables for homeowners
were constructed from the following questions:
Short-term: In the last 30 days, how many days did you work on the target
(last 30 days) activity in your historic or pre-1940 home?
On a typical day within the last 30 days, when you performed
the activity, about how many hours did you perform the
work?
Mid-term: During the past 12 months, in how many weeks did you
(last 12 months) spend any time performing the activity?
In a typical week in the past 12 months when you performed
the activity, about how many days did you perform the work?
The two questions related to short-term exposure of homeowners as a result of
conducting R&R activities were used to determine which type of exposure was most strongly
related to blood-lead concentration (duration of R&R project or hours of exposure). For each
target activity, relationships were examined between homeowner blood-lead concentration and
the number of days the target activity was conducted by the homeowner in their home, and the
total number of hours spent by the homeowner conducting the activity in the last 30 days
(calculated by multiplying the number of days by the hours spent on a typical day). This
distinction is important because it is believed that homeowners are thought to be exposed not
only by the actual work, but also by living in the home while the work is being performed. The
number of days in the last 30 days the homeowner performed the activity may account for
secondary exposure due to living in the home but may not account for the intensity of the
activities being performed. Conversely, the total number of hours spent performing the activity
may account for exposure due to performing the activity but may not account for the duration of
the R&R project. For example, consider two homeowners: the first performing some R&R every
day in the last 30 days but for only one hour a day (total of 30 hours), the second performing
R&R eight hours a day for one consecutive week (total of 56 hours). Univariate regressions and
plots did not provide insight into which measure was more related to homeowner blood-lead
concentrations. Therefore, the statistical analyses were conducted twice, once for each measure.
There were also two questions related to mid-term exposure of homeowners performing
R&R activities in their home: weeks in the last 12 months or days in the last 12 months
(calculated by multiplying weeks in the last 12 months by the number of days in an average week
where the R&R activity was performed). Again, an effort was made to determine which of the
two measures was the most related to homeowner blood-lead concentration. As with the
measures associated with short-term exposure, univariate regressions and plots did not provide
insight into determining which measure was more related to homeowner blood-lead
concentrations and the statistical analyses were performed once for each measure.
15
-------
2.4.1.2 Construction of Other Variables. Demographic variables such as age,
gender, race/ethnicity, and level of education were constructed from the questionnaire responses.
The questionnaires 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 or homeowner responded positively to one or more
of those questions then they were 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.
2.4.2 Descriptive Statistics
Descriptive statistics were calculated for each variable constructed in Section 2.4.1. For
continuous variables such as age, means and percentiles were calculated. For categorical
variables such as gender, the percentage of respondents in each category was calculated.
Geometric means and standard errors of blood-lead concentrations for each study group
and city were calculated. Additional tables were prepared to assess the variability in measured
blood-lead concentrations between duplicate blood samples and among duplicate chemical
analyses.
2.4.3 Exploratory Data Analysis for Ancillary Variables
Exploratory data analyses were performed to assess the relationships between blood-lead
concentration and various ancillary variables describing demographics, work 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 separately
for workers and homeowners and for each sampling frame.
2.4.4 Statistical Models
2.4.4.1 Blood-Lead Concentrations. The July 8,1994, WCBS QAPjP (also used for
this study) 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 4.5,10.6, and 20.8 ng/dL, respectively. An
ANOVA model appropriate for random effects was fit 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. Approximately
15 percent of the workers were selected for duplicate blood draws. An ANOVA model
appropriate for random effects was fitted to the subset of workers possessing two blood samples
16
-------
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.
2.4.4.2 Relationships Between Target Activities and Blood-Lead Concentrations.
Although the questionnaire for homeowners was based in part upon the worker questionnaire,
responses to similar questions may have different meanings as they relate to the two groups of
study participants. For example, most homeowners did not perform R&R activities all day long
while most professional workers did. Thus, responses to similar questions like: "How many
days did you perform the R&R activity?" may measure vastly different periods of exposure
between workers and homeowners. Therefore, the statistical analyses conducted for this study
were performed separately for workers and homeowners. However, a similar approach was used
to assess the relationships between the target activities and the blood-lead concentrations for
workers and homeowners. This approach is discussed below.
For each study group (workers and homeowners) a series of statistical models were fit to
the data to determine if there were any significant associations between blood-lead
concentrations and various types of work or target activities. For workers, the relationship
between blood-lead concentrations and potential lead exposure associated with R&R target
activities was investigated for exposure during the previous 30 days, exposure during the past
12 months, and historical exposure (years in career). For homeowners, the relationship between
blood-lead concentrations and potential lead exposure associated with performing the target
R&R activities was investigated for exposure during the previous 30 days and exposure during
the past 12 months. Multiple regression models were employed for both study groups to
examine these relationships. To simplify the regression models, results of multiple chemical
analyses and duplicate blood samples were averaged for each participant to provide a single
blood-lead concentration for each study participant.
Figure 1 displays the paradigm utilized for fitting the models to worker blood-lead
concentrations. Figure 2 shows the paradigm utilized for fitting the models to homeowner
blood-lead concentrations. The first step hi the model fitting for both workers and homeowners
(as shown hi the top box of both figures) was to conduct preliminary analyses to
1. Define the measures of exposure,
2. Verify the use of log-normal distribution for blood-lead concentrations, and
3. Select covariates for the statistical models.
The second step, presented on the left branches of Figure 1 and Figure 2, shows that
separate models were fit to the data for each target activity. Initially, linear regression models
were fit 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 (age, gender, ethnicity, smoking status, and room additions or renovation in
own home) as covariates. Finally, a linear regression model that incorporated the ancillary
17
-------
covariates and simultaneously investigated the potential for lead exposures within the past
month, the past year, and historically was fit to the data for each target activity.
The above analyses helped characterize the strength of the relationship between each
target activity and blood-lead concentrations for workers and homeowners. The final goal,
however, was to develop models (one for workers and one for homeowners) that explain 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 branches of Figure 1 and Figure 2, regression models that examined all of the target
activities simultaneously, were fit to the data. The initial models included effects for all five
target activities for each exposure period (short-term, mid-term, and long-term for workers;
short-term and mid-term for homeowners). For workers, these models were repeated with
worker group added to the model. Finally, an attempt was made to construct models for workers
and homeowners that would assess the effects of the exposure periods simultaneously for all of
the explanatory variables. However, correlations among the target activities and between the
exposure periods within a target activity were high for both workers and homeowners.
Therefore, as shown in Figures 1 and 2, only a subset of the variables for the various exposure
period and target activity combinations were included in the final models.
18
-------
Preliminary Analysis
Select Measures of Exposure
Select Co variates
Verify Lognormal Distribution for Blood-Lead
Analyses
for Each
Target Activity
Preliminary Analysis
• Select Measures of Exposure
• Select Covariates
• Verify Lognormal Distribution for Blood-Lead
Analyses
Employing All
Target Activities
CO
Without Covariates
for Each
Exposure Measure
Analyses
for Each
Target Activity
With Covariates and
Without Job Category
For Each Exposure
Measure
With Covariates
For Each
Exposure Measure
Employing All
Target Activities
Without Covariates
for Each
Exposure Measure
With Covariates and
With Job Category
For Each Exposure
Measure
With Covariates
Employing All
Exposure Measures
With Covariates
For Each Exposure
Measure
With Covariates
For Each
Exposure Measure
With Covariates and
With Job Category
Employing Selected
Exposure Measures
With Covariates
Employing Selected
Exposure Measures
With Covariates
Employing All
Exposure Measures
Figure 1. Paradigm for Fitting Statistical Models to
Worker Blood-Lead Concentrations
Figure 2. Paradigm for Fitting Statistical Models to
Homeowner Blood-Lead Concentrations
-------
3.0 RESULTS
The sections in this chapter discuss the results of recruitment, data collection, and
statistical analysis. It should be noted that the results presented in this chapter are based upon
questionnaires and venous blood-samples collected from professional R&R workers and
homeowners performing R&R in their own historic or older home. The recruitment approach
and the study cities were selected to maximize the likelihood that the study participants disturb
lead-based paint through their R&R activities. However, environmental samples were not
collected to examine the assumption that the participants actually did disturb lead-based paint.
3.1 RECRUITMENT RESULTS AND FIELD EXPERIENCES
3.1.1 Recruitment of Respondents
Recruitment activities were conducted in Charleston, SC, on April 7,1997, through
May 28,1997, in Savannah, GA, on April 27,1997, through May 29,1997, and in
Baltimore, MD, on June 11,1997, through July 16,1997. As a result of the recruitment
activities, 498 participants were screened for participation in the study. Through the screening
process, each participant was classified into one of the following four categories:
• Eligible Eligible participants were defined to be those participants who
were screened as eligible and scheduled for future data
collection.
• Eligible Refused These participants completed the screener and were eligible for
the study but refused to complete the main questionnaire.
• Refused Screener These are potential respondents who called in to the study line
but refused to participate in the screening/recruiting interview.
• Ineligible These respondents participated in the screening/recruitment but
were not eligible for the study (see Section 2.2.1.1).
As shown in Table 1, of the 498 respondents who were screened, 55 percent (274) were
determined to be eligible for the study. Approximately 74 percent (181) of the workers who
were screened were classified as eligible for the study while 37 percent (93) of the screened
homeowners were eligible for the study. This difference is likely due to the more restrictive
eligibility criteria for homeowners.
Although a majority of the 269 participants who were screened and scheduled for further
data collection completed the study, there was a small subset (31) who did not. Furthermore,
17 of the participants who were originally screened as eligible and completed the main
questionnaire actually were ineligible for the study at the time that the main questionnaire was
administered. One possible explanation for this is that for a few of these respondents, during the
delay between the time of the telephone screener and the time that the respondent completed the
questionnaire, a change occurred in the number of days worked in pre-1940 homes in the
last 30 days. Another, more likely explanation for misclassifying workers as eligible who were
20
-------
screened on-site is that the responses to the screener may not be as accurate as responses to the
main questionnaire (see Section 3.1.2). Eligible participants who completed the questionnaire
and the blood draw were classified as Complete. Questionnaire and blood results for the
17 participants who were originally screened as eligible but were later found to be ineligible were
not included in the statistical analyses.
Table 1. Summary of Recruitment Results by Study Group and City
Study Group
Workers
Homeowners
All
Participants
City
Baltimore
Charleston
Savannah
All Cities
Baltimore
Charleston
Savannah
All Cities
Baltimore
Charleston
Savannah
All Cities
Total
Screened
132
70
42
244
94
102
58
254
226
172
100
498
Eligible
101
47
33
181
39
25
28
92
140
72
61
273
Eligible
Refused
0
0
0
0
1
0
0
1
1
0
0
1
Ineligible
31
23
8
62
54
74
30
158
85
97
38
220
Refused
Screener
0
0
1
1
0
3
0
3
0
3
1
4
Total
Number
Completed
92
41
28
161""
34
22
26
82
-------
3.1.2 Field Experiences
Data collection sessions were held in three cities: Charleston, South Carolina; Savannah,
Georgia; and Baltimore, Maryland. Overall, the predominant mood of the study participants was
one of cooperation. Participants seemed interested in the study and in their own blood-lead
results. The following three sections discuss, in greater detail, the field experiences in each city.
3.1.2.1 Field Experiences in Charleston, South Carolina. Three data collection
sessions were held at Roper Hospital North from 4:00 p.m. to 8:00 p.m. on April 21, May 12,
and May 28,1997. The site, located in a suburban area north of Charleston, SC, required
participants to arrive by car. Hospital officials supplied a single conference room where study
participants checked in, completed the self-administered questionnaire (S AQ), and gave a sample
ofblood.
A two-person data collection team oversaw each data collection session. The study
manager checked in arriving participants, collected paperwork, and paid those who both
successfully completed the SAQ and gave a sample ofblood. A study supervisor provided SAQs
to participants, screened walk-ins (not pre-screened for eligibility by phone), field-edited
questionnaires, and prepared data for mailing and transport. Two phlebotomists drew blood and
labeled the blood samples. Because of the large number of participants expected during the last
two visits, an additional person was added to the data collection team to greet respondents, give
directions, and offer refreshments.
At each session, a pre-screened participant was directed into the conference room, where
he/she was then seated around a large conference table with about 10 seats. After the study
manager confirmed the appointment, the participant was seated in order to read and sign the
study consent form. After signing the consent the participant completed the SAQ, and then
moved to the end of the table behind a screen, where a phlebotomist drew a blood sample.
Afterward, participant and a study supervisor discussed any missing or unclear answers given in
the SAQ. They then returned to the study manager, who paid them and double-checked for
mailing address and telephone number changes.
Data collection proceeded in an orderly manner for all three data collections sessions.
Most respondents were orderly and kept their appointments, arriving alone or in small groups
of two to three. Renovation and remodeling workers tended to arrive earlier, at the end of their
workday, while homeowners usually came later in the evening. Since arrivals were spread over
the session, participants were able to begin paperwork and SAQs within minutes, and the
phlebotomists had no trouble taking samples in a reasonable amount of time. The average
respondent completed the SAQ in about 30 minutes, and the blood draw and payment process
took an additional 15 minutes.
At the end of each session, phlebotomists packed, refrigerated, and prepared blood
samples for shipping, according to the study protocol described in "Quality Assurance Project
Plan (QAPjP) for the R&R Worker Characterization and Blood-Lead Study," (July 8,1994).
However, CDC performance samples were unavailable for the April 21,1997, data collection
22
-------
session. The CDC performance samples were included in blood shipments for the remaining
data collection sessions.
Some small problems arose during the second and third sessions due to the large number
of respondents. Since the supervisor responsible for field-editing SAQs also performed several
other tasks (administering SAQs to poor readers and screening and enrolling walk-in
respondents), some SAQs were not completely field edited. In those cases, missing data were
retrieved later by phone. In addition, one respondent was unable to produce a blood sample
during their session.
There was very little activity in the hospital in the evenings and, other than the security
guard and phlebotomists, no other hospital staff were available on-site to access in case of any
sort of difficulty. The hospital liaison, through whom the data collection sessions were arranged,
was located at Roper's main facility several miles away. Since so few people were inside the
facility, all entrances but one were locked after 5:00 p.m. This caused some confusion when a
few respondents, arriving late in the evening and finding the front entrances locked, left without
participating in the study. Most, however, were later contacted and recruited for future sessions.
3.1.2.2 Field Experiences in Savannah, Georgia. Two data collection sessions were
held at Chatham County Health Department from 5:00 p.m. to 8:00 p.m. on May 14 and May
29,1997. The health department is located a few miles south of downtown Savannah, GA, and
most respondents arrived by car. A large classroom-style room was provided for check-in and
completing SAQs. Participants were led to a separate room for collection of the blood sample.
Collection sessions in Savannah, GA, were scheduled one or two days following those in
Charleston, SC, maximizing recruitment activities between the two cities and minimizing travel
costs. As a result, the same data collection team and collection procedures were used in both
Savannah, GA, and Charleston, SC. Two phlebotomists were hired through the health
department, and like the second and third Charleston, SC, sessions, an additional person was
added to the data collection team to escort study participants.
Because of the large, classroom style room, data collection sessions proceeded in an
orderly fashion, even though most participants arrived during the first 2 hours of each session in
groups of three to five. Although some were forced to wait in lines to check in, receive consent
forms and SAQs, give blood samples, and receive payment, almost 30 seats were available for
those filling out forms. Several family members who came along with screened respondents
were also accommodated. Also, only one respondent was screened on-site; the rest were
screened over the telephone.
Two participants refused to supply blood samples. Since both had already completed
SAQs, they were each paid $25.
Unlike the situation encountered in Charleston, SC, health department employees were
most cooperative and greatly facilitated the data collection in Savannah, GA. In particular, a
specific health department staffer was assigned as liaison each evening in case troubles arose. In
addition, maintenance staff were available to rearrange the room and supply needed furniture and
23
-------
materials. As in Charleston, SC, phlebotomists prepared and shipped blood samples, according
to the WCBS QAPjP protocol.
3.1.2.3 Field Experiences in Baltimore, Maryland. Data collection sessions were
held at Johns Hopkins University's Bayview Campus on June 26, May 15, and May 16,1997.
The data collection sessions began at 4:00 p.m. and lasted until 9:30 p.m. The campus, located
east of downtown near a large urban residential district, was accessed by respondents both by car
and by foot. Only a small conference room was available for check-in and completing SAQs;
blood samples were taken in another room down two long hallways. Because of limited space in
the conference room, and increasing numbers of expected participants, the lobby was used as a
waiting area during the last two data collection sessions.
Data collection procedures in Baltimore, MD, were essentially the same as those used in
the other cities. However, three changes were made to accommodate a larger number of
participants: (1) an additional editor and interviewer were added to the data collection team;
(2) participants were asked to wait in a reception area separate from where data collection was
being performed; and (3) a security guard was hired to ensure the safety of study staff.
There were some problems with data collection in Baltimore, MD. The most notable was
screening walk-in respondents for eligibility. During the first two data collection sessions in
Baltimore, MD, a small number of walk-in participants arrived and wanted to participate in the
study. These participants were screened on site, and, if eligible, allowed to participate in the
study on a first-come, first-serve policy. Those whose responses to the screening interview
indicated that they were ineligible for the study were not permitted to participate. Ineligible
participants left the data collection sessions without any problems. In addition, all eligible
participants were able to provide a sample of blood.
In the third Baltimore, MD, session, however, there were a large number of workers (up
to 40) who did not pre-screen for the study but who wanted to participate. A trained interviewer
administered the screener to these potential participants while they were waiting in the reception
area. However, it became apparent that some of the participants who were screened on-site were,
in fact, no't eligible for the study and were providing inaccurate information on the screener in
order to receive the $50 incentive. Blood and questionnaire information from these respondents
were not used in this study. Although every effort was made to include eligible walk-in
participants, due to their large numbers, limited supplies, and the inaccuracy in screener
responses, the inclusion of on-site screening for participation in Baltimore, MD, was curtailed.
Several problems related to refrigerating and shipping the collected blood samples needed
to be overcome. In contrast with practices at Charleston, SC, and Savannah, GA, the hospital
recommended that the samples not be stored at their facility overnight. Throughout the sample
collection period, temperatures in Baltimore, MD, exceeded 100 degrees at the time and the
study protocol needed to be slightly amended to ensure that the blood-samples did not overheat.
As at previous sites, phlebotomists packed samples each evening. Packages were then taken to a
separate facility and stored in a secured refrigerator until picked up by the delivery service
(usually the next business day). Care was taken to ensure that the boxes remained refrigerated at
all times.
24
-------
3.2 QUALITY CONTROL FOR BLOOD-LEAD
The laboratory analysis was conducted by the same laboratory that conducted the
analyses for the WCBS: ESA Laboratories. With a few exceptions, the laboratory and field
quality control (QC) measures used in the WCBS study were employed in this study. These
measures are discussed in detail in the July 8,1994, QAPjP for the WCBS. However, the
laboratory analysis protocol for this study differed from that used in the WCBS in the number
and frequency of analytic quality control samples.
The following provides a summary of the laboratory analysis protocol used for this study.
Following each field collection session, blood samples were sent to the laboratory for
analysis where the concentration of lead in each sample (duplicate analysis for each sample) was
determined using Graphite Furnace Atomic Absorption Spectroscopy (GFAAS) with Zeeman
background correction. The instrumental detection limit for this method was 1 ug/dL with a
linear operating range from 1- to 60-[ig/dL. The linearity of the method was based on a
four-point calibration curve using standards at nominal concentrations of 5,13,30, and
50 ug/dL. A correlation coefficient of greater than 0.995 was maintained throughout the
analyses.
As in the WCBS, to ensure the quality and consistency of the laboratory analysis results,
both internal laboratory and field QC measurements were taken. The internal (laboratory)
quality control consisted of blanks (matrix modifiers), continuing calibration reference materials,
and calibration check standards. Field measurements consisted of analytic blind samples (CDC
performance samples) and field duplicates (a second blood sample from a respondent). A
duplicate blood sample was collected from approximately 15 percent of the respondents7. In
addition, at least two chemical analyses were performed on each blood sample. Table 2
summarizes the analytical results for the Quality Control and CDC performance samples that
were collected.
The results from the internal quality control and performance samples indicate that the
data quality objectives were met and that the blood-lead data were accurate and reliable. All QC
samples met the precision criteria for this study and only two QC samples fell outside the
accuracy criteria for the study and both occurred with the lower Blood-Lead Level Laboratory
Reference System 1.9 ug/dL samples (BLLRS 1.9).
The duplicate blood samples were collected immediately following the collection of the primary
sample.
25
-------
Table 2. Summary of Laboratory Quality Control and CDC Performance Samples
Sample
Identification
Blank
BLLR 1 .9
BLLR 3.3
BioRad2
BioRadS
NISTSRM (Level 1)
CDC RS 1494
CDC RS 696
CDC RS 894
Accuracy
Criteria
< 1.0//g/dL
±25%
±25%
±25%
±25%
±10%
±20%
±20%
±20%
Precision
Criteria
N/A
1.14-2.66
2.64-3.96
26.8-35.3
47.0-61.4
3.31-6.06
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
Continuing Calibration
Reference Material
Samples
Calibration Check
Standard
Blind Performance
Blind Performance
Blind Performance
Reference
Lead Cone.
/yg/dL
0.0
1.9
3.3
30.4
55.3
5.01
4.5
10.6
20.8
Mean
Recovery
(%)
—
111.52
103.77
97.63
97.74
96.47
99.76
95.92
101.63
Duplicate blood samples were collected from 41 respondents. Each of the duplicate
samples were analyzed in the same manner as the regular blood samples and reference samples
(at least two measurements of blood-lead level per sample collection tube). Blood samples were
collected from 243 respondents. Table 3 presents the estimated log standard deviation
attributable to the variability between workers, between duplicate blood draws, and between
duplicate chemical analysis on the same blood draw, as estimated by a variance components
model.
As in the WCBS, the estimated variability between participants appeared to be an order
of magnitude (10 times) greater than the estimated variability attributed to either chemical
analysis or sampling variability. In addition, the magnitudes of the estimated variance
components relative to those for laboratory analysis or sampling variability were similar for
workers and homeowners. However, the estimate of the variability in the blood-lead levels
between homeowners was higher than the corresponding estimate for workers. This might be
due to the fact that some homeowners tend to perform R&R activities on a more sporadic basis
than others and at differing levels of intensity. For example, some homeowners perform a large
amount of R&R but only for a short period of time, while others perform only a few hours of
R&R, but over a longer duration.
26
-------
Table 3. Variance Component Estimates for Blood-Lead Concentrations
Source of Variation
All Particloants
Between Participants
Between Duplicate Blood Draws
Between Duplicate Chemical Analyses
°Wortt»r
(Field Samples)
logb/g/dL)
°W«k«
(Duplicates)
logfrg/dL)
0.7374
N/A
0.0448
0.7560
0.0400
0.0594
Courtly Control
(Reference Samples)
logU/g/dL)
0.7748
0.0463
Workers
Between Participants
Between Duplicate Blood Draws
Between Duplicate Chemical Analyses
Homeowners
Between Participants
Between Duplicate Blood Draws
Between Duplicate Chemical Analyses
0.6680
N/A
0.0378
0.6108
0.0368
0.0688
0.8544
N/A
0.0567
1 .0050
0.0487
0.0320
N/A
N/A
N/A
N/A
3.3 TELEPHONE SCREENING RESULTS
As discussed in Section 2.2.1, the first objective of the screener questionnaires was to
determine the eligibility of a potential respondent. Eligibility was determined using a hierarchial
series of questions (see Appendix A). That is, potential respondents answered questions until
they reached a question indicating ineligibility. For professional R&R workers, six questions
were used to determine eligibility. For homeowners, eight questions were used to determine
eligibility.
Table 4 presents a breakdown of the ineligible workers for each question used to
determine eligibility. As shown in the table, the predominant reason for ineligibility among
workers was related to the amount of R&R that was conducted in the last 30 days. Nearly
84 percent of all ineligible workers were ineligible because they either did not perform at least
10 days of R&R work hi the past 30 days (45.2%) or they did not perform at least 10 days
performing R&R in homes built before 1940 (38.7%). Moreover, this pattern was consistent
across the three cities.
Table 5 presents a breakdown of ineligible homeowners based on answers to questions
used to determine eligibility. As with the ineligible workers, most of the ineligible homeowners
were not eligible to participate in the study because they had not performed a significant amount
of R&R in their home. Approximately 32 percent of the ineligible homeowners were ineligible
because no R&R work was performed hi their home during the last 12 months. Additionally,
42.4 percent of the ineligible homeowners did not meet the study criterion corresponding to the
number of hours spent performing R&R hi the last 30 days (to be eligible, homeowners needed
to spend more than 20 hours in the last 30 days performing R&R).
27
-------
Table 4. Reasons for Worker Ineligibility
Reason for Ineligibility
Did not perform R&R for a living"1
Did not do hands-on R&R work
Spent less than 10 days (in the
last 30) performing general R&R
Spent less than 1 0 days (in the
last 30) performing general R&R in
homes built before 1 940
Spent less than 9 weeks in the last
1 2 months performing R&R in homes
built before 1 940
All
Screener
Question
Number
1
1a
2
2a
3
N/A
Baltimore,
MD
1 (3.2%)
1 (3.2%)
13 (41.9%)
14 (45.2%)
2 (6.5%)
31 (100%)
Charleston,
SC
1 (4.4%)
0 (0.0%)
13(56.5%)
7 (30.4%)
2 (8.7%)
23 (100%)
Savannah,
GA
2 (25.0%)
0 (0.0%)
2 (25.0%)
3 (37.5%)
1 (12.5%)
8 (100%)
Total
4 (6.5%)
1 (1.6%)
28 (45.2%)
24 (38.7%)
5(8.1%)
62 (100%)
(a) The homeowner screening questionnaire was also administered to these potential participants.
Table 5. Reasons for Homeowner Ineligibility
Reason for Ineligibility
Performed R&R for a living
No R&R performed in home in the
last 1 2 months
Did not live in home while R&R was
being performed
Home was built post- 1940
Did not perform a large portion of the
R&R work themselves
Spent less than 20 hours in the past
30 days performing general R&R
Performed some R&R in less than
9 weeks over the past 1 2 months
All
Screener
Question
Number
1
2
2a
2b
2c
2d,2e
3
N/A
Baltimore,
MD
0 (0.0%)
9 (16.7%)
9(16.7%)
0 (0.0%)
2 (3.7.%)
29 (53.7%)
5 (9.3%)
54(100%)
Charleston,
SC
0 (0.0%)
32 (43.2%)
7 (9.5%)
0 (0.0%)
5 (6.8%)
29 (39.2%)
1 (1.4%)
74 (100%)
Savannah,
GA •'-'
0 (0.0%)
9 (30.0%)
5(16.7%)
2 (6.7%)
4(13.3%)
9 (30.0%)
1 (3.3%)
30 (100%)
Total
0 (0.0%)
50(31.7%)
21 (13.3%)
2(1.3%)
11 (7.0%)
67 (42.4%)
7 (4.4%)
158(100%)
28
-------
The second objective of the screening interview was to collect information that could be
used to examine if there was a non-response bias. A non-response bias can occur if respondents
completing the study (study participants) have characteristics that differ from those that did not
participate (non-participants). Non-participants consist of those respondents that were screened
as eligible but: (1) refused the main questionnaire, (2) completed the questionnaire but did not
provide a blood sample, or (3) did not show up for their data collection session. There were
243 participants (161 workers and 82 homeowners) and 31 non-participants (20 workers and
11 homeowners). Information on demographics, R&R work history, and the R&R target
activities was collected during the screener interview.
Table 6 summarizes the demographic information for participants and non-participants.
Although there appear to be some differences between homeowners and professional workers,
participating workers have, on average, a similar age, race, and gender distribution as do
non-participating workers. Similarly, participating homeowners have approximately the same
age, race, and gender distribution as non-participating homeowners.
