EPA600/AP-93/001C
July 1993
Urban Soil Lead Abatement
Demonstration Project
Volume III. Part 1
Baltimore Report
Environmental Criteria and Assessment Office
Office of Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Printed on Recycled Pacer
-------
DISCLAIMER
This document is an internal draft for review purposes only and does not
constitute U.S. Environmental Protection Agency policy. Mention of trade, names or
commercial products does not constitute endorsement or recommendation for use;
11
-------
TABLE OF CONTENTS
Page
LIST OF TABLES vi
LIST OF FIGURES . . viii
ACKNOWLEDGEMENTS . xi
1. EXECUTIVE SUMMARY 1-1
1.1 STUDY DESIGN 1-1
1.2 ENVIRONMENTAL MEASURES 1-2
1.3 DEMOGRAPHIC AND BEHAVIORAL QUESTIONNAIRE ... 1-3
1.4 BIOLOGIC MEASURES 1-3
1.5 INTERVENTIONS 1-3
1.6 ANALYSIS 1-4
1.6.1 Results 1-6
1.7 CONCLUSIONS . 1-8
1.8 IMPLICATIONS 1-8
2. INTRODUCTION V ....... 2-1
2.1 HEALTH EFFECTS 2-1
2.2 BIOLOGICAL FATE AND METABOLISM OF LEAD 2-1
2.3 SOIL AND DUST LEAD AND THEIR RELATIONSHIP
TO BLOOD LEAD 2-2
2.4 BALTIMORE AS A STUDY SITE . 2-3
3. METHODOLOGY 3-1
3.1 PROTOCOL FOR STUDIES INVOLVING HUMAN
SUBJECTS 3-1
3.1.1 Confidentiality 3-1
3.1.2 Informed Consent . . . . 3-2
3.1.3 Ethical Considerations . . . . 3-2
3.2 STUDY DESIGN 3-2
3.2.1 Site Selection 3-3
3.2.2 Rationale for Study Site Criteria 3-5
3.2.3 Pre-study Data Gathering 3-6
3.2.4 Comparison of Study Communities ............. 3-6
3.2.5 Study Population . 3-8
3.2.6 Rationale for Study Subject Criteria 3-8
3.2.7 Sample Size Calculation 3-9
3.2.8 Comparison of Final Study Population 3-9
3.2.9 Attrition and Retention 3-10
-------
TABLE OF CONTENTS (cont'd)
I
?age
3.2.10 Community Outreach/Public Relations \ 3-13
3.2.11 Public Relations Officer . 3-13
3.2.12 Community Outreach Coordinator . 3-14
4. INTERVENTIONS 4-1
4.1 ENVIRONMENTAL MEASUREMENTS AND ANALYSIS '. . . 4-1
4.1.1 Son 4-1
4.1.2 Dust . . .... 4-3
4.1.3 Water : . ; 4-4
4.1.4 Exterior Paint 4-4
4.1.5 Interior Paint 4-4
4.1.6 Quality Assurance for Soil and Dust Sampling and
Analysis 4-5
4.2 DEMOGRAPHIC AND BEHAVIORAL QUESTIONNAIRE ... 4-5
4.3 BIOLOGICAL SAMPLING AND MEASURES v 4-6
4.3.1 Blood ; 4-7
4.3.2 Hand Lead Determinations . l 4-7
4.3.3 Quality Assurance and Control for Blood Lead ;
Measurements ;...... 4-8
4.4 DETAILED DESCRIPTION OF THE INTERVENTIONS ..... 4-9
4.4.1 Exterior Paint Stabilization :.;..... 4-9
4.4.2 Soil Abatement 4-10
4.4.3 Abatement Costs 4-10
5. ANALYSIS 51
5.1 DATA COLLECTION AND MANAGEMENT 5-1
5.2 RESULTS 5-2
5.2.1 Effect of Soil Abatement . . . . 5-2
5.2.2 Relationship to Blood Lead Level 5-2
6. DATA ANALYSIS 6-1
6.1 VARIABLE SELECTION 6-1
6.2 BIOLOGIC VARIABLES AND VARIABLES FROM THE
QUESTIONNAIRE 6-1
6.2.1 Blood Lead 6-1
6.2.2 Hand Lead 6-1
6.2.3 Age 6-8
6.2.4 Socioeconomic Status ...................... 6-8
6.2.5 Season 6-8
6.2.6 Mouthing Behavior 6-15
6.3 ENVIRONMENTAL VARIABLES i' 6-16
6.3.1 Abatement , . , . , 6-16
6.3.2 Soil Lead 6-16
iv
-------
TABLE OF CONTENTS (cont'd)
'age
6.3.3 Dust Lead 6-17
6.3.4 Exterior Paint . . 6-20
6.3.5 Interior Paint . , 6-20
7. STATISTICAL ANALYSIS 7-1
7.1 STATISTICAL ANALYSIS OF ENVIRONMENTAL
VARIABLES 7-1
7.2 STATISTICAL MODELS FOR BLOOD LEAD AND
HAND LEAD . 7.5
7.3 INTERPRETATION OF REGRESSION COEFFICIENTS ..... 7-10
7.4 RESULTS OF STATISTICAL ANALYSIS 7-11
7.4.1 Model 1 ..".... 7-11
7.4.2 Model 2 . 7-12
7.4.3 Model 3 . 7-23
7.4.4 Model 4 7-30
7.4.5 Model 5 7-30
7.5 IMPLICATIONS OF FINDINGS 7-46
7.6 CALL FOR FURTHER RESEARCH 7-50
8. REFERENCES 8-1
-------
LIST OF TABLES
Number
3-1
3-2
3-3
4-1
4-2
7-1
7-2
7-3
7-4
7-5
7-6
7-7
7-8
7-9
7-10
7-11
7-12
Characteristics of Study and Control Sites at Time of
Enrollment into Study
Characteristics of Final Study Population Based on Round 3
Study Data
Attrition and Recruitment Rounds 1 Through 6
Baltimore Paint Stabilization .
Baltimore Soil Abatement
Soil Statistics Before Intervention
Dust Statistics Before Intervention
Pre- and Post-intervention Soil Statistics
Dust Statistics for Control Group Before and After Soil
Abatement
Dust Statistics for Treatment Group Before and After
Soil Abatement
Dust Statistics for Properties with and Without Lead-based
Paint
Regression Coefficient for Direct Effect of Abatement on
Blood Lead Model 1
Regression Coefficient for Adjusted Effect of Abatement on
Blood Lead Model 2
Regression Coefficient for Effect of Age on Blood Lead
Model 2
Regression Coefficient for Effect of SES on Blood Lead
Model 2
Regression Coefficient for Effect of Season on Blood Lead
Model 2
Regression Coefficient for Effect of Log Hand Lead on
Blood Lead Model 2
Pag(
3-7
3-10
3-12
i
4-13
4-14
7-2
7-2
7-3
7-4
7-4
7-7
I
7-13
7-16
7-19
7-21
7-23
7-25
VI
-------
LIST OF TABLES (cont'd)
Number
7-13 Regression Coefficients for Effect of Abatement on Hand
Lead Model 3 . ". 7-27
7-14 Regression Coefficients for Adjusted Effect of Abatement on
Blood Lead Model 2 7-31
7-15 Regression Coefficients for Effect of Age on Hand Lead
Model 4 7-34
7-16 Regression Coefficients for Effect of Female Gender on Hand
Lead Model 4 7-36
7-17 Regression Coefficients for Effect of Season on Hand Lead
Model 4 7-38
7-18 Regression Coefficients for Effect of Dust on Hand Lead
Model 4 7-40
7-19 Regression Coefficients for Effect of Gender on Hand Lead
Model 5 . 7-42
7-20 Regression Coefficients for Effect of Age on Hand Lead
Model 5 7-44
7-21 Regression Coefficients for Effect of Season on Hand Lead
Model 5 7-46
7-22 Regression Coefficients for Effect of Dust on Hand Lead
Model 5 7-48
7-23 Regression Coefficients for Effect of Soil Lead on Hand Lead
Model 5 7-50
7-24 R-Squared Coefficient and Mean Square Error for Models with Log
(Blood Lead) as the Response Variable 7-52
Vll
-------
LIST OF FIGURES
Number
3-1
3-2
4-1
4-2
6-1
6-2
6-3
6-4
6-5
6-6
6-7
6-8
6-9
6-10
6-11
6-12
Baltimore study design
Recruitment and retention of participants , . . . ,
Schedule of project activites
Typical property diagram
Normal and log-transformed distributions for blood lead,
Round 1
Normal and log-transformed distributions for blood lead,
Round 2
Normal and log-transformed distributions for blood lead,
Round 3
Normal and log-transformed distributions for blood lead,
Round 4 ,
Normal and log-transformed distributions for blood lead,
Round 5
Normal and log-transformed distributions for blood lead,
Round 6
Normal and log-transformed distributions for band lead,
Round 1
Normal and log-transformed distributions for hand lead,
Round 2
Normal and log-transformed distributions for hand lead,
Round 3 ,
Normal and log-transformed distributions for hand lead,
Round 4
Normal and log-transformed distributions for hand lead,
Round 5 .
Normal and log-transformed distributions for hand lead,
Round 6
Page
• 3-4
3-12
4-2
4-11
|
6-2
6-3
6-4
6-5
6-6
6-7
6-9
6-10
6-11
6-12
6-13
6-14
vni
-------
LIST OF FIGURES (cont'd)
Number
6-13 Distribution of SES scores using Hollingshead Index 6-15
6-14 Tri-mean of pre- and postabatement soil lead concentrations
for control group 6-18
6-15 Tri-mean of pre- and postabatement soil lead concentrations
for treatment group . . 6-19
6-16 Pre- and postabatement dust lead load for control group,
all properties 6-21
6-17 Pre- and postabatement dust lead load for treatment group,
all properties 6-22
6-18 Pre- and postabatement dust lead load for control group 6-23
6-19 Pre- and postabatement dust lead load for treatment group 6-24
7-1 Correlation matrix of environmental variables 7-6
7-2 Model 1 results of effect of soil abatement on blood lead,
log transformed 7-14
7-3 Model 1 results of effect of soil abatement on blood lead 7-15
7-4 Model 2 results of effect of soil abatement on blood lead,
log transformed 7-17
7-5 Model 2 results of effect of soil abatement on blood lead 7-18
7-6 Model 2 results of effect of age on blood lead, log transformed . . 7-20
7-7 Model 2 results of effect of socioeconomic status on blood lead,
log transformed .' 7-22
7-8 Model 2 results of effect of season on blood lead, log
transformed 7-24
7-9 Model 2 results of effect of hand lead on blood lead, log
transformed 7-26
7-10 Model 3 results of effect of soil abatement on hand lead, log
transformed 7-28
IX
-------
LIST OF FIGURES (cont'd)
Number
7-11
7-12
7-13
7-14
7-15
7-16
7-17
7-18
7-19
7-20
Model 3 results of effect of soil abatement on hand lead
Model 4 results of effect of soil abatement on hand lead, log
transformed
Model 4 results of effect of soil abatement on hand lead
Model 4 results of effect of age on hand lead, log transformed . . .
Model 4 results of femal gender effect on hand lead, log
transformed
Model 4 results of effect of season on hand lead, log
transformed
Model 4 results of effect of dust lead on hand lead, log
transformed
Model 5 results of effect of gender on hand lead, log
transformed
Model 5 results of effect of age on hand lead, log transformed ...
Model 5 results of effect of season on hand lead, log
Page
7-29
7-32
7-33
7-35
I
7-37
7-39
7-41
7-43
7-45
transformed . . . 7-47
7-21 Model 5 results of effect of dust lead on hand lead, log
transformed . , 7-49
7-22 Model 5 results of effect of soil lead on hand lead, log
transformed 7-51
-------
ACKNOWLEDC
«
The Baltimore Soil Lead Abatement Demonstration Project was managed by the
Maryland Department of the Environment (MDE) and drew upon the technical and
epidemiological resources of the State's Lead Poisoning Prevention and the Environmental
Health Programs. Laboratory support was provided by an interagency agreement with the
Department of Health and Mental Hygiene Laboratories Administration.
The Principal Investigator, Katherine P. Farrell, M.D., M.P.H., was Assistant
Secretary for Toxics, Environmental Science and Health Administration TESH for the initial
two years of the project and worked as Director of Community Health Services at the Anne
Arundel County Department of Health for the final year. J. Julian Chisolm, Jr., M.D.,
Co-investigator, is the Director of the Lead Program at the John F. Kennedy Institute and an
Associate Professor of Pediatrics at The Johns Hopkins Medical Institutions. Charles A.
Rohde, Ph.D., Professor and Chairman Department of Biostatistics at The Johns Hopkins
University School of Hygiene and Public Health, and Boon P. Lim, M.D., M.P.H., MDE
Administrator for the Environmental Health Program, joined the team as Co-investigators
during the data preparation and analysis phase of the study.
The Project Manager was Merrill Brophy, M.S.N., R.N. Warren Strauss performed the
statistical analysis for the study.
Reginald Harris was an invaluable player during the development of the study design
and protocols. Dr. Richard Brunker, Region UJ of the Environmental Protection Agency
supplied technical guidance and support for the project too.
The goals of the project would not have been met without the dedication and
cooperation of the following: Denise Stanley, Outreach Coordinator; Rebecca Fahey,
Environmental Coordinator; and Laura Coleson, Biological Coordinator. Without their
enthusiastic and constant efforts, the project would not have succeeded.
The project received cooperation and assistance from the City of Baltimore, the
Property Owners Association, and the Park Heights and Walbrook Junction Community
Organizations. Special thanks are due the Liberty Medical Center who contributed clinic
space free of charge.
XI
-------
Above all we thank the children and their families who participated in the study.
* ;
Although the information in this document has been funded wholly or in part by the United
States Environmental Protection Agency under assistance agreement V-003409-01 to the
Maryland Department of the Environment, it may not necessarily reflect the views of the
Agency and no official endorsement is inferred.
Xll
-------
1. EXECUTIVE SUMMARY
Over the last 20 years, it has become increasingly clear that the health effects of
elevated blood lead levels in children are long term, if not permanent. Public health
programs have focused on lead paint as the most significant source of exposure. However,
other sources of lead exposure (air, water, and soil) continue to be of concern.
The importance of soil contamination, although recognized by the preventive
community, has never been quantitatively studied in terms of its impact. The 1986
Superfund Amendments and Reauthorization Act (SARA) provided funds for a national multi-
city study of the impact of abating residential lead contaminated soil on the blood lead levels
of children. Baltimore was one of three cities selected as a study site.
1.1 STUDY DESIGN
This study was designed to investigate the effect of soil abatement on children's blood
lead levels as a preventive strategy. The hypothesis to be tested was that a reduction in soil
lead levels would result in a statistically significant decrease in children's blood lead levels.
Neighborhoods to be used as study sites were selected based upon their having areas of
exposed soil around the house, a moderate risk for lead poisoning, a sufficient number of
participants to test the hypothesis, pre-1950 central city housing, comparable demographic
indicators, and primarily residential housing.
Although census tract data was an important factor, neighborhoods were selected to
permit the inclusion and exclusion of portions of census tracts that did not meet the study site
criteria. The communities of Lower Park Heights and Walbrook Junction were selected for
the project sites. Following the collection of baseline environmental and biological data,
Lower Park Heights was selected as the study area and Walbrook Junction as the control area
by randomization.
Subjects were enrolled by door to door recruitment based upon the following criteria:
(1) living in one of selected areas, (2) living in same house for at least 3 continuous months
and with no plans to move in the immediate future; and (3) between 6 mo and 6 years of
1-1
-------
age. Sample size calculations indicated that at least 88 children were needed in the control
area and 44 in the study area. To allow for attrition it was decided to enroll 400 children.
All environmental and biological sampling occurred after the child was enrolled in the
project. All laboratory analyses were conducted by the Maryland Department of Health and
Mental Hygiene Laboratories Administration. ;
1.2 ENVIRONMENTAL MEASURES
Soil sampling was done to characterize the potential exposure of participant children to
lead from soil and to document the effectiveness of abatement in reducing soil lead levels.
The soil sampling and analysis protocols were developed in conjunction with the EPA and
the demonstration project teams from Boston and Cincinnati. Using detailed property
diagrams and 15 cm soil coring apparatus, an average of nine composite surface (top 2 cm)
and nine deep (bottom 2 cm) core samples were taken at specified sites on the property.
Interior vacuum dust sampling was conducted to characterize the potential exposure of
children to lead from dust and to document whether there was any increase in interior dust
levels following paint stabilization and soil abatement. The dust sampling protocol was
developed by Dr. Thomas Spittler of EPA Region I. An average of 3 dust samples were
collected from the floors of the entry area(s) and two of the child's primary play areas.
Because of the small size of the dust samples, it was decided to analyze each sample both by
laboratory x-ray fluorescence (XRF) and by wet digestion atomic absorption
spectrophotometry (AAS).
Household first draw water samples from all faucets in the household were collected to
characterize the potential exposure of children from drinking water. Standard EPA water
sampling and analysis protocols were used.
Exterior paint samples were collected at the time of the first environmental visit to
examine the contribution of exterior paint to soil lead. Paint chip samples were collected
from painted surfaces and analyzed by XRF (Kevex). After the last biological testing
session, interior paint was analyzed for lead using portable XRF analyzers (Princeton Gamma
Tech XK3). Measurements were taken in the child's bedroom, kitchen, and living room on
1-2
-------
a painted wall surface and a painted wood surface (window, door). The sites of
measurement were indicated on a diagram of the room.
1.3 DEMOGRAPHIC AND BEHAVIORAL QUESTIONNAIRE
Within 1 week of enrollment, each child was scheduled for baseline biological testing
and interview data collection. The Biological Coordinator and staff administered a
questionnaire to the parent/guardian or primary caretaker to collect demographic data and
assess behaviors and other factors which influence the child's contact with various sources
of lead. Follow-up interviews were conducted with each round of biological testing.
1.4 BIOLOGIC MEASURES
Blood samples for blood lead level (whole blood), free erythrocyte protoporphyrin
(FEP), ferritin, and total iron binding capacity (TEBC) were collected 6 times throughout the
study. Blood lead levels were determined using graphite furnace atomic absorption and FEP
levels were determined by Chisolm's double extraction method. Quality control was
maintained by a strict internal quality control program and participation in the CDC quality
control system. The soil abatement intervention was conducted between rounds 3 and 4.
Hand wipes were obtained to determine lead dust levels on the child's hands at the time
of each blood collection. The protocols for sampling and analysis were developed by the
University of Cincinnati. Damage to the laboratory exhaust system, required a change from
the nitric acid/perchloric acid method of analysis to the hot nitric acid methodology after
round 2.
1.5 INTERVENTIONS
Houses with exterior leaded paint in both the study and control areas received exterior
paint stabilization the summer and fall of 1990. Paint stabilization consisted of wet scraping
the chipping, peeling paint followed by HEPA vacuuming the area. A primer and two coats
of latex paint were applied to all painted surfaces. The purpose of paint stabilization was to
1-3
-------
remove or encapsulate any chipping, peeling paint to prevent re-contamination of the abated
soil.
Soil was abated only in the study area during the summer/fall of 1990, within 1 week
of paint stabilization. Areas located within the property boundaries with a soil lead level
greater than 500 ppm lead were abated. Abatement consisted of removal of the top 6 in. of
soil and replacement with clean soil (less than 50 ppm lead), the area was then sodded or
seeded depending upon characteristics of the site.
During the exterior paint stabilization and soil abatement, household members were not
allowed on the property. Space was provided at a local community center for the families
during the work.
1.6 ANALYSIS
The purpose of the statistical analysis was to investigate the relationship between
children's blood lead levels and the measurable sources of lead to which, they was exposed.
Several models were selected to determine whether or not the intervention of removing lead
contaminated soil had any impact on the child's blood lead level.
A correlation analysis was performed on the four environmental variables (soil, dust,
exterior paint, and interior paint). This analysis demonstrated a strong relationship between
exterior paint and lead in soil, and between interior paint interior dust lead level. The
remaining correlation coefficients were not significant. The data analysis was conducted
using both SAS and GUM statistical software.
The natural log transformation was applied to the response variable (blood lead and
hand lead) to meet the assumption of normality necessary for linear regression. The
following models were performed on two different populations within the study: (1) children
who participated in all 6 rounds and (2) all children sampled.
The first model measures the direct effect of group assignment on the log of blood lead
in each round:
ij = boj Ti + by C£ + By
1-4
-------
where for the ith child in round j,
LPbBjj = Log of blood lead
Ti = 1 if in treatment group, otherwise 0
C{ = 1 if in control group, else 0
e^ = error term
This model computes a geometric mean and a standard error for each group. These can be
transformed back to the original scale of blood lead and the groups compared through use of
a two sample t-test.
The second multiple linear regression model uses the log of blood lead as the response
variable with group assignment, age, season, socio-economic status, and the interaction
between mouthing behavior and log of hand lead as covariates:
LPhBy = bojTi + bjjCi + bjyAgeg b3jSESi + b^Seasmiij +
where for the ith child in round j,
LPbBjj = Log of blood lead
TI = 1 if in treatment group, else 0
C£ = 1 if in control group, else 0
Age^ = Age
SBSi = Socio-economic status of family
Season^ = 1 if sampled in summer, else 0
LPBHly = log hand lead if he/she exhibits weak mouthing behavior, else 0
LPbH2jj = log hand lead if he/she exhibits strong mouthing behavior, else 0
e = error term
Similar to the first model, a geometric mean and associated standard error for blood lead are
produced that are comparable through use of t=tests.