Table 6. Summary of Demographic Information for Participants and Non-Participants
Variable Description
Sample Size
Age
Race
Sex
City
25th Percentile
Mean
75th Percentile
White
Black
Other
Male
Female
Baltimore, MD
Charleston.SC
Savannah, GA
Workers
Participants
161
32.5
37.2
42.0
65.2%
33.5%
1.2%
86.3%
13.7%
57.1%
25.5%
17.4%
Non-
Participants
20
29.5
37.2
44.5
75.0%
25.0%
0.0%
85.0%
15.0%
45.0%
30.0%
25.0%
Homeowners
Participants
82
31.0
40.0
46.0
79.3%
19.5%
1.2%
56.1%
43.9%
41.5%
26.8%
31.7%
Non-
Participants
11
33.0
43.0
52.0
81.8%
18.2%
0.0%
63.6%
36.4%
54.6%
27.3%
18.2%
All
Participants
243
32.0
38.2
44.0
70.0%
28.8%
1.2%
76.1%
23.9%
51.9%
25.9%
22.2%
Non-
Participants
31
30.0
39.2
45.0
77.4%
22.6%
0.0%
77.4%
22.6%
48.4%
29.0%
22.6%
Table 7 summarizes the responses for participants and non-participants in terms of their
general R&R work history. As shown in the table, with the exception of the number of weeks
performing R&R in the past 12 months, participating and non-participating workers have very
similar distributions of R&R work history. Thus, based on the results presented in Table 7, it
does not appear that non-participants were more likely to be exposed (because they perform more
29
-------
R&R work) than participants. Similarly, it does not appear that non-participating homeowners
were more likely to be exposed than participating homeowners.
Table 7. Summary of General R&R Work History of Participants and Non-Participants
Variable Description
Days spent performing R&R in
the last 30 days
Days, in the last 30 days,
spent performing R&R in
homes built before 1 940'"
Hours spent performing R&R
on a typical day where
performed R&R.
Weeks performed R&R in last
1 2 months
Years spent performing R&R
Days spent performing R&R in
residential buildings in
last 30 days
Days spent performing R&R in
non-residential buildings in
last 30 days
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
1-8 Weeks
9-26 Weeks
> 26 Weeks
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
Workers
Participants
20.0
23.9
28.0
18.0
21.4
25.0
N/A
5.0%
26.1%
68.9%
6.0
12.8
18.0
20.0
20.9
25.0
0
4.1
5
Non-
Participants
20.0
21.45
24.5
15.0
20.4
24.5
N/A
0.0%
10.0%
90.0%
6.0
13.7
19.0
15.0
18.9
24.5
0
2.4
0
Homeowners
Participants
N/A
8.0
14.2
20.0
4.0
5.6
7.0
8.5%
43.9%
47.6%
N/A
N/A
N/A
Non-
Participants
N/A
6.0
11.5
12.0
5.0
6.5
8.0
0.0%
18.2%
81.8%
N/A
N/A
N/A
(a) For homeowners, this questions pertains to R&R in their own home.
In addition to collecting information on demographics and general R&R work history, the
screener interview obtained information on the target activities8 that were performed in the past
30 days. Approximately the same percentage of participating and non-participating workers
performed hands-on cleanup, large structure removal, and paint removal/surface preparation.
More participants than non-participants removed carpets while a higher percentage of
The target activities include: large structure removal [LSR], carpet removal [CR], paint
removal/surface preparation [PR/PS], window/door casement removal [W/DCR], and cleanup
[CLNUP].
30
-------
non-participating workers remove windows or door casements. Figure 3 presents the percent of
participating and non-participating workers that performed each target activity in the last
30 days.
100
BO
80
70
80
i 60
i
; 40
30
20
10
0
CLNUP
CR
LBR PR/PS
I No
W/DCR IkrgMActMly
CLNUP = Cleanup; CR = Carpet Removal; LSR = Large Structure Removal; PR/PS = Paint
Removal/Prepare Surfaces; W/DCR = Window/Door Casement Replacement
Figure 3. Percent of Participating and Non-Participating Workers
Performing Target Activities*
Figure 4 presents the percent of participating and non-participating homeowners that
performed each target activity in the last 30 days. Nearly the same percentage of participating
and non-participating homeowners performed cleanup and paint removal/surface preparation
(two of the most common types of R&R performed by homeowners). A higher percentage of
non-participating homeowners removed carpets and window/door casements but not large
structures.
In summary, all three types of information collected during the screening interview were
similar between study participants and those that did not participate. That is, workers and
homeowners recruited and participating in this study did not appear to have different
characteristics than did non-participants.
31
-------
9
I
100
00
BO
70
80
60
• 40
30
20
10
CUNUP
Cfl
LBR
PHP8
W/DCfl
* CLNUP = Cleanup; CR = Carpet Removal; LSR = Large Structure Removal; PR/PS = Paint
Removal/Prepare Surfaces; W/DCR=Window/Door Casement Replacement
Figure 4. Percent of Participating and Non-Participating Homeowners
Performing Target Activities*
3.4 PARTICIPANT CHARACTERIZATION
The main questionnaires for this study collected information that can be used to
characterize study participants. Section 3.4.1 summarizes participant demographics while
Sections 3.4.2 and Section 3.4.3 summarize the R&R work history and R&R work practices of
study participants.
3.4.1 Demographics
Demographic information on the participants in this study is presented in Table 8.
Noteworthy points include:
• There were more workers from Baltimore, MD (92), than from Charleston, SC (41),
and Savannah, GA (28), combined. However, there were more homeowners from the
southern cities (48) than from Baltimore, MD (34). Further, there were twice as many
worker participants as there are homeowner participants.
• The sample of workers included approximately the same number of Carpenters (47),
Laborers (44), and Painters (44), and fewer "Other" workers (26).
32
-------
Table 8. Summary of Demographic Data
Variable Description
Sample
Size
Age
Gender
Race
Education
Number of
Children
Under Six
in Home
Union
Member
Age of
Home
Renovation
in Own
Home
During
Last 1 2
Months
Total
Baltimore, MD
Charleston, SC
Savannah, GA
25th Percentile
Mean
75th Percentile
Male
Female
White
Black
Other
Not A High
School
Graduate
High School
Graduate
More Than High
School
None
One
Two or More
Yes
No
Pre -1 940
1 940-1 978
Post- 1978
Yes
No
Worker Group
Carpenter
47
28
11
8
33.0
38.4
43.0
89.4%
10.6%
78.7%
17.0%
4.3%
12.8%
38.3%
48.9%
82.6%
10.9%
6.5%
2.1%
97.9%
53.2%
31.9%
14.9%
68.1%
31.9%
Laborer
44
30
10
4
33.0
36.1
41.0
90.9%
9.1%
56.8%
43.2%
0.0%
25.0%
50.0%
25.0%
65.1%
18.6%
16.3%
4.7%
95.3%
56.8%
29.6%
13.6%
61.4%
38.6%
Painter
44
20
12
12
27.0
36.4
42.0
70.4%
29.6%
59.1%
40.9%
0.0%
34.1%
25.0%
40.9%
81.4%
11.6%
7.0%
0.0%
100.0%
45.4%
27.3%
27.3%
52.3%
47.7%
Other
26
14
8
4
33.0
38.0
45.0
100.0%
0.00%
65.4%
34.6%
0.0%
23.1%
11.5%
65.4%
80.8%
11.5%
7.7%
7.7%
92.3%
69.2%
19.2%
11.5%
80.8%
19.2%
All
Workers
161
92
41
28
32.5
37.2
42.0
86.3%
13.7%
65.2%
33.5%
1.2%
23.6%
33.5%
42.9%
77.2%
13.3%
9.5%
3.1%
96.9%
54.7%
27.9%
17.4%
64.0%
36.0%
All
Homeowners
82
34
22
26
31.0
40.0
46.0
56.1%
43.9%
79.3%
19.5%
1.2%
4.9%
4.9%
90.3%
87.8%
4.9%
7.3%
NA
97.6%
1.2%
1.2%
100.0%
0.0%
All
Participants
243
126
63
54
32.0
38.2
44.0
76.1%
23.9%
70.0%
28.8%
1.2%
17.3%
23.9%
58.8%
80.8%
10.4%
8.8%
NA
69.1%
18.9%
11.9%
74.9%
25.1%
33
-------
• The overall age of workers (37.2 years) was very similar to the overall age of
participating homeowners (40 years). The average age within each worker group was
also similar.
• More males (76.1%) than females (23.9%) participated in the study. The ratio of
male to female participants (6.3 to 1) was greater for workers than for homeowners
(1.3 to 1). Among the worker groups, the ratio of males to females was smaller for
painters (2.4 to 1) than for any other worker group.
• An overwhelming majority (90.3%) of homeowners had more than a high school
education. However, only about 43 percent of workers had more than a high school
education. A higher percentage of workers in the "Other" worker group than hi any
other worker group had more than a high school education.
• The majority (80.8%) of participants did not have children under the age of six years
residing in their home. However, Laborers had a greater percentage of children under
the age of six living in their home (34.9%) than did any other worker group or
homeowners.
• Only a small fraction (3.1%) of workers belonged to a union.
• Over half (54.7%) of the participating workers lived in homes built before 1940.
3.4.2 R&R Work History
3.4.2.1 R&R Work History for Professional R&R Workers. Information on the R&R
work history of professional R&R workers was collected using a series of six questions for each
target activity (see Appendix A). Table 9 summarizes the responses among all workers to
questions related to R&R work history specific to each target activity. Table B-2 in Appendix B
presents a similar summary for each worker group.
• The sampled workers spent an average of 24 days performing R&R in the last 30 days.
On average, 21 days were spent performing R&R in historic homes or homes built before 1940.
Carpenters, Laborers, Painters, and Other workers spent about the same number of days
performing work in homes built before 1940.
Of the target activities, "hands-on" cleanup was the activity performed the most (average
of 19 days in all homes and average of 15 days hi pre-1940 homes). Following cleanup, workers
as a whole spent days performing (from most to least) paint removal/prepare surfaces, large
structure removal, window/door casement removal, and carpet removal. This pattern was
consistent for Carpenters, Laborers, and Other workers. As expected, Painters spent more days
performing paint removal than any other target activity except for cleanup.
34
-------
Table 9. Summary of Worker Responses for Questions Pertaining to R&R Target Activities
Variable
Description
Days performing
the activity in the
last 30 days
Days performing
the activity in
Pre-1940
housing in the
last 30 days
Days using a
respirator while
performing
activity
Number of
weeks spent
performing
activity in last
year
Number of years
spent performing
activity over
career
Number of
weeks spent
performing
activity in
average year
Statistic
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
<1 Week
1-8 Weeks
> 8 Weeks
25th Percentile
vlean
75th Percentile
<1 Week
1-8 Weeks
> 8 Weeks
Target Activities
General
R&R
20
23.93
28
18
21.37
25
(«>
(»)
10
0.0%
5.0%
95.0%
6
12.81
18
(<)
Hi
i«)
Large
Structure
5
12.81
20
2
10.80
20
0
5.05
10
13.7%
30.4%
55.9%
3
8.79
14
16.9%
36.4%
46.8%
Paint
Removal
10
15.91
24
5
13.65
20
0
6.66
10
6.8%
26.7%
66.5%
3
9.11
14
14.4%
26.8%
58.8%
Window
Replacement
0
7.40
11
0
6.24
10
0
3.17
2
25.5%
42.2%
32.3%
2
7.77
12
32.9%
33.6%
33.6%
Carpet
Removal
0
3.62
4
0
2.87
3
0
1.54
0
58.4%
27.3%
14.3%
1
5.25
8
53.7%
27.6%
18.7%
Cleanup
10
18.70
25
10
15.28
23
0
6.17
10
3.2%
30.4%
66.5%
4
10.13
15
14.7%
23.7%
61.5%
(a) This question was not asked for general R&R activity.
The number of weeks and the number of years spent conducting each target activity
appears to be strongly associated with the number of days spent performing the activity in the
last 30 days. For example, paint removal/preparing surfaces was the second highest activity in
terms of days performed, weeks performed, and years performed.
The number of weeks spent performing a particular activity in the past 12 months was
roughly the same as the number of weeks spent performing the activity in a "typical" year.
However, care should be taken when interpreting such results because of the difficulties inherent
with memory recall; workers may be biased toward responding as if the current year was a
"typical" year (i.e., work activities in the current year are more easily recalled than are work
activities in previous years).
35
-------
3.4.2.2 R&R Work History for Homeowners. Information on the R&R work history
of homeowners was collected using a series of eight questions for each target activity (see
Appendix A). Table 10 summarizes the responses to each of these questions by R&R target
activity.
Over their entire lives, homeowners have performed a wide mix of activities; more than
one-half of the homeowners indicated that they had performed each target activity at least once.
In addition, most of this work was conducted within the past five years. However, in the past
30 days, nearly all of the homeowners performed cleanup and paint removal but less than
one-half performed large structure removal, window/door casement removal, or carpet removal.
Homeowners worked an average of 14 days in the last 30 days. Further, they spent an
average of about 80 hours performing R&R in their home during the last 30 days. In terms of the
number of days spent performing an activity, homeowners spent the most number of days
performing cleanup followed by paint removal, large structure removal, window/door casement
replacement, and carpet removal. However, in terms of overall hours spent in the last 30 days,
homeowners spent more time performing paint removal/preparing surfaces than any other target
activity. Following paint removal/surface preparation, homeowners spent the majority of their
hours on: (from most to least) cleanup, large structure removal, window/door casement
replacement, and carpet removal.
Over the last 12 months, homeowners performed some R&R for an average of
6 hours per day, 3.5 days per week, and in 28 of the 52 weeks. In terms of weeks in which some
R&R was performed, homeowners performed cleanup the most followed by paint removal, large
structure removal, window/door casement replacement, and carpet removal. A similar pattern
was observed when examining the number of days worked over the last 12 months.
In addition to performing the target activities themselves, homeowners also indicated that
professional R&R workers have performed the target activities in their home. Forty-five percent
of homeowners had professional R&R workers perform large structure removal, paint removal,
and cleanup in their home. A smaller percent of homeowners had professional contractors
replace windows (40%) or remove carpets (15%). Most of the work performed by professional
R&R workers in homes owned by participant homeowners was performed within the last
12 months.
3.4.3 R&R Work Practices
Summary statistics on work practices are presented for each worker group, all workers,
and all homeowners in Table 11. Information on work practices included: tobacco use, respirator
use, hobbies with potential lead exposure, training and educational materials, hours of cleanup,
and type of paint removal. The following main points of interest summarize the statistics
presented in Table 11 for each of the items related to work practices:
36
-------
Table 10. Summary of Homeowner Responses for Questions Pertaining to R&R Target
Activities
Variable Description
Ever performed activity
Performed activity in
last 30 days
Days spent performing
activity in own home
during the last 30 days
Average number of
hours per day spent
performing activity
during the last 30 days
Average number of
hours spent performing
activity during the last
30 days""
Number of weeks
spent performing
activity in own home
during the last year
Number of days spent
performing activity in
tvoical week durinQ the
last year
Number of days spent
performing activity
during the last year161
Number of hours spent
performing activity on
a typical day during
the last year
Last time performed
activity
Contractor ever
performed activity
.ast time contractor
performed activity
Statistic
Yes
No
Yes
No
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
Never
Within 30 Days
Between 30 Days
and 1 Year
Between 1 and 5
Years
More Than 5 Years
Yes
No
Never
Within 30 Days
Between 30 Days
and 1 Year
Between 1 and 5
Years
More Than 5 Years
Target Activities
General
R&R
(c)
(cl
(O
(0
8
14.17
20
4
5.60
7
42
78.38
100
15
27.64
40
2
3.47
5
40
101.52
150
4
5.91
8
(0
Id
(cl
Id
Id
Id
Id
Id
Id
Id
Id
Id
Large
Structure
85.4%
14.6%
46.3%
53.7%
0
2.61
3
0
2.35
4
0
13.65
12
1
4.70
5
1
2.06
3
1
15.68
12
1
4.13
6
14.6%
46.3%
32.9%
4.9%
1.2%
44.4%
55.6%
55.6%
7.4%
22.2%
12.3%
2.5%
Paint
Removal
97.6%
2.4%
92.7%
7.3%
3
6.84
10
3
4.89
6
12
36.50
50
3
11.38
20
2
2.93
4
8
36.33
60
4
5.04
6
2.4%
92.7%
4.9%
0.0%
0.0%
45.1%
53.7%
53.7%
11.0%
20.7%
8.5%
4.9%
Window
Replacement
68.3%
31.7%
39.0%
61.0%
0
1.02
1
0
1.78
3
0
5.70
4
0
2.37
2
0
1.39
2
0
5.76
4
0
2.61
4
31.7%
39.0%
22.0%
6.1%
1.2%
39.5%
60.5%
60.5%
6.2%
19.8%
9.9%
3.7%
Carpet
Removal
56.1%
43.9%
14.6%
85.4%
0
0.39
0
0
0.54
0
0
2.09
0
0
0.88
1
0
0.68
1
0
2.81
1
0
1.46
2
43.9%
15.9%
25.6%
12.2%
2.4%
14.8%
85.2%
85.2%
2.5%
2.5%
3.7%
6.2%
Cleanup
100.0%
0.0%
97.6%
2.4%
4
9.41
15
1
2.50
3
8
25.51
26
3
13.78
22
2
2.83
4
6
48.56
60
1
2.91
4
0.0%
97.6%
2.4%
0.0%
0.0%
45.7%
54.3%
54.3%
14.8%
21.0%
7.4%
2.5%
(a) Calculated as the product of the number of days and the number of hours per day.
(b) Calculated as the product of the number of weeks and the number of days per week.
(c) This question was not asked for general R&R activity.
37
-------
Table 11. Summary of Responses for Questions Pertaining to Work Practices
Variable
Description
Sample Size
Use Tobacco
Products
Respirator
or
Dust Mask
Hobbies with Pb
Exposure
Received Pb
Training"*
Received
Educational
Material
Hours of Cleanup
Paint
Removal""
Variable Category
Yes
No
None
Dust Mask
Half Mask
Full Face Mask
Type C Supplied Air
Mask
PAPR
SCBA
No
Yes
Yes
No
Yes
No
None
< % Mrs/Day
K-1 Mrs/Day
1-4Hrs/Day
>4Hrs/Day
Dry Power-Sanding
Dry Hand-Sanding
Dry Scraping
Burning, Torching,
Heat Gun
Wet Scraping
Wet-Sanding
Chemical Stripping
Used Dust Collector
When Sanding
Carpenter
47
57.4%
42.6%
39.1%
39.1%
15.2%
10.9%
0.0%
2.2%
0.0%
29.8%
70.2%
29.8%
70.2%
38.3%
61.7%
0.0%
8.5%
31.9%
46.8%
12.8%
53.2%
76.6%
78.7%
31.9%
21.3%
19.1%
38.3%
25.5%
Laborer
44
65.9%
34.1%
32.6%
58.1%
11.6%
2.3%
0.0%
2.3%
0.0%
25.0%
75.0%
9.3%
90.7%
25.0%
75.0%
0.0%
2.3%
13.6%
50.0%
34.1%
65.9%
68.2%
68.2%
29.5%
38.6%
34.1%
47.7%
43.2%
Painter
44
72.7%
27.3%
37.2%
39.5%
7.0%
20.9%
2.3%
0.0%
0.0%
20.5%
79.5%
31.8%
68.2%
34.1%
65.9%
4.5%
9.1%
20.5%
52.3%
13.6%
63.6%
88.6%
79.5%
29.5%
29.5%
22.7%
43.2%
47.7%
Other
26
57.7%
42.3%
41.7%
41.7%
12.5%
4.2%
0.0%
0.0%
0.0%
42.3%
57.7%
26.9%
73.1%
34.6%
65.4%
3.8%
11.5%
7.7%
53.8%
23.1%
57.7%
73.1%
73.1%
38.5%
30.8%
19.2%
30.8%
30.8%
All Workers
161
64.0%
36.0%
37.2%
44.9%
11.5%
10.3%
0.6%
1.3%
0.0%
28.0%
72.0%
24.4%
75.6%
32.9%
67.1%
1.9%
7.5%
19.9%
50.3%
20.5%
60.2%
77.0%
75.2%
31.7%
29.8%
24.2%
41.0%
37.3%
All
Homeowners
82
30.5%
69.5%
56.8%
29.6%
13.6%
2.5%
0.0%
1.2%
0.0%
67.1%
32.9%
N/A
37.8%
62.2%
2.5%
0.0%
28.8%
55.0%
13.8%
57.3%
76.8%
82.9%
30.5%
9.8%
6.1%
36.6%
17.1%
(a) This question was not asked to homeowners.
(b) Percentages given represent the percentage of workers or homeowners that indicated they performed the activity.
38
-------
Overall, two-thirds of workers used tobacco products. Among the worker groups,
more painters (72.7%) used tobacco products than any other worker group. However,
only 30.5 percent of the homeowners used tobacco products.
The OSHA Lead in Construction Standard requires workers to wear an approved
respirator if exposed to an airborne concentration of lead of greater than
30 micrograms per cubic meter of air (30 ng/m3) calculated as an 8-hour
time-weighted average (TWA) as a result of performing R&R activities. However,
only 23.7 percent of workers reported using a respirator in the last 30 days. Similarly,
only 16 percent of homeowners used a respirator in the last 30 days. Homeowners
and workers in every work group used a dust mask more frequently than any type of
respirator (44.9 percent of workers and 29.6 percent of homeowners reported using
dust masks).
Several questions were asked of both workers and homeowners that pertain to their
participation in hobbies which have historically been associated with increased lead
exposure. These hobbies include: casting lead bullets or fishing sinkers, crimping
fishing sinkers onto the line with teeth, dismantling batteries, and working with
stained glass. Most workers (72%) did have at least one hobby where they could be
exposed to lead. Conversely, only (33%) homeowners had at least one hobby where
they could be exposed to lead.
The OSHA Lead in Construction Standard requires training of employees prior to
initial assignment to areas where there is a possibility of exposure over the
permissible exposure limit (PEL) of 50 ug/m3 averaged over an 8-hour workday.
However, more than two-thirds of the workers in each work group reported that they
had not received training for reducing lead exposure in the workplace. Over all four
worker groups, Laborers had the smallest percentage (9.3%) receiving training.
Similar patterns were observed with respect to workers receiving educational
materials (67 percent did not receive materials). Only 38 percent of homeowners
reported receiving any educational materials on reducing lead exposures.
All but a very small percentage of workers (2%) and homeowners (2.5%) performed
hands-on cleanup. About one-half of workers and homeowners spent 1-4 hours a day
performing cleanup. As expected, among the four worker groups, Laborers
performed the highest proportion of cleanup; approximately 84 percent performed this
activity for more than 1 hour a day.
Most of the participants (either workers or homeowners) performed dry sanding or
scraping for paint removal and the percentage of homeowners performing these
activities was approximately the same as the corresponding percentages of workers.
Also, the percentage of workers using dry scraping or dry sanding for paint removal
was similar for all four worker groups. Further, a similar percentage of workers and
homeowners used burning, torching, or a heat gun to remove paint (-30%). However,
workers and homeowners differed in the use of other methods for paint removal. In
particular, three times more workers than homeowners used wet scraping, four times
39
-------
more workers that homeowners used wet sanding, and two times as many workers as
homeowners used dust collectors when sanding.
3.5 OVERVIEW OF BLOOD-LEAD CONCENTRATIONS
This section presents the results of an initial examination of the blood-lead measurements
collected in this study. This analysis served two purposes. First, it provided detailed descriptive
statistics using the blood-lead concentrations of the study participants. These descriptive
statistics, presented in Section 3.5.1, were used as a guide when developing the statistical models
discussed in the next section.
The second purpose of this analysis was to place the blood-lead concentrations of the
study participants into perspective. The blood-lead concentrations of the study participants were
compared to workers participating in Phase n of the R&R study (WCBS). Also of interest was a
comparison of the blood-lead concentrations of study participants to those of the general, adult
U.S. population. Detailed comparisons to participants in the WCBS are presented in
Section 3.5.2; comparisons to the general U.S. population (as described by NHANES ffl,
Phase 2) are presented in Section 3.5.3. In both sections, the results indicate that workers and
homeowners performing R&R activities in historic or older homes are more exposed to lead than
either the general U.S. population or general R&R workers (i.e., workers not necessarily working
in historic or older homes).
3.5.1 Descriptive Statistics for the Blood-Lead Concentrations of Study Participants
Blood samples were obtained from 161 workers and 82 homeowners. The lead
concentration of each sample was measured in at least two separate laboratory analyses. The
results of multiple chemical analyses and duplicate blood samples were averaged for each
participant to provide a single blood-lead concentration for each participant.
As in the WCBS and in other recent studies, the distributions of blood-lead
concentrations for participants in this study were skewed. Figures 5 and 6 present histograms of
the blood-lead concentrations for workers and homeowners, respectively. Because of the
skewness of these distributions, a natural log transformation was employed. Normal and
log-normal probability plots are displayed in Figures D-l through D-4 in Appendix D.
Table 12 presents the geometric mean, standard error, and a 95 percent confidence
interval of the mean blood-lead concentrations for each city and study group. Table 13 presents
selected percentiles and the number of participants with blood-lead concentrations above 10,15,
and 20 ug/dL. Overall, the geometric mean blood-lead concentrations for workers and
homeowners were well below 10 ng/dL (5.73 jig/dL for workers and 4.45 ug/dL for
homeowners). However, 20.2 percent (49 of 243) of the study participants had a blood-lead
concentration greater than 10 ug/dL and 2.9 percent (7 participants) had blood-lead
concentrations greater than 25 ug/dL. In addition, three participants (all workers) had blood-lead
concentrations above 40 ug/dL.
40
-------
18
15
in
j> 12
i_
o
H- Q
o y
"c
0}
0)
D_
Si
\
\
0.5 1.0 2.0 4.0 8.0 16.0 32.0 64.0
Blood —lead Concentration (jU.g/dl_)
Figure 5. Histogram of Worker Blood-Lead Concentration (Semi-Logarithmic Scale)
in
o>
c
0>
o
0>
to
Q_
18
15
12
/
\
\
0.5 1.0 2.0 4.0 8.0 16.0 32.0 64.0
Blood—lead Concentration (/u-g/dL)
Figure 6. Histogram of Homeowner Blood-Lead Concentration (Semi-Logarithmic
Scale)
41
-------
Table 12. 95 Percent Confidence Intervals for Geometric Mean of Blood-Lead
Concentrations by City and Study Group
Study Group
Workers
Homeowners
All
Participants
City
Baltimore, MD
Charleston.SC
Savannah, GA
All Cities
Baltimore, MD
Charleston.SC
Savannah, GA
All Cities
Baltimore, MD
Charleston.SC
Savannah, GA
All Cities
N
92
41
28
161
34
22
26
82
126
63
54
243
Geometric Mean
(pg/dL)
5.54
6.30
5.57
5.73
3.84
5.28
4.66
4.45
5.02
5.93
5.11
5.26
log Std. Error
log Oug/dL)
0.061
0.118
0.156
0.053
0.148
0.170
0.174
0.094
0.062
0.097
0.116
0.048
95% Confidence
Interval (jjgldL)
(4.91, 6.25)
(5.00, 7.94}
(4.10, 7.56)
(5.16, 6.36)
(2.87, 5.14)
(3.79, 7.38)
(3.31, 6.56)
(3.70, 5.36)
(4.45, 5.66)
(4.90, 7.16)
(4.07, 6.42)
(4.79. 5.78)
Table 13. Selected Percentiles and Occurrence of Elevated Blood-Lead Levels by City and
Study Group
Study Group
Workers
Homeowners
All
Participants
City"
Baltimore
Charleston
Savannah
All Cities
Baltimore
Charleston
Savannah
All Cities
Baltimore
Charleston
Savannah
All Cities
N
92
41
28
161
34
22
26
82
126
63
54
^43_
25th
Percentile
3.77
3.75
3.07
3.65
2.30
2.95
2.17
2.75
3.30
3.25
2.45
3.20
Median
5.32
5.20
5.75
5.25
3.40
4.25
3.77
3.68
4.60
5.05
4.32
4.60
75th
Percentile
7.94
12.20
9.70
8.80
4.90
9.30
9.20
6.85
7.52
11.2
9.30
8.80
Number >
Wfjg/dL
(Percent)
13(14.1%)
12(29.3%)
7 (25.0%)
32(19.9%)
6(17.7%)
5 (22.7%)
6(23.1%)
17(20.7%)
19 (15.1%)
17(27.0%)
13(24.1%)
49 (20.2%)
Number >
15//g/dL
(Percent)
5 (5.4%)
7(17.1%)
3 (10.7%)
15(9.3%)
3 (8.8%)
3(13.6%)
2 (7.7%)
8 (9.8%)
8 (6.4%)
10(15.9%)
5 (9.3%)
23 (9.5%)
Number >
25/yg/dL
(Percent)
1 (1.1%)
1 (2.4%)
1 (3.6%)
3(1.9%)
1 (2.9%)
1 (4.6%)
2 (7.7%)
4 (4.9%)
2(1.6%)
2 (3.2%)
3 (5.6%)
7 (2.9%)
Baltimore, Maryland; Charleston, South Carolina; Savannah, Georgia
42
-------
Of the seven participants with blood-lead levels greater than 25 jig/dL, three were
workers and four were homeowners. Among the three workers: all were approximately the
same age (34,38, and 42 years); two of the three were white; two completed high school, and the
third graduated from college or technical school. One of the workers smoked and crimped
fishing sinkers onto the line with their teeth. Among the four homeowners: all were white;
highly educated (graduated college or technical school); but ranged in age from 35 years to
63 years old. None of the four homeowners smoked, only one had a hobby with potential
exposure (working with stained glass). Table 14 presents some of the R&R work characteristics
of the three workers and four homeowners with blood-lead concentrations above 25 ug/dL. As
shown in the table, all of these study participants performed a substantial amount of paint
removal.
The geometric mean blood-lead concentrations for workers were not significantly
different across the three cities (p-value = 0.584). Similarly, the geometric mean blood-lead
concentrations for homeowners were not significantly different across the cities
(p-value = 0.378). However, the geometric mean blood-lead concentrations for workers were
significantly (p-value = 0.012) greater than those for homeowners (on average 1.3 Hg/dL greater;
95 percent confidence interval [1.06,1.48]), Figures D-5 and D-6 in Appendix D display
side-by-side box plots of blood-lead concentrations for each city and study group. Despite
differences in the geometric mean, a similar percentage of workers and homeowners had
blood-lead concentrations above 10 ug/dL, 15 ug/dL, and 25 |xg/dL.
Each worker was assigned to one of four worker groups based upon their job title and job
activities (see Section 2.4.1). Table 15 presents descriptive statistics for each worker group and
Figures D-7 through D-l 1 in Appendix D display side-by-side box plots of blood-lead
concentrations for each worker group. The blood-lead concentrations among the worker groups
were on the borderline of being significantly different (p-value = 0.067).
3.5.2 Comparison to the WCBS
The design and results of the WCBS are described in detail in the published EPA Report,
"Lead Exposure Associated with Renovation and Remodeling Activities: Worker
Characterization and Blood-Lead Study," May 1997, EPA 747-R-96-006. Briefly, the WCBS
was a targeted survey of union carpenters and employees of independent contractors in two
cities: Philadelphia, Pennsylvania, and St. Louis, Missouri. Blood and questionnaire
information were collected from 585 participants (581 had complete blood and questionnaire
information).
3.5.2.1 Demographic Characteristics. The current study and the WCBS collected
information that was used to characterize study participants. Demographic characteristics were
similar for participants in this study and those participating in the WCBS. For example, workers
in the WCBS were on average 38 years old, predominantly white (83.8%), and high school
graduates (93.5%). In comparison, the average age of workers in this study was 37,70 percent
were white, and 82.7 percent were high school graduates. However, there was a higher
percentage of female workers (13.7%) in this study than in the WCBS (2.1%). Responses to
43
-------
Table 14. Characteristics of Study Participants with Blood-Lead Levels Above 25/sg/dL
Blood-
Lead
Level
Job Title
Workers
44.4
43.2
42.2
Owner
Paint
Company
Painter
Renovation
Contractor
Main Activities
Days Spent Performing Target Activity in Historic
Homes or Homes Built Prior to 1940
Large
Structure
Removal
Paint or
Preparing
Surface
Painting, sanding,
stripping
Painting, prep
Painting, remodeling
0
15
10
30
15
15
Window or
Door
Casements
Removal
Carpet
Removal
Clean-
up
0
15
5
0
0
0
30
30
20
Types of
Respirator
Used
Received
Educational
Materials
Received
Training
Half Mask
Dust
Mask
Dust
Mask
No
Yes
Yes
No
No
Yes
Homeowners
26.7
33.8
32.5
25.5
N/A
N/A
N/A
N/A
Burnt & scraped old
paint, installed closet,
refinished floors
Paint removal (including
burning and sanding),
plaster repair, painting
Removed plaster.
scraped and sanded
woodwork, removed tin
ceiling
Scrape and sand paint,
refinish wood, repair
window glazing
0
0
3
0
15
4
12
20
5
0
2
0
0
0
0
0
15
5
20
20
None
Dust
Mask
None
Dust
Mask
No
No
Yes
No
N/A
N/A
N/A
N/A
01
-------
Table 15. Descriptive Statistics of Blood-Lead Concentrations by Worker Group
Worker Group
Carpenter
Laborer
Painter
Other
N
47
44
44
26
25th
Percentlle
3.75
3.30
3.81
3.75
Median
6.85
4.15
6.12
5.17
75th
Percentile
10.60
7.12
9.32
7.60
Geometric
Mean
(pg/dL)
6.47
4.64
6.40
5.42
Log Std.
Error
log (/yg/dL)
0.098
0.094
0.117
0.102
95%
Confidence
Interval
(pg/dL)
(5.34, 7.84}
(3.86, 5.59)
(5.09, 8.05)
(4.44, 6.62)
questions on other characteristics, such as the age of home and whether there was renovation in
their own home, were similar between the two studies.
3.5.2.2 Work History. In the WCBS, as in this study, a series of three questions were
used to collect information on short-, mid-, and long-term work history for each target activity.
Figure 7 summarizes the work history for workers in the current study compared to those in the
WCBS. Overall, workers in the WCBS performed less R&R in older housing than did workers
in the current study.
Figure 7 presents comparisons between the short-term work activities of workers in the
WCBS and those in the current study. Workers in the current study performed, on average
during the last 30 days, 10 more days of general R&R in older housing than workers in the
WCBS. The greatest differences between workers in the WCBS and workers in the current study
were for the average number of days spent performing paint removal/surface preparation and
cleanup. The lower half of Figure 7 presents the average number of days that workers in the two
studies spent performing the target activities in older homes (pre-1950 for the WCBS, and
pre-1940 for the current study). The pattern of the average number of days worked in older
homes was similar to the average number of days spent performing R&R in any home.
Figure 8 displays the percentage of workers in the two studies that spent more than
8 weeks in the last 12 months performing a target activity. Again, a larger percentage of workers
in the current study spent more than 8 weeks performing the target activities than did workers in
the WCBS.
Workers in both studies performed the target activities for approximately the same
number of years.
45
-------
Average Number of Dayi Spent Pertaining R&R In Any Home
30-
23.9
tt.7
17.1
^
!
12,8
•ma
10-
03
7.4
33 3.0
Qeneml R&R LSR PH/BP W/DCR CR CLNUP Target
Cunw* IWV^J WCBS
Average Number of Deye Spent Performing R&R In Older
Current Study: Homes Butt Pre-1940
WCBS: Home* Butt Pre-1860
30
10-
21.4
153
13.7
10.7 103
4.4
1.4
QenereJ R&R LSR PR/8P W/DCR CR
Study I I Current IVWJ WCBS
CLNUP Target ActMty
Figure 7. Comparison of Short-Term Work History for Workers in the Current Study
(Phase IV) to Those in the WCBS (Phase II)
46
-------
That Spml Moo Than S WMta Performing Activity
100
90
SO
70-
60
60
40
30
20
10
B6.0
75.0
ee.s
865
ess
26.1
17.8
14.3
10.6
36.1
Qaneral R&R LSR PR/BP W/DCR CR
Study I I Currant rVWJ WCBS
CLNUP •togat ActMy
Number of Vtara Ovor Cvaar In WNch AdMty WM Parfomwd
10.1
10J5
0.1
7.1
7.B
Qanaral R4R LSR PR/SP WyDCR OR CLNUP target AcdvRy
Study I I Current fVAM WCBS
Figure 8. Comparison of Mid- and Long-Term Work History for Workers in the Current
Study (Phase IV) to Those in the WCBS (Phase II)
47
-------
3.5.2.3 Blood-Lead Concentrations
Table 16 presents the geometric means and 95 percent confidence intervals for all
581 workers that participated hi the WCBS and the 243 participants in the current study. The
geometric mean for homeowners in the current study was not significantly different from the
geometric mean for workers in the WCBS. While significantly different, the geometric mean
blood-lead concentration for workers in the current study was only 1.3 ug/dL (95 percent
confidence interval of 1.1 ug/dL to 1.4 ug/dL) higher than the geometric mean blood-lead
concentration for workers in the WCBS.
Carpenters were the only worker group where there was a significant difference between
the geometric mean blood-lead concentrations for workers hi the current study and those in the
WCBS. The geometric mean blood-lead for carpenters in the current study was 1.3 ug/dL
(95 percent confidence interval from 1.0 ug/dL to 1.6 fig/dL) higher than the geometric mean
blood-lead concentration of non-union carpenters in the WCBS.
Table 16. Geometric Means and 95 Percent Confidence Intervals for Blood-Lead
Concentrations of Participants from the WCBS and the Current Study
Description
Carpenter"'
Laborer
Painter
Other
Union Carpenter""
Drywall Worker""
Floor Layer""
Supervisor""
Window installer""
All Workers
All Homeowners
Current Study
N
47
44
44
26
N/A
N/A
N/A
N/A
N/A
161
82
Geometric
Mean (/sg/dL)
6.5
4.6
6.4
5.4
N/A
N/A
N/A
N/A
N/A
5.7
4.5
95%
Confidence
Interval
(5.3, 7.8)
(3.9, 5.6)
(5.1,8.1)
(4.4, 6.6)
N/A
N/A
N/A
N/A
N/A
(5.2, 6.4)
(3.7, 5.4)
WCBS
N
104
54
34
14
159
64
81
57
14
581""
N/A
Geometric
Mean (j/g/dL)
5.0
4.9
7.2
5.3
4.4
5.8
2.6
3.8
5.4
4.5
N/A
95%
Confidence
Interval
(4.5, 5.7)
(4.1, 5.7)
(5.8, 8.8)
(3.9, 7.3)
(4.0, 4.8)
(5.0, 6.8)
(2.3, 3.0)
(3.2, 4.4)
(3.9, 7.4)
(4.2, 4.7)
N/A
(a) This represents the geometric mean for non-union carpenters participating in the WCBS.
(b) Additional worker groups (union carpenter, drywall worker, floor layer, supervisor, and window installer) were used in
the WCBS.
Although differences in the geometric mean blood-lead concentrations were not
significantly different or slightly higher, a larger percentage of participants in the current study
had elevated blood-lead levels than did workers in the WCBS. Table 17 presents the percent of
participants with blood-lead levels above 10 ug/dL, 15 ug/dL, and 25 ug/dL in this study and
those in the WCBS. The distributions corresponding to the percent of workers and homeowners
48
-------
with elevated blood-lead levels were significantly different from those observed in the WCBS
(likelihood ratio p-values < 0.02). Moreover, as shown in Table 17, with the exception of
workers having blood-lead levels above 25 ug/dL, the percentage of homeowners and workers in
the current study with blood-lead levels greater than or equal to 10 ug/dL, 15 ^ig/dL, and
25 ug/dL was twice as large as the corresponding percentages for workers in the WCBS.
Table 17. Occurrence of Elevated Blood-Lead Levels for Participants in the Current Study
(Phase IV) and Workers in the WCBS (Phase II)
Blood-Lead Levels
Number and Percentage with
Blood-Lead Levels i10//g/dL
Number and Percentage with
Blood-Lead Levels * 1 5 //g/dL
Number and Percentage with
Blood-Lead Levels 2 25 //g/dL
Total Sample Size
Study
Current Study (Phase IV)
Homeowners
17
(20.7%)
8
(9.8%)
4
(4.9%)
82
Workers
32
(19.9%)
15
(9.3%)
3
(1.9%)
161
WCBS (Phase II)
52
(9.0%)
22
(3.8%)
7
(1.2%)
581
3.5.3 Comparison to NHANES III, PHASE 2
The design and results of the National Health and Nutrition Examination Surveys
(NHANES) are described in detail in two reports from the Centers for Disease Control and
Prevention (CDC):
Centers for Disease Control (1992), "Sample Design: Third National Health and
Nutrition Examination Survey. Series 2: Data Evaluation and Methods Research,"
National Center for Health Statistics, Centers for Disease Control and Prevention,
U.S. Department of Health and Human Services, DHHS Publication No. (PHS) 92-1887.
Centers for Disease Control (1994), "Plan and Operation of the Third National Health and
Nutrition Examination Survey, 1988-94," National Center for Health Statistics, Centers
for Disease Control and Prevention, U.S. Department of Health and Human Services,
DHHS Publication No. (PHS) 94-1308,1994.
Briefly, NHANES is conducted by the CDC's National Center for Health Statistics (NCHS), and
traces the health and nutritional status of the non-institutionalized, civilian, U.S. population.
Information is collected from both adults and youths using questionnaires and a standardized
physical examination (including a venipuncture blood sample). A complex survey design is
employed so that the resulting sample is nationally representative. The information presented in
this report was based upon the public use data set collected from adults (18 years of age or older)
as part of NHANES m, Phase 2.
49
-------
3.5.3.1 Blood-Lead Concentrations. Table 18 presents geometric means and
95 percent confidence intervals for blood-lead concentrations of study participants and those of
the adult U.S. population (as described by NHANES HI, Phase 2). As shown in the table, the
geometric mean blood-lead level for homeowners was significantly higher than the geometric
mean blood-lead level for the adult U.S. population (1.86 times higher with a 95 percent
confidence interval of 1.54 to 2.26). Similarly, the geometric mean blood-lead level for workers
was 2.4 times higher (95 percent confidence interval of 2.14 to 2.70) than the geometric mean
blood-lead level for the adult U.S. population. Similar differences exist irrespective of race and
gender.
Table 18. Geometric Means and 95 Percent Confidence Intervals for Blood-Lead
Concentrations of Participants in the Current Study and Adults in NHANES III,
Phase 2
Demographic
Characteristic
Race
Sex
Level
White
Non-White
Males
Females
All
Study
Current Study (R&R Phase IV)
Homeowners
4.89
(3.91,6.12)
3.11
(2.53, 3.82)
5.86
(4.52, 7.58)
3.14
(2.53, 3.89)
4.45
(3.70, 5.36)
Workers
5.76
(5.04, 6.59)
5.66
(4.79, 6.70)
6.09
(5.47, 6.77)
3.90
(2.83, 5.39)
5.73
(5.16, 6.36)
NHANES III,
Phase 2
2.34
(2.22, 2.47)
2.62
(2.44,2.81)
3.07
(2.90, 3.26)
1.89
(1.79, 2.00)
2.39
(2.27, 2.51)
Table 19 presents a comparison of the percentage of participants in this study at or above
10 ug/dL, 15 ug/dL, and 25 ug/dL, to the percentage of adults in the general U.S. population
with blood-lead concentrations at or above these thresholds. As shown in the table, a much
higher percentage of people participating in the study had elevated blood-lead levels compared to
the U.S. population (likelihood ratio p-value < 0.001).
50
-------
Table 19. Occurrence of Elevated Blood-Lead Levels for Participants in the Current Study
(Phase IV) and in the Adult U.S. Population
Blood-Lead Levels
Number and Percentage with
Blood-Lead Levels * 1 0 /sg/dL
Number and Percentage with
Blood-Lead Levels 2 1 5 fjg/dl
Number and Percentage with
Blood-Lead Levels 2 25 //g/dL
Total Sample Size
Study
Current Study (Phase IV)
Homeowners
17
(20.7%)
8
(9.8%)
4
(4.9%)
82
Workers
32
(19.9%)
15
(9.3%)
3
(1.9%)
161
NHANES III,
Phase 2
293
(3.5%)
74
(0.88%)
15
(0.18%)
8,457
Odds ratios examining the odds of blood-lead levels greater than 25 ng/dL among study
participants to that of adults in the general U.S. population are presented in Table 20. For both
workers and homeowners, the odds ratios were significantly greater than one. This implies that
participants in this study were more likely to have blood-lead levels above 25 jig/dL than the
general U.S. population. Specifically, workers were 10.68 times and homeowners were
28.86 times more likely to have blood-lead concentrations above 25 ng/dL than the general U.S.
population (95 percent confidence intervals were [3.06,37.28] and [9.37, 88.92], respectively).
Workers were 6.90 times more likely (95 percent confidence interval of 4.88 to 9.80) and
homeowners were 7.30 times more likely (95 percent confidence interval; 4.57 to 11.63) to have
blood-lead levels £ 10 ug/dL than adults in the general U.S. population.
Table 20. Odds Ratios Comparing Odds of Blood-Lead Concentrations Greater Than
25//g/dL Among Study Participants to the General U.S. Population
Study Group
Workers
Homeowners
NHANES III, Phase 2
No. with Blood-
Lead Levels
<25//g/dL
158
78
8,442
No. with Blood-
Lead Levels
225/ig/dL
3
4
15
Total Sample Size
161
82
8,457
Odds Ratio* and
(95% Cl)
10.68
(3.06, 37.28)
28.86
(9.37, 88.92)
-
These odds ratios compare the odds of blood-lead concentrations above 25 //g/dL for the study group referenced in
the row to the general U.S. population.
51
-------
3.6 STATISTICAL MODELING RESULTS
3.6.1 Statistical Model Building
Statistical models were used to investigate the relationships between log-transformed
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 activities. In addition to general R&R, lead exposure variables were
constructed for several R&R "target activities" including: large structure removal, paint
removal/preparing surfaces, window/door casement replacement, carpet removal, and cleanup.
For workers, three periods of exposure, short-, mid- and long-term, were captured by the
questionnaire for each target activity. For homeowners, information on only two periods of
exposure, short- and mid-term, were captured.
For workers, the number of days (in the last 30 days) an activity was performed in
historic homes or homes built before 1940 was selected as the measure of short-term exposure.
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 12 months. Long-term
exposure in each activity was measured by the number of years a worker performed the activity
over their career. Thus, a total of 18 variables, one for each combination of target activity and
exposure period, were used to characterize a worker's potential lead exposure resulting from
R&R.
For homeowners, both the number of days an activity (in the last 30 days) was performed
in their own home and the total number of hours (in the last 30 days) that the activity was
performed were used to characterize short-term exposure. Mid-term exposure was characterized
by the number of weeks in the past 12 months in which the target activity was performed and by
the total number of days that the target activity was performed during the past 12 months. Thus,
a total of 24 variables, two for each combination of target activity and exposure period, were
used to characterize a homeowner's potential lead exposure resulting from R&R.
3.6.1.1 Selection of Ancillary Covariates. Although the objectives of this study were
to investigate the relationship between blood-lead concentrations and the R&R target activities,
both the worker and homeowner questionnaires collected information on work practices and
demographics that might also be related to blood-lead levels. These variables were defined as
"ancillary variables" (i.e., variables that were not of immediate interest, but that needed to be
accounted for in the statistical models, such as gender and race). That is, the relationships
between the target activities and blood-lead concentrations need to be "adjusted" to account for
the effects of these variables.
In both the worker and homeowner questionnaires, information on two categories of
ancillary variables were collected:
• Variables related to the conduct of R&R
— Respirator use
— General work practices
52
-------
— R&R activity in own home (workers only)
— Previous lead training (workers only) and education
— R&R work conducted by contractors in home (homeowners only)
• Variables not related to the conduct of R&R
— Other occupations with potential lead exposure
— Age of home (workers only)
— Hobbies with potential lead exposure
— Race and ethnicity
— Age
— Tobacco use
— Education level
— City
The large number of candidate ancillary variables required an initial screening to select
covariates for use in the statistical models.
Each potential ancillary covariate was classified as being either discrete or continuous.
Scatter plots and regression lines were used to characterize the relationships between continuous
potential covariates and blood-lead concentrations (see Figure D-7 in Appendix D). Box plots
were used to characterize the relationship between categorical covariates and blood-lead
concentrations (see Figures D-8 through D-12 in Appendix D).
Whenever appropriate, the levels of some discrete variables were collapsed to improve
power for statistical tests. 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, a single variable was constructed to reflect the number of hobbies a
participant had with high potential for lead exposure.
The statistical significance of each candidate ancillary covariate's effect on blood-lead
concentration is displayed in Table C-l of Appendix C for workers and Table C-2 of Appendix C
for homeowners. Worker ancillary covariates that had significant relationships with blood-lead
were then simultaneously placed into a single model for worker blood-lead. Homeowner
covariates with significant relationships with blood-lead were then simultaneously placed into a
single model for homeowner blood-lead. The following variables were selected, based on these
analyses, for inclusion in covariate-adjusted models for workers and homeowners:
Workers Homeowners
1. Sex 1. Age
2. Room Additions or Renovation in Own Home 2. Sex
3. Smoke, Use Snuff, or Chewing Tobacco 3. Race
Table 21 provides the geometric mean blood-lead concentration and log (standard error) for each
level of the selected ancillary covariates.
53
-------
Table 21. Geometric Mean Blood-Lead Concentration and Log (Standard Error) for Each
Level of the Ancillary Covariates
Study Group
Workers
Homeowners
Variable
Sex
Room additions or renovation
in own home
Smoke, use snuff, or chewing tobacco
Age of participant
Sex
Race
Level
Male
Female
Yes
No
Yes
No
<31
31-40
>40
Male
Female
White
Non-White
Sample
Size
139
22
103
58
58
103
16
32
34
46
36
65
17
Blood-Lead (pg/dL)
Geometric
Mean
6.1
3.9
5.2
6.7
4.9
6.2
4.0
3.8
5.4
5.9
3.1
3.1
4.9
Log Std. Error
0.055
0.165
0.058
0.130
0.096
0.062
0.206
0.152
0.146
0.132
0.110
0.105
0.113
3.6.2 Blood-Lead Concentration Comparisons Using Covariate-Adjusted Model Results
The geometric mean blood-lead concentration of homeowners, workers, and each worker
group were estimated after adjusting for ancillary covariates. The covariate-adjusted geometric
means for each study group and worker group are presented in Table 22. These means were
calculated from a linear combination of the levels of each covariate with weights equal to the
percent of participants at each level. Separate models were used to estimate the adjusted
geometric means for each worker group and for workers and homeowners. The
covariate-adjusted geometric means were similar to the unadjusted means. Figure 9 displays the
95 percent confidence intervals for the covariate-adjusted geometric means.
54
-------
Table 22. Geometric Mean and 95 Percent Confidence Intervals for Each Study Group and
by Worker Group Based Upon Covariate-Adjusted Models
Description
Worker Group
Carpenters
Laborers
Painters
Other
All Workers
Homeowners
Geometric Mean
6.1
4.2
6.3
5.0
5.7
4.4
95% Confidence Interval
(5.1,7.4)
(3.5, 5.1)
(5.1,7.7)
(3.9, 6.4)
(5.2, 6.3)
(3.7, 5.2)
Painter
Carpenter
Other
Laborer
All Workers
All Homeowners
1-44 E-
5.0
42
1-44 E-
83
6.1
5.7
n-181 E-
4.4
n-82 E-
Blood-lead Concentration
8
Figure 9. 95 Percent Confidence Intervals for Geometric Mean of Blood-Lead
Concentrations for Each Worker Group and Study Group Based Upon
Covariate-Adjusted Models
55
-------
3.6.3 Relationship Between R&R Target Activities and Participant Blood-Lead
Concentrations
As discussed in Section 2.4.4, statistical models were used to examine the relationship
between the R&R target activities and the blood-lead concentrations of study participants.
Section 3.6.3.1 presents the results for workers, Section 3.6.3.2 presents the results for
homeowners.
3.6.3.1 Relationship Between R&R Target Activities and Worker Blood-Lead
Concentrations. The effect of each of the 18 target activity variables, unadjusted for covariates
or the conduct of other target activities, was estimated by fitting separate linear regression
models (one model for each target activity, exposure period, and study group). Estimated slopes,
standard errors, and p-values for each exposure measure are displayed in Table C-3 in
Appendix C.
With the exception of carpet and large structure removal, there was a statistically
significant positive relationship between worker blood-lead concentration and short-term conduct
in pre-1940 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
positive, with the exception of carpet removal. The apparent negative association between
blood-lead concentration and carpet removal should be interpreted with caution. It is important
to note that workers spending time performing carpet removal must spend less time performing
activities that may generate a greater exposure. Thus, a negative association may not imply that
carpet removal does not generate a lead exposure, rather that carpet removal does not generate as
much of an exposure as other target activities.
The relationship between cleanup and R&R activities presented in Table C-3 could
depend upon the specific type of cleanup performed (certain types of cleanup activities, such as
dry sweeping, may distribute lead-dust more than others). The questionnaire for this study did
not collect information on the specific type of cleanup performed. However, in the WCBS only
14.3 percent of all workers performed cleaning activities using a HEPA vacuum or wet mop with
TSP (even among these workers a large percentage also reported cleaning with a broom or
vacuum). Based upon anecdotal comments from respondents in this study, it is very likely that
this same result was true among participants in this study. In which case, the underestimation of
the relationship between cleanup and blood-lead concentrations would be minimized.
Many of the slope estimates presented in Table C-3 were positive and statistically
significant (p-value < 0.05). For each target activity, Table 23 presents the predicted changes in
blood-lead concentration for 10 days of work activity per month in pre-1940 buildings
unadjusted for other covariates9. The second column displays predicted blood-lead concentration
for each target activity, for workers who performed the activity for zero and 10 days in pre-1940
As shown in Table 9, the interquartile range for the number of days workers spent performing R&R
work in pre-1940 buildings was 7 days, and on average, workers spent 21 days performing R&R in
pre-1940 homes. However, 10 days were used for this table to allow for comparisons to the results of
the WCBS.
56
-------
houses during the previous month. The predicted increases in blood-lead concentration ranged
from 0.1 ug/dL for large structure removal to 0.9 ug/dL for window/door casement replacement.
The maximum predicted blood-lead concentration associated with 10 days of work in pre-1940
buildings, 6.1 ug/dL, was estimated for work related to window/door casement replacement. The
last row of Table 23 contains the results for conducting general R&R. Because all of the workers
spent more than 10 days performing general R&R in pre-1940 homes (this was one of the
eligibility criterion) the "Base Level" for general R&R represents 10 days of work in pre-1940
homes. As shown in Table 23, there was a significant predicted increase in blood-lead
concentration associated with performing 10 days versus 20 days of general R&R in pre-1940
homes (1.7 ug/dL).
The covariate-adjusted effects for each combination of target activity and exposure period
were also estimated using separate linear regression models. Estimated slopes, standard errors,
and p-values are shown in Table C-4 in Appendix C. With a few exceptions, the estimated
slopes for target activities were positive after adjusting for the effects of selected ancillary
covariates. The magnitudes of the covariate-adjusted slope estimates were typically less than the
unadjusted slope estimates.
The predicted changes in worker blood-lead concentrations, based on the
covariate-adjusted models, associated with 10 days of work in pre-1940 buildings are presented
in the third column of Table 23. The smaller slope estimates for the covariate-adjusted models
compared to the unadjusted models resulted in smaller predicted increases in worker blood-lead
concentrations. With the exception of general R&R, the estimated increase in blood-lead
concentration from the base level was very small (less than 1 ug/dL) for both the unadjusted and
covariate-adjusted models. The estimated increase for conducting 20 days of R&R instead of
10 days was 1.5 ug/dL with the covariate-adjusted model.
Table 23. Predicted Changes in Worker Blood-Lead Concentrations Associated with
10 Days of Work in Pre-1940 Homes
Target Activity
Large Structure
Removal
Paint Removal/
Prepare Surface
Window/ Door
Casement
Removal
Carpet Removal
Cleanup
General R&R"'
Based on Model Unadjusted for Covariates
Level When Worker Conducts
Base an Additional 10 Days per
Level1*1 -» Month of Activity
5.6 -» 5.7
4.8 -» 5.5"
5.2 -» 6.1 •
5.9-* 5.3
4.5 •» 5.3"
3.8 -» 5.5"
Based on Covariate-Adjusted Model
Level When Worker Conducts
Base an Additional 10 Days per
Level"1 •* Month of Activity
5.7 -» 5.7
4.7 -» 5.4"
5.3 -» 6.0
5.8 -» 5.4
4.4 -» 5.2"
4.0 -> 5.5"
(a) The Base Level for general R&R is 10 days, for all other activities the Base Level is zero days.