The third model evaluates the effects of group assignment on the log of hand lead:
where for the ith child in round j,
LPbHy = Log of hand lead
Tj = 1 if in treatment group, otherwise 0
Cj[ = 1 if in control group, else 0
e = error term
1-5
-------
The fourth model has log of hand lead as the response variable with group assignment,
age, season, gender, and interior dust as the covariates:
= bojTi + by-q + b2jAgeij b3jSexi + b^Season^ + b^-Dus^ + ey.
where for the ith child in round j,
LPbHy- = Log of hand lead
T£ = 1 if in treatment group, else 0
Cj = 1 if in control group, else 0
AgCy- = Age
Sexj = Socio-economic status of family
Season^ = 1 if sampled in summer, else 0
Dusty = log hand lead if he/she exhibits weak mouthing behavior, else 0
ey- = error term
The final model describes the association between the lead found on the hands of a child and
the sources of lead exposure measured within the child's home environment:
+ by-FemalCi + b2jAgeij b3jSeasoni + b4jDusty + b6jSoili +
where for the ith child in round j,
LPbHy- = Log of hand lead
Matej = 1 if male, else 0
Fematei = 1 if female, else 0
Agey- = Age
Seasonj = 1 if sampled in summer, else 0
Dusty = Measure of dust lead in home
Soily- = Measure of soil lead in home
e " = error term
1.6.1 Results
The statistical models were applied to the two populations to evaluate the potential bias
introduced by participant dropout. The regression coefficients were virtually identical
between the two populations for demographic, biological and environmental parameters,
indicating that the effect of participant dropout on the statistical models was negligible.
Although soil abatement in Baltimore did not result in the expected decrease of
1,000 ppm lead in soil, the average decrease in soil lead levels was 550 ppm (using the tri
mean measure). Based on the literature, the expected decrease in blood lead levels related to
1-6
-------
this magnitude of soil decrease would be 1 to 3 jug/dl. In round 4, 3 mo following
intervention, the blood lead levels of children in the study area decreased 1.5 /ig/dl and in
the control group decreased 1 /Kg/dl. This round of testing was in the winter months of
January through March, 1990. In round 5, May 22 through July 17, 1990, the blood lead in
the study group returned to the pre-abatement level and remained at that level in round 6.
However, the control group remained below the pre-abatement level throughout rounds
5 and 6. Due to the longitudinal aspect of this study, it is difficult to interpret these results,
however there does not seem to be any clinically significant reduction in children's blood
lead levels resulting from the soil abatement.
The effect of soil abatement on hand lead was also hard to interpret due to the
longitudinal aspect of the study. In the pre-intervention sampling rounds there was no
significant difference in hand lead levels between the treatment and control groups. In the
sampling round immediately following intervention, hand lead levels of children in the
control group were slightly less than those in the group which received soil abatement. This
round of sampling occurred during the winter months, when children are not outdoors much,
making it difficult to conclude if soil abatement had anything to do with this observation.
In the final two sampling rounds, which occurred during the spring and summer months, the
hand lead levels sharply increased in both groups. The hand lead concentration of children
in the abated group was lower than those of children in the control group for both of these
rounds. Although not statistically significant, this temporal trend may indicate a slight effect
of soil abatement on children's hand lead levels.
The regression coefficients of mouthing behavior as an effect modifier for hand lead in
the blood lead models indicated that children who exhibit stronger mouthing behavior will
have higher blood lead levels. This trend was apparent in all six sampling rounds. The use
of age as a continuous variable is questionable because the effect of age on blood lead may
not be linear. There seemed to be a negative effect of socio-economic status (as determined
by the HoUiagshead Four Factor Index) on blood lead levels throughout the study.
1-7
-------
1.7 CONCLUSIONS
Statistical analysis of the data from the Baltimore Lead in Soil Project provides ho
evidence that the soil abatement has a direct impact on the blood lead level of children in the
study. The statistical analyses to date have consisted of adjusted and unadjusted analysis for
selected covariates. It should be pointed out that the Baltimore study areas, both abated and
control, had lead based paint. Thus the conclusion might be more precisely stated as "in the
presence of lead based paint in the children's homes, abatement of soil lead alone provides
no direct impact on the blood lead levels of children".
1.8 IMPLICATIONS
The findings of this study might help avoid costly abatements of soil in cities, like
Baltimore, where the principal sources of lead exposure for children is lead in paint and lead
in household dust. Soil abatement for cities like Baltimore does not appear to be a cost
effective preventive strategy used alone, but it may well be an adjunct, in selected cases, to
the overall environmental management of children who become lead poisoned.
1-8
-------
2.0 INTRODUCTION
2.1 HEALTH EFFECTS
There has been a tremendous explosion in knowledge of the health effects of lead over
the past 20 years which has led to a progressive lowering of the blood lead levels considered
to be of concern. The Surgeon General Statement established an action level 40 /tg/dl in
1970. The Centers for Disease Control (CDC) lowered it to 30 jig/dl in 1975, to 25 jt*g/dl in
1985 and to 10 /tg/dl in 1991. The basis for these changes was findings of adverse effects at
lower and lower blood lead levels. It is now apparent that the blood lead threshold for
adverse neuro-developmental effect on the fetus and young children is 10 y^g/dl.
Furthermore, experimental evidence indicates that these effects are long lasting, if not
permanent (Air Criteria, ATSDR).
National Health and Nutrition Education Survey H (NHANES H) from 1976 to 1980
measured the distribution of blood lead concentrations in the United States population. The
mean blood lead level across the U.S. population in 1976, just prior to the decrease in leaded
gasoline, was found to be 15.9 /*g/dl. By 1980, with the continued decrease in the use of
leaded gasoline, the mean blood lead concentration had dropped to 9.6 jtg/dl (Annest, 1983).
Although urban poor minority children were found to have the highest risk of lead poisoning,
elevated blood lead levels were found across all social, geographic and racial groups.
2.2 BIOLOGICAL FATE AND METABOLISM OF LEAD
Absorption of ingested lead in children is more efficient than in adults. Absorption
rates are influenced by particle size (Barltrop and Meek, 1979) and nutritional factors
(Barltrop, 1974, 1975; Rosen, 1980).
Rabinowitz (1980) studied adults fed solutions of lead with and without food to
investigate the influence of food on lead absorption. He demonstrated that lead absorption
was reduced from 15 to 50% without food to 8 to 13 % with food. Bio-availability appears
to differ according to source and form of lead and is poorly understood.
2-1
-------
Dietary lead intake in excess of 5 /*g/ Kg of body weight/day in infants from birth to
2 years of age results in positive lead balance (Ziegler, et al., 1978). They also found an
inverse relationship between the amount of calcium in the diet and the absorption of lead.
Once absorbed, lead is distributed throughout the soft tissue and bone. There is a
continuous active process of absorption, tissue deposition and excretion. Deficiencies in
iron, calcium, and phosphorus are directly correlated with increased blood lead levels in
humans and experimental animals (Mahaffey, 1981; Mahaffey et al., 1980).
Under normal circumstances, the excretion of lead occurs 50-50 between the kidneys
and bowel (IRabinowitz, 1976).
2.3 SOEL AND DUST LEAD AND THEIR RELATIONSHIP TO
BLOOD LEAD
Whereas, air and food were significant sources of lead through the 1970's, these major
sources have been substantially reduced by reductions in lead gasoline and food. This led the
ADSTR to conclude in 1988, that as "persistent sources for childhood lead poisoning in the .
U.S., lead in paint and lead in dust and soil will continue as major problems into the
foreseeable future".
Lead poisoning in children was first reported in Australia by Gibson, et al. (1892).
Through experimentation and observation, Gibson concluded in 1904, "I believe and advance
a very strong plea for painted walls and railings as the source of the lead, and for the biting
of fingernails or sucking of fingers, as in a majority of cases, the means of conveyance of
the lead to the patient". Gibson's observation lay fallow for 70 years until Sayre, et al.
(1974) demonstrated an association between house dust and hand dust and blood lead in a
study of inner city and suburban homes.
Most studies in children during the last past 20 years have been around smelters and
mines. Roels, et al. (1980), in a study of school children who live near a smelter, reported
partial correlations between blood lead, hand lead, and air lead indicated that in the smelter
area the quantitative contribution of air lead to the children's blood lead levels is negligible
compared to hand lead. This relationship was found after air emissions from the smelter had
2-2
-------
been substantially decreased. These and other studies are extensively reviewed in the EPA
Air Quality Criteria for Lead (1986).
Ter Haar (1974) demonstrated a gradient in lead contamination between houses and the
nearest roadway suggesting that the house itself is also a source of lead in the soil. It is
generally believed that the major contributors to soil lead in cities are automobile exhaust and
exterior paint. Rabinowitz (1980) and Yaffe (1979) investigated sources of lead in blood by
use of stable isotope ratios distinguishing lead sources from paint and gasoline in a variety of
media, such as dust, soil, and blood. Whereas adult's blood lead resembled the isotope ratio
in airborne lead, the stable isotope ratios in children resembled interior household dust, in
some cases, and, in other cases, exterior paint and soil.
The EPA published a biokinetic model for lead (U.S. Environmental Protection
Agency, 1990). In this approach, amounts of lead in food, water, air and soil are calculated
from available data together with absorption factors for each. From this, total intake of lead
can be estimated and lead concentrations associated with various levels of intake can be
projected and modeled. Data from Binder, et al. (1986), in a study carried out in E. Helena,
Montana, appeared to validate this model. In these studies, the data suggested that children
might ingest 50 to 500 mg of soil per day.
Duggan and Inskip (1985) reviewed studies related to blood lead-soil lead ratio, or the
amount of increase in blood lead that can be attributed to a soil lead increase of 1,000 ppm.
They reported that this ratio is very variable between studies (range 1 to 9 >g pb/dl/blood
per 1,000 /*g pb/g soil). The ratio tended to be higher for younger children and lower for
older children. These data strongly suggest there may also be differences in the
bioavailability of lead from different environmental sources. Most of the studies they
reviewed were related to exposure to lead oxide dust among children living in the vicinity of
lead smelters. By contrast, there have been few studies of residential lead soils, away from
smelters (Stark, etal., 1982, Shellshear, 1975).,
2.4 BALTIMORE AS A STUDY SITE
Baltimore was one of the cities selected by an extensive review in 1987 to carry out one
of three linked studies on the issue of lead in soil and its impact on children's health.
2-3
-------
Baltimore City has a serious lead poisoning problem, which was first recognized in the
1930's as a result of an epidemic of lead poisoning related to burning battery casings for heat
during the depression (Williams, 1933). Following removal of wooden lead acid battery
casings from the market, lead poisoning cases continued to occur and it became apparent,
that among children, lead poisoning due to the ingestion of leaded paint was a serious
problem. The Baltimore City Health Department established a laboratory in 1935 ito provide
free diagnostic blood lead testing. This policy has continued through the years with a formal
screening program under CDC sponsorship being established in 1975.
Nearly all these children receive their excessive exposure through contact with lead
paint, contaminated indoor dust and possibly contaminated outdoor soil (Mielke, 1983).
Other sources of exposure in Baltimore are uncommon. The water supply is non-corrosive
and there are no uncontrolled emissions of the magnitude of some mining and smelting
towns. Few occupations involve lead exposure, principally auto body and radiatoi; repair and
the construction industry. As in other cities, occasional poisonings occur due to lead glazed
ceramics, fishing weights, and other unusual sources. But the most serious threat is the
continuing legacy of lead paint on older housing that becomes increasingly more available to
children as it ages, deteriorates, or during renovations.
Public health programs have focused on lead paint as the most significant source of
exposure and will continue to do so. The impracticality of widespread safe and permanent
abatements of the inside and outside of houses make it incumbent on us to consider what are
appropriate components of community based or individual approaches to reducing exposure
to lead.
The importance of soil contamination is of concern to preventive programs but has
never been quantitatively studied hi terms of its impact. Before launching into major
spending to abate lead in soil it is appropriate that we have an accurate picture of its impact
on prevention.
It is important to recognize that the impact of lead in soil may be quite different for
high risk children than it is for the general population of children. This study is designed to
examine only the impact of soil abatement as a preventive strategy. The study design does
not answer the next obvious question, which is whether soil abatement combined with paint
2-4
-------
abatement might be more effective than paint abatement alone in the management of poisoned
children.
2-5
-------
-------
3. METHODOLOGY
3.1 PROTOCOL FOR STUDIES INVOLVING HUMAN SUBJECTS
The Human Volunteers Committee Institutional Review Board (1KB) of the Department
of Health and Mental Hygiene reviewed the proposal submitted to the Environmental
Protection Agency along with all study protocols, the questionnaires and the consent forms.
Full approval was granted by the IRE in September, 1988. In addition, since Dr. J. Julian
Chisolm is a co-investigator and is on faculty at Johns Hopkins University, the proposal and
related materials were submitted for ERB review at Johns Hopkins. Full approval of this
group was granted in March, 1988, along with annual approval in 1989, 1990, and 1991.
3.1.1 Confidentiality
Particular attention was paid to the collection and handling of all personal, health
related or medical information. All such information was treated as confidential. Personal
identifiers were removed prior to processing the data and replaced with codes utilizing a
simple three digit sequential numbering system. Only coded information was entered into the
data base. Code keys were considered confidential. All confidential material was retained
under direct control of the investigation team.
Confidential medical information was accessible only to the members of the
investigation team and the subjects themselves or their physicians on receipt of a release of
information form signed by the subject specifying to whom the information should be
released.
Unless otherwise indicated (e.g., need for long term follow-up) confidential material
will be disposed of by shredding hard copies and deleting electronic data after a period of
5 years.
3-1
-------
3.1.2 Informed Consent
Informed consent was obtained prior to any procedure and participation was voluntary
throughout the project. No minor (under 18 years) was included in the project unless
informed consent was obtained from the parent or guardian. >
Information, provided to study subjects was in language likely to be understood by the
subject and was explained orally as well as in a written statement. There was an opportunity
to ask information of individuals familiar with the study and competent to address any related
issues and questions. No coercion was used by the investigation team when enrolling
subjects into the study.
3.1.3 Ethical Considerations
All procedures and interventions were reviewed for determination of safety to the
participant and community. No procedures were allowed which exposed human subjects to
risks. Because of the medical and developmental ramifications of elevated blood lead levels,
the parent or guardian of all children who had blood lead levels £: 25 /xg/dl (1985 CDC
action level) were notified. Information concerning the health and development effects of
lead poisoning was given to the parent/guardian and they were encouraged to seek medical
care. Upon informing the parents/guardians, the project informed the Baltimore City Health
Department Lead Poisoning Prevention Program of all elevated blood lead levels per
Maryland Regulations.
The children with elevated blood lead levels were not dropped from the study :but were
retained to monitor their lead levels.
3.2 STUDY DESIGN
The purpose of this project was to investigate the effects of removal and/or abatement
of lead contaminated soil with respect to childhood lead exposure. The hypothesis as stated
in the nuU was:
1. A significant reduction of lead (> 1,000 ppm) in residential soil accessible to
children will not result in a significant decrease (3 to 6 /*g/dl) in their blopd lead
levels. i
3-2
-------
Garden soil samples collected in Baltimore prior to the study and previous research
(Mielke, Chaney, 1982) suggested that soil lead levels could be decreased by 1,000 ppm
following abatement. However, soil samples collected during the study indicated lead
contamination was less than previous expected. Of the 204 properties tested for the study,
only 110 (54%) had soil lead levels above 1,000 ppm. For the remaining houses, abatement
would not achieve 1,000 ppm decrease in soil lead levels. In order to make use of the
voluminous amount of data collected in this project, the null hypothesis has been restated as
follows:
la A significant reduction of lead in residential soil accessible to children will not
result in a statistically significant decrease in their blood lead levels.
The final study design is illustrated in Figure 3-1. If soil abatement was found to be
associated with significant reduction in blood lead, similar treatment was planned for the
control area.
3.2.1 Site Selection
The sites for the project were selected based on the following criteria:
1. Identification as a moderate risk area for lead poisoning as determined by number
of hospitalizations for lead toxicity, lead screening results and/or predictions based
on the existence of risk factors.
2. Sufficient number of potential participants to test the hypothesis, based on birth
rates, power analysis and predicted recruitment and attrition rates.
3. Areas of exposed soil thought to be contaminated with lead at high concentrations
and accessible to children.
4. Pre-1950 central city houses in comparable condition as determined by drive-by
exterior inspections and housing census data.
5. Low livelihood of concurrent lead paint abatement projects performed with other
funding mechanisms.
6. Comparable socioeconomic class and other demographic indicators, as determined
by census data.
7. Areas should be non-contiguous.
8. Areas should be residential, single family housing and not near highways..
3-3
-------
Study Area
Recruitment
Biological
Sampling X 3
Blood Pb
FEP
Ferritin
Handwipe
Plant
Stabilization
Soil Abatament
(Onlyifsoil-Pb
>- 500 ppm)
Biological
Sampling X 3
Blood Pb
FEP
Ferritin
Handwipe
Data Analysis
Control Area
Environmental
Sampling
Ext. Paint
Water
Dust
Soil
Environmental
Sampling
Soil
Dust
Interior Paint
(If abatement is effective)
Report
Soil Abatement
(Only if soil-Pb
>s 500 ppm)
Report
Figure 3-1. Baltimore study design.
3-4
-------
Selection criteria applied to both the study and control areas.
During the initial phase of development of this project an intensive effort went into the
development of area profiles for site selection. Census tract data and information on lead
poisoning cases (also by census tract) formed the basis for initial site identification. Field
data based on drive through evaluations of the candidate areas helped narrow the field based
on accessibility of soil, confounding lead sources and similarities between study sites in terms
of layout and housing type.
Once the study sites had been identified, the boundaries of census tracts seemed
arbitrary and a decision was made to redraw the study neighborhoods based on homogeneity
and numbers of potential participants rather than the census tract borders. This permitted the
inclusion and exclusion of portions of census tracts based on their actual identification with a
neighborhood. The use of neighborhoods avoided having to exclude families simply because
they lived on one side of the street outside the census tract. It similarly avoided having to
include large areas with few children or little exposed soil.
3.2.2 Rationale for Study Site Criteria
Areas which were known to have no more than a moderate incidence of childhood lead
poisoning were desired to avoid the chance that other interventions would influence the
child's blood lead level during the period of the study. State of Maryland regulations require
an environmental evaluation and case management by the community health nurse, in concert
with the primary care provider, for blood lead levels ^25 /ig/dl, the 1985 Centers for
Disease Control (CDC) action level for lead poisoning. In addition, the primary care
provider is required to perform a thorough medical and nutritional evaluation. Thus, medical
and environmental interventions act as confounding variables. For the same reason, areas
that were expected to have lead paint abatement performed by other funding sources were
excluded from site consideration.
Communities in which the residences had accessible yards with at least some exposed
soil were considered because the children would have an opportunity for exposure to
contaminated soil. Single family housing with front and back enclosed yards were desired
because of increased likelihood of children playing in the immediate vicinity of their home.
Although most housing of this type in Baltimore is row or town houses, some communities
3-5
-------
have a mixture of row housing and semi- or detached housing. Houses built prior to 1950
tend to have the highest risk because of deterioration of lead paint over the years. !Such
properties are likely to be surrounded by soil contaminated with lead paint chips and dust
from deterioration of existing paint and from previous removal or renovation efforts.
A further consideration in selecting study and control areas is that they should be non-
contiguous communities. This avoids confounding results from crossover with children
moving from one area to another. Baltimore neighborhoods tend to be close knit ;
communities with extended families frequently residing within a few blocks of each other.
Care of children is often shared with grandparents or other relatives living close by.
Although families of low income rental units move frequently, they seldom move more than
a few blocks away.
Socioeconomic status and other demographics have been associated with blood lead
levels. Residential areas were selected to avoid the influence of heavy vehicular traffic and
heavy industry.
3.2.3 Prestudy Data Gathering
In order to select suitable study and control areas, the project drew upon previous
studies and data sources and performed soil sampling from candidate sites in Baltimore City.
Since Baltimore had been the site of an extensive soil study in the past, (Chaney and Mielke,
1982) the results of this study were used in the initial identification of possible sites.
Baltimore City, like many older urban centers, has a large number of housing units
painted with lead paint. A large body of information on patterns of lead poisoning in the city
from screening results, number of hospitalizations and previous studies were available in the
site selection process. In addition, data sources on risk factors such as socioeconomic status,
race, age, and housing were utilized.
3.2.4 Comparison of Study Communities
Based on the above information, the communities of Lower Park Heights and Walbrook
Junction were selected for the project sites.
The comparability of these two neighborhoods was borne out by baseline environmental
and biological data gathered in the course of the study. Mean soil lead levels were slightly
3-6
-------
higher for Walbrook Junction than for Lower Park Heights, but the difference was not
statistically significant (p = 0.08, two sample t test). All houses in both areas had lead-
based paint on the exterior. The remaining variables of interior dust and water were
comparable for the two areas. A comparison of community characteristics is found in
Table 3-1.