* Slope parameter estimate for days per month of activity was significant at the alpha = 0.05 level.
57
-------
Covariate-adjusted models that included all three exposure periods (short-, mid-, and
long-term) were also fit 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 C-5. Overall, the p-values associated with
the parameter estimates in Table C-5 are not consistent across the six target activities.
Short-term exposure (days pre-1940) appeared to be the best predictor for cleanup and general
R&R while mid-term exposure (weeks in last 12 months) appeared to be the best predictor for
carpet removal. None of the parameter estimates for long-term exposure (years in career) were
significantly associated with blood-lead levels after short and mid-term exposure measures were
included in the model. There were no significant relationships found at all for large structure
removal, paint removal/preparing surfaces, and window/door casement replacement. However,
the inconsistent results of these models are likely due to the high degree of correlation among the
three exposure period measures for each target activity.
3.6.3.2 Relationship Between R&R Target Activities and Homeowner Blood-Lead
Concentrations. As with workers, the effect of each of the 24 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 C-6 in Appendix C.
Very few target activities, for any exposure measure, were significantly related to
blood-lead concentrations. Both short-term exposures (days in last 30 days and total hours in last
30 days) for paint removal/surface preparation were significantly related to blood-lead levels.
General R&R as measured by hours in the last 30 days (short-term) and days in the last
12 months (mid-term) was significantly related to blood-lead levels. General R&R was the only
target activity where mid-term exposure was significantly related to blood-lead concentrations.
Estimated slopes, standard errors, and p-values are shown in Table C-7 in Appendix C for
covariate-adjusted models (a separate model for each combination of target activity and exposure
period measure). The estimated slopes for target activities were mostly positive after adjusting
for the effects of selected ancillary covariates. As with the unadjusted models, paint removal/
surface preparation and general R&R were the only two target activities with significant
relationship to blood-lead concentrations. Once adjustments for the ancillary covariates were
made, all exposure measures, both short- and mid-term, for paint removal/surface preparation
were significantly related to blood-lead levels (marginally significant for days in the last 30 days,
p-value = 0.054). A similar pattern was observed for general R&R (marginally significant for
days in the last 30 days, p-value = 0.077).
Table 24 presents the predicted changes in homeowner blood-lead concentrations for
selected days of work activity in pre-1940 buildings. The third column displays predicted
changes in blood-lead concentration for each target activity, for homeowners who performed the
activity for zero days and the average number of days for each target activity in their home
during the previous month. For example, the second row presents the predicted changes (from
3.3 to 4.5 ug/dL) in homeowner blood-lead concentrations associated with performing an
58
-------
additional three days of large structure removal. The fourth column displays the corresponding
predicted changes based upon the covariate-adjusted models.
Table 24. Predicted Changes in Homeowner Blood-Lead Concentrations Associated With
Changes in the Number of Days Spent Performing a Target Activity in Their
Pre-1940 Home
Target Activity
General R&R
Large Structure
Removal
Paint Removal/
Prepare Surface
Window/ Door
Casement Removal
Carpet Removal
Cleanup
Additional
Number of
Days**
15
3
7
1
0.5
10
Based on Model Unadjusted
for Covariatas
Leval When
Homeowner Conducts
Base Additional Days per
Level1" -» Month of Activity
3.3 -» 4.5
4.5 -» 4.5
3.3 -» 4.5"
4.4 -» 4.5
4.6 -» 4.4
4.0 -» 4.1
Based on Covariate-Adjusted
Model
Level When
Homeowner Conducts
Base Additional Days per
Level"' -# Month of Activity
3.3 -» 4.4
4.2 -» 4.4
3.5 -» 4.4
4.3 -» 4.4
4.4 -» 4.4
3.7 -» 4.4
(a) The Base Level is zero days for all comparisons.
* Slope parameter estimate for days per month of activity was significant at the alpha = 0.05 level.
* * The number of days is based upon the average number of days homeowners spent performing the target activity in their
home during the last 30 days.
As expected by the many insignificant estimated slope parameters, the predicted changes
to homeowner blood-lead levels were small. The largest predicted changes based upon the
additional number of days spent performing the target activity were for paint removal/surface
preparation (1.2 ug/dL for the unadjusted model and 0.9 ug/dL for the adjusted model) and
general R&R (1.2 ug/dL for the unadjusted model and 1.1 ug/dL for the adjusted model).
Similarly, the largest predicted changes when measuring short-term exposure by the total number
of hours (see Table 25) were also for general R&R and paint removal/surface preparation.
As for workers, covariate-adjusted models that included both exposure periods were also
fit for each target activity. In all, 12 models were fit: the first six models, one for each target
activity, included short- and mid-term exposure variables related to duration (i.e., days in the last
30 days and weeks in the last 12 months), the second six models, again one for each target
activity, included short and mid-term exposure variables related to intensity (i.e., total hours in
the last 30 days and days in the last 12 months). The estimated slopes, standard errors, and
p-values for these models are displayed in Table C-8 in Appendix C. As in the earlier models
that included only one exposure period per target activity, general R&R and paint
removal/surface preparation were the only target activities with exposure terms significantly
related to blood-lead. Considering the models based upon duration, it appears that mid-term
exposure (number of weeks in the past 12 months) was a better predictor of blood-lead than
short-term exposure (number of days in the last 30 days) for these two target activities.
59
-------
However, when considering the models based upon intensity, the parameter estimates were not
as consistent. Short-term exposure (total hours in the past 30 days) appeared to be a better
predictor for paint removal/surface preparation while mid-term exposure (days in the last
12 months) appeared to be a better predictor for general R&R.
Table 25. Predicted Changes in Homeowner Blood-Lead Concentrations Associated With
Changes in the Number of Hours Spent Performing a Target Activity in Their
Pre-1940 Home
Target Activity
General R&R
Large Structure
Removal
Paint Removal/
Prepare Surface
Window/ Door
Casement Removal
Carpet Removal
Cleanup
Additional
Number of
Hours'*
80
14
36
6
2
25
Based on Model Unadjusted for
Covariates
Level When
Homeowner Conducts
Base Additional Hours per
Level1" -» Month of Activity
3.0 -» 4.5"
4.6 -» 4.5
3.4 -» 4.5'
4.5 -» 4.5
4.5 -» 4.5
4.7 -» 4.5
Based on Covariate-Adjusted Model
Level When
Homeowner Conducts
Base Additional Hours per
Level"1 -» Month of Activity
3.1 -»4.5*
4.3 •» 4.4
3.5 -» 4.4"
4.3 -» 4.4
4.4 -» 4.4
4.2 •» 4.4
(e) The Base Level is zero hours for all comparisons.
* Slope parameter estimate for days per month of activity was significant at the alpha = 0.05 level.
* * The number of hours is based upon the average number of hours homeowners spent performing the target activity in
their home during the last 30 days.
3.6.4 Overall Statistical Models for Participant Blood-Lead Concentrations
The statistical models described in Section 3.6.3 helped to characterize the strength of the
relationship between each target activity and worker or homeowner blood-lead concentrations.
The following two sections present the results from developing statistical models that were used
to characterize how each of the target activities interacted in their association with blood-lead
concentrations while accounting for the effect of potential confounding ancillary covariates.
Section 3.6.4.1 presents the results for workers and Section 3.6.4.2 presents the results for
homeowners.
3.6.4.1 Overall Statistical Model for Worker Blood-Lead Concentrations. An initial
series of models were fit to the data to assess which of the specific target activities, if any, had
the largest impact on worker blood-lead concentrations during each exposure period. That is, for
each of the three exposure periods, the effects of all of the target activities (with the exception of
general R&R work) were estimated simultaneously using covariate-adjusted models. General
R&R work was not included in these models because it was not independent from the other
60
-------
target activities (e.g., the amount of general R&R must be at least as great as the amount of any
other target activity). Estimated slopes, standard errors, and p-values as presented in Table C-9.
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
was likely a result of the strong inter-correlations among the target activity variables within an
exposure period.
In the WCBS, it was observed that target activities varied across the worker groups and
that there were differences in the geometric mean blood-lead concentrations among the nine
worker groups10. In this follow-up study, however, the opposite was observed. That is, with a
few exceptions, workers in the four worker groups spent a similar amount of time performing the
target activities and differences in the geometric mean blood-lead concentrations among the
four worker groups were not significant (see Section 3.5). A series of models were fit to assess
the impact of worker group on the relationships between the target activities and worker
blood-lead concentrations. The estimated slopes, standard errors, and p-values for these models
are displayed in Table C-10.
After including worker group in the three exposure models, cleanup was the only target
activity significantly related to worker blood-lead concentrations in the short-term exposure
model. In the mid-term exposure model, paint removal/surface preparation and carpet removal
were significantly related to worker blood-lead concentrations. In the long-term exposure model,
carpet removal and cleanup were significantly related to worker blood-lead concentrations. With
the exception of the long-term exposure model where cleanup became significant, worker group
generally had only a small impact on the estimated slopes, standard errors, and p-values. This
result is illustrated in Table 26, which provides general F-tests for model effects after adjusting
for the effects of selected covariates and other effects included in the model.
The first row of Table 26 shows the R2 value (12.7%), F-test (7.58) and corresponding
p-value (<0.001) for a model that included only the selected ancillary covariates. The next three
rows correspond to models that for each exposure period, included the ancillary covariates and
exposure measures for the five target activities. The combined effect of all five target activities
was statistically significant for each exposure period, and accounted for between 8.6 and
13.9 percent of the variability in worker blood-lead concentrations after adjusting for the
covariates. The fifth 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 7.1 percent of the variability in blood-lead after adjusting for the covariates. The
sixth through eighth rows describe, for each exposure period, the combined effect of all five
target activities after adjusting for the effects of ancillary covariates and worker group. The
F-test for the combined exposure measures shows that the combined effects of target activities
are significantly related to worker blood-lead concentrations after adjusting for covariates and
worker group. The final three rows describe, for each exposure period, the effect of worker
10 The nine worker groups used in the WCBS were: (1) Union Carpenter, (2) Non-Union Carpenter,
(3) Laborer, (4) Painter, (5) Drywall Worker, (6) Window Installer, (7) Supervisor, (8) Floor Layer,
and (9) Other.
61
-------
group after adjusting for the effects of ancillary covariates and the combined effect of all five
target activities. For the short- and mid-term exposure models, the relatively small impact of
worker group is indicated by the marginally significant relationship between worker group and
worker blood-lead concentrations. In the long-term exposure model worker group appeared to be
more related to worker blood-lead concentrations and, this is reflected in the significance
(p-value <0.05) of the worker group term.
Table 26. General F-Tests for Model Terms on Worker Blood-Lead Concentrations, After
Adjusting for the Effects of Covariates and Other Terms Included in the Model
Base Model
Covariates
Covariates
Covariates +
Worker Group
Covariates + TA
(Days Pre-1 940)
Covariates + TA
(Weeks)
Covariates + TA
(Weeks)
Variables Under Investigation
Covariates
Target Activities ( Days Pre-1940)
Target Activities (Weeks)
Target Activities (Years)
Worker Group
Target Activities (Days Pre-1940)
Target Activities (Weeks)
Target Activities (Years)
Worker Group
Worker Group
Worker Group
R2
0.127
0.250
0.243
0.183
0.189
0.284
0.277
0.241
0.284
0.277
0.241
R%
—
0.139
0.134
0.086
0.071
0.120
0.110
0.091
0.045
0.045
0.071
F
7.58
4.80
4.69
2.75
3.92
3.97
3.66
2.88
2.30
2.34
3.63
P-Value
< 0.001
< 0.001
<0.001
0.021
0.010
0.002
0.004
0.017
0.080
0.075
0.015
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.
Since it appeared that both the target activities and, to some extent worker group, were
related to worker blood-lead concentrations, the final predictive model was developed using both
of these components. The first step in developing the final predictive model for workers was to
refine the models presented in Table C-10 for each exposure period. In particular, since there
was strong inter-correlation among the target activities, a series of stepwise models were fit to:
(1) determine which of the target activities were the most related to blood-lead concentrations,
and (2) reduce the number of terms in the models presented in Table C-10 by removing target
activities that were not significantly related to blood-lead.
The three models (one for each exposure period) presented in Table C-10 were used as
the initial models for developing the final predictive models. In the first iteration, the single
variable that was the least related to blood-lead was removed from the model (for short-term this
62
-------
was paint removal/surface preparation, and large structure removal for the mid-term and
long-term exposure models). In subsequent iterations, a single insignificant variable was
removed from the model. This process was repeated until all terms were significant. Each
removed variable was then re-entered into the final model for each exposure period to verify that
they were not significant. Table C-l 1 presents the slopes, standard error, and p-values for each
target activity in the final model for each exposure period.
The results for the iterative process were similar to those indicated by Table C-10 where
all target activities were fit simultaneously for each exposure period. That is, cleanup was the
only significant term retained in the short-term exposure model, paint removal/surface
preparation and carpet removal were the significant terms retained in the mid-term model, and
carpet removal was significant and included in the model for long-term exposure. Where the
results differ is in the significance of window/door casement replacement instead of cleanup in
the long term-exposure model.
The final predictive model for workers was developed using similar iterative steps. The
initial model included all terms from the three final models for each exposure period, the
ancillary covariates, and worker group. Only one term, years in career spent removing carpets,
was not significant and removed from the model. Thus, the final model included the following
activities from each exposure period:
• Short-term: Number of days, in the last 30 days, spent performing cleanup.
• Mid-term: Number of weeks spent performing paint removal/surface preparation
and the number of weeks spent performing carpet removal.
• Long-term: Number of years spent replacing window/door casements.
The model was further refined by examining the assumption of a common slope for all worker
groups for each of the target activities. That is, a model was fit with unequal slopes for each
worker group and activity and an F-test was performed to determine the activities, if any, for
which the assumption of common slopes did not appear to be valid. The assumption of equal
slopes appeared to be valid only for the number of weeks spent performing carpet removal in the
last 12 months. Different slopes for each worker group were used for the remaining three target
activities. Table C-12 presents the estimated parameters, standard errors, and p-values for the
final model.
To illustrate how worker blood-lead concentration is a function of the amount of time
spent performing cleanup, paint removal/surface preparations, carpet removal, or window/door
casement replacement, Table 27 displays the estimated blood-lead concentration (and 95 percent
confidence interval) associated with low, medium, and high exposure indices based on the final
model. The low, medium, and high exposure indices were based upon the 25th, 50th, and
75th percentiles of questionnaire responses from workers on these target activities and exposure
periods. For Carpenters, Laborers, and Painters, the model predicts an increase in blood-lead
concentrations of 1.7 ug/dL for Laborers, 3.3 ug/dL for Carpenters, and 5.7 (ig/dL for Painters.
However, the model predicts a decrease for the Other worker group of 1.5 ug/dL. This may be
63
-------
due to the fact that the Other worker group includes supervisors, managers, foremen, etc. and
time spent performing some of the activities may not be realistic of actual hands-on time.
Tukey's multiple comparison procedure for unbalanced data was employed to conduct
pairwise comparisons between the geometric mean worker blood-lead concentrations predicted
for each worker group at the Medium Exposure Index (see Table 27). The overall Type I error
rate for these simultaneous tests was ^0.10. Carpenters had a significantly higher geometric
mean blood-level than Laborers. All other comparisons were not significant.
Table 27. Predicted Worker Blood-Lead Concentrations Associated with Low, Medium, and
High Exposure Indices for Each Worker Group
Worker
Group
Carpenter
Laborer
Painter
Other
Low Exposure Index'"
0 Weeks CR,
1 0 Days of Cleanup,
5-8 Weeks PR/SP,
2 Years Over Career W/DCR
Geometric
Mean
5.5
4.2
4.7
6.8
95% C.I.
(4.1,7.3)
(3.3, 5.4)
(3.4, 6.6)
(5.0, 9.1)
Medium Exposure Index1"1
<1 WeekCR,
1 5 Days of Cleanup,
9-1 2 Weeks PR/SP,
6 Years Over Career W/DCR
Geometric
Mean
6.3
4.8
6.0
5.8
95% C.I.
(5.2, 7.6)
(4.0, 5.7)
(4.9, 7.4)
(4.6. 7.4)
High Exposure Index'01
1-4 Weeks CR.
23 Days of Cleanup,
>26 Weeks PR/SP,
12 Years Over Career W/DCR
Geometric
Mean
8.8
5.9
10.4
5.3
95% C.I.
(6.8, 11.3)
(4.5, 7.9)
(7.6, 14.1)
(3.5. 8.0)
The low, medium, and high exposure indices were based upon the 25th, 50th, and 75th percentiles of
questionnaire responses from workers on these target activities and exposure periods.
(a) Low Exposure Index =
(b) Medium Exposure
Index
(c)
High Exposure
Index
0 Weeks of Carpet Removal (CR) in the last 12 months,
10 days of Cleanup in the last 30 days,
5-8 Weeks of Paint Removal/Surface Preparation (PR/SP), and
2 Years Spent Performing Window/Door Casement Replacement (W/DCR) Over
Entire Career
< 1 Week of Carpet Removal (CR) in the last 12 months,
15 days of Cleanup in the last 30 days,
9-12 Weeks of Paint Removal/Surface Preparation (PR/SP), and
6 Years Spent Performing Window/Door Casement Replacement (W/DCR) Over
Entire Career
1-4 Weeks of Carpet Removal (CR) in the last 12 months,
23 days of Cleanup in the last 30 days,
> 26 Weeks of Paint Removal/Surface Preparation (PR/SP), and
12 Years Spent Performing Window/Door Casement Replacement (W/DCR) Over
Entire Career
An alternative predictive model based on the days, weeks, and years that a worker spent
performing general R&R activities was also investigated. The estimated parameters, standard
errors, and p-values for this alternative final model are presented in Table C-13. Unlike the final
predictive model, this model does not include interaction terms with worker group (they were not
significant at the 0.05 level). Further, the estimated parameters indicate that worker blood-lead
64
-------
concentrations are positively related to general R&R activity. However, this relationship was
significant only for short-term exposure and the alternative model did not predict worker
blood-lead concentrations as well as the final predictive model.
3.6.4.2 Overall Statistical Model for Homeowner Blood-Lead Concentrations. A
similar methodology to that employed in Section 3.6.4.1 for workers was used to develop an
overall predictive model for homeowner blood-lead concentrations. Therefore, an initial series
of models were fitted to the data to assess which of the specific target activities, if any, had the
greatest impact on homeowner blood-lead concentrations during each exposure period. That is,
for each of the two exposure periods (short- and mid-term), the effects of all of the target
activities (with the exception of general R&R work) were estimated simultaneously using
covariate-adjusted models. As with the models for workers, general R&R work was not included
in these models. However, two separate models were fit for each exposure period, one
corresponding to duration and one examining intensity. Estimated slopes, standard errors, and
p-values for the four models are presented in Table C-14.
When all of the target activities were considered simultaneously, only paint removal/
surface preparation had a significant impact on homeowner blood-lead concentrations. Mid-term
(weeks in last 12 months) paint removal/surface preparation and short-term (total hours in last
30 days) paint removal/surface preparation were significantly related to homeowner blood-lead
concentration. Again, this could have been due to the strong inter-correlations among the target
activity variables within an exposure period so additional steps were taken to develop the final
predictive models for each exposure period.
The four models presented in Table C-14 were used as the initial models for developing
the final predictive models for each exposure period (two models for each exposure period). In
the first iteration, the single variable that was the least related to blood-lead was removed from
the model and in subsequent iterations, a single insignificant variable was removed, in each
iteration, from the model. This process was repeated until all terms were significant. Each
removed variable was then re-entered into the final model for each exposure period to verify
insignificance. In all four models, paint removal/surface preparation was the only target activity
that was significantly related to homeowner blood-lead concentrations. Table C-15 presents the
slopes, standard error, and p-values for paint removal/surface preparation for the final models in
each exposure period.
After the final models for each exposure period were developed, additional models were
developed to determine which measure, intensity or duration, better predicted homeowner
blood-lead concentrations. Models corresponding to duration and intensity of paint removal/
surface preparation were developed by combining the short- and mid-term measures for duration
and intensity of paint removal/surface preparation into a model. That is, the number of days and
the number of weeks spent performing paint removal/surface preparation were included in one
model, and the number of hours in the last 30 days and the number of days in the last 12 months
were included in another model.
The final predictive model for homeowners included the significant measures in the
models from the preceding step. Thus, the final model contains both a measure of intensity (total
65
-------
hours in the last 30 days) and a measure of duration (weeks in the last 12 months). Table C-16
presents the slopes, standard error, and p-values for the final predictive model for homeowners.
Figure 10 illustrates the predicted increase in homeowner blood-lead concentrations as a function
of the number of hours spent in the last 30 days performing paint removal/surface preparation
and the number of weeks in the last 12 months in which some paint removal/surface preparation
was performed.
An alternative predictive model for homeowner blood-lead concentrations was developed
using general R&R. Since there were two measures of general R&R for each exposure period,
the same steps used to develop the final predictive model were employed. The final alternative
model, like the predictive model using paint removal/surface preparation, included a short-term
measure of intensity and a mid-term measure of duration. The slopes, standard error, and
p-values for general R&R duration and intensity are presented in Table C-17. This model
predicted homeowner blood-lead concentrations nearly as well as the final predictive model.
Predicted Bkxxl-
Lead Concentration
30 Total Hours Spent
25 In the Last 30 Days
Number of Weeks Spent
in the Last 12 Months
20
Figure 10. Predicted Increase in Homeowner Blood-Lead Concentrations as a
Function of Hours and Weeks Spent Performing Paint
Removal/Surface Preparation
66
-------
To illustrate the relationship between homeowner blood-lead concentrations and the
amount of general R&R work performed, Table 28 displays the estimated blood-lead
concentration (and 95 percent confidence interval) associated with low, medium, and high
indices based upon the alternative predictive model. The low, medium, and high exposure
indices were based upon the 25th, 50th, and 75th percentiles of the total hours and weeks spent
performing general R&R. The predicted homeowner blood-lead concentrations using paint
removal/surface preparation are also presented for comparison.
Table 28. Predicted Homeowner Blood-Lead Concentrations Associated with Low, Medium,
and High Exposure Indices Using General R&R
Model
Using General
R&R
Using Paint
Removal/Surface
Preparation1"
Low Exposure Index
42 Hours in the Last 30 Days
15 Weeks in the Last
12 Months
Geometric
Mean
3.4
3.5
95% C.I.
(2.7, 4.2)
(2.8, 4.4)
Medium Exposure Index
78 Hours in the Last 30 Days
28 Weeks in the Last
12 Months
Geometric
Mean
4.6
4.6
95% C.I.
(3.8, 5.4)
(3.9, 5.3)
High Exposure Index
100 Hours in the Last 30 Days
40 Weeks in the Last
12 Months
Geometric
Mean
5.8
5.5
95% C.I.
(4.6, 7.2)
(4.4. 6.7)
(a) The low, medium, and high exposure indices for paint removal were based upon the 25th, 50th, and 75th percentiles of
the total hours and weeks spent performing paint removal/surface preparation (12, 37, and 50 hours; 3, 12, and 20
weeks).
67
-------
4.0 SUMMARY AND CONCLUSIONS
This study surveyed two groups of persons (professional workers and homeowners) who
performed renovation and remodeling in historic homes or homes built before 1940. Information
from both groups of persons was collected in three cities (Charleston, South Carolina;
Savannah, Georgia; and Baltimore, Maryland). A total of 243 respondents completed
questionnaires and gave blood samples (161 workers, and 82 homeowners). The questionnaire
and blood-lead information was used to:
1. Characterize the types of work activities and work practices in which workers and
homeowners performing R&R in historic (or pre-1940) homes engaged
2. Determine if the blood-lead concentrations of homeowners differ from those of
workers
3. Determine if the blood-lead concentrations of workers in specific worker groups
differ after adjusting for potential confounding factors
4. Determine the relationship between blood-lead concentrations and work practices or
activities performed by workers or homeowners in old or historic homes, after
controlling for potential confounding factors
5. Determine if the blood-lead concentrations of participants in this study were
different from those of participants in the WCBS or adults in the general U.S.
population.
The overall results for the study are summarized below.
Questionnaires were collected from 246 participants (163 workers and 83 homeowners).
The sample was composed of individuals that were predominantly white (70%), male (77%), and
on average 39 years old. The percentage of study participants that were female was larger for
homeowners (44%) than for workers (14%). Although separate questionnaires were used for
workers and homeowners, both questionnaires captured information on how often each
participant performed general renovation and remodeling as well as specific target activities.
The workers spent, on average, 21 days performing general R&R in historic homes during the
last 30 days. Homeowners spent, on average, 14 days performing general R&R in their own
historic or pre-1940 home. The responses to the questionnaires also indicate that
• Workers spent more time performing R&R than did homeowners.
• Both workers and homeowners spent time performing a variety of R&R activities.
In particular, a large amount of time was spent performing large structure removal
and paint removal/surface preparation.
Despite the OSHA Lead in Construction Standard, only 23.7 percent of workers
reported using a respirator in the last 30 days. Similarly, only 16 percent of
68
-------
homeowners used a respirator in the last 30 days. Homeowners and workers in
every work group used a dust mask more frequently than any type of respirator
(44.9 percent of workers and 29.6 percent of homeowners reported using dust
masks).
• The majority of workers (64%) used tobacco products, but only 30 percent of
homeowners used tobacco products.
• Over two-thirds of the workers had hobbies that had some potential for lead
exposure (e.g., working with stained glass windows, casting lead bullets or
sinkers, etc.). Conversely, only one-third of homeowners had similar hobbies.
• Despite the OSHA Lead in Construction Standard, the majority of workers had not
been trained (76%) nor received educational information (67%) on lead exposure
due to R&R work. Similarly, 62 percent of homeowners had not received
information on lead (homeowners were not asked about training).
• Over 75 percent of workers and homeowners reported using dry sanding/scraping to
remove paint. About one-third of homeowners and workers reported using
chemical stripping or buming/torching/heat gun to remove paint. Three times more
workers than homeowners used wet scraping, four times more workers than
homeowners used wet sanding, and two times as many workers as homeowners
used dust collectors when sanding.
Blood samples were collected from 161 workers and 82 homeowners. Overall, the
geometric mean blood-lead concentrations were well below 10 ug/dL: 5.7 jig/dL for workers and
4.5 ug/dL for homeowners. Approximately 20 percent of the study participants had blood-lead
concentrations above 10 ug/dL. Also, 2.9 percent had blood-lead concentrations above
25 ug/dL, and three study participants had blood-lead levels above 40 ug/dL. Additional
blood-lead results indicate that
• The distributions of blood-lead concentrations for workers and homeowners were
approximately log-normal. The geometric mean blood-lead concentrations were
5.7 ug/dL for workers and 4.5 ug/dL for homeowners with geometric standard
deviations of 1.95 ug/dL and 2.3 ug/dL, respectively.
• Geometric mean blood-lead concentrations for workers was significantly greater
than those for homeowners (on average 1.3 ug/dL greater with a 95 percent
confidence interval of 1.1 ug/dL to 1.5 ug/dL).
• Covariate-adjusted geometric mean blood-lead concentrations among the worker
groups ranged from 4.2 ug/dL for Laborers to 6.3 ug/dL for Painters. The
covariate-adjusted geometric mean blood-lead concentrations were not significantly
different among the worker groups.
69
-------
• The predicted blood-lead concentrations for each worker group, using a medium
exposure index and parameter estimates from the final predictive model, were
significantly different. In particular, the predicted blood-lead concentration for
Carpenters (6.3 ng/dL) was significantly higher than the predicted blood-lead
concentration for Laborers (4.8 ng/dL).
• Three ancillary variables were related to worker blood-lead concentrations: gender,
tobacco use, and whether room additions or renovations have been made in the
worker's own home. For gender and tobacco use, the effect of estimated parameters
were as expected: male workers and tobacco users had higher blood-lead
concentrations. However, workers who had room additions or renovations made in
their own homes had lower blood-lead concentrations.
• Three ancillary variables were related to homeowner blood-lead concentrations:
gender, race, and age of homeowner. For age and gender, the estimated effect of
each factor was as anticipated: older males had higher blood-lead concentrations.
However, white homeowners had higher blood-lead concentrations than black
homeowners.
For each study group (workers and homeowners) a series of statistical models were fit to
the data to investigate the relationship between blood-lead concentrations and various types of
work or target activities. Initially, separate models were fitted to the data for each target activity
and study group. In all, the models indicated that there was a significant relationship between the
conduct of certain R&R activities and blood-lead concentrations. Specific results based on the
separate, covariate-adjusted, models fit to each target activity and study group are presented
below for each study group.
For workers:
The number of days a worker spent performing general R&R, paint removal/surface
preparation, and cleanup were significantly related to increases in worker
blood-lead concentrations. However, the estimated increases in predicted
blood-lead concentration associated with performing 10 additional days of paint
removal/surface preparation or cleanup in homes built pre-1940 were small:
0.7 ug/dL and 0.8 ug/dL, respectively.
Mid-term exposure, as measured by the number of weeks worked in the past
12 months, was significantly related to increases in worker blood-lead
concentrations for general R&R work and paint removal/surface preparation, and
decreases in worker blood-lead concentrations for carpet removal.
The number of years workers spent performing general R&R was significantly
related to increases in worker blood-lead concentrations.
70
-------
For homeowners:
• The number of days a homeowner spent performing paint removal/surface
preparation was marginally related to increases in homeowner blood-lead
concentrations. As with workers, the estimated increase was small: 0.9 |ig/dL.
• The total number of hours a homeowner spent performing general R&R and paint
removal/surface preparation was significantly related to increases in homeowner
blood-lead concentrations. The estimated increase in homeowner blood-lead
concentrations associated with 80 additional hours of general R&R in a
30-day period was 1.4 ug/dL. The estimated increase associated with 36 additional
hours of paint removal/surface preparation in a 30-day period was 0.9 ug/dL.
• The number of weeks in the last 12 months in which some general R&R or paint
removal/surface preparation was performed was significantly related to increases in
homeowner blood-lead concentrations.
• Significant increases in homeowner blood-lead concentrations were related to the
total number of days in the last 12 months in which some general R&R or paint
removal/surface preparation was performed.
Final models were developed for each study group that included ancillary covariates and
selected target activities. For workers, the final model included at least one target activity for
three exposure periods (short-term, mid-term, and long-term). For homeowners, two alternative
models were developed: one based upon general R&R and the other based upon paint removal/
surface preparation. Results based upon the final models for each study group are presented
below.
For workers:
Although some of the variability in worker blood-lead concentration was explained
by worker group, most of the variability was explained by the combined effect of
the number of days spent performing cleanup, the number of weeks spent
performing paint removal/surface preparation, the number of weeks spent
performing carpet removal, and the number of years in their R&R career where
some time was spent replacing window or door casements.
For Carpenters, Laborers, and Painters, the number of days spent performing
cleanup and the number of years where time was spent replacing window/door
casements were significantly related to increases in blood-lead concentrations.
These same activities for the Other worker group were related to decreased
blood-lead concentrations.
For Carpenters, Painters, and the Other worker group, the number of weeks spent
performing paint removal/surface preparation was significantly related to increased
71
-------
blood-lead concentrations. For Laborers, this activity was related to decreased
blood-lead concentrations.
For all worker groups, the number of weeks spent performing carpet removal was
significantly related to decreases in blood-lead concentrations.
An empirically based index of low, medium, and high potential for lead exposure
was estimated based upon the 25th, 50th, and 75th percentiles of questionnaire
responses from workers on the target activities and exposure periods included in the
final model. For Carpenters, Laborers, and Painters, the model predicts an increase
in blood-lead concentrations from the low to the high index; 1.7 jig/dL for Laborers,
3.3 ug/dL for Carpenters, and 5.7 ug/dL for Painters. However, the model predicts
a decrease for the Other worker group of 1.5 ug/dL. This may be due to the fact
that the Other worker group is largely composed of supervisors, managers, foremen,
etc. and time spent performing some of the activities may not be realistic of actual
hands-on time. Further, it may be that the longer this group of persons work in their
current capacity instead of performing the hands-on work, there would be less
exposure to lead.
For homeowners:
Two models were developed for homeowners. The first was developed using the
five target activities, the second using general R&R. The models fit the homeowner
blood-lead concentrations equally well and explained a similar amount of
variability.
Of the five target activities, paint removal/surface preparation was the single target
activity that explained the most variability in blood-lead levels. Furthermore, no
other target activity had a significant effect over-and-above the effect of paint
removal/surface preparation.
Significant increases in homeowner blood-lead concentrations were related to
increases in short-term intensity (total hours in the last 30 days) and mid-term
duration (weeks in the last 12 months) of conducting paint removal/surface
preparation.
Significant increases in homeowner blood-lead concentrations were related to
increases in short-term intensity (total hours in the last 30 days) mid-term duration
(weeks in the last 12 months) of conducting general R&R.
Low, medium, and high exposure indices were developed from the 25th, 50th, and
75th percentiles of general R&R and paint removal/surface preparation. For the
model using general R&R, the low exposure index was defined to be 42 hours
performing general R&R in the last 30 days and 15 weeks in the last 12 months.
The high exposure index was defined as 100 hours and 40 weeks. For the model
using paint removal/surface preparation, the low index was defined to be 12 hours
72
-------
and 3 weeks, the high index was 50 hours and 20 weeks. Regardless of the model,
increases were predicted in homeowner blood-lead concentrations between the
two indices compared: 2.4 ug/dL for the model with paint removal/surface
preparation and 2.0 ug/dL for the model using general R&R.
The WCBS was a targeted survey of union carpenters and employees of independent
contractors in two cities: Philadelphia, Pennsylvania, and St. Louis, Missouri. Blood and
questionnaire information were collected from 585 participants (581 had complete blood and
questionnaire information). Participants in this study, by design, tended to work in older homes
more than participants in the WCBS did. Thus, it would seem reasonable to assume that
participants in the current study are at higher risk for lead exposure than workers that participated
in the WCBS (i.e., because older homes are more likely to contain lead-based paint). Indeed, the
geometric mean blood-lead level for workers in this study was significantly greater than that of
workers in the WCBS. However, the geometric mean blood-lead level was only 1.3 ug/dL
(95 percent confidence interval from 1.1 ug/dL to 1.4 ug/dL) higher than the geometric mean
blood-lead concentration of workers in the WCBS. The geometric mean blood-lead level for
homeowners in the current study was not significantly different from that of workers in the
WCBS; however, a significantly larger proportion of individuals in this study had elevated
blood-lead concentrations (20 percent above 10 ug/dL compared to 9.0 percent in the WCBS).
In relation to adults in the general U.S. population, the geometric mean blood-lead level
for the study participants was significantly higher than that of adults in the general U.S.
population regardless of race or gender. Furthermore, a significantly larger proportion of
individuals in this study have blood-lead concentrations greater than 10 ug/dL (~20 percent
above 10 ug/dL compared to 3.5 percent in the general U.S. population). Finally, odds ratios
comparing the odds of elevated blood-lead levels among study participants to those of adults in
the general U.S. population were significantly greater than one. In particular, workers were
10.68 times more likely to have blood-lead concentrations above 25 ug/dL than the general U.S.
population. Homeowners were even more likely (28.86 times) to have blood-lead concentrations
above 25 ug/dL than the general U.S. population.
In summary, the results of this study do not contradict those of the WCBS. Overall, the
results demonstrate that there is a significant relationship between the conduct of certain R&R
activities and blood-lead concentrations. Further, for some workers or homeowners, conducting
these activities results in elevated blood-lead concentrations.
73
-------
This page intentionally left blank.
74
-------
APPENDIX A
QUESTIONNAIRES
A-1
-------
This page intentionally left blank.
-------
»n«.l l-l LI I I I I
RECORD #: loll I
WORKER CHARACTERIZATION AND BLOOD-LEAD STUDY
TELEPHONE INTERVIEW SCREENER FOR R&R WORKERS
Hello, my name is
_. Thank you for calling the Study Line. The Battelle
Memorial Institute is conducting a study of people engaged in renovation and remodeling activities
in historic homes. The study is being done for the U.S. Environmental Protection Agency. I'd like
to ask a few questions about renovation and remodeling work you have done. 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?
1 a. Do you do hands-on work? For example do you do
painting, surface preparation such as sanding and
scraping, carpentry, demolition, cleanup or
installation?
IF R ANSWERS "NO" TO Q1 OR Ola. STOP.
YES (CONTINUE)
NO (HH Q2) . .
YES (CONTINUE)
NO (STOP) . .
1
2
1
2
2. 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.
NUMBER OF DAYS
a. [IF Q2 > 0] Of these days worked, how many
were in historic homes, or homes built before
1940?
# DAYS WORKED
3. Altogether in the last 12 months, how many weeks
did you spend doing renovation and remodeling work
in historic homes, or homes built before 1940?
NONE
LESS THAN 1 WEEK
1 - 4 WEEKS
5 - 8 WEEKS ,
9-12 WEEKS
(ABOUT 2 - 3 MONTHS) . .
13-26 WEEKS
(ABOUT 4 - 6 MONTHS) . .
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
01
02
03
04
05
06
07
yi»cr»«n.v1
rav. 10/28/97
-------
IF R ANSWERS 10 DAYS OR MORE TO Q2a AND 9 WEEKS OR MORE TO Q3. CONTINUE.
4. How many years altogether have you earned your
living by working in the renovation and remodeling
industry?
NUMBER OF YEARS
5. In the last 30 days, how many days have you worked
in residential buildings such as homes or apartments?
DAYS IN HOMES
6. In the last 30 days, how many days have you worked
in non-residential buildings such as offices, schools or
government buildings?
DAYS IN NON-RES.
7A. In the last 30 days, have you spent time removing
carpets?
7B. In the last 30 days, have you spent time removing
windows or door casements?
7C. In the last 30 days, have you spent time maintaining,
repairing or cleaning heating, ventilation or air
conditioning systems?
7D. In the last 30 days, have you spent time removing
large structures such as making openings for large
windows or doorways, tearing down ceilings, putting
up walls or removing kitchen cabinets?
7E. In the last 30 days, have you spent time removing
paint or preparing surfaces?
7F. In the last 30 days, have you spent time doing hands-
on cleanup, where you cleaned up the dirt, dust and
debris caused by renovation and remodeling
activities?
YES
NO
YES
NO
YES
NO
, 1
, 2
, 1
2
1
2
YES
NO .
YES
NO .
YES
NO .
1
2
1
2
1
2
8. In your own home, have any room additions or any
major remodeling changes been made in the last 12
months?
YES (CONTINUE) 1
NO (SKIP TO 0.9) 2
8a. Did you live in the house while any of the work was
being performed?
YES
NO .
1
2
9. Approximately what year was your house built?
BUILT SINCE 1990
1
BUILT BETWEEN
1978 AND 1990 2
BUILT BETWEEN
1940 AND 1978 3
BUILT BEFORE 1940 4
rw. 1028/97
-------
10. What is your current job title and what are your main activities at work?
JOB TITLE:
MAIN ACTIVITIES:.
11. Now I'd like to ask you just a few more questions
about yourself. How old are you?
AGE
12. What is your race and ethnic group?
PROBE FOR HISPANIC ORIGIN
SPECIFY:
WHITE. NOT HISPANIC 01
WHITE, HISPANIC 02
BLACK. NOT HISPANIC 03
BLACK. HISPANIC 04
AMERICAN INDIAN . . . 05
ASIAN/PACIFIC ISLANDER 06
OTHER . . . (SPECIFY) 07
13. RECORD GENDER WITHOUT ASKING. IF POSSIBLE.
IF NOT, ASK "Are you male or female?"
MALE 1
FEMALE 2
rav. 1028/97
-------
That completes this portion of the interview. The answers you just gave indicate that you are
eligible for the study. I would like to schedule an appointment for you to complete a short
questionnaire about your renovation and remodeling work and to have a small blood sample taken.
(IF ASKED, IT IS ABOUT A TABLESPOON.) The blood sample will be analyzed for lead levels
only. We will pay you $50 for your cooperation.
We will be holding the data collection in your city on two dates from 4:00 pm to 8:00 pm at
XXXX. (REFER TO SCHEDULE) On which date could you attend? Approximately what time could
you arrive?
RECORD APPOINTMENT INFORMATION ON NEXT PAGE.
May I have your name, address and phone number?
RECORD IDENTIFYING INFORMATION ON NEXT PAGE.
Thank you once again for taking the time to speak with me. If you have any questions or
concerns, the study manager, John Egel, would be happy to speak with you. You can reach him
at 1-800-444-5234, ext. 104.
rw. 10/28/97
-------
ID*:U-U-I I I I I
RECORD APPOINTMENT INFORMATION HERE.
DATE:
TIME:
LOG:
CONFIRM APPOINTMENT TIME AND DATE.
RECORD IDENTIFYING INFORMATION HERE.
NAME:
ADDRESS:
PHONE ( )
RECORD FINAL SCREENING DISPOSITION
Eligible, scheduled 01
Eligible, refused main study .... 02
Not eligible 03
Refused screener 04
Other/specify OS
SPECIFY:
yucrMnvl
wv. 1WZM7
-------
This page intentionally left blank.
-------
IDf;U-U-l 1 I I
RECORD #:
DATE:L_U.U_J.LLJ
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.
Thank you for your help in this important research project.
-------
-1-
SECTION A: DEMOGRAPHICS
A1. How old are you?
AGE
A2. What is your race and ethnic group?
(Check only one.)
WHITE. NOT HISPANIC D 01
WHITE. HISPANIC DO,
BLACK. NOT HISPANIC D „
BLACK. HISPANIC D „
AMERICAN INDIAN D w
ASIAN/PACIFIC ISLANDER . . . D M
OTHER . . . (SPECIFY) DO,
SPECIFY:
A3. Are you male or female?
MALE D,
FEMALE D j
A4. How much schooling have you had? Check the
highest level completed.
GRADES 1 THROUGH 8 LJ ,
SOME HIGH SCHOOL D ,
HIGH SCHOOL/GED D,
APPRENTICESHIP TRAINING . . D 4
SOME COLLEGE/TECH SCHOOL D 5
COLLEGE/TECH SCHOOL GRAD D „
AS. How many children under age 6 currently live with
you in your home?
ff CHILDREN UNDER 6
A6. Do you belong to a union?
YES
NO .
n,
n,
If yes, SPECIFY
QUEST.Y8.v3
rev. 1008/97
-------
-2-
SECTION B: 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.
B1. How old is your home? If you've lived in more than BUILT SINCE 1990 D,
one home during the last 12 months, how old is the
home you lived in the longest? B^EEN 19?8 AND 19go ^
BUILT BETWEEN
1940 AND 1978 D,
BUILT BEFORE 1940 Q4
B2. Have any room additions or any major remodeling YES D,
changes been made to your home in the last 12 N0 Q
months?
IF YOU ANSWERED NO TO QUESTION B2, SKIP TO SECTION C - WORK HISTORY.
A. Was the inside of the house remodeled in the last YES D,
12 months? N0 fjj
B. Was the exterior of the house remodeled in the YES D,
last 12 months? NO Qa
C. Was paint stripped or sanded during the YES D,
remodeling? N0 Q
D. Did the remodeling include major work on the YES D,
kitchen or the bathroom? NQ Q,
E. Did you do any of the work yourself? YES U,
NO D,
If you did any of the work yourself, please describe what you did on the
SPECIFY line below.
SPECIFY —
~~ " OFFICE USE
F. Did you live in the house while the work was YES LJ,
being done? NO d
QUEST_YS.v3
-------
-3-
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.
SECTION C: WORK HISTORY
C1. What is your current job title and what are your main activities at work?
JOB TITLE:
MAIN ACTIVITIES:.
OFFICE USE
C2. During the last 30 days, how many days did you
spend doing any kind of renovation and remodeling
work?
days
C3. During the last 30 days, how many days did you
spend renovating or remodeling historic homes, or
homes built before 1940?
days
C4. During the last 30 days, how many days did you
work in residential buildings (homes, apartments)?
.days
C5. During the last 30 days, how many days did you
work in non-residential buildings (offices, schools,
government buildings)?
days
C6. Altogether in the last 12 months, how many weeks
did you spend doing renovation and remodeling?
NONE
LESS THAN 1 WEEK
1 - 4 WEEKS ,
5 - 8 WEEKS
9-12 WEEKS
(ABOUT 2 - 4 MONTHS) . .
13-26 WEEKS
(ABOUT 4 - 6 MONTHS) . .
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
Do,
n«
n«
Do.
Do.
Do.
.0.
C7. For how many years altogether have you done
renovation and remodeling work?
years
QUEST.YS.v3
wv. 102*97
-------
-4-
C8. Altogether in the last 12 months, how many weeks
did you spend doing renovation and remodeling in
historic homes, or homes built before 1940?
NONE
LESS THAN 1 WEEK
1 - 4 WEEKS
5 - 8 WEEKS
M
9-12 WEEKS
(ABOUT 2 - 3 MONTHS)
13-26 WEEKS
(ABOUT 4 - 6 MONTHS)
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
..D
..D.,
..n.
•0.
• D.
C9. During the last 30 days, how many days did you
spend doing any kind of renovation and remodeling
work in your own home?
days
C10. Altogether in the last 12 months, how many weeks
did you spend doing renovation and remodeling in
your own home?
NONE Da,
LESS THAN 1 WEEK D,,
1 -4 WEEKS D,,
5-8 WEEKS DO,
9-12 WEEKS
(ABOUT 2 - 3 MONTHS) DH
13 -26 WEEKS
(ABOUT 4 - 6 MONTHS) DN
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) . . DOT
QUEST_YS.v3
nv. 100M7
-------
-5-
SECTION D: 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
• removing walls
• removing kitchen cabinets
D1. In the last 30 days, how many days did you remove
large structures?
days
IF YOU ANSWERED NO TO QUESTION D1, SKIP TO QUESTION D4.
D2. In the last 30 days, on how many days did you use a
respirator while you removed large structures?
days
D3. In the last 30 days, how many days did you remove
large structures from historic homes or homes built
before 1940?
days
D4. Altogether in the past 12 months, how many weeks
did you spend removing large structures?
NONE
LESS THAN 1 WEEK
1 - 4 WEEKS
5 - 8 WEEKS
9- 12 WEEKS
(ABOUT 2 - 3 MONTHS) . .
13-26 WEEKS
(ABOUT 4 - 6 MONTHS) . .
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
.DO.
.DC.
05. Think about all the years you've done renovation or
remodeling. In how many of these years did you
remove large structures at least some of the time?
years
IF YOU ANSWERED '0 YEARS' TO QUESTION D5, SKIP TO SECTION E.
QUEST.Y8.v3
OH. 1028/87
-------
-6-
06. Think about all the years you've removed large
structures. In an average year, how many weeks did
you spend removing large structures?
LESS THAN 1 WEEK Q ,
1 - 4 WEEKS D,
5 - 8 WEEKS D,
9- 12 WEEKS
(ABOUT 2 - 3 MONTHS) D 4
13-26 WEEKS
(ABOUT 4 - 6 MONTHS) D,
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) . . D,
D7. When was the last time you removed large
structures?
WITHIN THE LAST 30 DAYS . . D,
MORE THAN 30 DAYS AGO . . D,
MORE THAN A YEAR AGO . . . D,
MORE THAN 5 YEARS AGO . . D«
OUEST.YS.v3
-------
-7-
SECTION E: PAINT REMOVAL AND SURFACE PREPARATION - By surface preparation, we
mean activities such as sanding, scraping, torching, or floor refinishing. (Include work done inside
and outside.)
E1. During the last 30 days, how many days did you days
remove paint or prepare surfaces?
IF YOU ANSWERED NO TO QUESTION E1, SKIP TO QUESTION E6.
E2. During the last 30 days, how many days did you days
remove paint or prepare surfaces in historic homes,
or homes built before 1940?
E3. During the last 30 days, how many days did you use days
a respirator when you removed paint or prepared
surfaces?
E4. During the last 30 days how many days did you days
remove paint or prepare surfaces inside?
E5. When you removed paint in the last 30 davs. how many
days did you use the following?
A. dry power sanding? days
B. dry hand sanding? days
C. dry scraping? days
D. burning, torching, or a heat gun? days
E. wet scraping? days
F. wet sanding? days
G. chemical stripping? days
H. a dust collector when you sanded? days
QUEST.YS.v3 nn. 1CV2W97
-------
-8-
Altogether in the past 12 months, how many week
did you spend doing paint removal or surface
preparation?
NONE
LESS THAN 1 WEEK
1 - 4 WEEKS
5 - 8 WEEKS
Do,
n.
n.
n
9-12 WEEKS
(ABOUT 2 - 3 MONTHS) . .
13-26 WEEKS
(ABOUT 4 - 6 MONTHS) . .
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
E7. Think about all the years you've done renovation and
remodeling. In 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.)
years
IF YOU ANSWERED '0 YEARS' TO QUESTION E7, SKIP TO SECTION F.
E8. 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?
LESS THAN 1 WEEK
1 - 4 WEEKS
5 - 8 WEEKS
9-12 WEEKS
(ABOUT 2 - 3 MONTHS) . .
13-26 WEEKS
(ABOUT 4 - 6 MONTHS) . .
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
D,
n,
a,
a
a,
a,
E9. When was the last time you did any paint removal or
surface preparation?
WITHIN THE LAST 30 DAYS . . D,
MORE THAN 30 DAYS AGO . . D ,
MORE THAN A YEAR AGO . . . D ,
MORE THAN 5 YEARS AGO . . D 4
OUEST_Y3.v3
-------
-9-
SECTION F: WINDOW OR DOOR CASEMENT REPLACEMENT
F1. In the last 30 days, how many days did you remove
windows or door casements?
days
IF YOU ANSWERED NO TO QUESTION F1, SKIP TO QUESTION F4.
F2. In the last 30 days, how many days did you use a
respirator while you removed windows or door
casements?
days
F3. In the last 30 days, how many days did you remove
windows or door casements from historic homes, or
homes built before 1940?
days
F4. Altogether in the past 12 months, how many weeks
did you spend removing windows or door casements?
NONE
LESS THAN 1 WEEK
1 - 4 WEEKS
5 - 8 WEEKS
Do,
9- 12 WEEKS
(ABOUT 2 - 3 MONTHS) . .
13-26 WEEKS
(ABOUT 4 - 6 MONTHS) . .
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
Do.
Do.
D
07
F5. Think about all the years you've done renovation or
remodeling. In how many of these years did you
remove windows or door casements at least some of
the time?
years
IF YOU ANSWERED '0 YEARS' TO QUESTION F5, SKIP TO SECTION G.
F6. 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?
LESS THAN 1 WEEK
1 -4 WEEKS
5 - 8 WEEKS ,
9-12 WEEKS
(ABOUT 2 - 3 MONTHS) . .
13-26 WEEKS
(ABOUT 4 - 6 MONTHS) . .
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
D,
n.
n,
n.
n.
QUEST.Y8.v3
mv. 1U2M7
-------
F7.
-10-
When was the last time you removed windows or
door casements?
WITHIN THE LAST 30 DAYS .. D,
MORE THAN 30 DAYS AGO . . D,
MORE THAN A YEAR AGO . . . Dj
MORE THAN 5 YEARS AGO . . D 4
OUEST.Y8.v3
.100WT
-------
-11-
SECTION G: CARPET REMOVAL
G1. In the last 30 days, how many days did you remove
carpet?
days
IF YOU ANSWERED NO TO QUESTION G1, SKIP TO QUESTION G4.
G2. In the last 30 days, how many days did you use a
respirator while you removed carpet?
days
G3. In the last 30 days, how many days did you remove
carpet from historic homes, or homes that were built
before 1940?
days
G4. Altogether in the last 12 months, how many weeks
did you spend removing carpet?
NONE
LESS THAN 1 WEEK
1 - 4 WEEKS ,
5 - 8 WEEKS
Do,
Do,
D.
Do.
9-12 WEEKS
(ABOUT 2 - 3 MONTHS) . .
13-26 WEEKS
(ABOUT 4 - 6 MONTHS) . .
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
D,
G5. Think about all the years you've done renovation or
remodeling. In how many of these years did you
remove carpet at least some of the time?
years
IF YOU ANSWERED '0 YEARS' TO QUESTION G5, SKIP TO SECTION H.
G6. Think about all the years you've removed carpet. In
an average year, how many weeks did you spend
removing carpet?
LESS THAN 1 WEEK
1 - 4 WEEKS
5 - 8 WEEKS
9-12 WEEKS
(ABOUT 2 - 3 MONTHS) . .
13-26 WEEKS
(ABOUT 4 - 6 MONTHS) . .
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
D,
a
a,
D4
a.
QOE8T.Y8.v3
r*v. 1U2W7
-------
-12-
G7. When was the last time you removed carpet?
WITHIN THE LAST 30 DAYS . . Q,
MORE THAN 30 DAYS AGO . . D,
MORE THAN A YEAR AGO . . . D,
MORE THAN 5 YEARS AGO . . D 4
QUEST.YS.v3
-------
-13-
SECTION H: CLEANUP — By cleanup work, we mean the hands-on 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 doing the work.
H1. During the last 30 days, how many days have you
spent doing hands-on cleanup?
days
IF YOU ANSWERED NO TO QUESTION H1, SKIP TO QUESTION H6.
H2. In the last 30 days, about how many hours in a
typical day did you spend doing hands-on cleanup?
LESS THAN 54 HOUR/DAY . . . D ,
54 TO LESS THAN 1 HOUR/DAY D ,
1 TO 2 HOURS/DAY D ,
3 TO 4 HOURS/DAY ... D 4
MORE THAN 4 HOURS/DAY . . D 8
H3. In the last 30 days, how many days did you spend
doing hands-on cleanup in historic homes or homes
built before 1940?
days
H4. In the last 30 days, how many days did you use a
respirator while doing hands-on cleanup work?
days
H5. Altogether in the past 12 months, how many weeks
did you spend doing hands-on cleanup?
NONE
LESS THAN 1 WEEK
1 - 4 WEEKS
5 - 8 WEEKS
D
D
9-12 WEEKS
(ABOUT 2 - 3 MONTHS) . .
13-26 WEEKS
(ABOUT 4 - 6 MONTHS) . .
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
D
07
H6. Think about all of the years you have done renovation
and remodeling. In how many years altogether have
you spent doing hands-on cleanup work at least some
of the time?
years
QUE8T_Y8.v3
-------
-14-
IF YOU ANSWERED '0 YEARS' TO QUESTION H6, SKIP TO SECTION I.
H7. Think about all the years you've done hands-on
cleanup. In an average year, how many weeks did
you spend doing hands-on cleanup?
LESS THAN 1 WEEK D ,
1 - 4 WEEKS D ,
5-8 WEEKS D,
9-12 WEEKS
(ABOUT 2 - 3 MONTHS) D 4
13-26 WEEKS
(ABOUT 4 - 6 MONTHS) D ,
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS) . . D ,
H8.
When was the last time you did any hands-on
cleanup work?
WITHIN THE LAST 30 DAYS . . D ,
MORE THAN 30 DAYS AGO . . D ,
MORE THAN A YEAR AGO . . . D ,
MORE THAN 5 YEARS AGO . . D 4
QUEST.YS.v3
r*v. 10(2*87
-------
-15-
SECTION I: TOBACCO USE
11. Do you currently use snuff or chewing tobacco?
YES
NO .
D,
n,
12. Do you currently smoke?
YES
NO .
D,
D,
If you do not smoke, go to J1. If you do smoke, please answer questions A through C
below.
A. Do you smoke while you work?
YES
NO .
a
a,
B. Do you smoke while on break?
YES
NO .
a,
a,
C. Do you carry your cigarettes, cigars, pipe or pipe YES
tobacco in your shirt or pants pocket at work? NO .
n,
n,
SECTION J: NON-WORK ACTIVITIES
J1. Do you cast lead into bullets or fishing sinkers?
YES
NO .
D,
D,
J2. Do you go fishing more than 10 times a year?