TABLE 3-1. CHARACTERISTICS OF STUDY AND CONTROL SITES
AT TIME OF ENROLLMENT INTO STUDY
Lower Park
Heights
(study area)
Soil Lead Level
(Tri mean PPM, XRF
Analysis)
Dust Lead Level
Floors**
(XRF Analysis)
Exterior Paint
(Mean mg/crn)
Water - First Draw
(Mean /tg/1)
% Owner Occupied
Premises ***
Mean
S.D.
N*
Mean
S.D.
N
Mean
S.D.
N
Mean
S.D.
N
33.5
= 546
= 326
= 112
= 778
= 1287
= 115
= 4.71
= 3.95
= 109
= 7.83
= 15.85
= 133
Walbrook Junction
(control area)
660
384
92
775
985
97
5.01
6.15
94
5.66
9.89
112
45.9
Total
598
357
204
777
1156
212
4.95
5.08
203
6.84
13.5
245
38.7
*N = Number of housing units
** = Total dust per area sampled (4 feer)
*** = At time of intervention
Random allocation of the areas to study or control status was made, by the toss of a
com, after the collection of baseline environmental and biological data to avoid selection
bias. Lower Park Heights was selected, in January 1990, as the study site where soil
abatement was to be carried out.
3-7
-------
3.2.5 Study Population
Door to door recruitment in the areas was done to enroll study subjects. Criteria for
participation in the project was based upon the following child characteristics:
1. Child must live in one of the selected sites. ,
2. Child must have lived in the same house for at least 3 continuous, months and
the family had no plans to move in the next year:
3. Child must have been between 6 mo and 6 years of age at the time: of
enrollment in the study. Special emphasis was placed on recruiting children
under 3 years of age.
3.2.6 Rationale for Study Subject Criteria
Three month residency at the current address was requked to ensure that the child's
baseline blood lead levels reflect the current residence's environment. Intention to remain in
current residence was selected to minimize attrition following enrollment in the study.
Children were not excluded if they attended day care or were enrolled in pre-kindergarten
programs.
Because of their increased vulnerability to the effects of lead, children between 6 mo
and 6 years were recruited. There was a special interest to identify and recruit children
under the age of 3 years, because these children will be in the population at risk throughout
the study and still have a high degree of hand-to-mouth activity. Children in this age group
also spend more time in the home environment.
During the enrollment phase of the study, no attempt was made to limit the number of
children recruited per housing unit. Some of the housing units contain 2 to 3 single mothers
with 1 to 2 children each. There were also some parents with several children and some
multi-generational families in which grand-mother and mother would each have a child in the
study. Throughout the study (rounds 1 through 6), the percent of housing units with more
than one child remained consistent around 60 % and the number of housing units with up to
3 children averaged 93%.
3-8
-------
3.2.7 Sample Size Calculation
Sample size calculations were based upon the following assumptions:
- the lead abatement intervention is the only change in the two populations
- at least one-half of the properties would have soil concentrations high enough to
warrant abatement, so that the effect of abatement would be felt by at least half of
the study area population (Ns)
- a = 0.05 (one tail test)
- 1-B = 0.80 (power of test)
- a « 42 (variance) S.D. = 6.5 /tg/dl
= 3/tg/dl (protocol)
The sample size formula for control population is:
Nc = (Za+ZK)2 a (K+D/K
Thus, in the control group Nc = 88 and in the study group Ns = 44. It was hoped
that the study would have at least 200 participants in the study, 100 from each area.
To allow for 20% attrition each year over the three years of the study, it was decided to
enroll 400 participants at the beginning of the study. (See actual attrition rates in section on
final study population.)
3.2.8 Comparison of Final Study Population
At the time of enrollment into the study, children in Walbrook Junction were slightly
older than those in Lower Park Heights. However, at the time of round 3 blood screening,
immediately prior to the intervention, and for children who remained in the study for its
duration, the two groups were similar in age. There was no significant difference in the ages
between those that stayed in the study versus those that dropped before round 3 testing.
The mean blood lead levels and ferritin levels were similar for the two populations
both at time of enrollment and at the time of round 3 testing.
Socioeconomic level according to the Holliiigshead Four Factor Index varied from the
time of enrollment into the study and round 3 of testing. At the time of enrollment into the
study there was no statistical difference in the two groups socioeconomic level. However by
round 3, the socioeconomic level for Walbrook Junction was higher than that for Lower Park
Heights. Those who dropped from the study in Lower Park Heights had a higher level of
3-9
-------
maternal education than those who remained in the study. This education difference in the
drop out group did not occur in Walbrook Junction.
The characteristics of the final study population are presented in Table 3-2.
TABLE 3-2. CHARACTERISTICS OF FINAL STUDY POPULATION
BASED ON ROUND 3 STUDY DATA
Lower Park
Heights
(study group)
Age in Months
Blood Lead Level
(pg/dl)
Ferritin Level
% Male**
% Black
% Class 5, SES***
Mean
S.D.
N*
Mean
S.D.
N
Mean
S.D.
N
N
= 47.2
= 22.3
= 154
= 11.1
= 6.5
= 154
= 23.5
= 18.7
= 148
52.6
= 152
100
49.8
Walbrook Junction
(control group)
50.1
18.8
116
10.2
5.4
116
22.9
14.7
107
40.4
114
100
51.9
Total
48.4
20.9
270
1.0.7
6.1
.270
23.3
17.1
255
47.4
255
100
50.8
* N = Number of children enrolled in round 3.
** p = 0.047.
*** According to Hollingshead Four Factor Index of Socio-economic Status.
3.2.9 Attrition and Retention
In the first round of biological sampling held between August 22, 1999 and
December 2, 1988, 408 children were recruited into the study (212 in Park Heights or the
study area and 196 in Walbrook Junction or the control area). By round 2 held February 2,
1989 and August 15, 1989, 100 (24.5 %) children were lost to the project because of lack of
interest/refusal to participate or moving out of the study area. During round 2, 14 additional
children were enrolled in the study.
3-10
-------
Because of the time lag in implementing the paint stabilization and soil abatement, an
additional round of biological testing was conducted from January 22, 1990 to August 13,
1990, and prior to the interventions occurring. At this time, 270 children were tested, a loss
of an additional 102 (31.6%) children. Once more, this loss was due primarily to moving
out of the area and refusal to participate or be tested. An additional 50 children were
enrolled in the study during this time. No new recruitment or enrollment occurred after this
round.
At the beginning of round 3 and continuing throughout the study, an intensive program
to retain study participants was conducted by the Outreach Coordinator and her assistant.
The focus of their campaign was to increase the participant's interest and commitment to the
project. This was achieved by increased personal contact with the participants and their
landlords (see section on Community Outreach).
Interventions, paint stabilization in study and control areas and soil abatement in study
area, took place from April 27, 1990 through December 10, 1990. Round 4 biological
testing took place three months after the interventions between January 2, 1991 and
March 26, 1991. Despite the above intensive efforts, 73 children (27%) were lost to the
study primarily because of refusal of the landlord to participate in the study. An added
benefit to the increased contact and outreach activities was the shorter time period necessary
for each biological clinic because of less missed appointments.
Round 5 biological testing occurred between May 22, 1991 and July 19, 1991.
One-hundred-ninety-three children were seen, an attrition rate of only 4% (8 children). Four
children who were tested in round 3 but were unable to come to round 4 were also seen.
The final round of testing occurred from August 19, 1991 to September 30, 1991.
For this round 185 children were tested, with an attrition rate of 5.7% (11 children). Once
more, 3 children who had been seen in rounds 3 and 4 but not round 5 were tested.
The overall attrition rate for the project with an initial enrollment of 408 children,
a loss of 294 children and an addition of 71 children was 54.6% (see Table 3-3 and
Figure 3-2).
3-11
-------
TABLE 3-3. ATTRITION AND RECRUITMENT ROUNDS 1 THROUGH 6
ROUND
1
2
3
TOTAL SEEN
408
322
270
ATTRITION
0
100
102
% LOSS
0
24.5
31,6
GAIN
14
50
INTERVENTION
4
5
6
197
193
185
73
8
11
27.0
4.0
5.7
0
4
3
500
400
300
200
100 ^^^
•••
•BiiiliBlSSiiiii iiBBaSi
BiHifii
_^^^MBimB pgaaa
2
Fall '88 Winter '89
• 1991 Maryland DepL of tha Environment
3
Winter '90
4
Winter'91
5 6
Spring '91 Summer '91
Number of Children
Figure 3-2. Recruitment and retention of participants.
3-12
-------
3.2.10 Community Outreach/Public Relations
The success of this project depended upon good relationships between the project and
the participants, as well as the maintenance of a positive image for the project in the eyes of
the communities directly impacted by the project. This image depended not only on what
was done, but on how it was presented. Public Relations and Community Outreach positions
were developed to create an awareness of lead hazards and the measures that reduce lead
exposure in general and the specific activities being undertaken as part of the project in
particular.
3.2.11 Public Relations Officer
The Public Relations Officer was responsible for marketing the project to the public,
especially the populations that directly contributed to the success of the project.
Prior to the start of the project, The Maryland Department of the Environment began
to increase efforts in the public relations and awareness areas in order to lay the groundwork
for the project itself. These activities included an opening ceremony press conference to
announce the award of the grant; the declaration by the Governor of Maryland designating
May 15 through 22, 1988, as Lead Poisoning Prevention Week, creating a forum for a
variety of publicity/education efforts; and the creation of a logo and slogan to increase name
recognition of the project.
Once the study neighborhoods were selected, interaction with the neighborhood groups
escalated. Meetings were held with community organizations, political leaders and church
groups to enroll participants and with community and coalition groups to enlist support for
the project.
Staff training in community relations was done to assure that the project maintained a
positive image in the community. Tee shirts, jackets and caps with the project logo were
worn by all staff in the field to increase project visibility.
The Public Relations Officer also made arrangements for media coverage of the
project. Television stations were contacted and arrangements made for visits to the
environmental and biological testing sites. Continuing stories on the project were done by
two television stations in the area. A series of newspaper articles on lead hazards included a
section on the project.
3-13 • , '
-------
Television and radio public service spot announcements were produced with the
cooperation of a local community college's tele-communication students. They were aired by
the television and radio stations during the initial recruitment and enrollment period of the
study.
Some Baltimore landlords were hesitant to allow their rental properties to be in the
study out of concern that participation in the study would leave them vulnerable to law suits
related to the possibility of lead poisoning. The project solicited the cooperation of the
Baltimore Property Owners Association (POA) to support the project among its members..
Meetings were held with reluctant landlords both in groups and individually to explain the
project and to decrease concerns related to liability. The Project Manager spoke at POA
training sessions concerning lead abatement of properties. Letters were also sent to all
property owners with property in the study explaining the project and answering legal
questions that had been raised.
3.2.12 Community Outreach. Coordinator
The Community Outreach Coordinator worked with the Public Relations Officer and
the Environmental Health Aides in the recruitment and retention of participants in the study.
The coordinator acted as a liaison between the communities in the study and the project.
The Outreach Coordinator conducted training sessions with all project staff in outreach
techniques and worked with them on how to handle difficult situations. The Outreach
Coordinator, Environmental Health Aides and other staff were then assigned to door-to-door
recruitment at alternate hours (evenings and weekends) in addition to normal business hours.
Every house in each community was contacted by someone in the study. If the mother of an
eligible child was not at home on the first visit, literature on the project was left and a return
visit was scheduled. Within 3 mo, 408 children were enrolled into the study. !
The Outreach Coordinator also met individually with landlords who were reluctant to
participate in the study to explain the project and the benefits to the property owner. Photos
of before and after paint stabilization were used to help convince rental property owners of
the benefits of participation. Of the properties enrolled in the study, 75% were as a direct
result of this intensive person to person campaign.
3-14
-------
Once the subjects were enrolled, emphasis was placed upon keeping clinic
appointments and retention in the study. The Outreach Coordinator placed emphasis on the
positive aspects of the mother's concern for her child's present and future health.
Additional recruitment took place prior to abatement in the study area because attrition
had been a problem, mainly due to families moving away. Pre-abatement data (biological
and environmental) was collected on the new participants.
In order to encourage the attendance by the participants at scheduled clinics, an
incentive plan was developed. Children who had blood tests performed received a tee-shirt
and sticker with the project's logo. A variety of incentives were offered to the
parents/guardians for bringing their children to the clinic and included the following:
• One month passes on the Metro Transportation System for all mothers who
brought their children in during a particular month.
• Food coupons to local grocery stores for mothers who brought their children in
during a particular clinic.
• Coupons to be redeemed for turkeys at a local super market for all participants
for Thanksgiving, 1989.
• Shoe coupons for children's shoes for children who attended a particular clinic
session.
• Educational toys and books for all children who attended a particular clinic
session.
• Social events (parties and skating events with refreshments and entertainment)
for all participants and their family members who remained in the study.
• Drawings for prizes to family if child attended clinic on first scheduled visit.
The Outreach Coordinator also worked with families to prevent eviction from houses
in which paint stabilization and soil abatement had been performed but biological monitoring
had not yet been completed. She worked with landlords, social agencies and
church/community support groups to obtain assistance for the families in meeting their rent
and electricity obligations. It was necessary to offer rent assistance to ten families in the
study area who were notified of eviction intent by their landlords. This was a one time
3-15
-------
payment and required attendance at a budget planning session conducted by the Outreach
Coordinator.
At the completion of the study, vacuum cleaners were given to all households that
participated in the entire study.
3-16
-------
4. INTERVENTIONS
Project Timeline presented in Figure 4-1.
4.1 ENVIRONMENTAL MEASUREMENTS AND ANALYSIS
4.1.1 SoH
Soil sampling was conducted to characterize the potential exposure of participant
children to lead from the soil and to document the reduction in soil lead levels following soil
abatement. Initial soil sampling was started August, 1988. However, because of a drought
that year in the Baltimore area, the soil sampling was postponed until after the initial
enrollment and blood testing of participant children. Soil sampling resumed November,
1988, and continued through February, 1989, because of the unusually mild winter. Follow-
up sampling was conducted within 1 week of soil abatement to document the decrease in soil
lead levels.
The soil sampling and analysis protocol was developed in conjunction with the EPA and
the demonstration project teams from Boston and Cincinnati (see Appendix A). A decision
was made to use X-Ray Fluorescence (XRF) for soil analysis following participation in the
Round Robin Study to evaluate the effectiveness of XRF analysis versus wet digestion
Atomic Absorption Spectrometry (AAS).
The soil sampling process is summarized below. The protocols for soil sampling and
soil analysis by XRF are in Appendix A. After a child was enrolled into the study, a
detailed drawing was made of the property that showed the boundary of the lot, the buildings
on the lot, the position of sidewalks and other paved areas, and the position of play areas,
if known.
Using the diagrams, composited soil samples were taken throughout the property. The
large area pattern of soil testing was utilized on most of the properties in the study. A line
20 in. from the base of the foundation into the soil area and running the length of the
foundation was measured and marked with stakes. One composite sample was collected
along this foundation line and one was collected along the boundary if the yard was less than
10 ft. wide. If the property was more than 16 ft. wide, an additional composite sample was
4-1
-------
1988
1989
.1990
1991
1992
Soil
Sampling
Paint
Sampling
Dust
Sampling
Wator
Sampling
Biological
Sampling
Paint
Stabilization
Son
Abatement
I - Initial
-Adjusted
- Control area
if indicated
Figure 4-1. Schedule of project activites.
collected midway between the foundation line and the boundary. Ten randomly selected
15 cm. (6 in.) long core samples were collected on each line.
From each core sample, the top 2 cm. and the bottom 2 cm. were put into separate
bags labeled "top" and "bottom". The tops/bottoms from each line were composited and
identified as a single sample. All soil samples were transported to the State Of Maryland
Department of Health and Mental Hygiene (DHMH) Laboratories Administration and
analyzed by XRF.
Following soil abatement, composite soil samples were taken from at least three sites
(foundation, mid-yard and boundary) hi the abated areas of each property using the above
method.
4-2
-------
4.1.2 Dust
Household dust sampling was conducted to characterize the potential exposure of
children to lead from dust and to document whether there was any increase in interior dust
lead levels following paint stabilization and soil abatement. The dust sampling was carried
out at the time of the initial environmental visit and within 1 week following soil abatement
in the study area. In the control area, second dust samples were collected after stabilization
but with no time limit.
The dust sampling protocol was developed by Dr. Thomas Spittler of Region I of the
Environmental Protection Agency. Based on the recommendation of Dr. Spittler and the
results of the Round Robin Study, the study initially planned on analyzing all dust samples by
XRF. Because of difficulty with analysis of the small sample size (<50 mg.), it was
decided to analyze each sample first by XRF and then by wet digestion AAS. The dust
sampling protocol and each method of dust analysis is in Appendk A .
For this study, the household dust samples were defined as the samples that represent
dust most likely to impact on a child's hands during indoor activity. During the
environmental visit, a sketch of the approximate layout of the residence was made and
sampling sites were selected and indicated on the diagram. The areas targeted for dust
sampling were the main entrance to the household and the two areas most frequently used for
play activities of the child or children. Additional areas for sampling that could be selected
include secondary entrances to the household, additional areas of activity frequented by the
children, and sources of accumulation of dust within the household (rugs and upholstered
furniture).
The Sirchee-Spittler Hand Held Dust Vacuum unit, which is a dust buster that had been
modified to catch the dust sample in a fine mesh stainless steel screen, was used to obtain the
samples. At each sample site, a 4 x 4 foot sample area was measured and marked with
masking tape. The dust sample was taken from the marked area. The sample was
transported in an upright sealed paper envelope to the DHMH Laboratory Administration for
analysis.
4-3
-------
4.1.3 Water
Household water samples were collected to characterize the potential exposure of
children to lead in drinking water. First draw water samples were collected from all
drinking water faucets in the household by Environmental Health Aides. Health Aides
visually inspected the sink area for indications of prior water use. If water had been used
that morning, another appointment was made for water sample collection. If unable to obtain
a first draw water sample after two tries, a non-first draw sample was collected. This was
indicated on the sample sheet and in the data file. The samples were transported to DHMH
Laboratories Administration for analysis. The water collection and analysis protocols are in
Appendix A.
4.1.4 Exterior Paint
Exterior paint samples were collected to determine the contribution of exterior paint to
the soil lead level at the time of the first environmental visit. Sample locations of painted
surfaces that were peeling, chipped or cracked were identified. A sample approximately
2.0 in. in diameter of all paint down to the substrate was obtained from each exterior surface
that was chipping or peeling. This sample was taken to DHMH Laboratories Administration
for analysis by XRF. The exterior paint sampling and analysis protocols are in Appendix A.
4.1.5 Interior Paint
At the end of the biological sampling, portable XRF analyzers were used to identify
interior lead paint to characterize the exposure to ulterior leaded paint. Measurements were
taken in the child's bedroom, kitchen and living room or other area identified by the parent
as a primary child play area. One measurement was taken in each room on a painted wall
surface and one on a painted wood surface (window, door frame). The sites of measurement
were indicated on a sketch of the room. The interior paint sampling protocol is in
Appendix A. Information on interior lead paint was shared with the Baltimore City Health
Department LPPP.
4-4
-------
4.1.6 Quality Assurance for Soil and Dust Sampling and Analysis
A quality assurance plan for the sampling and analysis of environmental samples was
developed by the Baltimore Soil Lead Abatement Demonstration Project and the State of
Maryland Department of Health and Mental Hygiene Laboratories Administration. (See
Appendix A) It includes a description of the proper procedures for sampling, sample custody,
equipment calibration and analysis, internal quality control checks and corrective actions.
The laboratory also participated in an external quality control plan for soil and dust
analysis that was supervised by Dr. Harold Vincent of EPA/EMSL Las Vegas. Audit
samples for soil and dust were developed by EMSL and inserted in the analysis stream by the
Environmental Coordinator (Appendix B). The project is awaiting final biweight
distributions for the soil and dust audit samples from EMSL to determine fraction of results
outside of the analytical acceptance windows at the 95 % prediction interval. The windows
are to be derived by EMSL using a modification of the EPA's biweight program. These
windows will apply to the XRF and AAS determination of lead. Lead values outside the
given ranges for the audit sample will affect assignment of a flag for data obtained for the
related sample group.
4.2 DEMOGRAPHIC AND BEHAVIORAL QUESTIONNAIRE
Within 1 week of enrollment, each child was scheduled for baseline blood testing and
data collection. These were performed at space donated by The Liberty Medical Center for
the Lead in Soil Clinic. The Biological Coordinator and staff (all Registered Nurses with
public health experience) administered a questionnaire to the child's parent/guardian or
primary care giver. The questionnaire was designed to (1) collect demographic information
to characterize the study population and (2) assess behaviors and other factors that influenced
the child's contact with various sources of lead. Drs. Katherine Farrell and Julian Chisolm
initiated the development of the questionnaire. Dr. Edmund Maes of the Centers for Disease
Control reviewed the questionnaire and his comments were incorporated into the final copy.
Copies of the questionnaire are included in Appendix E.