YES
NO .
D,
D,
If you answered no, skip to question J3. If you do go fishing more than 10 times a year,
answer question A below.
A. Do you crimp fishing sinkers onto the line with
your teeth?
YES
NO .
D,
D,
J3. Do you dismantle car or truck batteries?
YES
NO .
D,
n,
J4. Do you work with stained glass?
YES
NO .
D,
a,
-------
-16-
SECTION K: MEDICAL HISTORY
K1. Have you ever been diagnosed by a health care
professional as having an elevated lead level in your
blood?
YES
NO .
n,
n,
K2. In the last 12 months, have you been diagnosed by a
health care professional as having anemia?
YES
NO .
D,
K3. Has anyone living in your household ever had their
blood tested for lead? Check all that apply.
YES. MYSELF Q,
YES, OTHER ADULT D,
YES. CHILD D4
NO D,
K4. Has anyone living in your household ever been
identified by a health care professional as having an
elevated blood lead level? Check all that apply.
YES. MYSELF D,
YES, OTHER ADULT D,
YES. CHILD D4
NO D,
SECTION L: OTHER ISSUES
L1. When you used a respirator in the last 30 days, what
type did you use? Check all that apply.
DIDN'T USE RESPIRATOR DOI
DUST MASK Da
HALF MASK Dn
FULL FACE MASK D M
TYPE C SUPPLIED AIR MASK . D M
PAPR DO.
SCBA DO,
L2. Has your home ever been tested for lead paint?
YES
NO .
D,
a
QUEST_YS.v3
-------
-17-
L3. Do you think that lead poses a potential problem for YES D,
you at work? N0 Q t
L4. Have you ever received any pamphlets or other YES D,
information about how to reduce potential lead NO D
exposures when conducting renovation and
remodeling work?
L5. Have you ever received any training about how to YES D,
reduce potential lead exposures when conducting N0 d
renovation and remodeling work?
L6. Are there other ways you think you may be exposed
to lead that we haven't asked about?
L7. Is there anything else you'd like to tell us, or do you
have any additional comments you would like to
make?
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, or would like the results of this
study, you can call John Egel at 1-800-444-5234, ext. 104.
QUEST.Y8.v3
-------
. I l-l I.I I I I J
RECORD #: loll J
WORKER CHARACTERIZATION AND BLOOD-LEAD STUDY
TELEPHONE INTERVIEW SCREENER FOR HOMEOWNERS
Hello, my name is
_. Thank you for calling the Study Line. The Battelle
Memorial Institute is conducting a study of people engaged in renovation and remodeling activities
in historic homes. The study is being done for the U.S. Environmental Protection Agency. I'd like
to ask a few questions about renovation and remodeling work you have done on your home. 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?
YES ... (R&R SCREENER) 1
NO (CONTINUE) 2
2. Have any room additions or any remodeling changes
been made to your home in the last 12 months?
YES (CONTINUE) 1
NO (STOP) 2
2a. Did you live in the house while any of the work was
being performed?
YES (CONTINUE) 1
NO (STOP) 2
2b. Approximately what year was your house built?
BUILT SINCE 1990 (STOP) 1
BUILT BETWEEN
1978- 1990...(STOP) 2
BUILT BETWEEN
1940 AND 1978..(STOP) 3
BUILT BEFORE 1940 4
2c. Did you do a large portion (more than 25%) of the
hands-on work yourself? For example did you do
surface preparation (PROBE: for example sanding and
scraping), painting, carpentry, demolition, cleanup or
installation?
YES
NO .
(CONTINUE)
. (STOP) ..
1
2
rav. KV28/97
-------
2d. In the last 30 days, how many days did you spend
doing major renovation and remodeling work in your
home?
NUMBER OF DAYS
HOURS PER DAY
2e. In the last 30 days, about how many hours did you
spend doing renovation and remodeling on a typical
day?
IF ANSWER TO Q2d MULTIPLIED BY THE RESPONSE TO Q2e > 20 HOURS, CONTINUE.
3. During the last 12 months, in how many weeks did
you do some kind of major renovation and remodeling
activities in your home?
1 - 4 WEEKS
5 - 8 WEEKS
9-12 WEEKS
(ABOUT 2 - 3 MONTHS)
1
2
13 - 26 WEEKS
(ABOUT 4 - 6 MONTHS) . .
MORE THAN 26 WEEKS
(MORE THAN 6 MONTHS)
IF ANSWER TO 03 > 9 WEEKS, CONTINUE.
4. Do you routinely perform major renovation and
remodeling activities in homes other than your own?
(For example, in friends' and relatives' homes.)
4a. In the last 30 days, how many days did you spend
doing any kind of renovation and remodeling work in
homes other than your own?
YES (ANSWER 4a) 1
NO (SKIP TO 5a) 2
NUMBER OF DAYS
5a. In the last 30 days, have you spent time removing
carpets?
5b. In the last 30 days, have you spent time removing
windows or door casements?
5c. In the last 30 days, have you spent time maintaining,
repairing or cleaning your heating, ventilation or air
conditioning system?
5d. In the last 30 days, have you spent time removing
large structures such as making openings for large
windows or doorways, tearing down ceilings, putting
up walls or removing kitchen cabinets?
5e. In the last 30 days, have you spent time removing
paint or preparing surfaces?
5f. In the last 30 days, have you spent time doing the
hands-on cleanup, where you cleaned up the dirt,
dust and debris caused by the renovation and
remodeling activities?
YES
NO .
YES
NO .
YES
NO .
YES
NO .
YES
NO .
YES
NO .
1
2
1
2
1
2
1
2
1
2
1
2
rav. KV2U97
-------
6. Now I'd like to ask you just a few questions about yourself. What is your current job title
and what are your main activities at work?
JOB TITLE:
MAIN ACTIVITIES:.
7. How old are you?
AGE
8. What is your race and ethnic group?
PROBE FOR HISPANIC ORIGIN
SPECIFY:
WHITE. NOT HISPANIC 01
WHITE, HISPANIC 02
BLACK. NOT HISPANIC 03
BLACK, HISPANIC 04
AMERICAN INDIAN 05
ASIAN/PACIFIC ISLANDER 06
OTHER . . . (SPECIFY) 07
9. RECORD GENDER WITHOUT ASKING, IF POSSIBLE.
IF NOT, ASK "Are you male or female?"
MALE 1
FEMALE 2
That completes this portion of the interview. The answers you just gave indicate that you are
eligible for the study. I would like to schedule an appointment for you to complete a short
questionnaire about your renovation and remodeling work and to have a small blood sample taken.
(IF ASKED, IT IS ABOUT A TABLESPOON.) The blood sample will be analyzed for lead levels
only. We will pay you $50 for your cooperation.
We will be holding the data collection in your city on two dates from 4:00 pm to 8:00 pm at
XXXX. (REFER TO SCHEDULE) On which date could you attend? Approximately what time could
you arrive?
RECORD APPOINTMENT INFORMATION ON NEXT PAGE.
May I have your name, address and phone number?
RECORD IDENTIFYING INFORMATION ON NEXT PAGE.
Thank you once again for taking the time to speak with me. If you have any questions or
concerns, the study manager, John Egel, would be happy to speak with you. You can reach him
at 1-800-444-5234, ext. 104.
rav. KV2OT7
-------
-------
This page intentionally left blank.
-------
in* I LI I.I I I I I
RECORD #: loll I
DATErLU-LLJ-LU
MM DD YY
HOMEOWNER EXPOSURE STUDY
-------
This questionnaire collects basic information about you, your renovation and
remodeling activities, and some of your hobbies and other 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.
Thank you for your help in this important research project.
-------
-1 -
SECTION A: DEMOGRAPHICS
A1. How old are you?
AGE
A2. What is your race and ethnic group?
(Check only one)
WHITE, NOT HISPANIC D 01
WHITE, HISPANIC H^
BLACK, NOT HISPANIC DO,
BLACK, HISPANIC D „
AMERICAN INDIAN DO,
ASIAN/PACIFIC ISLANDER ...Do,
OTHER . . . (SPECIFY) D „
SPECIFY:
A3. Are you male or female?
MALE D,
FEMALE D ,
A4. How much schooling have you had? Check the
highest level completed.
GRADES 1 THROUGH 8 D ,
SOME HIGH SCHOOL D 2
HIGH SCHOOL/GED D ,
APPRENTICESHIP TRAINING . . D «
SOME COLLEGE/TECH SCHOOL D 8
COLLEGE/TECH SCHOOL GRAD De
A5. How many children under age 6 currently live with
you in your home?
# CHILDREN UNDER 6
SECTION B: 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.
B1. How old is your home? If you've lived in more than
one home during the last 12 months, how old is the
home you lived in the longest?
BUILT SINCE 1990 D ,
BUILT BETWEEN
1978 AND 1990 D,
BUILT BETWEEN
1940 AND 1978 D,
BUILT BEFORE 1940
B2. Have any room additions or any major remodeling
changes been made to your home in the last 12
months?
YES
NO .
Was the inside of the house remodeled in the
last 12 months?
YES
NO .
QUEST.Y3.HH
rev. 10(28/97
-------
-2-
B. Was the exterior of the house remodeled in.
the last 12 months?
YES
NO .
D,
C. Was paint stripped or sanded during the
remodeling?
YES
NO .
D. Did the remodeling include major work on the
kitchen or the bathroom?
YES
NO .
E. Did you do a large portion (more than 25%)
of the hands-on work yourself?
YES
NO .
If you did any of the work yourself, please describe what you did on the
SPECIFY lines below.
SPECIFY:
OFFICE USE
F. Did you live in the house while any of the
work was being done?
YES
NO .
n,
D,
QUESTYS.HH
rav. 1008/87
-------
-3-
SECTION C: RENOVATION AND REMODELING WORK HISTORY
C1. During the last 30 days, how many days did you days
spend doing any kind of renovation and remodeling
work in your home?
C2. During the last 12 months, in how many weeks did weeks
you do some renovation and remodeling in your
home?
C3. in the past 12 months have you done renovation and YES D,
remodeling work in homes other than in your own NO D,
residence? (For example friends and relatives)
IF YOU ANSWERED NO TO QUESTION C3, THEN SKIP TO QUESTION C4.
A. During the last 30 davs. how many days did days
you work in homes OTHER THAN YOUR
OWN that were built before 1940?
B. During the last 30 davs. how many davs did days
you work in homes OTHER THAN YOUR
OWN that were built in or after 1940?
C. During the last 12 months, in how many weeks
weeks did you do some renovation and
remodeling in homes OTHER THAN YOUR
OWN that were built before 1940?
D. During the last 12 months, in how many weeks
weeks did you do renovation and remodeling
in homes OTHER THAN YOUR OWN that
were built in or after 1940?
C4. During the last 30 davs. about how many hours did hours per day
you spend working on a typical renovation and
remodeling work day? (In your own home, or any
other home.)
C5. During the last 12 months, how many davs did you days per week
do renovation and remodeling work in a typical week?
C6. During the last 12 months, how many hours did you hours per day
do renovation and remodeling work in a typical
renovation and remodeling work day?
QUEST_YS.HH rav. 1008/97
-------
-4-
SECTION D: 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
• removing walls
• removing kitchen cabinets
D1. Have you ever removed large structures (as defined YES D,
in the instructions above}? NO Q,
IF YOU ANSWERED NO TO QUESTION D1, THEN SKIP TO QUESTION D10.
D2. During the last 30 davs. did you remove large YES D,
structures from a home? N0 Q
IF YOU ANSWERED NO TO QUESTION D2, THEN SKIP TO QUESTION D6.
D3. In the last 30 davs. how many davs did you remove
large structures from:
a) Your home? days
b) OTHER homes (not yours) built before days
1940?
days
c) OTHER homes (not yours) built in or after
1940?
D4. On a typical day within the last 30. when you hours
removed large structures, about how many hours did
you perform the work?
D5. In the last 30 davs. on how many of those days did days
you use a respirator while removing large structures?
D6. During the past 12 months, in how many weeks did weeks
you spend any time removing large structures?
IF YOU ANSWERED '0 WEEKS' TO QUESTION D6, THEN SKIP TO QUESTION D9.
D7. In a typical week in the past 12 months when you days per week
removed large structures, on about how many davs
did you perform the work?
QUeST_YS.HH «W. 108*97
-------
-5-
D8. On a typical day in the past 12 months when you
removed large structures, about how many hours did
you perform the work?
hours per day
D9. When was the last time you removed large
structures? (Include only when you did the work
yourself.)
WITHIN THE LAST 30 DAYS . . D ,
MORE THAN 30 DAYS AGO . . D ,
MORE THAN A YEAR AGO . . . D ,
MORE THAN 5 YEARS AGO . . D 4
D10. Has someone else, such as a contractor, ever
removed large structures from your home?
YES
NO .
D,
D,
IF YOU ANSWERED NO TO QUESTION 010, SKIP TO SECTION E.
011. When was the last time someone else, such as a
contractor, removed large structures from your
home?
WITHIN THE LAST 30 DAYS . . D ,
MORE THAN 30 DAYS AGO . . D ,
MORE THAN A YEAR AGO . . . D a
MORE THAN 5 YEARS AGO . . D 4
QUEST.YS.HH
iwv. 10128/97
-------
-6-
SECTION E: PAINT REMOVAL AND SURFACE PREPARATION - By surface preparation, we
mean activities such as sanding, scraping, torching, or floor refinishing. (Include work done inside
and outside.)
E1 . Have you ever removed paint or prepared surfaces YES
for painting? NO
IF YOU ANSWERED NO TO QUESTION E1, THEN SKIP TO QUESTION E12.
E2. During the last 30 davs. did you remove paint or YES D
prepare surfaces for painting? N0 Q
IF YOU ANSWERED NO TO QUESTION E2, THEN SKIP TO QUESTION E8.
E3. In the last 30 davs. how many days did you remove
paint or prepare surfaces for painting in:
a) Your home? days
b) OTHER homes (not yours) built before days
1940?
days
c) OTHER homes (not yours) built in or after
1940?
E4. On a typical day within the last 30. when you hours
removed paint or prepared surfaces for painting,
about how many hours did you perform the work?
E5. During the last 30 davs. on how many davs did you days
use a respirator while removing paint or preparing
surfaces for painting?
E6. During the last 30 days when you removed paint or days
prepared surfaces, how many davs did you do this
inside?
QUEST_YS.HH *•". 1WW/S7
-------
-7-
E7. When you removed paint in the last 30 days, how many
days did you use the following?
A. dry power sanding?
days
B. dry hand sanding?
days
C. dry scraping?
days
D. burning, torching, or a heat gun?
days
E. wet scraping?
days
wet sanding?
days
G. chemical stripping?
days
H. a dust collector when you sanded?
days
E8. During the past 12 months, in how many weeks did
you spend any time doing paint removal or surface
preparation?
weeks
IF YOU ANSWERED '0 WEEKS' TO QUESTION E8, THEN SKIP TO QUESTION E11.
E9. In a typical week in the past 12 months when you
removed paint or prepared surfaces for painting, on
about how many days did you perform the work?
days per week
E10. In a typical day in the past 12 months when you
removed paint or prepared surfaces for painting,
about how many hours did you perform the work?
hours per day
E11. When was the last time you did any paint removal or
surface preparation? (Include only when you did the
work yourself.)
WITHIN THE LAST 30 DAYS . . D ,
MORE THAN 30 DAYS AGO . . D ,
MORE THAN A YEAR AGO . . . D ,
MORE THAN 5 YEARS AGO . . D 4
E12. Has someone else, such as a contractor, ever
removed paint or prepared surfaces for painting in
your home?
YES
NO .
D,
QUEST_YS.HH
rav. 10OW97
-------
-8-
IF YOU ANSWERED NO TO E12. SKIP TO SECTION F.
E13. When was the last time someone else, such as a
contractor, removed paint or prepared surfaces for
painting in your home?
WITHIN THE LAST 30 DAYS . . D ,
MORE THAN 30 DAYS AGO . . D ,
MORE THAN A YEAR AGO . . . D ,
MORE THAN 5 YEARS AGO . . D4
QUEST_YS.HH
. 1008197
-------
-9-
SECTION F: WINDOW OR DOOR CASEMENT REPLACEMENT
F1. Have you ever removed any windows or door
casements?
YES
NO .
D,
IF YOU ANSWERED NO TO QUESTION F1, THEN SKIP TO QUESTION F10.
F2. During the last 30 days, did you remove windows or
door casements?
YES
NO .
n,
IF YOU ANSWERED NO TO QUESTION F2, THEN SKIP TO QUESTION F6.
F3. In the last 30 days, how many days did you remove
windows or door casements from:
a) Your home?
days
b) OTHER homes (not yours) built before
1940?
c) OTHER homes (not yours) built in or after
1940?
days
days
F4. On a typical day within the last 30. when you
removed windows or door casements, about how
many hours did you perform the work?
hours
F5. In the last 30 days, on how many of those days did
you use a respirator while removing windows or door
casements?
days
F6. During the past 12 months, in how many weeks did
you spend any time removing windows or door
casements?
weeks
IF YOU ANSWERED '0 WEEKS' TO QUESTION F6, THEN SKIP TO QUESTION F9.
F7. In a typical week in the past 12 months when you days per week
removed windows or door casements, on about how
many days did you perform the work?
OUEST.YS.HH
rev. KV2W97
-------
-10-
F8. On a typical day in the past 12 months when you
removed windows or door casements, about how
many hours did you perform the work?
hours per day
F9. When was the last time you removed windows or
door casements? (Include only when you did the
work yourself.)
WITHIN THE LAST 30 DAYS . . D,
MORE THAN 30 DAYS AGO .. D,
MORE THAN A YEAR AGO . . . D,
MORE THAN 5 YEARS AGO . . D4
F10. Has someone else, such as a contractor, ever
removed windows or door casements from your
home?
YES
NO .
D,
D,
IF YOU ANSWERED NO TO QUESTION F10, SKIP TO SECTION G.
F10. When was the last time someone else, such as a
contractor, removed windows or door casements
from your home?
WITHIN THE LAST 30 DAYS .. D,
MORE THAN 30 DAYS AGO . . D,
MORE THAN A YEAR AGO . . . D,
MORE THAN 5 YEARS AGO .. D4
QUEST_Y8.HH
-------
-11-
SECTION G: CARPET REMOVAL
G1. Have you ever removed carpet?
YES
NO .
D,
D,
IF YOU ANSWERED NO TO QUESTION 61, THEN SKIP TO QUESTION G10.
G2. During the last 30 days, did you remove carpets?
YES
NO .
D,
D,
IF YOU ANSWERED NO TO QUESTION G2, THEN SKIP TO QUESTION G6.
G3. In the last 30 days, how many days did you remove
carpet from:
a) Your home?
days
b) OTHER homes (not yours) built before
1940?
days
c) OTHER homes (not yours) built in or after
1940?
days
G4. t On a typical day within the last 30. when you
removed carpet, about how many hours did you
perform the work?
hours
G5. In the last 30 davs. on how many of those days did
you use a respirator while removing carpet?
days
G6. During the past 12 months, how many weeks did you
spend removing carpet?
weeks
IF YOU ANSWERED '0 WEEKS' TO QUESTION G6, THEN SKIP TO QUESTION G9.
G7. In a typical week in the past 12 months when you
removed carpet, on about how many davs did you
perform the work?
days per week
QU6ST.YS.HH
r«». KV28/97
-------
-12-
G8. In a typical day in the past 12 months when you
removed carpet, about how many hours did you
perform the work?
hours per day
G9. When was the last time you removed carpets?
(Include only when you did the work yourself.)
WITHIN THE LAST 30 DAYS . . Q,
MORE THAN 30 DAYS AGO . . D,
MORE THAN A YEAR AGO . . . D,
MORE THAN 5 YEARS AGO . . D4
G10. Has someone else, such as a contractor, ever
removed carpets from your home?
YES
NO .
B;
IF YOU ANSWERED NO TO QUESTION G10, SKIP TO SECTION H.
G11. When was the last time someone else, such as a
contractor, removed carpets from your home?
WITHIN THE LAST 30 DAYS .. D,
MORE THAN 30 DAYS AGO . . D,
MORE THAN A YEAR AGO . .. D,
MORE THAN 5 YEARS AGO . . D4
QUEST.YS.HH
.102M7
-------
KECOKD02
SECTION H: CLEANUP — By cleanup work, we mean the hands-on 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 doing the work.
H1 . Have you ever done any hands-on cleanup (as YES ................... D ,
defined in the above instructions)? NO .................... D ,
IF YOU ANSWERED NO TO QUESTION H1, THEN SKIP TO QUESTION H10.
H2. In the last 30 days, have you ever done any hands-on YES D,
cleanup? NO D,
IF YOU ANSWERED NO TO QUESTION H2, THEN SKIP TO QUESTION H6.
H3. In the last 30 days, how many days did you do
hands-on cleanup in:
a) Your home? days
b) OTHER homes (not yours) built before days
1940?
days
c) OTHER homes (not yours) built in or after
1940?
H4. On a typical day within the last 30. when you did hours
hands-on cleanup, about how many hours did you
perform the work?
H5. In the last 30 davs. how many days did you use a days
respirator while doing hands-on cleanup?
H6. Altogether in the past 12 months, how many weeks weeks
did you spend doing hands-on cleanup?
IF YOU ANSWERED '0 WEEKS' TO QUESTION H6, THEN SKIP TO QUESTION H9.
H7. In a typical week in the oast 12 months when you days per week
did hands-on cleanup, on about how many davs did
you perform the work?
QUEST_Y3.HH rev. 10*2097
-------
-14-
H8. In a typical day in the past 12 months when you did
hands-on cleanup, how many hours did you spend
doing the work?
hours per day
H9. When was the last time you did hands-on cleanup?
(Include only when you did the work yourself.)
WITHIN THE LAST 30 DAYS . . D ,
MORE THAN 30 DAYS AGO . . D ,
MORE THAN A YEAR AGO . . . D ,
MORE THAN 5 YEARS AGO . . D 4
H10. Has someone else, such as a contractor, ever done
cleanup of renovation and remodeling work in your
home?
YES
NO .
n,
IF YOU ANSWERED NO TO QUESTION H10, SKIP TO SECTION I.
H11. When was the last time someone else, such as a
contractor, did cleanup of renovation and remodeling
work in your home?
WITHIN THE LAST 30 DAYS . . D ,
MORE THAN 30 DAYS AGO . . D 2
MORE THAN A YEAR AGO . . . D ,
MORE THAN 5 YEARS AGO . . D 4
QUEST_YS.HH
wv. 1008/97
-------
-15-
KICOKD02
SECTION I: TOBACCO USE
11. Do you currently use snuff or chewing tobacco? YES i
NO D,
12. Do you currently smoke? YES D,
NO D2
If you do not smoke, go to J1. If you do smoke, please answer questions A through C
below.
A. Do you smoke while you do renovation and YES D,
remodeling work? NO D,
B. Do you smoke while taking a break from YES D,
renovation and remodeling work? NO D,
C. Do you carry your cigarettes, cigars, pipe or YES D,
pipe tobacco in your shirt or pants pocket NO D,
while you do renovation and remodeling
work?
SECTION J: HOBBIES
J1. Do you cast lead into bullets or fishing sinkers? YES D,
NO D,
J2. Do you go fishing more than 10 times a year? YES D,
NO D,
If you answered no, skip to question J3. If you do go fishing more than 10 times a year,
answer question A below.
A. Do you crimp fishing sinkers onto the line YES D,
with your teeth? NO D 2
J3. Do you dismantle car or truck batteries? YES D,
NO D,
J4. Do you work with stained glass? YES D,
NO D,
«WT.V8.HH
mv. KVM/97
-------
-16-
SECTION K: MEDICAL HISTORY
K1. Have you ever been diagnosed by a health care
professional as having an elevated lead level in your
blood?
YES
NO .
K2. In the last 12 months, have you been diagnosed by a
health care professional as having anemia?
YES
NO .
D,
n,
K3. Has anyone living in your household ever had their
blood tested for lead? Check all that apply.
YES, MYSELF D ,
YES, OTHER ADULT D 3
YES, CHILD D4
NO D,
K4. Has anyone living in your household ever been
identified by a health care professional as having an
elevated blood lead level? Check all that apply.
YES, MYSELF D ,
YES, OTHER ADULT D ,
YES, CHILD D «
NO D,
SECTION L: OTHER ISSUES
LI. When you used a respirator in the last 30 days, what
type did you use? Check all that apply.
DIDN'T USE RESPIRATOR D 0,
DUST MASK DM
HALF MASK D M
FULL FACE MASK DO,
TYPE C SUPPLIED AIR MASK . D M
PAPR DO.
SCBA Dm
L2. Has your home ever been tested for lead paint?
L3. Do you think that lead poses a potential problem for
you when you are doing renovation and remodeling
work?
YES
NO
YES
NO
D,
D,
n,
D,
QUeST_YS.HH
rav. 10(28/97
-------
-17-
L4. Have you ever received any pamphlets or other YES D,
information about how to reduce potential lead N0 [jj
exposures when conducting renovation and
remodeling work?
L5. Are there other ways you think you may be exposed
to lead that we haven't asked about?
KECOKD02
L6. Is there anything else you'd like to tell us, or do you
have any additional comments you would like to
make?
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, or if you would like the results of
this study, you can call John Egel at 1-800-444-5234, ext. 104.
. 10/28/97
-------
This page intentionally left blank.
-------
APPENDIX B
ADDITIONAL QUESTIONNAIRE TABLES
B-1
-------
This page intentionally left blank.
-------
Table B-1. Main Activities and Job Titles for Each Worker Group
Job Tide
'" Mali* ActTvftTas
Main ActJvitfei for Carpenters"
Technician
Carpenter and Plasterer
Carpenter
Carpenter
Home Improvement Worker
Carpenter
Carpenter
Carpenter/welder
Carpenter
Carpenter
Carpenter
Renovate Old House
Total Restoration
Restoration
Do Work and Watch Other Workers
Remodel & Repairs
Repair Old Windows and Doors; ornamental Plaster
Carpenter
Carpentry
Building; remodeling
Demolition
Restoring Old Houses
Home Renovation & Remodeling
All Phases of Remodeling
Renovation
Framing, Trim, Repair, Etc...
Renovations; painting
Jill of All Things
Remodeling (Framing Drywall Trim Plumb. Electric); Demolition
Carpentry; prep; painter
Rehabbing Rental Properties
Replace Molding; scrape; paint
Renovate Homes; plumbing
Residential Remodeling
Renovation; cleaning; painting
Carpentry Repairs, Additions
Hanging Windows & Doing Floors & Doors
Maintenance & Repairs
Demolition; new Additions; HVAC Installation
Painting; drywall; framing
Painting; scraping
Improvement
Home Improvement
Missing
Renovations and Restorations
B-2
-------
Table B-1. Main Activities and Job Titles for Each Worker Group (Continued)
Job Tide
I
; Main Activities
Carpenter
Builder
Carpenter
Carpenter & Mason Tender
Carpenter
Removing Lead Paint
Remodeling
Home Improvement Contractor
General Home Improvements
Remodeling; replace Window, Doors; tear out Walls, Floors
Remodel/take down Walls & Windows; replace Walls; replace Windows
Paint; masonry; demo
Remodeling Homes
Everything
Remodeling; tearing Paneling; removing Paint
Missing
Missing
Home Improvement
Main Activities for Laborers
Demolition
Helper
Carpenter Helper
Laborer
Building Laborer
Carpenter Helper
Laborer
Tearing down Foundation
Tearing Down; building up
Whatever Boss Says-clean up
Clean up
Labor-clean up
Clean up
Laborer
Drive-clean up
Cleaning
Cleanup; nail Studs
Form Builder
Window & Doors, Roors
Painting; sheet Rock; laying Pipes
Help Scrape Paint; put up Walls; put down Carpet
Laborer-construction
Pulling Wood-ceilings Etc.