Two questionnaires were utilized in the course of the study. The first questionnaire
obtained data on the child's previous health status. It was felt that this information would
4-5 .
-------
not change over time and the data was not .gathered on subsequent visits to the clinic. The
follow-up questionnaire was administered at all remaining blood collecting sessions to verify
demographic data and to identify behavioral changes over time.
The interviewing staff was supervised by the Biological Coordinator. Each interviewer
was observed during the first few interviews and at regular intervals throughout each
screening session. Immediate feedback was given. The Biological Coordinator was available
for guidance and interpretation of questions/responses throughout each screening session.
The original Biological Coordinator remained with the study for its duration.
4.3 BIOLOGICAL SAMPLING AND MEASURES
Blood samples for blood lead level (whole blood), free erythrocyte protoporphyrin
(FEP), ferritin and total iron binding capacity (TIBC) and hand wipes for lead were collected
6 times throughout the study. The FEP, ferritin, and TIBC were collected as nutritional
status indicators. All biological sampling took place at the clinic site donated by Liberty
Medical Center. The sampling schedule for blood and handwipes was the same for the study
and control groups and was as follows:
ROUND DATES OF SAMPLING SESSIONS/INTERVENTIONS
Pre-abatement
1 August 22, 1988 to December 2, 1988
2 February 2, 1989 to August 15, 1989
3 January 22, 1990 to August 13, 1990
Interventions
Paint stabilization in study and control areas
April 27, 1990 to November 10, 1990
Soil abatement in study area
September 4, 1990 to December 10, 1990
Post-abatement
4 January 3, 1991 to March 26, 1991
5 May 22, 1991 to July 19, 1991
6 August 19, 1991 to September 30, 1991
4-6
-------
4.3.1 Blood
During the blood sampling, approximately 5 ml of blood was drawn from the
antecubital vein by a trained pediatric phlebotomist. Two ml of blood were utilized for
blood lead level and FEP analysis. The remaining blood was centrifuged in the clinic and
the plasma extracted for ferritin and TIBC analysis. All blood samples were cooled at
collection and were transported to the State of Maryland Department of Health and Mental
Hygiene Laboratories Administration (DHMH) Laboratories Administration within 24 h of
collection.
All laboratory results were reviewed within one day of receipt from the laboratory and
health care providers were notified of the results. If the blood lead level was 25 /^g/dl or
higher, the child was referred to the Baltimore City Health Department Lead Poisoning
Prevention Program and followed according to Maryland state law. During the course of the
study, no environmental or medical interventions were reported by the parent/guardian.
All blood analysis was performed by the DHMH Laboratories Administration
personnel. The laboratory meets stringent performance criteria including experience in
performing biologic analysis for health studies, participation in proficiency testing programs,
continuous OSHA certification and a detection limit of 1 /*g/dl for blood lead.
DHMH Laboratories determine blood lead levels using graphite furnace atomic
absorption (Pruszkowski, Cornick, and Slavin, 1983) and FEP using double extraction
method (Chisolm and Brown, 1975). Ferritin was determined by Abbott's Ferricyme
Enzyme Immunoassay (Forman and Parker, 1980) and TIBC by Radioactive Energy
Attenuation (REA) using Abbott TDX analyzer (Shaffar and Stroupe, 1983).
4.3.2 Hand Lead Determinations
Hand wipe samples were obtained each time blood samples were collected. Health
Aides were trained and supervised by the Biological Coordinator in the proper hand wipe
collection protocol (see Appendix A). To assess the extent of any contamination during
sampling, six wipes from each container opened were handled to simulate wiping the child's
hands. These were analyzed to determine background wet wipe lead levels.
Each set of hand wipes was transported to DHMH Laboratories Administration for
analysis. The Cincinnati perchloric acid analysis of hand wipes protocol was utilized for
4-7
-------
rounds 1 and 2 of screening. However, damage to the laboratory exhaust system by the
acid, precluded further analysis using this methodology, therefore, starting with round 3 and
continuing for subsequent rounds, 1M hot HNO3 was used in analysis (see Appendix A).
The total quantity of lead was reported in /*g per pair of hands.
In order to address concerns that some parents might wash children's hands before
bringing them to the clinic, elbow wipes were collected for comparison in round 1. This
decision was based on the premise that mothers might wash the child's hands but not wash
the whole arm up to the elbow. The resulting analysis found no indication that children's
hands had been recently washed. Elbow wipes were not used in subsequent rounds.
4.3.3 Quality Assurance and Control for Blood Lead Measurements
The State of Maryland DHMH Laboratories Administration maintained strict internal
quality control for their blood lead analysis. Calibration curves are composed of a minimum
of a blank and three standards. A calibration curve is made every hour of continuous sample
analysis. A minimum of one blank per sample batch was analyzed to determine if
contamination or any memory effects were occurring. Check standards were analyzed after
every 15 samples. One duplicate sample was run for every 10 samples. A duplicate sample
is a sample brought through the whole sample preparation process. Spiked samples or
standard reference materials were periodically employed to ensure the correct procedures
were followed and that all equipment was operating correctly.
The laboratory also participated in the external quality control system for blood analysis
developed and overseen by Dr. Daniel Paschal of the Centers for Disease Control. The
project followed the guidelines of the CDC quality assurance standards (Appendix B).
Analysis of Baltimore's bench and blind data by Dr. Paschal indicated that there were no
statistically significant trends with time and that the laboratory detection limits were
appropriate and precise (see Appendix B).
The protocol also includes the results of the initial characterization of the four whole
blood pools used in this project. Each laboratory was individually compared as to within-run
precision, among runs precision and total precision. The laboratory detection limits, using
the definition of the limit of detection as 3 SD (wr) developed by Winefordner, were also
compared. The conclusions stated by CDC were that: (1) comparable values were obtained
4-8
-------
in all laboratories, (2) laboratory data for blood lead were produced from analytical systems
in statistical control (as defined by Shewhart); and (3) no statistically significant time trends
were observed in the data (i.e., the difference in pre- and post-abatement blood lead values
are real and not the product of unstable analytical systems) (Centers for Disease Control,
1991).
4.4 DETAILED DESCRIPTION OF THE INTERVENTIONS
4.4.1 Exterior Paint Stabilization
Houses with exterior leaded paint in both the study and control area received exterior
paint stabilization during the summer and fall of 1990. Paint stabilization consisted of wet
scraping the chipping, peeling lead paint followed by HEPA (High Efficiency Particle
Accumulator)-vacuuming the area to capture the leaded paint chips and dust. Paint scrapings
and debris were sealed in plastic bags at least 6 mils thick and disposed of in a municipal
landfill according to the regulations in COMAR 26.02.07.07. A primer and 2 coats of latex
paint were applied to all painted surfaces within 48 hours of the scraping and vacuuming.
The purpose of the paint stabilization was to safely remove or encapsulate any chipping,
peeling paint to prevent re-contamination of the abated soil.
Precautions were taken to avoid contamination of surrounding areas by covering the
immediate ground and neighboring porches and yards with protective plastic; taping all
windows and doors of the house, and strict worker safety guidelines. Occupants of
neighboring properties were contacted, given an explanation of the activities and requested to
stay out of the work area and to keep their doors and windows closed. These activities were
monitored by project staff to ensure the safety of the participating family, neighborhood
residents, and workers.
Residents were encouraged not to be on site during this process. To provide the
* "families with somewhere to go during the paint stabilization process, space was obtained at a
local community center for the duration of the interventions. Families were transported to
and from the center in project vans, were provided age appropriate activities throughout the
day and were fed a snack or lunch, depending upon the time frame, while there.
4-9
-------
4.4.2 Soil Abatement
Soil was abated in the study group during the summer/fall of 1990, within 1 week after
the paint stabilization of the house. Detailed diagrams were made of each property (see
Figure 4-2) with fixed boundaries such as fences, hedges, sidewalks and house foundation
indicated. These boundaries defined areas which were identified as area A, B, C, etc. Each
area was treated as a unit in the soil abatement process. If the area/unit had soil sample
results of greater than 500 ppm lead, the entire area was abated. Separate areas of the
property which had sample results less than 500 ppm were not abated. Thus, if area A as
defined by the house, 2 sidewalks, and side fence was the front yard and had one or more
soil lead levels greater than 500 ppm, the entire area was abated.
The purpose of the soil abatement was to remove lead contaminated soil on the study
properties and to provide a barrier between lead contaminated soil and the child. Abatement
was performed by removal of the top 6 in. of ground coverings and soil. This soil was
replaced with previously tested soil containing less than 50 ppm lead. The replacement soil
was analyzed for the contractor for lead and other metals by an independent laboratory
(Business Industrial Safety Supplies [BISS] in Baltimore, Maryland) using wet digestion
AAS.
The area was then sodded or seeded depending upon characteristics of the site.
Families were given printed material on how to care for their newly abated lawns. Areas of
bare soil with lead levels less than 500 ppm were prepped and seeded to provide ground
cover. If the properly did not have exterior water access, the sites were watered weekly by a
private contractor during the summer dry spells.
Residents were discouraged from being on site during the soil abatement process. The
temporary relocation to a community center described above was utilized.
j
4.4.3 Abatement Costs
The costs of paint stabilization and soil abatement activities are included primarily as an
indication of feasibility for other investigators and for public health interventions. It should
be stressed, however, that costs are likely to vary greatly between areas and from project to
project based on local factors, constraints on purchasing and procurement, and the size of the
proposed abatement.
ii
4-10
-------
20'-0n
CONCRETE WALK
g
2572 DRUID PARK DRIVE
(2 STORY ROW HOUSb)
J
CURB
DESCRIPTION
ABATEMENT METHOD : REMOVE AND DISPOSE OF TOP 6 INCHES OF SOIL,
REFILL WITH "CLEAN" MATERIAL AS SPECIFIED.
AREA:"A" = 336SF
"B"= 67 SF
"Cn = 414SF
COVER: COVER AREAS "A"
THRU "C" WITH SOD
Figure 4-2. Typical property diagram.
STATE OF MARYLAND
DEPARTMENT OF THE ENVIRONMENT
LEAD PAINT STABILIZATION PROJECT
IN BALTIMORE CITY
4-11
-------
Costs can be broken down into those for exterior paint stabilization and those for soil
abatement. Further breakdown indicates costs related to sampling, analysis,
stabilization/abatement and replacement of soil with landscaping. These costs reflect the
unique conditions under which the abatements were conducted for the study.
The study incurred a variety of costs that may or may not be applicable to other
projects. These included costs related to mowing grass and removing debris before work
could start ($10,010.00), watering newly sodded or seeded yards during 1990 summer
drought ($6,000.00), alternate housing during the paint stabilization and soil abatement phase
($2,388.33) and food for participants and families while in alternate housing ($205.00).
Many of these expenses would not be applicable to the activities taken as part of the
environmental management of a child with elevated blood lead levels. ;
Four contracts were ultimately developed for the purpose of the exterior paint
stabilization in both areas and soil abatement in the study area. Because of the lack of
experience within the project with contract development related to these activities, contracts
1, 2, and 3 were developed by the Beavin Company Architectural/Engineering (A/E) firm for
$39,715.38. The firm reviewed contract 4 for the project for $411.33. The total cost for
the development of the four contracts was $40,126.71. Based upon experience in this area,
this cost may not be applicable in future soil abatements in this or other communities.
For the exterior paint stabilization of 125 houses in the study and control houses the
average cost per house was $3166.41 with a range of $366.00 to $5178.00. The total cost
for exterior paint stabilization was $248,087.29. Table 4-1 presents a detailed cost
breakdown for paint stabilization. (See Appendix H for bid schedule of units covered in
contract.) Force account work for repair, removal or replacement of painted surfaces such
as porches or steps cost an additional $6,858.43. ; '
The average cost per house of soil abatement in the study areas was $2163.39 with a
range of $600.00 to $4891.33. For the 63 houses in the study area, the total cost of soil
abatement was $136,293.62. Table 4-2 presents a detailed cost breakdown for soil
abatement. Additional costs of $2516.74 covered force account work such as removal and
replacement of fences.
4-12
-------
TABLE 4-1. BALTIMORE PAINT STABILIZATION
1. Paint Sampling and Analysis
Labor, Sampling 125 properties $ 2,820.00
Sample Collection Knives 160.03
Sample Collection Envelopes 72.98
Miscellaneous Supplies 200.00
Analysis of approximately 858 exterior samples 8580.00
Total Paint Sampling and Analysis 11,833.01
2. Contract Development and Supervision
Engineering Design and Supervision $20,063.36
Salaries
Administrative Contract Specialist (6 mo) 12,000.00
4 Enviornmental Health Aides (6 mo) 29.800.00
Total Contract Development and Supervision 61,863.36
3. Stabilization Contracts
Mobilization 65,018.00
125 Properties @ $1,984.70 Each 248.087.29
Total Stabilization Contracts 313,105.29
4. Miscellaneous Extra Costs
Pre-Stabilization Yard Cleaning
Total Miscellaneous Extra Costs 9.000.00
BALTIMORE PAINT STABILIZATION GRAND TOTAL $395,801.66
Total Stabilization Contract Work 125 Properties $248,087.29
Average Cost Per Property $3,166.41
Average Cost for Windows Per Property $511.00
Average Cost for Doors Per Property $ 89.00
Average Cost for Other Exterior Work $1,055.86
4-13
-------
TABLE 4-2. BALTIMORE SOIL ABATEMENT
Soil Sampling and Analysis
Labor, Sampling 70 properties
4 Core Sampling Tubes @ $507.08 each
4 Core Sampling Tubes @ $57.87 each
1200 Polyethylene Bags @ $646.80 per 1000
Miscellaneous Supplies
Analysis of 4330 samples
EPTOX Analysis 70 properties
Total Soil Sampling and Analysis
Contract Development and Supervision
Engineering Design and Specification
Salaries
Environmental Coordinator (6 mo)
2 Environmental Health Aids (6 mo)
Total Contract Development and Supervision
Abatement Contract
Delineation 70 © $600.00 each
63 Properties @ $1,478.43 each
70 Properties (<500 ppm) Landscape Work
Total Abatement Contract
Miscellaneous Extra Costs
Pre-Abatement Yard Cleaning
Vehicle Storage During Abatement
Hoses and Sprinklers
Post-Abatement Yard Maintenance
Total Miscellaneous Extra Costs
$ 1,656.00
2,028.30
231.48
776.16
211.71
43,300.00
9.520.00
57,723.65
20,063.36
16,000.00
16.000.00
52,126.71
42,0(30.00
93,141.09
14.394.29
149,535,38
1,010.00
490.00
50.00
6.000.00
7,550.00
BALTIMORE SOIL ABATEMENT GRAND TOTAL
.$245,222-41
Total Abatement Contract Cost for 63 Properties
Average Abatement Contract Cost for 63 Properties
Average Total Cost Per Property (63)
Total Cubic Yards Soil Abated
Cost Per Cubic Yard of Soil Abated
Total Cubic Yard Soil Replaced
$136,293.62
$2,163.39
$ 3,892.42
902.04 CY
964.78 CY
$ 341.93
4-14
-------
5. ANALYSIS
5.1 DATA COLLECTION AND MANAGEMENT
Within the project, data were generated and maintained in distinct groups:
environmental (pre-abatement soil, dust, exterior paint, and water analysis and post-
abatement soil and interior paint analysis), biological (blood and hand lead analysis), and
questionnaire data. Two of the three data groups, biological and questionnaire, were
repeated every time blood was drawn. Each blood sampling is referred to as a round. There
were six rounds in the study. Within the environmental data set, soil and dust samples were
obtained before and after abatement. Interior and exterior paint, and water samples were
obtained once through-out the study prior to abatement.
Data from the laboratories and the questionnaire were coded onto data entry forms.
The original report sheets and questionnaires were stored in separate files.
Quality control and assurance measures included checking all completed questionnaires
and forms manually for accuracy and completeness. Standard data entry validation tools
(double entry, range checks, etc.) were used for all data sets created through data entry.
Any problems were resolved on an ongoing basis. Data was backed up daily and archived
weekly in an off-site location.
The study data base consists of data files that were created by data entry using the
dBase M database management system. All dBase HI files were converted into Statistical
Analysis System (SAS) data sets for data management and analysis.
Information can be combined from various files by the use of the key fields of PROPID
and ID. PROPID is the unique seven digit identifier for a property in the study and ID is the
unique three digit identifier for each child. All the Environmental databases contain the
PROPID and all Biological databases contain the ID fields. The Questionnaire database
contains both the PROPID and the ID fields and is, therefore, the link between the Biological
and Environmental data files. All files can be merged through the Questionnaire files.
A more detailed description of the data management plan is included in Appendix C.
5-1
-------
5.2 RESULTS
5.2.1 Effect of Soil Abatement
Soil abatement in Baltimore, while effective in reducing surface lead levels to, 0 to
50 ppm, did not achieve the desired decrease in soil lead levels of 1,000 ppm. This is
because of the lower soil lead levels encountered prior to abatement. The average decrease
in soil lead levels was 550 ppm (using tri mean measure). Based on the literature, the
expected decrease in blood lead levels related to this magnitude of soil lead decrease might
have been in the 1 to 3 /tg/dl range. However, even prior to abatement, a relationship
between soil lead levels and blood lead levels for the children in the study was not found.
This may be because the soil contamination in Baltimore was clearly related to proximity to
lead painted surfaces of houses and was not uniform as it would be in cities where the source
of lead was air deposition from a stationary point source, such as, a smelter.
The soil lead concentrations decreased with distance from the house. These areas might
have been preferred as play areas for some children. Although the study included data on
f'
time spent outdoors and/or at other properties, insufficient information was available to
distinguish between time spent playing within 2 ft. of the house or farther away.
The abatements were carried out on the immediate property included in the project.
Most abatements were of single properties in an area with few abatements carried out on
several contiguous properties. Very few (1 or 2) children used public playgrounds!for play.
These tended to be older children and the play areas were surfaced with concrete rather than
soil.
The properties in the project generally had enclosed yards. This allowed for little
crossover to "next-door" yards for play. The front yards were more likely to be open, but
children usually played in enclosed areas. Even the open yards tended to have clearly
demarcated property markers (i.e., hedges and small fences between properties).
5.2.2 Relationship to Blood Lead Level
Statistical analysis of the data from the Baltimore Lead in Soil Project provides no
evidence that the soil abatement has a direct impact on the blood lead level of the children in
the study. The analysis to date has consisted of an unadjusted analysis and analysis adjusting
for selected covariates. Both analyses indicate no significant difference between the abated
5-2
-------
area and the control area insofar as blood lead levels are concerned. The principal reason
for this finding appears to be the low levels of soil lead found in the area under study.
5-3
-------
-------
6. DATA ANALYSIS
6.1 VARIABLE SELECTION
The purpose of the statistical analysis was to investigate the relationship between a
child's blood lead concentration, and the measurable sources of lead that the child was
exposed to during each round of sampling. These models were then used to determine
whether or not the experimental treatment of removing contaminated soil had any impact on
the blood lead concentration of the children involved in the study. The models that were
selected for presentation in this paper excluded many of the variables that were measured
throughout the experiment. A list of all the variables measured during the experiment can be
found in the Data Management Plan in Appendix C. Following is a rationale and brief
description for the variables that were used in the analysis.
6.2 BIOLOGIC VARIABLES AND VARIABLES FROM THE
QUESTIONNAIRE
6.2.1 Blood Lead
Blood lead concentration, as measured in micrograms of lead per deciliter, was
designated as the response variable in the multiple linear regression models. The distribution
of blood lead concentration was skewed to high values in each of the sampling rounds
throughout the experiment. This made it necessary to pursue an appropriate transformation
of the blood lead data to make its distribution appear more normal. The natural log
transformation was selected, thus inferences from the regression models can be interpreted in
terms of the geometric mean of the blood lead. The distribution of the original and log
transformed blood lead concentration for each round is presented in Figures 6-1 to 6-6.
6.2.2 Hand Lead
Similar to the distribution of blood lead, hand lead measured in micrograms of lead per
hand wipe sample had a distribution that was skewed to high values in each sampling round.
6-1
-------
25 -
20 -
1 35 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49
Blood Lead (ng/dl)
25 -I
0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 3.1 3.3 3.5 3J7 3.9
Log Blood Lead (ng/dl)
Figure 6-1. Normal and log-transformed distributions for blood lead, Round 1.
6-2
-------
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49
Blood Lead (ng/dl)
25 H
0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2£ 2.7 2.9 3.1 3.3 3.5 3.7 3.9
Log Blood Lead (ng/dl)
Figure 6-2. Normal and log-transformed distributions for blood lead, Round 2.
6-3
-------
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49
Blood Lead (ng/dl)
CD
I
Q
o.
25 H
20
15
10-
5 -
0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 3.1 3.3 3.5 3.7 3.9
Log Blood Lead (jig/dl)
Figure 6-3. Normal and log-transformed distributions for blood lead, Round 3.
6-4
-------
1357
11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49
Blood Level (ng/dl)
25 H
0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 3.1 3.3 3.5 3.7 3.9
Log Blood Level (ng/dl)
Figure 6-4. Normal and log-transformed distributions for blood lead, Round 4.