Scrape Paint; clean up
Cleans Up; tears down
Cleaning; demolishing Houses
Demolition; hauling
Demo; drywall & Insulation; paint
B-3
-------
Table B-1. Main Activities and Job Titles for Each Worker Group (Continued)
•:;:H::;:;:::?-«::Jbb:11tle &-:
Labor/helper
Laborer
Temporary Labor
Renovation Helper
Laborer
Laborer
General Laborer Helper
Temporary Laborer
Laborer
Maintenance Associate-
laborer
Labor & Helper
Laborer
(Main Activities
Taking down Windows; helping Paint
Painting
Demolition; clean up
Clean up
All Phases of Rehab-remodel
All Activities
Remove; rebuild Walls
Demolition
Demolition; clean up
Paint; clean; sheetrock
Construction
Gutting out Houses
Scrape Walls, Duct Work; removing Frames; clean up
Gral Repairs-remodeling
Demolition
Demolition
Clean up
Prep; paint; carpet
Demolition
Demolition
Painting
Assistant Laborer
Demolition
^ 3 Main Activities for Paint*r»x^^v >*^' -*•-<' > '
Painting; sand Blasting
Painting
Painting
Prep Work for Painting; painting
Painting; sanding & Stripping
Painting; supervising
Prep Work for Painting; cleanup
Prep Work; spray Paint and Painting
Paint; prep; demolition
Painting; sanding; blowing Ceiling
Painting; preparation to Paint
B-4
-------
Table B-1. Main Activities and Job Titles for Each Worker Group (Continued)
Job Tide
Main Activities
Painter
Painter Remodeling
House Painter
Painter
Painting
Painting
Light Painting; cleanup; strip Paint off Windows
Paint Houses
Sanding off Paint
Grinding Pressure Washing; painting
Grinding, Pressure Washing; painting
Tearing Out; cleaning Up; painting
Preping; painting
Prep; painting; glazing & Window Replacement
Paint; prep
Painter Prep
Grinding, Sanding, Scraping & Pressure Washing; painting
Remodeling; painting; plaster
Painting; remodeling (Most Everything That Goes along W/it)
Painting; remodeling
Plaster; painting; remodeling
Painting; remodeling
Scraping, Sanding; caulking. Glazing
Preparing Surfaces for Paint; vinyl Siding
Painting; sanding
Paint; clean up after Remodeled
Demolition; painting
Clean up Sometime Office; painting
Clean Up; paint Removal; paint
Painting; demolition; sanding
Painter; drywall Rnisher
Demolition; painting
Clean Up; paint; sand
Painting
Painting
Plaster; paint Removal/prep
Painting; sand Blasting; roofing
Paint; remove Old Paint
Scrape Paint, Remove Chipping _
B-5
-------
Table EM. Main Activities and Job Titles for Each Worker Group (Continued)
Job Title
Main Activities
Main Activities for Other Workers
President of R&r Company
Registered Architect
Ornamental Plasterer
Demolition Foreman
Plumber
Millworker
Small Business Owner
Warehouse Manager for
Chas. Repairs
Construction Inspector
(Commercial)
Plumber, Remodeling
Lead Man
Job Forman
Lead Abatement Worker
Laborer
Dry Wall Labor
Plumber
Owner
Lead Paint Technician
Home Improvement
Contractor
Iron Worker
President-property Mgmt.
Co.
Construction Worker
Brick Layer-owner
Handy Man Pipe Fitter
Bricklayer
Run Field Operations-mostly Finishing Work
Construction Admin/ep Production/design
Restoring Ornamental Plasterwork
Tearing down
Cut out Old Pipes & Put in New
Repairing Shutters; removing Old Paint
Heating/air Conditioning
Material Procurement; jobsite Safety; clean up
Assure Const. To Specs/plans Sign off Insp. If Pas
Plumbing
Tell People What to Do
Sheetrocking; painting; refurbushing of Wood Floors, Mantles, & Stairways
Removal; washdown Hepa-vac; painting
Sheetrock
to Get up Dry Wall
Install Copper Water Pipe & Waste Pipes
Run Concrete; home Improvement Bus.
Removal of Lead from Window Sills, Doors, Etc.
HVAC; plumbing; drywall
Drywalls; building Decks
Welding
Renovates
Construction
Cleaning Brick
Cutting Old Pipe Out; remodeling Old Houses
Cleaning Brick, Masonry; lead Paint Removal
B-6
-------
Table B-2. Summary of Responses for Questions Pertaining to R&R Target Activities for Each Worker Group
{< ;
<$$'* ^^&L' W&f/-" *.'-, . . .. f
Days performing the activity In the last 30
days
Daya performing tha activity in Pre-1940
housing In the last 30 daya
Daya using a respirator while performing
activity
Number of weeks spent performing activity
in last year
Number of years apent performing activity
over cereer
Number of weeks apent performing activity
In average year
25th Percentile
Mean
76th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
< 1 Week
1-8 Weeks
> 8 Weeks
25th Percentile
Mean
75th Percentile
< 1 Week
1-8 Weeks
> 8 Weeks
'•jjJMfr '''-'/' ' *' / ''''''" ' Target Activities T ,,/^"', ',- ', ', ',,,
^||i|i,
Replace fniJRyt'
^ CARPENTER fy", ,' :/*'
-------
Table B-2. Summary of Responses for Questions Pertaining to R&R Target Activities for Each Worker Group (Continued)
' ' ' '"' f
™'< " Variable Description
f
, 'statistic^
$*y*f&^ - -'„ Target Activities '-/'"'-
Ganerol
> fi&R-
5twge
Structure
Paint
Removal
Window
Replacement
Carpet
Removal
Cleanup
.i,i& «*$8"i ' "' - ' * &* . "£,''> '"LABORER
Days performing the activity in the last 30
days
Days performing the activity In Pre-1 940
housing in the last 30 days
Days using a respirator while performing
activity
Number of weeks spent performing activity
in last year
Number of years spent performing ectivity
over career
Number of weeks spent performing activity
in average year
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
< 1 Week
1-8 Weeks
> 8 Weeks
25th Percentile
Mean
75th Percentile
< 1 Week
1-8 Weeks
> 8 Weeks
20
22.86
27
18
20.14
23
* -- w :
t W
w ,, ,
0.0%
9.1%
90.9%
5
10.84
15
W '<
f
M -
W '
7
14.07
21
4
11.26
20
0
6.07
10
9.1%
29.5%
61.4%
3
7.56
12
23.3%
26.6%
51.2%
6
14.02
20
5
12.36
20
0
7.59
13
9.1%
29.5%
61.4%
2
7.74
12
19.5%
26.8%
53.7%
2
8.23
12
0
7.07
10
0
3.91
5
22.7%
43.2%
34.1%
2
6.79
10
38.5%
20.5%
41.0%
0
4.73
6
0
4.11
5
0
2.45
1
47.7%
29.5%
22.7%
1
4.67
7
42.9%
28.6%
28.6%
11
18.68
25
9
15.27
20
0
6.95
15
2.3%
20.5%
77.3%
2
7.12
10
9.8%
22.0%
68.3%
•p
GO
-------
Table B-2. Summary of Responses for Questions Pertaining to R&R Target Activities for Each Worker Group (Continued)
>« •>
•** - \
\
x- - ..
'<,.:;. Variable Description
^ , % /£&
£ ,•** *
v ,^
- "- Statistic
'^iV Target Actfyjjies *
Qaneral,
R&R s
PAINTER
11 Days performing the activity in the last 30
II j-.,.,
II aays
I
| Days performing the activity in Pre-1 940
housing in the last 30 days
Days using a respirator while performing
activity
I Number of weeks spent performing activity
in last year
Number of years spent performing activity
over career
Number of weeks spent performing activity
In average year
25th Percentile
Mean
75th Percentile
25th Percentite
Mean
75th Percentile
25th Percentite
Mean
75th Percentile
< 1 Week
1-8 Weeks
> 8 Weeks
25th Percentile
Mean
75th Percentile
< 1 Week
1-8 Weeks
> 8 Weeks
23
24.75
29
17
21.80
27
; w
w
M
f^
0.0%
4.5%
95.6%
4
11.95
17
W
i
^m
f
*
u
Large
Structure
Paint ,
Removal
;^flndow
Replacement
Carpet
Removal
Cleanup
•> % '
0
9.48
15
0
7.93
13
0
3.95
7
29.5%
29.5%
40.9%
2
6.66
10
21.1%
44.7%
34.2%
15
20.07
25
10
16.43
24
0
6.16
10
0.0%
13.6%
86.4%
4
10.39
15
4.5%
15.9%
79.5%
0
5.82
10
0
5.07
10
0
3.70
6
34.1%
45.5%
20.6%
1
6.74
9
47.2%
33.3%
19.4%
0
3.98
6
0
2.91
4
0
1.64
0
63.6%
25.0%
11.4%
0
4.53
8
62.1%
24.1%
13.8%
10
17.80
25
10
15.16
21
0
5.00
10
2.4%
33.3%
64.3%
3
9.84
15
14.3%
16.7%
69.0%
CO
CD
-------
Table B-2. Summary of Responses for Questions Pertaining to R&R Target Activities for Each Worker Group (Continued)
*
•. * f
Variable Description
Statistic
> \ , - Tflr0e$ Activities , - - \ s s *
Genera} ^
R*R *•
' J ^ i$- & \ v^,- <, , "• - " -. .. v OTHER ••
Days performing the activity in the last 30
days
Days performing the activity In Pre-1940
housing In the lest 30 days
Days using a respirator while performing
activity
Number of weeks spent performing activity
in last year
Number of years spent performing activity
over career
Number of weeks spent performing activity
In average year
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
25th Percentile
Mean
75th Percentile
< 1 Week
1-8 Weeks
> 8 Weeks
25th Percentile
Mean
75th Percentile
<1 Week
1-8 Weeks
> 8 Weeks
20
23.08
28
15
20.50
25
w
w
MS
0.0%
3.8%
96.2%
7
14.62
20
w
w
Ul
^ large
/Structure
%
Paint
Removal
Window
Replacement
Carpet
Removal
cleanup
-"7" '••*vr
-------
This page intentionally left blank.
-------
APPENDIX C
STATISTICAL MODEL TABLES
c-i
-------
This page intentionally left blank.
-------
Table C-1. Summary of Relationships Between Covariates and Blood-Lead Concentrations for Workers
.• x- x f'&f- *j&* <$*->%' fv
'I *%>'<5. "A
' - <> '"
Covariate Description f: ••-
Worker's Age
Race (All Levels)
Race (White/Other)
Gender
Education (All Levels)
Education (High School or
Less/More than High School)
Union Status (Yes/No)
Age of Home (Pre 1940, 1940-
1978, Post-1978)
Room Additions or Renovation in
Own Home (Yes/No)
Cast Lead into Bullets or Fishing
Sinkers (Yes/No)
Dismantle Truck Batteries
(Yes/No)
Work With Stained Glass
(Yes/No)
Smoke, Use Snuff, or Chewing
Tobacco (Yes/No)
Fishing (No/Yes, Crimp Sinkers
with Teeth/Yes, Do Not Crimp
Sinkers)
< , " '" Combined -> '
F
1.51
1.47
0.02
8.58
0.46
0.01
0.19
1.22
5.13
0.02
1.32
1.28
4.65
0.52
P-Value
0.222
0.203
0.876
0.004
0.806
0.908
0.665
0.299
0.025
0.897
0.253
0.260
0.033
0.594
Baltimore. ..
*:• -
3.12
0.86
0.70
8.09
0.10
0.16
0.14
1.41
1.02
1.35
4.22
1.17
2.85
0.13
,P-Value
0.081
0.510
0.404
0.006
0.992
0.691
0.710
0.250
0.315
0.249
0.043
0.282
0.095
0.878
Charleston «sv,"<*
'•*' ..
F -~
1.75
6.36
9.37
4.97
1.33
1.75
—
1.11
4.63
1.55
0.32
0.25
4.36
2.63
4p?VfiiuV^
0.193
0.004
0.004
0.032
0.278
0.194
—
0.338
0.038
0.220
0.577
0.621
0.043
0.085
Savannah
F
1.02
2.37
2.37
0.05
1.45
2.08
—
0.14
0.20
0.37
1.83
<.0001
0.14
0.02
P-Value
0.322
0.136
0.136
0.833
0.253
0.162
—
0.870
0.662
0.547
0.188
0.991
0.716
0.983
o
Kl
-------
Table C-1. Summary of Relationships Between Covariates and Blood-Lead Concentrations for Workers (Continued)
*&r
- *s
'<*-.*
-• - Covariate Description
Any Exposure Through Non-
IWork Activities (Yes/No)
Number of Non-Work Activities
Respirator Use (None/Respirator)
Respirator Use (None/Dust
Mask/Other Respirator)
I Large Structure Removal:
Number of Days Used Respirator
In Last 30 Days
Paint Removal and Surface
Preparation: Number of Days
Used Respirator In Last 30 Days
I Window or Door Casement
Replacement: Number of Days
Used Respirator In Last 30 Days
Carpet Removal: Number of Days
Used Respirator In Last 30 Days
Cleanup: Number of Days Used
|| Respirator In Last 30 Days
Received Information on
Reducing Lead Exposures
(Yes/No)
Received Training on Reducing
Lead Exposures (Yes/No)
Combined
F
1.22
0.673
0.02
0.50
0.77
1.00
0.55
2.89
0.635
0.14
0.72
P-Value
0.271
0.416
0.875
0.606
0.383
0.320
0.460
0.091
0.427
0.708
0.397
Baltimore
F
1.38
5.47
0.32
0.16
0.36
0.158
0.03
2.39
0.005
0.03
0.36
P-Value
0.243
0.028
0.573
0.852
0.550
0.693
0.858
0.126
0.945
0.860
0.551
Charleston
F
1.93
1.67
0.80
2.93
0.51
0.20
0.34
0.89
0.003
0.11
0.61
P~Value
0.173
0.223
0.377
0.066
0.481
0.654
0.565
0.352
0.954
0.743
0.439
Savannah ,
F
0.30
3.09
0.42
0.29
3.53
1.11
11.81
0.44
3.70
0.48
0.03
P-Value
0.591
0.153
0.522
0.752
0.072
0.302
0.002
0.514
0.065
0.494
0.864
o
u
-------
Table C-2. Summary of Relationships Between Covariates and Blood-Lead Concentrations for Homeowners
- Covarfate Description
Worker's Age
Race (All Levels)
Race (White/Other)
Gender
Education (All Levels)
Education (High School or
Less/More than High
School)
Age of Home (Pre 1 940,
1940-1978, Post-1978)
Room Additions or
Renovation in Own Home
(Yes/No)
Done Renovation or
Remodeling in Other
Homes (Yes/No)
Cast Lead into Bullets or
Fishing Sinkers (Yes/No)
Dismantle Truck Batteries
(Yes/No)
Work With Stained Glass
(Yes/No)
Combined
',-,F
7.75
1.61
3.90
12.26
1.01
0.95
0.58
4.59"1
0.45
0.02""
2.95"
1 .09(d)
PA/aiue %
0.007
0.194
0.052
0.001
0.395
0.333
0.563
0.035
0.503
0.898
0.090
0.299
Baltimore
<>>F
9.34
0.23
0.23
6.71
0.06
0.01
0.82
1.69
1.1
—
4.20
0.01
P-Value
0.004
0.636
0.636
0.014
0.980
0.914
0.451
0.202
0.295
—
0.049
0.919
Charleston X'*X T^,
- , •> f - .
0.55
0.99
0.99
1.01
0.90
— 1»
—
™~
0.13
—
~
0.55
>P-Vafue^
0.467
0.331
0.331
0.328
0.354
~"
—
"™
0.721
—
--
0.468
Savannah -
F
1.53
1.34
3.27
3.14
1.42
1.56
—
2.98
1.88
0.04
—
5.95
P-Value
0.224
0.286
0.083
0.089
0.263
0.224
--
0.097
0.183
0.842
—
0.023
o
-------
Table C-2. Summary of Relationships Between Covarlates and Blood-Lead Concentrations for Homeowners (Continued)
x >w >,§?Xv,V . ,
^ ^: lp^
Covarinta DeibriDtlon
Smoke, Use Snuff, or
Chewing Tobacco
(Yes/No)
1 Fishing (No / Yes, Crimp
Sinkers with Teeth / Yes,
1 Do Not Crimp Sinkers)
Any Exposure Through
Non-Work Activities
(Yes/No)
Contractor Did Large
Structure Removal in
Home
Contractor Did Paint
Removal / Surface
Preparation
Contractor Removed
Window or Door
Casements
Contractor Removed
Carpet
Contractor Performed
Cleanup
Respirator Use
(None/Respirator)
Respirator Use (None/Dust
II Mask/Other Respirator)
•. f+nri
uon
F
1.42
0.4218'
0.04
1.94
0.31
2.84
1.33
1.57
3.05
1.60
nblned
P-Value
0.238
0.657
0.845
0.168
0.732
0.096
0.252
0.213
0.085
0.209
Baltimore ,x< f%«
F
0.38
~
2.00
0.13
0.17
0.18
2.78
0.04
1.04
0.51
P-Value
0.541
--
0.167
0.722
0.842
0.610
0.105
0.837
0.315
0.606
, *** - Charleston
^l^iH* x
1.89
0.26
0.46
1.78
<0.01
3.02
1.47
0.01
0.85
0.69
N3&&&SXS
P-VahuF®'
0.185
0.770
0.505
0.197
0.953
0.097
0.240
0.940
0.369
0.515
- "• -- „ Savannah
F " '
2.54
0.72
0.25
2.55
1.84
3.41
2.46
4.13
2.96
1.43
P-Value
0.124
0.498
0.621
0.123
0.188
0.077
0.130
0.053
0.098
0.260
o
ui
-------
Table C-2. Summary of Relationships Between Covariates and Blood-Lead Concentrations for Homeowners (Continued)
fvrry^;;^
Covariafe Description
Large Structure Removal:
Number of Days Used
Respirator In Last 30 Days
Paint Removal and Surface
Preparation: Number of
Days Used Respirator In
Last 30 Days
Window or Door
Casement Replacement:
Number of Days Used
Respirator in Last 30 Days
Carpet Removal: Number
of Days Used Respirator In
Last 30 Days
Cleanup: Number of Days
Used Respirator In Last 30
Days
Received Information on
Reducing Lead Exposures
(Yes/No)
Combined - " -
F
0.53
2.93
0.743
0.03
1.41
0.20
P-Value
0.469
0.091
0.391
0.867
0.239
0.656
Baltimore
F
0.18
4.69
0.581
<*_
3.87
0.35
P-Value
0.671
0.038
0.452
•—
0.058
0.556
^ s Charleston ^ ^ ,
F
0.90
1.11
0.34
0.25
0.33
0.19
P-Value
0.354
0.305
0.566
0.623
0.572
0.669
<, * Savannah %*4"1
F
0.10
1.13
0.04
™"
0.04
<0.01
P-Vaiue
0.758
0.299
0.842
""
0.845
0.953
o
O)
(a) Only three homeowners indicated that there were no room additions made to their home in the last year.
(b) Only two homeowners indicated that they cast lead into bullets or fishing sinkers.
(c) Only two homeowners dismantled truck or car batteries.
(d) Only three homeowners indicated that they work with stained glass.
(e) Only two homeowners indicated that they lived in home built after 1940.
(f) AN homeowners in Charleston had more than a high school education.
(g) Only six homeowners indicated that the went fishing more than 10 times a year (and only two indicated that they crimp sinkers onto the line with
their teeth).
-------
Table C-3. Summary of Univariate Relationships Between Conduct of Target Activity and Worker Blood-Lead
Concentrations for Each Target Activity*
jj*^' , *
'''••.•"• I
Yl| - '
^. Target Activity
General R&R
Large Structure Removal
Paint Removal/ Surface Preparation
Window/Door Casement Replacement
Carpet Removal
Cleanup
< Short Terra
(Days of Pre-1940 Activity) ^
Beta
0.0354
0.0019
0.0126
0.0158
-0.0094
0.0160
Std.
Error
0.0088
0.0059
0.0060
0.0079
0.0104
0.0057
P-value
0.000
0.751
0.039
0.047
0.367
0.006
Mid-Term
(Weeks In Last 12 Months)
Beta
0.1390
0.0005
0.0906
0.0416
-0.0749
0.0570
Std.
Error
0.0585
0.0285
0.0304
0.0281
0.0343
0.0330
P-value
0.019
0.985
0.003
0.141
0.031
0.086
Long Term
(Years over Career)
Beta
0.0192
0.0144
0.0163
0.0145
-0.0055
0.0164
Std. .
Error
0.0065
0.0074
0.0079
0.0077
0.0085
0.0072
P-value
0.003
0.055
0.039
0.063
0.516
0.024
9
-4
There are a total of 18 models being fit (one for each target activity and exposure period combination).
Table C-4. Summary of Univariate Relationships Between Conduct of Target Activity and Worker Blood-Lead
Concentrations for Each Target Activity (Adjusted for Ancillary Covariates)*
a
&S 44
, # -r
jj^Sf "^ *' *•*?* ^'£.
Target Activity * ^ ^%
General R&R
Large Structure Removal
Paint Removal/ Surface Preparation
Window/Door Casement Replacement
Carpet Removal
Cleanup
^^ Short farm
:XV (Days of Pre-1940 Activity)
f
Beta
0.0323
0.0011
0.0141
0.0135
-0.0066
0.0177
Std.
•• Error
0.0085
0.0056
0.0058
0.0077
0.0101
0.0054
P'ValUBv
0.000
0.852
0.016
0.080
0.513
0.001
Mid-Term
(Weeks In Last 12 Months)
,»$
- Beta ;
0.1693
-0.001 1
0.0839
0.0225
-0.0809
0.0515
|LStd«
>x Error
0.0549
0.0274
0.0289
0.0279
0.0326
0.0316
P-value
0.002
0.969
0.004
0.421
0.014
0.106
Long Term v
(Years over Career)
•. Beta
0.0156
0.0097
0.0137
0.01 1 1
-0.0089
0.0124
Std. Error
0.0066
0.0075
0.0076
0.0076
0.0082
0.0071
£'
: P-value
0.019
0.198
0.073
0.149
0.277
0.080
* There are a total of 18 models being fit (one for each target activity and exposure period combination).
-------
Table C-5. Summary for Each Target Activity of the Covariated Adjusted Relationship Between Worker Blood-Lead
Concentration and the Combined Effect of All Three Exposure Period Measures'
'<,,
*<•* , **f #* -
, TWQrt Activity^
General R&R
Large Structure Removal
Paint Removal/Surface Preparation
Window/Door Casement Replacement
Carpet Removal
Cleanup
Short Term - **<|&,
(Dnys1nLa«3Q)^|||,
, Beta
0.0209
-0.003
0.0436
0.0025
-0.065
0.0113
k Std,
En/or
0.0123
0.0218
0.0184
0.0444
0.0545
0.0124
P'Valua
0.094
0.899
0.020
0.956
0.240
0.364
^' Mld-T»rm
q ^(W0eks In Last Year)
" SfvjT
Bet*
0.0125
0.001 6
0.0148
-0.0089
-0.0296
0.0001
Std, *
Errpr »
0.0067
0.0144
0.0095
0.0192
0.0266
0.0070
-. ">
•>
P'value
0.068
0.912
0.122
0.642
0.269
0.989
Shprt Term
- (Hours. In Last 3.0 Pays)
\ •.
' Beta
0.0049
-0.0016
0.0077
-0.0007
-0.0069
-0.0015
^std,,;
Error
0.0019
0.0036
0.0028
0.0057
0.0071
0.0029
^& * ; v
; •* t *
P'Vdue
0.010
0.653
0.007
0.899
0.336
0.594
Mid Term „
(OnyslnLwtl^lVJpntlw^;
Beja;
0.0030
0.0016
0.0028
-0.0028
-0.0050
0.0009
* Stdn
Crrof
0.0012
0.0034
0.0024
0.0052
O.OOS5
0.001 5
P-value
0.017
0.642
0.242
0.594
0.367
0.551
* There are a total of 24 models being fit (one for each target activity and exposure period combination).
-------
Table C-7. Summary of Univariate Relationships Between Conduct of Target Activity and Homeowner Blood-Lead
Concentrations for Each Target Activity (Adjusted for Ancillary Covariates)'
^ j-,»
H'
<$| T*g« Activity '•><"?
General R&R
Large Structure Removal
Paint Removal/Surface Preparation
Window/Door Casement Replacement
Carpet Removal
Cleanup
Short Term „
^ (Day* In Last 30) ^
> f ^
%Betfc
0.0199
0.0161
0.0328
0.0181
-0.038
0.0162
Std.
' Errpr
0.0111
0.0213
0.0167
0.0416
0.0512
0.0110
P'Voluetl
0.077
0.4SO
0.054
0.665
0.457
0.145
Mid-Term
||;;; (W«ek«lnLe*tY«r)
PL
0.0157
0.0059
0.0199
-0.0008
-0.0229
0.0049
Std,
Error
0.0059
0.0145
0.0083
0.0192
0.0249
0.0063
P-valuB
0.010
0.686
0.020
0.969
0.361
0.433
Short Term
IHowr? In la$t 30 D«y»)
Beta
0.0046
0.0012
0.0069
0.0020
-0.0051
0.0017
SMI.
prior
0.0017
0.0035
0.0024
0.0053
0.0066
0.0027
P-ValMe
0.007
0.726
0.006
0.704
0.446
0.521
'J9Wr<
*&t< Mid Term
-(Days In Lost 12 Month?)
f V
^5" '
Beta
0.0034
0.0025
0.0045
-0.0011
-0.0043
0.0020
<,$|d-
y Error
0.001 1
0.0033
0.0021
0.0049
0.0051
0.001 3
P-valua
0.002
0.461
0.039
0.827
0.401
0.121
o
,' Xfe,,.i \ •*
f " ' TBrfl«tAcdvttyT«^
General R&R
Large Structure Removal
Paint Removal/Surface Preparation
Window/Door Casement Replacement
Carpet Removal
CleanuD
Short Tnrm ^
•"&& (0«y« In L«»t 30) P
^•ta
0.0130
0.0105
0.0207
0.0278
0.0347
0.0154
Std.
Error
0.0113
0.0278
0.0180
0.0498
0.1342
0.0119
P-valulf
0.251
0.708
0.255
0.578
0.797
0.198
a^. - ?r' MW-T«rm ^}S
it>
-------
Table C-9. Summary for Exposure Period of the Covariate Adjusted Relationship Between Worker-Blood Lead
Concentration and the Combined Effect of All Five Target Activities Exposure Variables'
^•*s-'»
%• /
>> ^
vX ' ^
Target Activity " ' "
General R&R
Large Structure Removal
Paint Removal/ Surface Preparation
Window/Door Casement Replacement
Carpet Removal
Short Term
(Day? of Pr«'1940 Activity)
^ <
Beta
-0.0125
0.0102
0.0152
-0.0237
0.0212
^:std« ,
'Error x
0.0064
0.0064
0.0086
0.0106
0.0064
P-value
0.052
0.111
0.080
0.027
0.001
Mid-Term , '
(Week? in Last 12 Months)
; Beta
-0.0204
0.0952
0.0472
-0.1315
0.0049
Std.
Error
0.0308
0.0310
0.0355
0.0362
0.0054
• -< ^ <
P-vaiueV
0.509
0.003
0.186
0.000
0.363
Long Term , ^
(Years ovsr Career)
Beta
-0.0032
0.0109
0.0220
-0.0253
0.0597
Std.
- Error
0.0159
0.0106
0.0155
0.0100
0.0321
P-value
0.839
0.304
0.156
0.013
0.065
o
o
* There are a total of three models being fit. Each model represents a single period of exposure and the table reads down the three columns.
Table C-10. Summary for Exposure Period of the Relationship Between Worker Blood-Lead Concentration and the Combined
Effect of All Five Target Activities Exposure Variables, After Adjusting for the Effects of Covariates and
Worker Group*
>. S ^f r •. -f ^
„* - '.
.%r • , -*. ' t •
*H < ,
~#£ Target Activity
General R&R
Large Structure Removal
Paint Removal/ Surface Preparation
Window/Door Casement Replacement
Carpet Removal
Short Term
(Days of Pre-1940 Activity) >
Beta
-0.0115
0.0074
0.0151
-0.0191
0.0204
Std,
Error
0.0064
0.0064
0.0086
0.0106
0.0064
P-value
0.074
0.253
0.082
0.074
0.002
VMid-Terrrj^Lj^
(Weelcs in Last 12 Months.)^,
>t$
H
Beta *'
-0.0221
0.0859
0.0391
-0.1173
0.0063
$Xc».|/V
V ***S8&^S'
^ Erro?ff
0.0313
0.0327
0.0356
0.0366
0.0054
;; pivalue
0.481
0.010
0.273
0.002
0.241
(Years oveiTOaffliflil^x
-\-
' Beta,
-0.0003
0.0054
0.0212
-0.0272
0.0672
Std-
Error
0.0163
0.0110
0.0152
0.0099
0.0319
«.-S3S>
P-value..-
0.984
0.629
0.164
0.007
0.037
There are a total of three models being fit. Each model represents a single period of exposure and the table reads down the three columns.
-------
Table C-11. Summary for Exposure Period of the Relationship Between Worker Blood-Lead Concentration and the Effect of
Significant Target Activities Exposure Variables, After Adjusting for the Effects of Covariates and Worker
Group*
: ^ ^ V£% ,
•".•••• i
* ^ s. •• y^vj > ^v-. -. •"
• . '- - , -"' , £<
Target Activity x$
General R&R
Large Structure Removal
Paint Removal/ Surface Preparation
Window/Door Casement Replacement
Carpet Removal
^< Short Term ^ ^<
(Days of Pre-1 940 Activity)
''• 4 ^
^ B0ta
--
~
~
~
0.017
$td, '•-
Error -
»
—
-
~
0.005
>^-
P-value
--
—
~
-
0.002
Mid-Term < I
(Weeks in Last 12 Months)
> Beta
—
0.099
—
-0.101
-
std,
, Error
-
0.030
«
0.033
~
^ _. f
P-vafu*
-
0.001
-
0.003
~
; -\ , Jlong Term ^
"' ' (Years over Career)
'J ', •• ^
* ^$% -1 -
TB«t^^'
—
—
0.023
-0.026
-
std,
Error
-
-
0.009
0.010
-
P-value
-
—
0.012
0.007
-
o
* The results from three models are being presented, each model represents a single period of exposure and the Table reads down the three columns.
-------
Table C-12. Final Predictive Model for Worker Blood-Lead Concentrations
, Variable
Intercept
Gender
Room Additions in
Own Home
Use Tobacco
Products
Worker Group
Carpet Removal
Worker Group *
Days Spent
Performing Cleanup
Worker Group *
Weeks Spent
Performing Paint
Removal/Surface
Preparation (PR/SP)
Worker Group *
Years Spent
Performing
Window/Door
Casement
Replacement
(W/DCR)
Description
Male
Female1"
Yes
No111
No
Yes"1
Carpenter
Laborer
Painter
Other1"
Weeks Spent in Last 1 2 Months
Carpenter * Cleanup days
Laborer * Cleanup days
Painter * Cleanup days
Other * Cleanup days
Carpenter • PR/SP Weeks
Laborer * PR/SP Weeks
Painter *PR/SP Weeks
Other • PR/SP Weeks
Carpenter * W/DCR Years
Laborer * W/DCR Years
Painter * W/DCR Years
Other * W/DCR Years
Effect on Log
(blood-fead)
1.632
0.480
—
-0.345
—
-0.181
—
-0.516
-0.195
-1.101
—
-0.141
0.019
0.022
0.037
-0.017
0.089
-0.079
0.160
0.121
0.023
0.058
0.010
-0.010
Standard
Error
0.358
0.141
—
0.010
-
0.094
—
0.428
0.411
0.537
—
0.033
0.010
0.011
0.010
0.012
0.054
0.060
0.068
0.064
0.012
0.018
0.014
0.014
P-Value
< 0.001
<0.001
—
<0.001
—
0.056
—
0.230
0.636
0.041
—
<0.001
0.050
0.052
< 0.001
0.142
0.098
0.188
0.020
0.062
0.050
0.002
0.468
0.498
(a) Reference level
C-12
-------
Table C-13. Alternative Predictive Model for Worker Blood-Lead Concentrations Based
Upon General R&R
. Variable
Intercept
Gender
Room Additions in
Own Home
Use Tobacco
Products
Worker Group
Short-Term Exposure
Mid-Term Exposure
Long-Term Exposure
4&&r%, ,->y^'^ '-"
- • -^vft-l <-^
-------
Table C-14. Summary for Exposure Period of the Covariate Adjusted Relationship Between Homeowner Blood-Lead
Concentration and the Combined Effect of All Five Target Activities Exposure Variables*
, *k,
^Target Activity
Large Structure Removal
Paint Removal/Surface Preparation
Window/Door Casement
Replacement
Carpet Removal
Cleanup
***££l£m
Bat«
0.0053
0.0265
0.0332
-0.085
0.0080
'Std,
* Error'
0.0272
0.0195
0.0717
0.0724
0.0124
P.
value
0.845
0.179
0.645
0.243
0.620
Mid-Term
(Week* in Ust Your)
,^4-
/!•»•>
0.0026
0.0255
-0.0023
-0.0390
-0.0044
Std.
Error
0.0168
0.0106
0.0259
0.0317
0.0072
P-
valua
0.881
0.018
0.930
0.223
0.543
^ ' " Short Term <
riHours tolas* 30 PaV»r
Beta
0.0005
0.0083
0.0045
-0.0178
0.0004
Std. ,
Error
0.0049
0.0028
0.0082
0.0095
0.0028
,v,
value
0.915
0.004
0.587
0.066
0.898
Mid Term
(Days In Last 12 Months)
^ Beta.
0.0020
0.0042
0.0058
-0.0166
0.001 1
Std, I
Error ""
0.0047
0.0027
0.0158
0.0149
0.0016
P-
value
0.668
0.127
0.714
0.269
0.477
o
• There are a total of four models being fit (two for each exposure period). The table reads down the four columns.
Table C-15.
Summary for Exposure Period of the Covariate Adjusted Relationship Between Homeowner Blood-
Lead Concentration and the Effect of Significant Target Activities Exposure Variables'
>>
»
- Activity &
Paint Removal/ Surface
Preparation
Exposure Period
Short-Term Exposure
Mid-Term Exposure
"" >
Measurement
Days in last 30
Total Hours in Last 40 days
Weeks in Last 1 2 Months
Total Days In Last 1 2 Months
s * \ ^«s -statistic;; ^r ' «- *^ISx\" „
Beta ,
0.033
0.007
0.020
0.005
•• ^ '•?>\¥
st^&nr <«&1
0.017
0.002
0.008
0.002
\ "^ P^Valua
0.054
0.006
0.020
0.039
' Four models are being represented in the table which reads across the four rows.
-------
Table C-16. Final Predictive Model for Homeowner Blood-Lead Concentrations
•^
' Variable
Intercept
Gender
Ethnicity
Age of Homeowner
Paint Removal/
Surface Preparation
Paint Removal/
Surface Preparation
Description
Male
Female1"
Non-White
White'11
Total hours spent in the last 30
days
Number of weeks in the last 1 2
months
{> *,«>X-- **?*
Effect dhs Log:
». . . 'J#^AUy3&
(blood-leadl «
0.011
0.585
—
-0.313
—
0.022
0.006
0.014
Standard
I Error
0.327
0.165
—
0.212
—
0.007
0.003
0.009
P-Value
0.972
<0.001
—
0.144
—
0.002
0.033
0.122
(a) Reference level
Table C-17. Alternative Predictive Model for Homeowner Blood-Lead Concentrations
Using General R&R
Intercept
Gender
Ethnicity
Age of Homeowner
General R&R
General R&R
^j^^^^^^f^°^^^;^^&
Male
Female"1
Non-White
White1"
Total hours spent in the last 30
days
Number of weeks in the last 1 2
months
Effect on Log
, {blood-lead}
-0.375
0.545
—
-0.324
—
0.025
0.013
0.004
Standard
, Error .
0.393
0.168
—
0.217
—
0.007
0.006
0.002
P-Value
•, \
0.343
0.002
—
0.140
—
0.001
0.038
0.035
(a) Reference level
C-15
-------
APPENDIX D
ADDITIONAL FIGURES
D-1
-------
This page intentionally left blank
-------
?
K>
60
50
40
30
20
10
0 0.1
2.3 15.9 50.0 84.1 97.7
Cumulative Normal Probability
99.9 100
Figure D-1. Normal Probability Plot of Worker Blood-Lead Concentrations (straight line indicates normality).
-------
9
u
64
32
16
8
\
o 2
CD
0.1
2.3 15.9 50.0 84.1
Cumulative Normal Probability
97.7 99.9 100
Figure D-2. Semi-log Probability Plot of Worker Blood-Lead Concentrations (straight line indicates normality).
-------
60
50
40
30
20
I
o 10
CO
0
0.1
2.3 15.9 50.0 84.1 97.7
Cumulative Normal Probability
99.9 100
Figure D-3. Normal Probability Plot of Homeowner Blood-Lead Concentrations (straight line indicates normality).
-------
o
Ul
64.01
32.0
16.0
8.0
4.0
S 2.0
I
i
ffi 1.0
0.5
0.1
2.3 15.9 50.0 84.1 97.7
Cumulative Normal Probability
99.9 100
Figure D-4. Semi-log Probability Plot of Homeowner Blood-Lead Concentrations (straight line indicates normality).
-------
9
o>
64.0
32.0
IT
T3
^ 16.0
!8.0
4.0
f 2.0
^
s 1.0
0.5
+ i
I
I
.
r I
i
1 1
1
,
1
*
*
i
i
1
i
+
x/
Baltimore
Charleston
Savannah
All Cities
Figure D-5. Boxplot of Blood-Lead Concentrations for Each Study Group and City.
-------
8
64.0
32.0
16.0
8.0
4.0
"8
¥ 2.0
m 1.0
0.5
S
Figure D-6. Boxplot of Blood-Lead Concentrations by Worker Group.
-------
9
00
64.0
32.0
I 16'°
I
8.0
4.0
m
fc
2.0
1.0
0.5
* -i, * ^
** ** *
*gS
3F ^
* * *
*
*
10 20 30 40 50 60 70 80 90
AGE
* * « Observed
Predicted
Figure D-7. Scatterplot and Predicted Homeowner Blood-Lead Concentration by Age.
-------
CD
ft
64.0
32.0
16.0
8.0
4.0
2.0
1.0
0.5
X
Baltimore
Charleston
Savannah
All Cities
Figure D-8. Boxplot of Homeowner Blood-Lead Concentration by City and Race.
-------
_ 64.0
^^
o*
-^ 32.0
|
1 16*°
§
-a 8.0
T
9 "§4.0
0 55
| 2.0
1 1.0
0.5
+
i
i
l>,
i
I
+
/- X <** X
I
+
^ ^
N.^ vy^ sy »!r
^r ^r >r
-------
64 \
32
16
°TD o
(8 o
i
m 4
i
YES NO
Baltimore
YES NO
Charleston
YES NO
Savannah
YES NO
All Cities
Figure D-10. Boxplot of Worker Blood-Lead Concentration by City and Incidence of Home Renovation/Remodeling in
Last 12 Months.
-------
64
32
16
8
CQ 4
i
YES NO
Baltimore
YES NO
Charleston
YES NO
Savannah
YES NO
All Cities
Figure D-11. Boxplot of Worker Blood-Lead Concentration by City and Tobacco Use.
-------
64
32
16
8
CD
Baltimore
/•
Charleston
Savannah
All Cities
Rgure D-12. Boxplot of Worker Blood-Lead Concentration by City and Gender.
-------
APPENDIX E
RECRUITMENT MATERIALS,
HUMAN SUBJECTS APPROVAL, AND
INFORMED CONSENT
E-1
-------
This page intentionally left blank.
-------
BAII tlit
NEWS
RELEASE
Robin Yoem
Telephone 6UU24.5544
WilKopp
Ittephoi
krtomrt
SBaitene
For Immediate Release
BATTELLE TO EXAMINE BLOOD-LEAD LEVELS
IN RENOVATORS AND REMODELERS OF HISTORIC HOMES
Battelle is conducting a study of the potential lead exposure hazards associated with
renovating and remodeling old or historic homes.
The study, funded by the U.S. Environmental Protection Agency, will be performed in
Charleston, S.C, Savannah, Ga., Baltimore, and Washington, D.C These cities were selected
because of the large numbers of old homes and renovations projects.
Ben Pierce, a Battelle researcher and head of the study, said the object of the study is to
determine whether any particular work activities or practices are associated with elevated blood-
lead levels.
Pierce is seeking homeowners and workers to participate in the study. Participants are
asked to fill out a questionnaire and allow a small blood sample to be taken. Those who
participate will receive a stipend of $50.
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.
Renovators ingest lead through airborne particles in the workplace but can reduce the
likelihood of lead exposure by taking precautions such as wearing a respirator, wetting down
surfaces before working, avoiding dry power sanding, sealing off work areas, and washing
hands before eating.
Interested candidates should contact Battelle Study Manager Joan Huber al 80C.444.5234.
Founded in 1929, Battelle is an international technology organization based in
Columbus, Ohio. With laboratories and offices around die world, Battdle serves industry and
government by developing, commercializing and managing technology. Battelle's primary
business areas are environmental, medical, national defense, transportation, and commercial and
Industrial technology.
f. r/W*"fa>'i
SOS King Awnu»
Colunbui. OWo 43201-2693
Pa* |6U) 42*3889 (614)424.;
TOTfiL P. 132
-------
pBatreiie
Renovation and Remodeling
Lead Exposure Study
... Purring Technology 7b Work
Centers for Public Health
Research and Evaluation
401 N. Lindbergh Boulevard, Suit* 330
St. Louis. Missouri 63141 -7839
Telephone (3 U) 993-5234
Fax (314)993-5163
Who's
conducting
the study and
why?
Who can
participate?
The Battelle Memorial Institute is studying the potential lead exposure hazard
associated with renovation, remodeling, restoration and preservation in old or
historic homes. The study is being conducted on behalf of the U.S.
Environmental Protection Agency (EPA). The objectives of this study are (1) to
determine the relationship (if any) between renovation and remodeling
activities in historic homes and lead exposure to those doing the work, and (2)
to gather information on the types of work activities and work practices
engaged in by people who do renovation and remodeling of old or historic
homes. The study will take place in Charleston, Savannah and Baltimore.
Participants in the study will be people who perform the hands-on work of
renovation, remodeling, restoration and preservation in old or historic homes.
We are looking for workers who make their living at renovation and
remodeling, as well as homeowners who are themselves renovating or
remodeling their houses.
What will a
study
participant
do?
Is the data
collection
confidential?
When a prospective study participant calls the 800 number below, we will ask a
few questions to determine if he or she is eligible for the study. If so, we will
then ask him/her to come to Johns Hopkins University on June 26 or July 15 at
5501 Hopkins Bayview Circle for a 45-minutc data collection session. He/she
will complete a questionnaire about his/her renovation and remodeling work
history. After completion of the questionnaire, a staff member of the hospital
blood lab will collect a small blood sample (approximately 1 tablespoon). The
blood sample will later be analyzed for lead content only. We will pay the
participant S50 as compensation for completion of the questionnaire and the
blood sample.
Extensive safeguards will be used to protect the confidentiality of information
obtained from this study. A participant identification number will be assigned
to each questionnaire and blood sample and will be the only identifier
associated with that information.
Who do I call
for more
information?
You can call Battelle in St Louis. MO. The toll-free number is 1-800-444-5234.
What is the
Battelle
Memorial
Institute?
Battelle is a research institute founded in 1929 and headquartered in Columbus,
Ohio. It has offices around the world engaged in a broad range of science-
based services in the areas of commercial and industrial technology, health,
environment, national security, and transportation.
-------
ATTENTION: RENOVATORS AND
REMODELERS!
DO YOU WORK IN THE HOME RENOVATION AND
RESTORATION INDUSTRY?
OR
DO YOU LIVE IN AN OLD OR HISTORIC HOME THAT
YOU ARE CURRENTLY RENOVATING OR
RESTORING?
YOU MAY BE ELIGIBLE FOR AN IMPORTANT STUDY
ON LEAD EXPOSURE!
Workers or homeowners that are renovating/restoring old or historic
homes are needed for a landmark research project being conducted by
Battelle for the U.S. Environmental Protection Agency. This study will
explore the potential for lead exposure among homeowners or workers
that have performed remodeling, renovation, and restoration work in old
or historic homes.
Eligible respondents will be paid $50 for their participation.
For more information on how you can participate in this study, please
call toll free at 1-800-444-5234
-------
DO YOU OWN AN OLD OR
HISTORIC HOME? YOU MAY
BE ELIGIBLE FOR AN
IMPORTANT STUDY ON
LEAD EXPOSURE.
Owners of old or historic homes are needed
for an important research project being con-
ducted by the Battclle Memorial Institute for
the U.S. Environmental Protection Agency.
This study will explore the potential for lead
exposure among odulc homeowners who hove
performed remodeling, renovation and res-
toration work on their homes.
Eligible homeowners will receive S50 for
their participation.
"You could be eligible if your home meeu the
study criteria, and you have performed reno-
vation and remodeling work on iU
For more information on how you can par-
ticipate in this study, please call toll free at
1.8OM44-5234 Monday through Friday 9:00
am - 6:00 pm.
HOME RENOVATION AND
RESTORATION WORKERS!
Workers in the home renovation and restoration in-
dustries are needed for a landmark research project
being conducted by the Battelle Memorial Institute
for the U.S. Environmental Protection Agency. This
study will explore the potential for lead exposure
among carpenters and laborers who work in old or
historic homes.
Eligible workers will be paid$SQ for their participa-
tion.
You could be eligible if you work in the preserva-
tion, restoration, renovation and remodeling of old
or historic homes.
For more information on how you can participate in
this study, please call toll free at 1-800-444-5234
Monday through Friday 9:00 am - 6:00 pm.
-------
I. *^U^M*fr3wFfatflf(~il- -T.Vl'*'f'V -wV '•'•'' •>'•' •'SV -w-fc.Ci-!- •' «''—:'- -i'-V V-
if ^wt^tonpfe
eriovators
>*
I
I
"y,
»
.:!~:*' fiS you woricljjhj home rf novation and remodel jng industry?''
%%m$ rusrw?" i:v ' 4,OR - -w . : •'. '. -. - • •
. . . . 9
y^reyou. renovating or remodeling your home?
.^ypH'mpyifjP"«jiglt»l^fo study,on lead, exposure?
;Wqrl^p|ji 'and Homeowners who are- renovating/restoring: old or historic
homes '.kfe. needed for a landmark research project being conducted by
Battelle fojr'Uie U.S. EPA;:This study will explore the potential for lead expo-
Jsury'amdftg'jieoplc i\vho*%are perfomiing7rem,odpliijgVren6vation and
restora-
i'-'- • —
how'yoircan participate, please call .toll-free af,.
^^JL
y -'-V
i\-v
-------
Chattel !e
... Putting Technology To Work
Survey Research Associates". Inc.
TOO Ccpitola Drive. Suite 301
Durham. North Carolina 27713
Tel 919-544-3717
Fax 919-544-0830
February 6,1997
Patricia. Henderson,
401 N. Lindbergh Btvd, Suite 330
St. Loms, Missouri 63141-7816
Dear Ms. Henderson;
I have reviewed the revisions as requested by the Battrile/SRA Institutional Review Board (KB) for
the study protocol gntrrfprf "Wodcer Characterization. & Blood-Lead. Study of Renovation &
Remodeling Workers in Historic Homes" j(FG002889YS) and find the revisions acceptable. You also
have made a minor modification in the study population to include homeowners who are doing their
own. renovations. Thfs modification is acceptable. You are therefore granted apptuv«d to proceed
mljg
As with an SRA studies, mis study wfll be subject to an armnaT IBB review at the end of next year.
We wiH **nd you me necessary form, for »minaT review at the appropriate nrnft, la the meantime,
should any changes occur in your protocol or questionnaire, please infimn tne 1KB. Similarly, the
IRB needs to be notified in the event of any injury or unexpected outcome arising from mis study.
I wish you the best m your study.
Sincerely,
Margaret R. Pennybackcr, Ph_D.
IRB
cc: Charlotte Coley
Suf/OY Besscrcn A»ooc'w. «rrc. ir c fjosdlcr/ of Sctfatts Memcricf /rHnfure
-------
02/25/97 i4:6i 'O j»-.o»o-*«- ..-- -
Survey Research Associates, Inc. (SRA)
100 Capftola Drive, Suite 301
Durham, NC 27713
INSTITUTIONAL REVIEW BOARD NOTICE OF APPROVAL
PROJECT DIRECTOR: Patricia M. Henderson
TITLE: Worker Characterization & Blood-Lead Study of Renovation & Remodeling Workers
in Historic Homes \ __
CLIENT: EPA & Sattelle/ESTD PROTOCOL DATE: 12/4. 796
SRA PROJECT CODE: SG002889YS or PROPOSAL NUMBER:
NATURE OF REVIEW: (check <>»•}
FULL — MEETING DATE: 12 7il8 796
EXPEDITED (specify reason}'. , .
EXEMPT (specify cosoni:
TYPE OF APPROVAL; (check onel
PRELIMINARY. SCHEDULE NEXT REVIEW PRIOR TO INVOLVEMENT OF HUMAN SUBJECTS.
PRETEST/PILOT TEST. SCHEDULE NEXT REVIEW PRIOR TO FULL IMPLEMENTATION.
FULL IMPLEMENTATION.
RENEWAL
AMENDMENT DATED _./ /
Please note the following requirements:
PROBLEMS OR ADVERSE REACTIONS: If any problems in treatment of human subjects or
unexpected adverse reactions occur as a result of this study, you. must notify the IRB
Chairperson immediately.
CHANGES IN PROTOCOL: If there are significant chsr.gss in procedures or study protocol, you
must notify the IRB Chairperson before they are implemented.
RENEWAL: You are required to apply for.renewal of approval at least annually for as long as
the study is active. Your next review date should be on or before 12 718 797 .
I^B Chairperson ' Date
Barbara A. Moser
Print or Type Name
X Copy of approved Informed Consent attached.
cc: Project Director
IBs Administrator
-------
la
Survey Research Associates, Inc.
ICG Ccsi.'Olc Drive. Suite 301
Surna-5. MoMh Carolina 27X13
fcl 919-544-3717
=cx 9I9-M4 0830
December 20, 1996
Patricia Henderson
401 N. Lindbergh Blvd. Sufta 330
St. Louis, Missouri 63141-7815
Dear Ms. Henderson;
The Battelle/SRA Institutional Review Board (IR8) has reviewed the protocol for the study
entitled "Worker Characterization & BloadrLead Study of Renovation &. Remodeling .Workers In
Historic Homes* (SG0028S9YS) end requests some revisions prior to granting you approval to
proceed'with this study.
Please revise the following:
1. Add an IRB contact to the informed consent.
2. Delete the statement 'in keeping with legal requirements' form the informed
consent, as it is unclear whet this! means.
t3. Please checfc what steles require reporting of bicod-lead levels and what, is done
"with this information; and inctudeithis information in th« informed consent If ths
process differs from state to stats, please develop appropriate informed consents for
each.
4. On your protocol application, stipend grvart to subjects can net be considered a
benefft of the study. P'lease revise that page and delete thai reference.
5, State the amount of blood to ;be drawn in lay terms, i.e., teaspoons vs. ml.
The Board also had several comments/suggestions:
1. Define OSHA.
2. On page 3 of the protocol application, please else check the box for biological
specimens.
c?*s; ire- «• c fttf wWcr/ of A£tsXa .Vc.Ttf.~S i
-------
3. The directions for the self-administered questionnaire reference an advance letter,
but one was not included. Please revise the directions to delete this reference.
Work can not begin on joint SRA/CPHRE projects until both the SRA IRB and the Battelle
Human Subjects Committee (HSC) have reviewed the protocol and approved ft. I will forward
the necessary materials to the Battalia HSC for review after the above revisions have been
made and will notify you as soon as I recewe the decision of the BatteQe HSC,
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, jn 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,
Barbara A. Moser
IRB Chair
cc: M.R. Pennybacker
C.H. Coley
-------
STUDY ED:
ENVIRONMENTAL PROTECTION AGENCY WORKER CHARACTERIZATION AND
BLOOD-LEAD STUDY OF RENOVATION AND REMODELING WORKERS
IN HISTORIC HOMES
INFORMED CONSENT DOCUMENT
The U.S. Environmental Protection Agency (EPA) has contracted with Battelle/Survey Research Associates
(Battelle/SRA) to study the lead exposure hazard associated with renovation and remodeling in historic HOMES.
The objectives of this study are to (1) characterize the relationship (if any) between renovation and remodeling
activities in historic HOMES 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 mis study, you will be asked to complete a questionnaire mat includes information
relevant to lead exposures on (a) work history (both current and long term); (b) personal characteristics and habits
related to lead exposure; (c) non-work activities; and (d) 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 mat it will take approximately 40 minutes 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 in historic HOMES 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 I I
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 ug/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 measure hi adults greater than or equal to 25 ug/dl be reported to
the state registry. If your blood-lead measure is greater man or equal to 25 ug/dl, we will attempt to notify you of
this regardless of whether you requested your results. We will also pay you S25.00 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
participant's names and their participant identification number will not be released outside of Battelle/SRA.
If you have any questions or comments regarding this study, or if you experience and difficulties as a result
of participation in this study, please contact: Mr. Bennett Pierce, Battelle, 21 15 East Jefferson Street, Suite 400,
Rockville Maryland 20852. His phone number is (301) 770-2280. You may also call the Battelle/SRA Study
Manager, John Egel, at (3 14) 993-5234.
YOU HAVE THE RIGHT TO WITHDRAW FROM PARTICD7ATION 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
blood sample to be collected.
I , UNDERSTAND THE NATURE OF THIS STUDY AND AGREE
TO PARTICIPATE.
SIGNATURE
-------
50272-101
REPORT DOCUMENTATION
PAGE
1. REPORT NO.
EPA-747-R-99-001
3. Recipient's Accession No.
4. Title and Subtitle
"Lead Exposure Associated with Renovation and Remodeling Activities:
Phase IV, Worker Characterization and Blood-Lead Study of R&R Workers Who
Specialize in Renovation of Old or Historic Homes"
5. Report Date
March 1999
6.
7. Author(s)
Bennett Pierce, Patricia Henderson, Joan Huber, and Patrick Kiser
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.
G003470-05
11. Contract(C) or Grant(G) No.
(C) 68-D5-0008
12. Sponsoring Organization Name and Address
U.S. Environmental Protection Agency
Office of Pollution Prevention and Toxics (7401)
401 M Street, S.W.
Washington, D.C. 20460
13 Type of Report & Period Covered
Final Report
14.
15. Supplementary Notes
16. Abstract (Limit 200 words)
(Section 402(c)) of Title IV of the Toxic Substances Control Act, enacted in 1992, required EPA to conduct a study of
lead exposure associated with renovation and remodeling activities (R&R Study). This report documents Phase IV of that
study.
Phase IV was designed as a follow-on study to Phase II. Questionnaire and blood-lead measurements were collected
from 161 R&R workers and 82 homeowners at high-risk for lead-exposure because they perform R&R in older homes that
typically contain lead-based paint. Target R&R activities examined in Phase IV included removal of large structures
(demolition), window replacement, carpet removal, paint removal/surface preparation, and post-activity cleanup.
Overall, the blood-lead data indicate that Phase IV study participants were more highly exposed to lead than R&R
workers who do not specialize in R&R of older homes. The geometric mean blood-lead concentrations were 5.73 //g/dL for
workers and 4.45 //g/dL for homeowners.
Statistical models indicate that there was a significant relationship between the conduct of certain R&R activities and
blood-lead concentrations. Among homeowners, paint removal/surface preparation was positively associated with
blood-lead concentrations For workers, the models indicated that worker blood-lead concentrations were associated with:
cleanup, paint removal/surface preparation, carpet removal, and replacing window or door casements.
17. Document Analysis
a. Descriptors: Blood-lead concentration, renovation and remodeling, workers, homeowners, survey
b. Identifiers/Open-Ended Terms: Section 402, Renovation and Remodeling Study
c. COSATI Field/Group
18. Availability Statement
Release Unlimited
19. Security Class (This Report)
Unclassified
20. Security Class (This Page)
Unclassified
21. No. of Pages
181
22. Price
SeeANSI-239.18)
OPTIONAL FORM 272 (4-77)
(Formerly NTIS-35)
Department of Commerce
------- |