6-5
-------
25.1
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49
Blood Level (ng/dl)
25 1
I
Q>
EL
0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 3.1 3.3 3.5 3.7 3.9
Log Blood Level (ng/dl)
Figure 6-5. Normal and log-transformed distributions for blood lead, Round 5;
6-6
-------
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 46 47 49
Blood Level (ng/dl)
CO
§
-------
The natural log transformation was once again selected for this variable. The distribution of
the original and log transformed hand lead is presented in Figures 6-7 to 6-12.
6.2.3 Age
The age of each child was measured in years to the nearest decile for use as a covariate
in the multiple regression models. The distribution of age seemed approximately uniform
throughout the experiment. Age is known to have an effect on both blood lead and hand
lead, although this effect may not be functionally linear. Age was therefore broken into four
groups [(0-1), (1-2), (2-3), and (3+)] for use as covariates in the regression models.
Dichotomous indicator variables were fit in the regression models for the first three age
groups, and a linear term for age was included in the models for those children who were
older than three years.
6.2.4 Socioeconomic Status
SES as measured by the Hollingshead Four Factor Index (Appendix F) was computed
for the family of each child in each round of sampling based on questions from the parental
interview. The measure of SES was then averaged for use as a covariate. The distribution
of SES is presented in Figure 6-13.
6.2.5 Season
Season is a dichotomous variable indicating whether or not the blood sample was drawn
between the months of November and March. This variable was included in the regression
models because children in Baltimore do not typically spend much time outside during the
winter months, thus the lead exposure sources may change with season. It has also been
I
documented that the blood lead concentration of children is usually higher in the summer
months. Another approach to adjusting for seasonal variation within regression models is to
include a fourier transformation of the sample date. This approach would be useful in a
longitudinal analysis of the data set, but there is not enough seasonal variation within each
. !
round of sampling to justify this approach in the cross sectional models.
6-8
-------
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Hand Lead (jig)
0>
25
20
15 -
0- 10 -
5 -
0.1 0.3 0.6 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 3.1 3.3 3.5 3.7 3.9 4.1 4.3 4.5 4.7 4.9 5.1 5.3 5.5 5.7 5.9
Log Hand Lead
Figure 6-7. Normal and log-transformed distributions for hand lead, Round 1.
6-9
-------
CD
en
.2
I
£
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Hand Lead (jig)
0.1 03 0.5 0.7 0.9 1.1 1.3 IS 1.7 1.9 2.1 2.3 2.5 2.7 2.9 3.1 33 3.5 3.7 &9 4.1 43 4.5 4.7 4.9 5.1 5.3 5.5 5.7 5.9
Log Hand Lead (p.g)
Figure 6-8. Normal and log-transformed distributions for hand lead, Round 2.
6-10
-------
1
1
-------
0 S 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Hand Lead
25 -,
CD
PJ
I
0.1 03 O.G 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 3.1 3.3 3J5 3.7 33 4.1 45 4.5 4.7 4.9 5.1 5.3 5.5 5.7 5.9
Log Hand Lead
Figure 6-10. Normal and log-transformed distributions for hand lead, Round 4.
6-12
-------
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Hand Lead
O)
25 1
20
15 H
10 -
0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 3.1 3.3 3.5 3.7 3.9 4.1 4.3 4.54.74.9 5.1 5.3 5.5 5.7 5.9
Log Hand Lead (pg)
Figure 6-11. Normal and log-transformed distributions for hand lead, Round 5.
6-13
-------
0 5 10152025303540455055606570 75 80859095 100
Hand Lead
25 -
20 -
f 15 -I
1
CL 10 '
5 -
0.1 03 0.6 0.7 05 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 3.1 3.3 3.5 3.7 3.9 4.1 4.3 4.5 4.7 4.9 5.1 55 5.5 5.7 5.9
Log Hand Lead
Figure 6-12. Normal and log-transformed distributions for hand lead, Round 6.
6-14
-------
0)
9
g
(D
Q_
25
20
15
10
5
0 5 10 15 20 25 30 35 40 45 50 55 60 65
SES - Hollingshead Index
Figure 6-13. Distribution of SES scores using Hollingshead Index.
6.2.6 Mouthing Behavior
Mouthing behavior is a dichotomous variable created by combining the parental
response from the following two questions in the questionnaire:
401. How often does the child put her/his fingers in her/his mouth?
1 = a lot } SOME
2 = just once in a while}
3 = almost never } NEVER
6-15
-------
402. How often does the child put toys and things that are not food into her/his mouth?
1 = a lot } SOME
2 = just once in a while}
3 =» almost never } NEVER :
^ FINGER 401
TOYS SOME
Mouthing Behavior =? 402 NEVER
STRONG
WEAK
WEAK
WEAK
This variable was designed to be used as an effect modifier in the regression model. It
allows the differences in the effect of hand wipes on blood lead according to mouthing
behavior to be seen.
6.3 ENVTRONMENTAL VARIABLES
6.3.1 Abatement
This is a dichotomous variable which indicates whether or not the property received the
experimental treatment of soil abatement following the third round of biological sampling.
6.3.2 Soil Lead
Many top-soil samples were taken from each property and analyzed for lead content by
the method of X-ray fluorescence in units of micrograms of lead per gram of soil. The XRF
analysis result was multiplied by the constant derived for Baltimore from the IntercaUbration
Study conducted by the EPA and cities involved in the study (Appendix G).
Combining the measurements of these samples into a single soil exposure variable for
each property was necessary for use as a covariate in the statistical models. The number of
i
soil samples collected differed from property to property, based both on the size of each
yard, and the number of defined areas in the yard available for sampling.
Summary statistics of the lead concentration measurements of surface soil were
produced for each property. They are as follows:
6-16
-------
1) Mean 4) Upper Quartile 7) Foundation Median
2) Median 5) Lower Quartile 8) Foundation Maximum
3) Maximum 6) Foundation Mean
These summary statistics were used as variables in a principal components analysis, to find
which ones characterize most of the variability of soil within a property. This analysis led to
the selection of the tri-mean of soil lead concentration as a covariate for the regression
models where,
Tri-Mean = (Lower Quartile + 2*Median + Upper Quartile) / 4
The distribution of soil lead concentration for both treatment and control groups is
displayed in Figures 6-14 and 6-15.
6.3.3 Dust Lead
Interior dust samples were collected from several different rooms in each child's home.
Each dust sample was weighed in units of milligrams per surface area sample (16 ft2), then
analyzed for lead concentration by method of X-ray Fluorescence (ppm) and then later
measured by Atomic Absorption Spectrometry (ppm). The dust results were also multiplied
by the appropriate constant for Baltimore derived from the Intercalibration Study
(Appendix G).
Once again, the problem of combining these samples into one variable which indexes
lead content in interior dust of a given property arose. The mass of dust collected in each
sample was highly variable, making it difficult to directly compare measurements in units of
parts per million. It was decided to select a measurement for dust lead that represented the
average amount of lead in dust per surface area sampled in each property. The dust variable
was computed as a weighted average of the dust lead concentration measurements, each
measurement weighted by the mass of sample collected. The XRF measurements, were
selected because they correlated slightly better with blood lead than did the AAS
measurements.
6-17
-------
0 100200300400600*00700800900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000
Preabatement Soil Lead (ppm)
25 H
0 100200S00400600«00700800900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000
Post Abatement Soil Lead (ppm)
Figure 6-14. Tri-mean of pre- and postabatement soil lead concentrations for control
group.
6-18
-------
25 1
0 10020030040050060070)800800 1000 1100 1200 1300 1400 1500 1600 1700 1800 1000 2000
Preabatement Soil Lead (ppm)
25 H
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Post Abatement Soil Lead (ppm)
Figure 6-15. Tri-mean of pre- and postabatement soil lead concentrations for treatment
group.
6-19
-------
DUST = N'E; _, fXRR-') (WEIGHT^) i = l,.. ,# OF PROPERTIES
j i—j y j
Cj = 1 (WEIGHTj) j==l,..,N
Mix
!> =#-OF DUST SAMPLES
COLLECTED FROM PROPERTY
The distribution of dust lead concentrations per unit area sampled is displayed in
Figures 6-16 to 6-19.
6.3.4 Exterior Paint
Paint Chips were collected from the e cterior of each home prior to paint stabilization,
and were analyzed by XRF in units of ppm. The tri-mean of the exterior paint
measurements was selected as a covariate through use of principal components analysis.
6.3.5 Interior Paint
Measurements of micrograms of lead per centimeter squared were taken from painted
surfaces inside each home using a portable X-ray fluorometer. The maximum reading was
used to indicate whether or not a property contained interior lead based paint. A maximum
measurement of 1.5 milligrams per centimeter square or greater thus indicates that there is
some lead based interior paint within the property. The cut-off point of 1.5 mg/cm2 was
used to dichotomize interior paint for i.ome of the statistical analyses. Interior paint
measurements were not taken from evjry property involved in the study because of an
' j..
inability to access the house during the interior paint lead analysis tune period.
6-20
-------
25 -i
0 100200300400600600700800800 1000 1100 1200 1300 MOO 1500 1600 1700 1800 1900 2000
Preabatement Dust Lead (ng Pb/m2)
-------
25
20 4
0 100200300400600900700800800 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000
Preabatement Dust Lead (ng Pb/m2)
0 100200300400600600700800900 1000 1100 1200 1300 14001500 1600 1700 1800 1900 2000
Postabatement Dust Lead (jig Pb/m2)
Figure 6-17. Pre- and postabatement dust lead load for treatment group, all properties.
6-22
-------
100 200 300 «» 500 800 700 800 900 1000 1100 1200 1300 1400 1600 1600 1700 1800 1900 2000
Preabatement Dust Lead
Pb/m2)
0 100200300400500600700800900 1000 1100 1200 13OO 1400 1500 1600 1700 1800 1900 2000
Postabatement Dust Lead (ng Pb/m2)
Figure 6-18. Pre- and postabatement dust lead load for control group. Data are for
those properties that were sampled both before and after intervention.
6-23
-------
25
0 100200300400500600700800900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000
Preabatement Dust Lead (\ig Pb/m2)
25 H
O 100200300400500600700800800 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000
Postabatement Dust Lead (ng Pb/m2)
Figure 6-19. Pre- and postabatement dust lead load for treatment group. Data are for
those properties that were sampled both before and after intervention.
6-24
-------
7. STATISTICAL ANALYSIS
The intention of the statistical models was to evaluate the effects of soil abatement on
blood lead while still taking into consideration the pathway through which a child is likely to
develop lead poisoning:
DETERIORATION
1
Interior Paint = = = > Dust
Exterior Paint = = = > Soil
t
EROSION
AGE AND SEASON
Hand Lead = = = = > Blood Lead
Iff
GENDER MOUTHING SES
BEHAVIOR
This pathway suggests that environmental lead exposure sources influence blood lead through
the hand to mouth activity of the child. The lead found on the hands of a child results from
contact with lead contaminated dust and/or soil. The pathway also indicates that the erosion
of interior and exterior paint contribute to the lead contained in dust and soil.
7.1 STATISTICAL ANALYSIS OF ENVIRONMENTAL VARIABLES
The proposed pathway for lead poisoning in children indicates that dust lead and soil
lead are the primary exposure sources in a child's environment. It is therefore important to
show that the dust lead and soil lead levels were comparable between the treatment and
control groups prior to intervention.
Question 1 - Were the soil lead levels comparable in the treatment and control groups prior
to intervention?
7-1
-------
A two sample t-test yields a t-statistic of 0.049 with 202 degrees of freedom, indicating that
there is no difference in soil lead between the treatment and control groups (Table 7-1) prior
to intervention.
TABLE 7-1. SOIL STATISTICS BEFORE INTERVENTION
.
Group N Mean se j
Treatment 57 503.6 268.2
Control 147 501.3 312.1
Question 2 - Were the dust lead levels comparable in the treatment and control groups prior
to intervention?
•
A two sample t-test yields a t-statistic of 0.851 with 210 degrees of freedom, indicating that
there is no difference in dust lead between the treatment and control groups (Table 7-2) prior
to intervention. The distributions of pre intervention soil and dust lead for treatment and
I; ,
control groups are presented in Figures 6-14 to 6-19.
TABLE 7-2. DUST STATISTICS BEFORE INTERVENTION
Group N Mean se
Treatment 57 2,869.4 380.1
Control 155 1,902.8 152.8 '.
Following intervention, properties which received the experimental treatment of soil
abatement were re-sampled for quality control reasons. These measurements of post
intervention soil lead can be used to demonstrate whether or not the project achieved a
significant reduction in soil lead in the treatment properties.
7-2
-------
Question 3 - Was there a significant reduction in soil lead in the properties that received the
experimental treatment of soil removal?
A one sample t-test on the difference yields a t-statistic of 13.15 with 56 degrees of freedom,
indicating that there is a statistically significant difference in pre-intervention and post
intervention soil lead levels (Table 7-3) in the treatment group.
TABLE 7-3. PRE- AND POST-INTERVENTION SOIL STATISTICS
Group N Mean se
Pre Intervention 57 503.6 268.2
Post Intervention 57 33.6 34.9
Difference 57 470.1 269.8
The distribution of post intervention soil lead for the treatment groups is presented in
Figure 6-15.
Data for evaluating whether soil recontamination has occurred in the abated properties
has not been collected. Post intervention soil samples were not taken from control properties
under the assumption that the soil lead concentration in these areas would remain stable.
The experimental treatment of soil abatement may have an impact on the interior dust
lead levels. Interior dust samples were collected following intervention in several properties
in both the treatment and control groups in an effort to address this issue.
Question 4 - Was there a significant reduction in ulterior dust lead in properties that
received soil abatement?
A one sample t-test on the difference yields a t-statistic of 1.32 with 39 degrees of freedom,
indicating that there is no significant difference in pre intervention and post intervention dust
lead levels (Table 7-4) in the treatment properties.
7-3
-------
TABLE 7-4. DUST STATISTICS FOR CONTROL GROUP
BEFORE AND AFTER SOIL ABATEMENT
Group
N
Mean
se
Pre Intervention
Post Intervention
Difference
40
40
40
1,751.4
1,108.7
642.7
3,169.1
1,480.7
3,089.2
Question.! - Was there a significant reduction in interior dust lead in the control
properties?
A one sample t-test on the difference yields a t-statistic of 2.25 with 32 degrees of freedom,
indicating that there is a statistically significant difference in pre intervention and post
intervention dust lead levels (Table 7-5) in the control properties.
TABLE 7-5. DUST STATISTICS FOR TREATMENT GROUP BEFORE
AND AFTER SOIL ABATEMENT
Group
Pre Intervention
Post Intervention
Difference
N
33
33
33
Mean
1,784.9
976.0
808.8
se
2,340.9
928.9
2,068.8
The Distributions of pre and post intervention dust lead for both treatment and control
properties that were sampled following abatement is presented in Figures 6-16 and 6-19.
A correlation analysis was performed on the four environmental variables (soil, dust,
I
exterior paint and interior paint) to evaluate the first part of the proposed pathway. The
correlation coefficient between soil and exterior paint was +0.18 (p-value 0.0134), and the
correlation coefficient between dust and interior paint was +0.22 (p-value 0.0240). The
remaining correlation coefficients were not significant at the alpha=0.05 level. This analysis
confirmed the belief that erosion of exterior paint contributes to lead in soil, while
7-4
-------
deteriorating interior paint is positively associated with lead in interior dust. The correlation
matrix is presented in Figure 7-1.
The measurement for interior paint was easily dichotomized into an indicator of
whether or not the property had lead based paint on interior surfaces (a measurement of
>j
interior paint greater than 1.5 mg/cnar was taken as a positive result). If the proposed
pathway is to be upheld, one would expect higher interior dust lead levels in the houses that
tested positive for interior lead paint.
Question 6 - Is the interior dust lead concentration higher in those properties that tested
positive for interior lead based paint?
A one sample t-test on the difference yields a t-statistic of 2.76 with 103 degrees of freedom,
indicating that there is a statistically significant difference in dust lead concentration between
properties that test positive and properties that test negative for interior lead based paint
(Table 7-6).
7.2 STATISTICAL MODELS FOR BLOOD LEAD AND HAND LEAD
The experimental treatment in the Baltimore Study was designed to eliminate the bottom
half of the pathway through exterior paint stabilization and soil abatement. The following
statistical models evaluate the effect that this treatment had on both hand lead and blood lead
of children participating in the study. Both response variables for these models, blood lead
and hand lead, appear to be distributed log-normal as mentioned earlier. This leaves two
logical statistical approaches for modeling these data:
1) Apply a natural log transformation to the response variable and model the
data through multiple linear regression with additive errors.
2) Use the untransformed response variable in a multivariate normal
generalized linear model with a log link function.
The errors associated with this model are multiplicative. The main difference between
these two statistical approaches is in the interpretation of the regression coefficients. The
7-5
-------
CORRELATION ANALYSIS:
Four Variables: Soil
Exterior Paint
Dust
Interior Paint
Variable N
SOIL 204
DUST 212
INTERIOR PAINT 106
EXTERIOR PAINT 203
Simple Statistics
Mean Std. Dev.
597.80991
776.90932
5.37013
4.85391
357.11916
1156
4.02194
5.07530
Sum
121953
164705
569.23333
985.34388
Pearson Correlation Coefficients
Prob. > |R| under Ho: Rho=0
Number of Observations
SOIL
DUST
INTERIOR
PAINT
EXTERIOR
PAINT
SOIL
1.00000
0.0
204
0.08666
0.2236
199
0.17189
0.0856
101
0.17724
0.0134
194
DUST
0.08666
0.2236
199
1.00000
0.0
212
0.22203
0.0228
105
0.05334
0.4566
197
INTERIOR
PAINT
0.17189
0.0856
101
0.22203
0.0228
105
1.00000
0.0
106
-0.02332
0.8152
103
EXTERIOR
PAINT
0.17724
0.0134
194
0.05334
0.4566
197
-0.02332
0.8152
103
1.00000
0.0
203
Figure 7-1. Correlation matrix of environmental variables.
7-6
-------
TABLE 7-6. DUST STATISTICS FOR PROPERTIES
WITH AND WITHOUT LEAD-BASED PAINT
Group
Positive
Negative
N
65
40
Mean
1,432.0
627.7
se
2,136.7
754.9
first approach was selected for presentation in this report, although the statistical results for
both approaches is presented in Appendix I.
The following cross sectional models were performed on two different populations from
within the study; the first consisting only of children who consistently participated in all six
rounds of the study, and the second consisting of all children sampled throughout the
experiment. This helped to evaluate the sensitivity of the statistical models to participant
dropout.
The first model measures the direct effect of group assignment (treatment or control) on
the log of the blood lead in each round: LPbBy = b0jTj + byCj + ey where for the ith
child, in round j,
= Log of blood lead of child(i) in round j
T| = 1 if child(i) is in treatment group, else 0
Cj = 1 if child(i) is in control group, else 0
ey = error term for child(i) in round j.
Essentially, this model computes a geometric mean and standard error for each group. These
can be transformed back to the original scale of blood lead, and the two experimental groups
can be compared through use of a two sample t-test.
Next is a multiple linear regression model which uses the log of blood lead as the
response variable with group assignment, age, season, socio economic status, and the
interaction between mouthing behavior and log of hand lead as covariates. LPbBy = bQjTj
+ bjjCi + b2jAgeOij + b3j Agel^ + b4j Age2£j + b5j Age3jj + b6j SESj + b?j Season^
+ bgj LPbHly + b9j LPbH2ij + e^ where for the ith child in round j,
7-7
-------
LPbBj: = Log of blood lead of child© in round j
Tj = 1 if child© is in treatment group, 0 else
C: = 1 if child(i) is in control group, 0 else
AgeOj: = 1 if child© is age(O-l) in round j, 0 else
Ageljj = 1 if child© is age(l-2) in round j, 0 else
Age2j| = 1 if child© is age(2-3) in round j, 0 else
AgeS-j = (Age-3) if child© is age(3+) in round j, 0 else
SESj = Socio economic status of child©'s family
Season:: = 1 if child© was sampled in the summer of round j, 0 else
LPbHlj: = Log hand lead for child© in Round j, if he/she exhibits weak mouthing
behavior, 0 else
LPbH2j: = Log hand lead for child© in Round j, if he/she exhibits strong mouthing
behavior, 0 else
e.. = error term for child© in round j
** Note - Season was only included as a covariate in the first two rounds.
li.
This model allows the assessment of the effects of soil abatement after adjusting for
covariates described in the pathway that may influence a child's blood lead concentration.
Similar to the first model, a geometric mean and associated standard error for blood lead are
produced which are comparable through the use of t-tests.
The third model evaluates the effects of group assignment on the log of hand lead.
This model is identical to the first model except for the change in response variable:
LPbHj: = b0: Tj + by Cj + e^ where for the ith child, in round j,
LPbHjj = Log of hand lead of child© in round j
Tj = 1 if child© is in treatment group, else 0
C| = 1 if child© is in control group, else 0
ej: = error term for child© in round j.
\-
The fourth model has log of hand lead as the response variable with group assignment,
age, season, gender, and interior dust as covariates: LPbHy = bqjTj + bjjCj + b^j
AgeOjj + b3j Agely + b4j Age2ij + b5j AgeS^ + b6j SeXi + b7j Season^ + bgj Dustj +
ej: where for the ith child in round j,
LPbHj: = Log of hand lead of child© in round j
Tj = 1 if child© is in treatment group, 0 else
C| = 1 if child© is in control group, 0 else
AgeOy = 1 if child© is age(0-l) in round j, 0 else
7-8 -
-------
Agely = 1 if child(i) is age(l-2) in round j, 0 else
Age2tj = 1 if child(i) is age(2-3) in round j, 0 else
AgeSy = (Age-3) if child© is age(3+) in round j, 0 else
Sexj = 1 if child(i) is female, 0 else
Season^ = 1 if child(i) was sampled in the summer of round j, 0 else
Dustj = Measure of dust lead in child(i)'s home
e^j = error term for child(i) in round j
** Note - Season was only included as a covariate in the first 2 rounds. Rounds 3 through 6
had no seasonal variation among the participants.
This model allows us to evaluate the effects of soil abatement after adjusting for the
covariates described in the pathway that are thought to influence a child's hand lead
concentration.
The final model presented has log of hand lead as the response variable with gender,
age, season, interior dust lead, and soil lead as covariates. LPbHjj = bQ: Malej + b^
Femalej + b2j AgeO^ + b3j Ageljj + b4j Age2jj + b5j AgeSy + bgj Season^ + b7j Dustj
+ bgj Soilj + ey where for the ith child in round j,
LPbHy = Log of hand lead of child(i) in round j
Malej = 1 if child(i) is male, 0 else
Female^ = 1 if child(i) is female, 0 else
AgeOy = 1 if child(i) is age(0-l) in round j, 0 else
Agely = 1 if child(i) is age(l-2) in round j, 0 else
Age2y = 1 if child(i) is age(2-3) in round j, 0 else
Age3jj = (Age-3) if child(i) is age(3+) in round j, 0 else
Season^ == 1 if child(i) was sampled in the summer of round j, 0 else
Dust| = Measure of dust lead in child(i)'s home
Soilj; = Measure of soil lead in child(i)*s home in round j
e^ = error term for Child(i) in Round j
** Note - Season was only included as a covariate in the first two rounds, Rounds 3 through
6 had no seasonal variation among the participants.
The measure of soil lead changes in round 4 for properties which received soil
abatement.
This model describes the association between the lead found on the hands of a child,
and the sources of lead exposure measured within that child's home environment.
7-9
-------
Regression coefficients and their corresponding confidence intervals from these models
are presented and discussed in the Results section. These models contain only those
covariates which were consistently statistically significant in each of the rounds. Models that
included other potential confounders which were measured throughout the study were
explored at great length. The data analysis was conducted using both SAS and GLIM
statistical software. The methods used (correlation analysis, multiple linear regression, two
sample t-tests, and confidence intervals) are described in standard statistical text books.
7.3 INTERPRETATION OF REGRESSION COEFFICIENTS
The above cross sectional statistical models used the log transform of either blood or
head lead measurements as the response variable in a multiple linear regression model. The
regression coefficients derived from these models have additive effects on the log
transformed response variable. When the estimates are converted back to the original unit of
measure, these parameter estimates are interpreted as having multiplicative effects: log(Yj) =
E B:'Xj + GJ, where i = 1 to n (sample size) and j = 0 to p (number of parameters)
t = exp (S B4 >Xi)*exp(ei)
= exp (B*0 Xi0)*exp(B*1 Xn)* ... *exp(B*p Xip)*exp(ei)
* !
The regression coefficients (B ) for this multiplicative model are computed by exponentiating
the regression coefficients from the additive model on the log scale. I
B*J =exp(Bj)
The associated standard errors for regression coefficients from the multiplicative model are
calculated by using the delta method:
se(B*j ) = exp(Bj) * a * {X'Xy}-1/2
7-10
-------
Models for Comparison with Boston and Cincinnati Projects
A second set of statistical models were explored and evaluated for the purpose of
comparing and contrasting the findings of the Baltimore study with the results of the Boston
and Cincinnati projects. These models use post intervention log blood lead as a response
variable, and a summary of the pre intervention log blood lead along with group assignment
as covariates. These models were fit cross sectionally for each post intervention round of
sampling, and similar models were also fit for log hand lead.
7.4 RESULTS OF STATISTICAL ANALYSIS
Regression coefficients and their associated standard errors for both the additive and
multiplicative interpretations of the statistical models are presented in Tables 7-7 through
7-23. The regression coefficients computed in each round for each covariate were also
graphically displayed in the form of 95% Confidence Intervals across the time line of the
experiment. The vertical line in the center of these graphs represents the time at which soil
abatement occurred during the experiment. When appropriate, a zero line is drawn to help
ascertain whether or not the corresponding regression coefficients are statistically different
from zero at the a = 0.05 significance level.
The statistical models were applied to two different populations within the experiment
to evaluate the potential bias introduced from participant dropout. The graphs of the
regression coefficients were virtually identical between these two populations, indicating that
the effect of participant dropout on the statistical models was negligible. These graphs are
presented in Figures 7-2 through 7-22.
7.4.1 Model 1
Geometric means of blood lead and their associated standard errors were calculated in
each round for children for both the treatment and control groups in this model. This
measures the direct effect of group assignment on blood lead. The geometric means of the
two groups seem almost identical to each other in the first three rounds, thus there is no
statistical difference between the treatment and control groups prior to intervention. This is
evidence that the two groups were comparable from the start of the study. Following the
7-11
-------
intervention procedure of soil abatement, there is no significant difference between the abated
group and the control group as measured by this model. Contrary to the hypothesis of
interest, it seems as if children in the control group have slightly lower blood lead
concentration in the rounds following intervention, although this pattern is not statistically
significant. The regression coefficients, associated standard errors and confidence intervals
from model 1 are presented in Table 7-7 and Figures 7-2 and 7-3.
7.4.2 Model 2
This model calculates the geometric means and associated standard errors for group
assignment after adjusting for variables that are thought to influence a child's blood lead
concentration. The regression coefficients for the effect of group assignment in this model
are similar to those found in the first model. There is no detectable difference in group
assignment in the rounds prior to and following intervention (Table 7-8 and Figures 7-4 and
7-5).
The effect of age on blood lead was determined by creating 3 indicator variables for the
age groups (0-1), (1-2) and (2-3), and including a linear term for children over the age of 3.
The intercept for the (0-1) age group was significantly negative through the rounds of
sampling that included these very young children. This can be explained by the fact that
infants demand constant supervision from their caretakers. The (1-2) and (2-3) age groups
were positive in some rounds and negative in others. The linear term attached to children
over the age of 3 was consistently negative and significant in the latter part of the
experiment, indicating that blood lead decreases with age after a child is 3 years old
(Table 7-9 and Figure 7-6).
Socioeconomic status consistently has a statistically significant negative effect on the
blood lead of children participating in our study. Thus the blood lead of a child is inversely
related to the level of education and profession of the parent(s) (Table 7-10 and Figure 7-7).
Season was added as a covariate in the first 2 rounds, as these sampling rounds took
place over a change of season. The literature suggests that blood lead concentration is
typically higher in the months when children actively play outdoors. The effect of season on
blood lead is statistically significant only in the first round of the study (Table 7-11 and
Figure 7-8).
7-12
-------
TABLE 7-7. REGRESSION COEFFICIENT FOR DIRECT EFFECT
OF ABATEMENT ON BLOOD LEAD MODEL 1
Children Present In All Six Rounds of Sampling
Log (Blood Lead)
Round
1
2
3
4
5
6
Group
Treatment
Control
Treatment
Control
Treatment
Control
INTERVENTION
Treatment
Control
Treatment
Control
Treatment
Control
All
EST
2.473
2.428
2.383
2.327
2.245
2.228
2.117
2.028
2.179
2.088
2.234
2.123
Children Sampled
SE
0.06491
0.05300
0.06220
0.04971
0.06584
0.05638
0.06830
0.05825
0.06743
0.05903
0.06549
0.06120
Throughout Experiment
Blood Lead
EST
11.8580
11.3362
10.8374
10.2472
• 9.4404
9.2813
8.3062
7.5989
8.8375
8.0688
9.3371
8.3562
Log (Blood Lead)
Round
1
2
3
4
5
6
Group
Treatment
Control
Treatment
Control
Treatment
Control
INTERVENTION
Treatment
Control
Treatment
Control
Treatment
Control
EST
2.405
2.361
2.338
2.313
2.257
2.254
2.169
2.036
2.259
2.106
2.305
2.134
" SE
0.06001
0.03592
0.05857
0.03355
0.05313
0.04314
0.05309
0.05339
0.05682
0.05584
0.05360
0.05423
EST
11.0784
10.6015
10.3605
10.1047
9.5544
9.5258
8.7495
7.6599
9.5735
8.2153
10.0242
8.4486
SE
0.76970
0.60082
0.67408
0.50939
0.62156
0.52328
0.56731
0.44263
0.59591
0.47630
0.61149
0.51140
Blood Lead
SE
0.66482
0.38081
0.60681
0.33901
0.50762
0.41094
0.46451
0.40896
0.54397
0.45874
0.53730
0.45817
7-13
-------
1
3.0
>
2.5
2.0
'1.5
Children Who Participated in all Six Rounds
Intervention
1 2 3 4 56
Round of Sampling
All Children Sampled Throughout the Study
Intervention
I3'0 '
I
§2.5 •
&
1
12.0
Regression (
bi
*• ** „
i
** ^ **'
1 23 4 5 6
Round of Sampling
O = Abate
D = Control
Figure 7-2. Model 1 results of effect of soil abatement on blood lead, log transformed.
Bars show 95% confidence interval on regression coefficient.
7-14
-------
213
111
110
75
5= Q
S
^ 8
J>6
Children Who Participated in all Six Rounds
Intervention
2 3 4 5
Round of Sampling
6
I14
£12
111
Ho
75
9
8
o
O
6
All Children Sampled Throughout the Study
Intervention
O
345
Round of Sampling
6
O - Abate
n = Control
Figure 7-3. Model 1 results of effect of soil abatement on blood lead. Bars show 95%
confidence interval on regression coefficient.
7-15
-------
TABLE 7-8. REGRESSION COEFFICIENT FOR ADJUSTED EFFECT
OF ABATEMENT ON BLOOD LEAD MODEL 2
Children Present In All Six Rounds of Sampling
Round
1
2
3
4
5
6
Round
1
2
3
4
5
6
Group
Treatment
Control
Treatment
Control
Treatment
Control
INTERVENTION
Treatment
Control
Treatment
Control
Treatment
Control
All
Group
Treatment
Control
Treatment
Control
Treatment
Control
INTERVENTION
Treatment
Control
Treatment
Control
Treatment
Control
Log
EST
2.319
2.341
2.388
2.361
2.069
2.073
1.977
1.998
1.710
1.549
1.956
1.854
(Blood Lead)
SE
0.1947
0.1832
0.1562
0.1590
0.1728
0.1712
0.1669
0.1623
0.1743
0.1879
0.2066
0.2186
Blood Lead
EST
10.1655
10.3916
10.8917
10.6015
7.9169
7.9486
7.2210
7.3743
5.5290
4.7068
7.0710
6.3853
SE
1.97922
1.90375
1.70128
1.68565
1.36804
1.36081
l'
1.20519
1.19685
0.96370
0.88440
1.46087
1.39583
Children Sampled Throughout Experiment
Log
EST
2.333
2.327
2.441
2.431
2.202
2.213
2.094
2.094
1.874
1.665
2.095
1.954
(Blood Lead)
SE
0.1306
0.1265
0.1209
0.1197
0.1378
0.1349
0.1304
0.1283
0.1512
0.1662
0.1631
0.1736
EST
10.3088
10.2472
11.4845
11.3702
9.0431
9.1431
8.1173
8.1173
6.5143
5.2857
8.1254
7.0569
Blood Lead
SE
1.34633 ,
1.29626
1.38848
1.36102
1.24614
1.23340
1.05850
1.04145
0.98496
0.87848
1.32526
1.22507
7-16
-------
Children Who Participated in all Six Rounds
Intervention
TJ
O
s3.0 i
s
I 2.5
c •
12.0
1
I1-5
§1-0
c
_ i—
5 C
i C-
3 <-
CC I
1
^) r
J L
1
2
-J
I
L (
) E
]
T T T
T ^ E]
I" ^
--
I I I 1
3 456
Round of Sampling
§3.0
m
12.5
- c
12.0
1.5
'1.0
All Children Sampled Throughout the Study
Intervention
3 4
Round of Sampling
6
O = Abate
D = Control
Figure 7-4. Model 2 results of effect of soil abatement on blood lead, log transformed.
Bars show 95% confidence interval on regression coefficient.
7-17
-------
I 15-
? 14-
1 13
1 12
5 11
I 10
i § 1
^«r
fi
Children Who Participated in ail Six Rounds
Intervention
DC
2345
Round of Sampling
All Children Sampled Throughout the Study
Intervention
i 1 —i
2 3 4
Round of Sampling
O- Abate
D = Control
Figure 7-5. Model 2 results of effect of soil abatement on blood lead. Bars show 95%
confidence interval on regression coefficient.
7-18
-------
TABLE 7-9. REGRESSION COEFFICIENT FOR EFFECT
OF AGE ON BLOOD LEAD MODEL 2
Children Present In All Six Rounds of Sampling
Log (Blood Lead)
Round
1
2
3
4
5
6
Group
AgeO
Agel
Age 2
Age 3
Agel
Age 2
Age 3
Agel
Age 2
Age 3
INTERVENTION
Age 2
Age 3
Age 3
Age 3
All
EST
-0.46810
0.11890
0.13090
0.02635
0.00754
0.21740
0.02353
-0.49180
0.03714
-0.04629
-0.13210
-0.05259
-0.04856
-0.04667
SB
0.17720
0.13280
0.12460
0.05892
0.13890
0.12720
0.05385
0.46130
0.13620
0.04060
0.34310
0.02992
0.02675
0.02956
Blood Lead
EST
0.62619
1.12626
1.13985
1.02670
1.00757
1.24284
1.02381
0.61152
1.03784
0.95477
0.87625
0.94877
0.95260
0.95440
SE
0.11096
0.14957
0.14203
0.06049
0.13995
0.15809
0.05513
0.28210
0.14135
0.03876
0.30064
0.02839
0.02548
0.02821
Children Sampled Throughout Experiment
Log (Blood Lead)
Round
1
2
3
4
5
6
Group
AgeO
Agel
Age 2
Age 3
AgeO
Agel
Age 2
Age 3
AgeO
Agel
Age 2
Age 3
INTERVENTION
Agel
Age 2
Age 3
Agel
Age 2
Age 3
Age 2
Age 3
EST
-0.65580
-0.01758
0.03895
-0.04762
-0.51210
-0.14680
0.05951
-0.03420
-0.18430
0.07009
0.12180
-0.04249
0.02029
-0.06191
-0.07006
-0.02518
0.08798
-0.06385
-0.10740
-0.06745
SE
0.12450
0.09485
0.09171
0.03830
0.26700
0.10140
0.09507
0.03748
0.17720
0.13960
0.10790
0.03370
0.21100
0.15040
0.02446
0.26480
0.15340
0.02378
0.15940
0.02294
Blood
EST
0.51903
0.98257
1.03972
0.95350
0.59924
0.86347
1.06132
0.96638
0.83169
1.07260
1.12953
0.95840
1.02050
0.93997
0.93234
0.97513
1.09197
0.93815
0.89817
0.93477
Lead
SE
0.06462
0.09320
0.09535
0.03652
0.16000
0.08756
0.10090
0.03622
0.14737
0.14974
0.12188
0.03230
0.21532
0.14137
0.02280
0.25822
0.16751
0.02231
0.14317
0.02144
7-19
-------
Children Who Participated in all Six Rounds
Intervention
I O.Si
? 0.4
i 0.2-
fn n
U.U
JB -°-2
1 -0-4
I -0.6
I -0.8 •
I -1-°
1 -1-2-
1-1.4-
gj • * *
en
.,..........;: -i ,
— -j-r -
T123
1
0
1
i
«K
.1... .
?i
3
1
i
t
^«
ic
>3
t
1
J- I I
3 3 3
%
/
t i
••* ^^
234
Round of Sampling
All Children Sampled Throughout the Study
Intervention
8 0.5 1
S 0.4
§ °-3
1 II
1 -Ol2-
1 -0.3
ill
g ,Q*8
^ -0.9
§ II"?
f
V
•)
J
"<
>3
i
/
i
'
.
>3
•
- 1
\
Li
23
:
"
-I
3
2
:
i
',
T :
Ls
>
i
.1
3
>
234
Round of Sampling
0 = AgeO
1=Age1
2-Age2
3 = Age3
Figure 7-6. Model 2 results of effect of age on blood lead, log transformed. Bars show
95% confidence interval on regression coefficient.
7-20
-------
TABLE 7-10. REGRESSION COEFFICIENT FOR EFFECT
OF SES ON BLOOD LEAD MODEL 2
Round
1
2
3
DSTTERVENTTON
4
5
6
Children Present In All Six Rounds
Log (Blood Lead)
EST SE
-0.009112 .003973
-0.009250 .004070
-0.006410 .004198
-0.007688 0.004291
-0.004526 0.003968
-0.012690 0.004315
of Sampling
Blood
EST
0.99093
0.99079
0.99361
0.99234
0.99548
0.98739
Lead
SE
.0039370
.0040325
.0041712
.0042581
.0039501
.0042606
All Children Sampled Throughout Experiment
Round
1
2
3
INTERVENTION
4
5
6
Log (Blood Lead)
EST SE
-0.009923 .002417
-0.010820 .002662
-0.007053 .003221
-0.008668 .003405
-0.007217 .003319
-0.012290 .003347
Blood
EST
0.99013
0.98924
0.99297
0.99137
0.99281
0.98779
Lead
SE
.0023931
.0026334
.0031984
.0033756
.0032951
.0033061
Hand lead is positively associated with blood lead, as indicated by this model.
Mouthing behavior acts as an effect modifier for hand lead in this model, so that the effect of
hand lead is calculated separately for the two types of mouthing behavior. Although there is
no significant difference between these types of behavior, it is comforting to observe that the
effect of hand lead on blood lead is consistently higher for those children who exhibit strong
mouthing behavior (Table 7-12 and Figure 7-9).
7-21
-------
Children Who Participated in all Six Rounds
Intervention
1 0.0051
CQ
A f\f\f\
0.000
| -0.005
f-0.010
1 -0.01 5
g -0.020
& -0.025
r !"
is
1 23456
Round of Sampling
AH Children Sampled Throughout the Study
Intervention
| 0.005
•3 n nnr\
C U.UUU
1-0.005
1-0.010
1 -0.01 5
1-0.020
f-0.025
1
2
> (
3
t
t (
5 €
i
Round of Sampling
Figure 7-7. Model 2 results of effect of socioeconomic status on blood lead, log
transformed. Bars show 95% confidence intervals on regression coefficient.
I"
7-22
-------
TABLE 7-11. REGRESSION COEFFICIENT FOR EFFECT
OF SEASON ON BLOOD LEAD MODEL 2
Children Present In All Six Rounds of Sampling
Round
1
2
Log
EST
0.13700
0.03492
(Blood Lead)
SE
0.08284
0.07738
All Children Sampled Throughout
Round
1
2
Log
EST
0.13910
0.00414
(Blood Lead)
SE
0.05343
0.05630
Blood
EST
1.14683
1.03554
Experiment
Blood
EST
1.14924
1.00415
Lead
SE
0.095003
0.080130
Lead
SE
0.061404
0.056534
7.4.3 Model 3
As in the first model, Geometric means of hand lead and their associated standard
errors are calculated in each round for both the treatment and control groups in this model.
This measures the direct effect of group assignment on hand lead. The geometric means of
the two groups seemed to be almost identical to each other in the first 3 rounds, indicating
that there is no detectable difference between the two groups in hand lead measure prior to
intervention. In round four, the children in the control group on average had slightly lower
hand lead concentration than children in the treatment group. This difference is not
statistically valid, and the hand lead measurements in round four were taken between the
months of January and February, when children have little exposure to soil in the city of
Baltimore. Rounds five and six took place during the summer. Although the difference is
not statistically valid, the hand lead measurements of children in the abated group are lower
on average in the last two rounds of sampling (Table 7-13 and Figures 7-10 and 7-11).
7-23
-------
Children Who Participated in all Six Rounds
Intervention
1 0.31
f 0.26
I °'21
g 0.16
€ 0.11
| 0.06
f 0.01
| -0.04
1 -0.09
g -0.14
1
i
I
- , ;.
*" i
E
i
i i i i i :•
23456
Round of Sampling
All Children Sampled Throughout the Study
Intervention
1 0.28
I 0.23
J 0.18
| 0.13
1 0.08
1 0.03
f-0-02
|-0.07
Jt-0.1?
cc **• ••-'
I
2
! 3
i
i i . i
45 6
Round of Sampling
Kgure 7-8. Model 2 results of effect of season on blood lead, log transformed. Bars
show 95% confidence interval on regression coefficient.
I
7-24
-------
TABLE 7-12. REGRESSION COEFFICIENT FOR EFFECT
OF LOG HAND LEAD ON BLOOD LEAD MODEL 2
Children Present In All Six Rounds of Sampling
Log (Blood Lead)
Round
1
2
3
4
5
6
Round
1
2
3
4
5
6
Group
Weak
Strong
Weak
Strong
Weak
Strong
INTERVENTION
Weak
Strong
Weak
Strong
Weak
Strong
All
Group
Weak
Strong
Weak
Strong
Weak
Strong
INTERVENTION
Weak
Strong
Weak
Strong
Weak
Strong
EST
0.06375
0.14500
0.02039
0.07505
0.15650
0.23520
0.19030
0.28240
0.25380
0.38870
0.26090
0.35800
SB
0.06111
0.05798
0.04701
0.04542
0.05300
0.05273
0.05733
0.05888
0.04902
0.05584
0.06144
0.07089
Blood Lead
EST
1.06583
1.15604
1.02060
1.07794
1.16941
1.26516
1.20961
1.32631
1.28891
1.47506
1.29810
1.43047
SE
0.06513
0.06703
0.04798
0.04896
0.06198
0.06671
0.06935
0.07809
0.06318
0.08237
0.07976
0.10141
Children Sampled Throughout Experiment
Log
EST
0.11390
0.16500
0.06129
0.10270
0.08384
0.18150
0.16470
0.27190
0.24530
0.37150
0.24120
0.33010
(Blood Lead)
SE
0.04152
0.03896
0.03494
0.03445
0.04169
0.04068
0.04659
0.04401
0.04527
0.05017
0.04998
0.05402
EST
1.12064
1.17939
1.06321
1.10816
1.08745
1.19901
1.17904
1.31246
1.27800
1.44991
1.27278
1.39111
Blood Lead
SE
0.04653
0.04595
0.03715
0.03818
0.04534
0.04878
0.05493
0.05776
0.05786
0.07274
0.06361
0.07515
7-25
-------
I 0.5-
m
| 0.4
I 0.3
I 0.2-
§ 0.1
o
| 0.0
I -0.1
Children Who Participated in all Six Rounds
Intervention
I E
1 2345
Round of Sampling
6
All Children Sampled Throughout the Study
Intervention
0.48
0.43
0.38
0.33
0.28
0.23
0.18
0.13
0.08
0.03
-0.0?
c
' E
)
3
(•
\
i
1
T T '-
p
L
}
n
O
I - I
2 3
Round
D
-,-
--
G]
T T E3
' T B e ^ el :
G -L '
i-
1
• i i • • i
4 5 6
of Sampling
O = Weak Mouthing
Behavior
D = Strong Mouthing
Behavior
Figure 7-9. Model 2 results of effect of hand lead on blood lead, log transformed. Bars
show 95% confidence interval on regression coefficient.
7-26
-------
TABLE 7-13. REGRESSION COEFFICIENTS FOR EFFECT
OF ABATEMENT ON HAND LEAD MODEL 3
Children Present In All Six Rounds of Sampling
Round
1
2
3
4
5
6
Round
1
2
3
4
5
6
Group
Treatment
Control
Treatment
Control
Treatment
Control
INTERYENTION
Treatment
Control
Treatment
Control
Treatment
Control
All
Group
Treatment
Control
Treatment
Control
Treatment
Control
INTERVENTION
Treatment
Control
Treatment
Control
Treatment
Control
Log (Hand Lead)
EST SE
2.198 0.09658
2.136 0.07886
2.187 0.12480
2.191 0.09972
2.053 0.10970
2.021 0.09397
1.878 0.10300
1.565 0.08781
2.387 0.10210
2.616 . 0.08934
2.352 0.08260
2.451 0.07719
Children Sampled Throughout Experiment
Log (Hand Lead)
Hand Lead
EST
9.0070
8.4655
8.9084
8.9442
7.7912
7.5459
6.5404
4.7827
10.8808
13.6809
10.5066
11.5999
EST SE EST
2.095 0.09132
2.090 0.05466
2.157 0.11280
2.278 0.06462
1.899 0.08919
1.958 0.07243
1.859 0.08674
1.544 0.08724
2.347 0.08131
2.627 0.07992
2.374 0.07138
2.448 0.07222
8.1254
8.0849
8.6452
9.7571
6.6792
7.0851
6.4173
4.6833
10.4542
13.8322
10.7403
11.5652
SE
0.86989
0.66759
1.11177
0.89191
0.85470
0.70909
0.67366
0.41997
1.11093
1.22225
0.86784
0.89540
Hand Lead
SE
0.74202
0.44192
0.97517
0.63051
0.59572
0.51318
0.55664
0.40857
0.85003
1.10547
0.76664
0.83524
7-27
-------
I 3.0
3
« 2.5
I 2.0
1.5
1.0
Children Who Participated in all Six Rounds
Intervention
2345
Round of Sampling
3.0
2.5
2.0
EC
All Children Sampled Throughout the Study
Intervention
1 ' T~
3 4 5
Round of Sampling
O - Abate
D = Control
-T~
6
Figure 7-10. Model 3 results of effect of soil abatement on hand lead, log transformed.
Bars show 95% confidence interval on regression coefficient.
7-28
-------
I M
•5 15
J3 14
5 13
£ 12
14
Children Who Participated in all Six Rounds
Intervention
0
1 1 T
345
Round of Sampling
i 16
i 15
E 14
S 13
S 12
= 11
i 10^
i 9
I 8
6
5
4
3
All Children Sampled Throughout the Study
Intervention
e
I
34
Round of Sampling
O = Abate
D - Control
Figure 7-11. Model 3 results of effect of soil abatement on hand lead. Bars show 95%
confidence interval on regression coefficient;
7-29
-------
7.4.4 Model 4
This model is similar to the second model, in that it computed geometric means and
associated standard errors for group assignment after adjusting for variables that are thought
to be influence a child's hand lead concentration. The regression coefficients for the effect
of group assignment in this model are similar to those found in the third model, although the
associated standard errors are somewhat larger. There is no detectable difference between
the two groups in the rounds prior to intervention. The differences observed in the effect of
group assignment on hand lead following soil abatement were not statistically significant
(Table 7-14 and Figures 7-12 and 7-13).
There was a positive association between age and hand lead measurements in the
sampling rounds of the experiment prior to intervention. Although this effect was not
statistically valid, one interpretation of this result is that as children are more active and are
supervised less as they grow older (Table 7-15 and Figure 7-14).
Gender was added as a covariate for hand lead, because it was observed that; the female
participants seemed to have better personal hygiene. The regression coefficient for female
gender in this model was consistently negative as expected (Table 7-16 and Figure 7-15).
Hand lead was measured over a change in season for the first 2 rounds, therefore a
covariate which indicates this seasonal change was included in the model. This covariate
was not significant in either round for this model (Table 7-17 and Figure 7-16).
A measure of interior dust lead concentration was added as a covariate in this model.
The effect of dust lead on hand lead is consistently positive throughout the experiment. This
effect seems weaker in those rounds that took place over the summer, indicating that the
children probably spend less time inside their houses in these rounds (Table 7-18 and
I
Figure 7-17).
7.4.5 ModelS
The final model attempts to measure the effect of both soil and dust lead concentration
on hand lead measures independent of the effect of group assignment. Soil concentrations
measured before intervention were used as covariates in rounds 1 through 3 in the abated
properties. Following intervention, soil samples from the abated properties were collected
and analyzed for use as the soil covariate in the abated properties for the remaining 3 rounds.
7-30
-------
TABLE 7-14. REGRESSION COEFFICIENTS FOR ADJUSTED EFFECT
OF ABATEMENT ON BLOOD LEAD MODEL 2
Children Present In
Round
1
2
3
4
5
6
Round
1
2
3
4
5
6
Group
Treatment
Control
Treatment
Control
Treatment
Control
INTERVENTION
Treatment
Control
Treatment
Control
Treatment
Control
All
Group
Treatment
Control
Treatment
Control
Treatment
Control
INTERVENTION
Treatment
Control
Treatment
Control
Treatment
Control
Log
EST
2.498
2.394
1.933
1.979
2.239
2.267
1.789
1.531
2.192
2.456
2.422
2.556
All Six Rounds of Sampling
(Hand Lead)
SE
0.1937
0.1739
0.2311
0.2214
0.1974
0.1870
0.1752
0.1692
0.1764
0.1704
0.1567
0.1557
Hand Lead
EST
12.1582
10.9572
6.9102
7.2355
9.3839
9.6504
5.9835
4.6228
8.9531
11.6581
11.2684
12.8842
SE
2.35503
1.90546
1.59695
1.60194
1.85239
1.80463
1.04830
0.78218
1.57933
1.98654
1.76575
2.00607
Children Sampled Throughout Experiment
Log
EST
2.298
2.259
1.794
1.902
2.141
2.187
1.691
1.444
2.232
2.521
2.402
2.505
(Hand Lead)
SE
0.1451
0.1313
0.1794
0.1611
0.1643
0.1516
0.1560
0.1574
0.1428
0.1464
0.1313
0.1362
EST
9.9543
9.5753
6.0135
6.6993
8.5079
8.9084
5.4249
4.2376
9.3185
12.4410
11.0452
12.2436
Hand Lead
SE
1.44436
1.25700
1.07881
1.07925
1,39785
1.35052
0.84628
0.66700
1.33068
1.82137
1.45024
1.66757
7-31
-------
Children Who Participated in all Six Rounds
Intervention
3 4
Round of Sampling
All Children Sampled Throughout the Study
Intervention
3 4
Round of Sampling
O = Abate
D = Control
Figure 7-12. Model 4 results of effect of soil abatement on hand lead, log transformed.
Bars show 95% confidence interval on regression coefficient.
7-32
-------
§ 20
i 18
I 16
5 14
| 12
1 10
| 8
o 6
I 4
i 2
CD f.
» 0
Children Who Participated in all Six Rounds
Intervention
234 5 6
Round of Sampling
All Children Sampled Throughout the Study
Intervention
£ 20 1
e
5 18
1 16
1 14
s 12
1 10
tss CJ
ft O '
o 6
I 4
OS ^^
CO ^
5 0
DC v
5 E
I
1
]
C
)C
i
2
/•
C
0
•v r
:) t
I
3
i
j
j
T
T ^ O ^
<:;>
z
X (|
1 1 1
4 5 6
Round of Sampling
O = Abate
D = Control
Figure 7-13. Model 4 results of effect of soil abatement on hand lead. Bars show 95%
confidence interval on regression coefficient.
7-33
-------
TABLE 7-15. REGRESSION COEFFICIENTS FOR EFFECT
OF AGE ON HAND LEAD MODEL 4
Children Present In All Six Rounds of Sampling
Log (Hand Lead)
Round
1
2
3
4
5
6
Round
1
2
3
4
5
6
Group
AgeO
Agel
Age 2
Age 3
Agel
Age 2
AgeS
Agel
Age 2
Age 3
INTERVENTION
Age 2
Age 3
Age 3
Age 3
All
Group
AgeO
Agel
Age 2
Age3
AgeO
Agel
Age 2
AgeS
AgeO
Agel
Age 2
AgeS
EtflERVENTION
Agel
Age 2
AgeS
Agel
Age 2
AgeS
Age 2
AgeS
EST
-1.10200
-0.33540
0.00674
-0.03076
-0.15050
0.16700
0.20380
-0.79020
•0.31750
-O.00479
-0.13830
0.04095
0.08735
0.03363
SB
0.25130
0.19940
0.19320
0.09023
0.26730
0.24980
0.10150
0.79460
0.23140
0.06893
0.55240
0.04757
0.04513
0.03881
Hand
EST
0.33221
0.71505
1.00676
0.96971
0.86028
1.18175
1.22605
0.45375
0.72797
0.99522
0.87084
1.04180
1.09128
1.03420
Lead
SE
0.08348
0.14258
0.19451
0.08750
0.22995
0.29520
0.12444
0.36055
0.16845
0.06860
0.48105
0.04956
0.04925
0.04014
Children Sampled Throughout Experiment
Log
EST
-1.21400
-0.41660
-0.07236
0.01890
-0.72620
-0.11540
0.06398
0.21540
-0.68540
-0.45940
-0.21580
0.01716
0.83000
0.35360
0.04633
0.87800
0.01320
0.06734
0.39450
0.03421
(Hand Lead)
SE
0.18400
0.14850
0.14810
0.06098
0.50740
0.19240
0.18190
0.06869
0.29330
0.22920
0.17980
0,05653
0.37990
0.27510
0.04440
0.43680
0.25880
0.03758
0.24260
0.00335
Hand
EST
0.29701
0.65928
0.93020
1.01908
0.48374
0.89101
1.06607
1.24036
0.50389
0.63166
0.80590
1.01731
2.29332
1.42419
1.04742
2.40608
1.01329
1.06966
1.48364
1.03480
Lead
SE: • ;;-
0.05465
0.09790
0.13776
0.06214
0.24545
0.17143
0.19392
0.0$520
0.14779
0.14478
0.14490
0.05751
1
0.87123
0.39179
0.04651
1.05098
0.26224
0.04020
0.35993
0.03451
7-34
-------
Children Who Participated in all Six Rounds
Intervention
1 1
3
•5
| 0
_c
i -1
«|
o ^
c •*
_O j;
1- "
ST -3
|- .
T _L
. . T!T- - - - "J T I.
ft .. o T
-•io 13 --3
T 23 12 •
-1 1 2
- -
0
1
T
f-i- -r
_
2
s. , ; ' . . " •
(g- - I.I 1 1
1.2 3 45 6
Round of Sampling
All Children Sampled Throughout the Study
| Intervention
1 2|
at
5
I 1
"c .
1
| -1
"55
i
o> O •
tn ^^«
U. Till ...j...
T13 fl3 T il
23 ..^2 ..-I3
I1 I1
0 h
*^
^
-- - -•
^3 1{3 23
2
£C -I' III
123456
Round of Sampling
0 = AgeO 2 = Age2
1=Age1 3 = AgeS
Figure 7-14. Model 4 results of effect of age on hand lead, log transformed. Bars show
95% confidence interval on regression coefficient.
7-35
-------
TABLE 7-16. REGRESSION COEFFICIENTS FOR EFFECT OF
FEMALE GENDER ON HAND LEAD MODEL 4
Children Present In All Six Rounds of Sampling
Round
1
2
3
INTERVENTION
4
5
6
Log (Hand Lead)
EST SE
-0.1454 0.11490
-0.3607 0.14870
-0.4888 0.13740
-0.2887 0.13180
-0.3327 0.13110
-0.4055 0.10930
Hand
EST
0.86468
0.69719
0.61336
0.74924
0.71699
0.66664
Lead \
SE
0.09935
0.10367
0,08428
0,09875
0.09400
0.07286
All Children Sampled Throughout Experiment
Round
1
2
3
INTERVENTION
4
5
6
Log (Hand Lead)
EST SE
-0.1462 0.08502
-0.1378 0.10600
-0.4666 0.10530
-0.1999 0.12200
-0.2834 0.11110
-0.3392 0.09927 ,
Hand
EST
0.86398
0.87127
0.62713
0.81881
0.75322
0.71234
Lead
SE
0.07346
0^09235
0:06604 >i
0:09990
0:08368
0107071
The measure of soil lead concentration for properties in the control group remained the same
for all six rounds of the experiment. This model also adjusts for covariates that are thought
to influence hand lead.
The effect of gender on hand lead in this model is similar to that found in the previous
model; hand lead is higher on average for the male participants (Table 7-19 and
Figure 7-18).
7-36
-------
Children Who Participated in all Six Rounds
Intervention
S 1.0O
I 0.75
| 0.50
1 °-25
c f) rin
| 1.00
f 0.75
1 0.50
| 0.25
1 o.oo
TS
1 -0.25
o
1 -0.50
i -0.75
1-1.00
.
• - ^^ if
Intervention
T
--
-•
-1-
1 23 4 5 6
Round of Sampling
Figure 7-15. Model 4 results of femal gender effect on hand lead, log transformed.
Bars show 95% confidence interval on regression coefficient.
7-37
-------
TABLE 7-17. REGRESSION COEFFICIENTS FOR EFFECT
OF SEASON ON HAND LEAD MODEL 4
Children Present In All Six Rounds of Sampling
Round
1
2
Log (Hand Lead) Hand Lead
EST SE EST
-0.11910 0.12610 0.88772 JO.
0.05406 0.14940 1.05555 0.
SE
11194
15770
All Children Sampled Throughout Experiment
Round
1
2
Log (Hand Lead) Hand Lead
EST SE EST
0.02264 0.08617 1.02290 0.
0.12090 0.10570 1.12851 0.
SE
08814
11928
Age is positively associated with hand lead (Table 7-20 and Figure 7-19). Change in
season has no an effect on the hand lead concentration of children in the study (Table 7-21
and Figure 7-20).
The effect of dust lead concentration on hand lead was consistently positive throughout
.
the experiment. This effect is markedly higher in the sampling rounds that took place during
the winter months (Table 7-22 and Figure 7-21).
The effect of soil lead concentration on hand lead was inconsistent throughout the
experiment. The inconsistency could not be attributed to season in the same way that the
effect of dust lead was explained (Table 7-23 and Figure 7-22).
The R-squared coefficient and estimated mean square error from these 5 models are
presented in Table 7-24.
The results of the models applied to the Baltimore data for the purpose of comparison
to the Boston and Cincinnati results revealed no statistical difference between treatment and
control groups in any of the rounds following intervention. The statistical output from these
models are in Appendix I.
The reader will notice that there are different numbers of children included in each of
the cross sectional models, even in the population which consisted of children who were
7-38
-------
"8
Children Who Participated in all Six Rounds
Intervention
1 0.41
o>
3 0.3
1 0.2
1 0.1-
^ f\ f\
§ 0.0
1 -0.1
3 -0.2
1 -0.3
2
S* -04
c u-^
i
«
i ii
1 2 345 6
ftound of Sampling
All Children Sampled Throughout the Study
Intervention
| 0.351
g> 0.30
^ 0.25
| 0.20
= 0.15
I 0.10
1 0.05
5? /\ f\f\
§ O.OQ
§ -0.05
§ -0.10
1* _n i c-
3456
Round of Sampling
Figure 7-16. Model 4 results of effect of season on hand lead, log transformed. Bars
show 95% confidence interval on regression coefficient.
7-39
-------
TABLE 7-18. REGRESSION COEFFICIENTS FOR EFFECT OF
DUST ON HAND LEAD MODEL 4
Children Present In All Six Rounds of Sampling
Round
1
2
3
BSTTERVENTION
4
5
6
Log (Hand Lead)
EST SE
.00000710 .00003667
.00014850 .00004625
.00008436 .00004288
.00008650 .00004040
.00007582 .00004064
.00000500 .00003345
Hand
EST
1,00000710
1.00014851
1.00008436
1.00008650
L00007582
1.00000500
•
Lead
' SE.
.000036670
.000046257
.000042884
.000040403
.000040643
.000033450
All Children Sampled Throughout Experiment
Round
1
2
3
INTERVENTION
4
5
6
Log (Hand Lead)
EST SE
.00003876 .00003166
.00015090 .00003866
.00008175 .00003749
.00008103 .00003922
.00006544 .00003703
.00001513 .00003226
Hand
EST ;
1.00003876
1.00015091
1.00008175
1.00008103
1.00006544
1.00001513
Lead
SE
.000031661
.000038666
.000037493
.000039223
.000037032
.000032260
7-40
-------
1 0.00025
0.00020
0.00015
0.00010
0.00005
0.00000
-0.00005
Children Who Participated in all Six Rounds
Intervention
1 2 3 4 5 ; 6
Round of Sampling
All Children Sampled Throughout the Study
Intervention
1 0.00025
f 0.0002C
To
I 0.00015
| 0.0001 C
S 0.00005
1 O.OOOOC
to
s.
£ -0.00005
1 234 56
Round of Sampling
Figure 7-17. Model 4 results of effect of dust lead on hand lead, log transformed. Bars
show 95% confidence interval on regression coefficient.
7-41
-------
TABLE 7-19. REGRESSION COEFFICIENTS FOR EFFECT
OF GENDER ON HAND LEAD MODEL 5
Children Present In All Six Rounds of Sampling
Round
1
2
3
4
5
6
Round
1
2
3
4
5
6
Group
Male
Female
Male
Female
Male
Female
INTERVENTION
Male
Female
Male
Female
Treatment
Control
All
Group
Male
Female
Male
Female
Male
Female
INTERVENTION
Male
Female
Male
Female
Male
Female
Log (Hand Lead)
EST SE
2.424 0.1928
2.277 0.1889
1.792 0.2417
1.441 0.2434
2.224 0.2230
1.741 0.2160
1.672 0.1716
1.360 0.1835
2.293 0.1729
1.969 0.1903
2.414 0.1527
1.997 0.1657
Children Sampled Throughout Experiment
Log (Hand Lead)
EST SE
2.281 0.1372
2.135 0.1370
1.762 0.1717
1.626 0.1750
2.199 0.1782
1.730 0.1648
1.609 0.1539
1.399 0.1595
2.316 0.1425
2.037 0.1489
2.400 0.1283
2.053 0.1344
Hand Lead
EST
11.2909
9.7474
6.0014
4.2249
9.2442
5.7030
5.3228
3.8962
9.9046
7.1635
11.1786
7.3669
EST
9.7865
8.4570
5.8241
5.0835
9.0160
5.6407
4.9978
4.0511
10.1351
7.6676
11.0232
7.7912
SE
2.17689
1.84128,
1.45055
1.02835
; 2.06146
;- 1.23186
-. . 1
0.91339
0.71495
1.71251
1.36322
1.70697
1.22070
Hand ] Lead
1 SE
, 1.34270
, , 1.15862
0.99999
6.88961
.1,60665
0.92958
0.76916
; 0.64616
.' • ; • 1.44425
1.14170
1.41427
; 1.04714
7-42
-------
I 3.0
^ 2.5
1
2 2.0
1.5
0.5
0.0
Children Who Participated in all Six Rounds
Intervention
1 2 34 5 6
Round of Sampling
13
3
I 3.0
•5 2-5
1 2.0
| 1.5
1 1.0
I 0.5
I 0.0
All Children Sampled Throughout the Study
Intervention
3 4
Round of Sampling
(
6
O =Male
a = Female
Figure 7-18. Model 5 results of effect of gender on hand lead, log transformed. Bars
show 95% confidence interval on regression coefficient.
7-43
-------
TABLE 7-20. REGRESSION COEFFICIENTS FOR EFFECT
OF AGE ON HAND LEAD MODEL 5
Children Present In All Six Rounds of Sampling
Round
1
2
3
4
5
6
Round
1
2
3
4
5
6
Group
AgeO
Agel
Age 2
Age 3
Agel
Age 2
Age3
Agel
Age 2
Age 3
INTERVENTION
Age 2
Age 3
Age3
Age 3
All
Group
AgeO
Agel
Age 2
Age 3
AgeO
Agel
Age 2
AgeS
AgeO
Agel
Age 2
Age 3
INTERVENTION
Agel
Age 2
Age 3
Agel
Age 2
Age 3
Age 2
Age3
Log (Hand Lead)
EST SE
-1.10600 0.25210:
-0.33970 0.20150
-0.01190 0.19250
-0.03152 0.09115
-0.19250 0.26700
0.14940 0.24820
0.18480 0.10160
-0.77470 0,79260
-O.31930 0.23090
-0.00567 0.06877
-0.22040 0.55860
0.04482 0.4817
0.08537 0.04570
0.03344 , 0.03862,
Children: Sampled Throughout Experiment
Log: (Hand Lead)
EST SE
-1.21200 Oi.18500.
-0.41590 0.14950
-0.07558 0,14800
0.01779 0.06142
-0.72440: 0:50570
-0.12660 0,19190
0.06258 0,18100
0.20830 0.06867
-0.69160 0.29410
-0.47870 0:22670
-0:21700 0.17990
0.01706 0.05657
0.87800 0:38380
,0,41440. Oi27630
0.04575 0,04488
0.86730 0.44460
-0.01182 0.26340
0.07011 0.03820
0.39450 0.24150
0.03247 0.03332
Hand
EST
0.33088
0.71198
0.98817
0.96897
0.82489
1,16114
1.20298
0.46084
0.72666
0.99434
0.80220
1.04584
1.08912
1.03401
Hand
EST
0.29760
0,65975
0,92721
1.01795
0.48462
0188109
1,06458
1.23158
0.50077
0.61959
0.80493
1.01721
2.40608
E51346
1.04681
2.38047
0.98825
1.07263
1.48364
1.03300
Lead
SE
0.08341
0,14346
0.19022
0.08832
0.22025
0.288195
0,12222,
0.36526.
0.16779
0.06838
1
0,44811
0.05038
; , 0.04977
Or.03993
Lead!
SE
0.05506
0.09863
0.13723
0,06252
0.24507
0,16908
0.19i69
0.08457
0,14728
0:14046
0,14481
0.05754
0.92345^
' 0.41817
0.04698.
1.05836
0.26030
0.04097
0.35830
0.03442
7-44
-------
Children Who Participated in all Six Rounds
Intervention
* 1 1
0)
«• n
yj ^j
"c
c
I -1
«|
0 .0
c "
_o
i
& -3
(
•
T
-l| •
I2 :
t
• t
-£
L3
)
T^
| 3
2
i
/•
-r j
3 3 1
i
1 2 3 4 5 6
Round of Sampling
All Children Sampled Throughout the Study
Intervention
1 2 \
5?
3
•
C
1 o
_£
|f|---
ji
J
0 ]
FlI1
Tl3"""i
J-2
1
i
i i
12
rT-i—
i3
2
.-...^l........^! ^..._._...
2
iiit
3456
Round of Sampling
0 = AgeO
1 = Age1
2 = Age2
3 = Age3|
Figure 7-19. Model 5 results of effect of age on hand lead, log transformed. Bars show
95% confidence interval on regression coefficient.
7-45
-------
TABLE 7-21. REGRESSION COEFFICIENTS FOR EFFECT
OF SEASON ON HAND LEAD MODEL 5
1
Children Present In All Six Rounds of Sampling
Round
1
2
Log
EST
-0.08192
0.08260
(Hand Lead)
SE
0.11830
0.14960
All Children Sampled Throughout
Round
1
2
Log
EST
0.02684
0.13030
(Hand Lead)
SE
0.08542
0.10560
Hand
EST
0.92135
1.08611
Experiment
Hand
EST
1.02720
1.13917
Lead
SE
0.16900
0.16248
Lead
SE
0.08774
0.12030
present in all six rounds of sampling. This is largely a result of missing or unavailable data
for the models (mostly destroyed or inadequate samples).
7.5 IMPLICATIONS OF FINDINGS
If it can be assumed that the dose response relationship between soil lead and blood lead
is sigmoidal in shape, it would be expected that the reductions achieved at the low levels in
this study would not result in statistically significant reductions in blood lead levels.
The findings of this project may help avoid costly abatements of soil in cities where, like
Baltimore, the principal source of lead exposure for children is paint in and around their
houses along with the resulting house dust. There may be individual children, however, who
are more likely to benefit from soil abatement; i.e., those who have unusually strong
I
mouthing behavior or pica. For these children, however, environmental controls, while
important, must be supplemented with attention to hygiene, nutrition, and any underlying
behavioral or medical problems. Abatement of lead paint problems and of lead dust levels in
7-46
-------
Children Who Participated in all Six Rounds
0?
3 Intervention
1 OA]
5?
3 0.3
I 0.2
1 0.1
1 0.0
1 -0.1
f -0.2
| -0.3
i? -0.4
1 2 3 45 6
Round of Sampling
All Children Sampled Throughout the Study
Intervention
| 0.351
g> 0.30
•s 0.25
| 0.20
= 0.15
1 0.10
1 0.05
§ n nn.
Q U.UU
I -0.05
1 -0.10
1 23456
Round of Sampling
Figure 7-20. Model 5 results of effect of season on hand lead, log transformed. Bars
show 95% confidence interval on regression coefficient.
7-47
-------
TABLE 7-22. REGRESSION COEFFICIENTS FOR EFFECT OF
DUST ON HAND LEAD MODEL 5
Children Present In All Six Rounds of Sampling
Round
1
2
3
INTERVENTION
4
5
6
Log (Hand Lead)
EST SE
.00000730 .00003677
.00014840 .00004587
.00008422 .00004286
.00008800 .00004094
.00007411 .00004116
.00000390 .00003329
Hand
EST
1.00000730
1.00014841
1.00008422
1.00008800
1.00007411
1.00000390
Lead
SE '
.000036770
i
.000045877
.000042864
.000040944
.000041163
.000033290
^,— All Children Sampled Throughout Experiment
• •
Round
1
2
3
I^^mVENTION
4
5
6
Log (Hand Lead)
EST SE
.00004023 .00003176
.00014350 .00003862
.00008190 .00003760
.00008444 .00003958
.00006301 .00003762
.00001454 .00003216
Hand
EST
1.00004023
1.00014351
1.00008190
1.00008444
1.00006301
1.00001454
Lead
SE
.000031761
.000038626
.000037603
.000039583
.000037622
.000032160.
their houses would probably take priority over soil abatement, which might be limited to
areas of obvious contamination where the child is known to spend time.
Soil abatement for cities like Baltimore does not appear to be a cost effective preventive
strategy used alone, but it may weE be an adjunct, in selected cases, to the overall
environmental management of children who become lead poisoned.
7-48
-------
I 0.0002£i
3 0.0002(1
•5
I 0.0001 fi
1 O.OOOKi
I o.oooof;
3 0.0000(r
1 -0.00
£ -0.0001
Children Who Participated in all Six Rounds
Intervention
1 23456
Round of Sampling
| 0.0002£
f 0.0002C
S 0.0001 £
1 0.0001 c
I O.OOOOf
I o.ooooc
f-0.00006
I -0.0001 C[
All Children Sampled Throughout the Study
Intervention
123456
Round of Sampling
Figure 7-21. Model 5 results of effect of dust lead on hand lead, log transformed. Bars
show 95% confidence interval on regression coefficient.
7-49
-------
TABLE 7-23. REGRESSION COEFFICIENTS FOR EFFECT OF
SOIL LEAD ON HAND LEAD MODEL 5
Round
1
2
3
INTERVENTION
4
5
6
Round
1
2
3
INTERVENTION
4
5
6
Children Present In All Six Rounds
Log (Hand Lead)
EST SE
-0.00000500 ,00002036
0.00036780 .00002507
0.00006254 .00002586
-0.00010790 .00002379
0.00019070 .00002413
0.00033070 .00001967
All Children Sampled Throughout
Log (Hand Lead)
EST SE
-0.00002800 .00001307
0.00024190 .00001573
0.00005640 .00002096
-0.00017300 .00002306
0.00024000 .00002168
0.00025980 .00001899
of Sampling
Hand
EST
0.99999500
1.00036787
1.00006254
0.99989211
1.00019072
1.00033075
Experiment
Hand
EST
0.99997200
1.00024193
0.99994360
0.99982701
1.00024003
1.00025983
Lead i
SE
.00020360
.00025709
.00025862
! ,'
.00013070
.00015734
i1
.00020959
Lead
SE
.00023787
.00024135
.00019677
; r - •
.00023056
.00021685
.00018995
7.6 CALL FOR FURTHER RESEARCH
•
The models that were applied to the Baltimore data were cross sectional by sampling
round, and do not combine all the data from the study into one model. One of the
assumptions of linear regression models is that each response is independent and identically
distributed from a known distribution. The multiple measurements taken on each child
throughout the experiment violates the assumption of independence. Some of the children in
7-50
-------
I 0.00100
J* 0.00075
I 0.00050
= 0.0002
•§
O.OOOOC
-0.00025
-0.00053
-0.0007
'-0.001 OQ
Children Who Participated in all Six Rounds
Intervention
i 0.001 OG
j? 0.00075
1 0.0005G
5 0.0002E
~ OOOOOC
I -0.0002E
0-0.0005C
|-aooo7s
I-0.0010C
I ^
„
:
1 2 3 4 5 6
Round of Sampling
All Children Sampled Throughout the Study
Intervention
123456
Round of Sampling
Figure 7-22. Model 5 results of effect of soil lead on hand lead, log transformed. Bars
show 95% confidence interval on regression coefficient.
7-51
-------
TABLE 7-24. R-SQUARED COEFFICIENT AND MEAN SQUARE ERROR
FOR MODELS WITH LOG (BLOOD LEAD) AS THE RESPONSE VARIABLE
Model 1 - Population 1
Round
1
2
3
4
5
6
N
140
136
130
133
136
133
DF
138
134
128
131
134
131
R-Sguare
0.002
0.003
0.000
0.008
0.008
0.012
Variance
0,2359
0.2051
0.2384
0.2612
0;2683
0.2659
Model 1 - Population 2 -.'.'.••
Round
1
2
3
4
5
6
N
273 "
255
229
175
173
170
DF
271
253
227
173
171
168
R-Square
0.001
0.000
0.000
0.018
0.021
0.029
Variance
0.2593
0.2162
0.2569
0.2480
0.2744
,'
0.2471
Model 2 - Population 1
Round
1
2
3
4
5
6
N
140
136
130
133
136
133
DF
130
127
122
126
130
127
R-Square
0.230
0.105
0.193
0.204
0.324
0.251
Variance
0.1931
0,1943
0.2020
. 0.2179 .
0.1885
0.2080 . "' .
7-52
-------
TABLE 7-24 (cont'd). R-SQUARED COEFFICIENT AND MEAN SQUARE ERROR
FOR MODELS WITH LOG (BLOOD LEAD) AS THE RESPONSE VARIABLE
Model 2 - Population 2
Round
1
2
3
4
5
6
N
273
255
229
175
173
170
DF
263
245
220
167
165
163
R-Square
0.296
0.133
0.162
0.279
0.359
0.325
Variance
0.1883
0.1936
0.2222
0.1886
0.1863
0.1771
Model 3 - Population 1
Round
1
2
3
4
5
6
N
140
136
130
133
136
133
DF
138
134
128
131
134
131
R-Square
0.002
0.000
0.000
0.039
0.021
0.006
Variance
0.5224
0.8253
0.6623
0.5937
0.6146
0.4230
Model 3 - Population 2
Round
1
2
3
4
5
6
N
273
255
229
175
173
170
DF
271
253
227
173
171
168
R-Square
0.000
0.003
0.001
0.036
0.034
0.003
Variance
0.6005
0.8017
0.7239
0.6621
0.5620
0.4381
7-53
-------
TABLE 7-24 (cont'd). R-SQUARED COEFFICIENT AND MEAN SQUARE ERROR
FOR MODELS WITH LOG (BLOOD LEAD) AS THE RESPONSE VARIABLE
Model 4 - Population 1
Round
1
2
3
4
5
6
N
140
136
130
133
136
133
DF
131
128
123
127
131
128
R-Square
0.176
0.153
0.132
0.111
0.107
0.104
Variance
0.4544
0.7319
0.5983
0.5668
0.5732
0.3902
Model 4 - Population 2
Round
1
2
3
4
5
6
N
273
255
229
175
173
170
DF
264
246
221
168
166
164
R-Square
0.211
0.157
0.160
0.100
0.117
0.080
Variance
0.4862
0.6978
0.6256
0.6368
0.5294
0.4141
Model 5 - Population 1
Round
1
2
3
4
5
6
N
140
136
130
133
136
133
DF
131
128
123
127
131
128
R-Square
0.171
0.166
0.132
O.087
0.084
0.113
Variance
0.4568
0.7209
0.5983
0.5822
0,5881
0.3863
Model 5 - Population 2
Round
1
2
3
4
5
6
N
273
255
229
175
173
170
DF
264
246
221
168
166
164
R-Square
0.211
0.162
0.159
0.082
0.088
0.085
Variance
0.4864
0.6934
0.6260
0.6496
\
0.5467
0.4121
7-54
-------
the Baltimore study also reside in the same household, and are likely to be correlated as well
(Figure 6). While combining the data from all six rounds of sampling, these correlation
structures must be taken into consideration. Another statistical issue is that the time of
sampling was not consistent throughout the experiment. There were also children who were
present in most of the rounds of sampling but may have missed 1 or 2 rounds (Figure 6).
The data from measurements on these children may provide additional information for
explaining the effects of soil abatement. Research is now being conducted in an effort to
combine the data longitudinally while addressing these issues.
7-55
-------
-------
8. REFERENCES
Agency for Toxic Substances and Disease Registry: The Nature and Extent of Lead Poisoning in Children. U.S.
Public Health Service, U.S. Department of Health and Human Services 1988
Annest JL, Pirlde JL, Makuc D, Neese JW, Bayse DD, Kovar MG: Chronological Trend in Blood Lead Levels
between 1976-1980. New England Journal of Medicine 1983;308:1373-1377
Barltrop D, Khoo HE: The influence of Nutritional Factors on Lead Absorption. Postgraduate Medical Journal
1975;51:795-800
Barltrop D, Meek F: Effect of Particle Size on Lead Absorption from the Gut. Archives of Environmental Health
1979:34:280-285
Barltrop D, Strehlow CD, Thorton I, Webb JS: Significance of High Soil Lead Concentrations for Childhood
Lead Burdens. Environmental Health Perspectives 1974;7:75-82
Binder S, Sokal D, Maughan D: Estimating Soil Ingestion: The use of tracer elements in estimating the amount
of soil ingestion by young children. Archives of Environmental Health 1986;41:341-345
Centers of Disease Control: External Quality Assurance/Quality Control. May, 1991.
Chisolm JJ, Jr, Brown DH: Micro-scale photofluorometric determination of "free erythrocyte porphyrin"
(protoporphyrin TX). Clinical Chemistry 1975; 21:1669-1682
Duggan MJ, Inskip MJ: Childhood Exposure to Lead in Surface Dust and Soil: A Community Health Problem.
Public Health Review 1985;13:l-54
Environmental Protection Agency: Air Quality Criteria for Lead. Pub. No. 600/8-83/028bf. Research Triangle
NC: EPA 1986:1-4
Forman DT, Parker SL: The measurement and interpretation of Serum Ferritin. Annals of Clinical and
Laboratory Science 1980;10: 345-350
Gibson JL, Love W, Hardine D, Bancroft P, Turner AJ: Notes on Lead Poisoning as observed among children in
Brisbane. 1892; Transcription of the 3rd Intercolonial Medical Congress:76-83
Mahaffey KR: Nutritional Factors in Lead Poisoning. Nutrition Reviews 1981;39:353-362
Mahaffey KR, Rader JI: Metabolic Interactions: Lead, Calcium, and Iron. Annals of the N.Y. Academy of Science
1980;355:285-297
Mielke HW, Anderson JC, Berry KJ, Mielke PW, Chaney RL, Leech M: Lead Concentrations in Inner-City
Soils as a Factor in the Child Lead Problem. American Journal of Public Health 1983;73,12:1366-1369
Pruszkowski E, Cornick GR, Slavin W: Blood lead determination with the platform furnace technique. Atomic
Spectroscopy 1983;4:59-61
Rabinowitz MB, Kopple JD, Wetherill GW: Effects of Food intake and fasting on Gastrointestinal lead
absorption in humans. American Journal of Clinical Nutrition, 1980;33:1784-1788
8-1
-------
Rabinowitz MB, Wetherill GW, Kopple ID: Kinetic Analysis of Lead Metabolism in Healthy Humans. The
Journal of Clinical Investigation 1976;58:260-270
Rods HA, Buchet J-P, Lauwerys R, Bruaux P, Claeys-Thoreau F Lafontaine A, Verduyn G: Exposure to Lead
by the Oral and the Pulmonary Routes of Children Living in the Vicinity of a Primary Lead Smelter.
Environmental Research 1980;22:81-94
I .
Rosen JF, Chesney RW, Hamstra A, DeLuca HF, Mahaffey KR: Reduction hi 1,25-DihydroxyvitaminD hi
Children with Increased Lead Absorption. New England Journal of Medicine 1980;302:1128-1131
Sayre JW, Chamey E, Jaroslav V, Pless IB: House and Hand Dust as a Potential Source of Childhood Lead
Exposure. American Journal of Disease in Children 1974;127:167-170
Shaffar M, Stroupe SD: A general method for routine clinical chemistry on the Abbott TDx analyzer. Clinical
Chemistry 1983;129:1251 (abstract)
Shcllshear ED, Jordan LD, Hogan DJ, Shannon FT: Environmental Lead Exposure in Christchurch Children: Soil
Lead a Potential Hazard. New Zealand Medical Journal 1975;81:382-386 \
Stark AD, Quah RF, Meigs JW, DeLouise ER: The Relationship of Environmental Lead to Blood-Lead Levels hi
Children. Environmental Research 1982;27:372-383
Ter Haar, G. and Aronow, R: New information on Lead hi Dirt and Dust as related to the Childhood Lead
Problem. Environmental Health Perspective 1974;7:83-89
USEPA: Uptake/BioMnetic Model. ECAO/CIN/September 1, 1990.
I
Williams H, Schulze WH, Rothchild HB, Brown AS, Smith FR, Jr.: Lead Poisoning from the Burning of Battery
Casings. Journal of the American Medical Association 1933;100:1485-1489
Zicglcr EE, Edwards BB, Jensen RL, Mahaffey KR, Fomon SJ: Absorption and Retention of Lead by Infants.
Pediatric Research 1978;12:29-34
8-2
-------
APPENDICES
Baltimore Soil Lead Abatement
Demonstation Project
Final Report
Submitted April 20, 1992
to
The Environmental Protection Agency
by
Katherine P. Parrel I, M.D., M.P.H.
J. Julian Chisolm, M.D.
Charles A. Rohde, Ph.D.
Boon P. Urn, M.D,, M.P.H.
Merrill C. Brophy, M.S.N., R.N.
and
Warren J. Strauss, B.S.
-------
Appendix A
Protocols
PAGE
Blood Sampling •, • • «A-1
Blood Analysis, Graphite Furnace AAS .A-3
Handwipe Sampling ««-A-16
Handwipe Analysis, Nitric Acid/Perchloric Acid Digestion....A-17
Handwipe Analysis. Nitric Acid Digestion —A-21
Soil Sampling • :• • -A-25
Soil Analysis, XRF Protocol A-28
Soil Quality Assusrance Plan • A-32
Household Dust Sampling ,• • .A-34
Household Dust Analysis, XRF Protocol . A-35
Household Dust Analysis, Wet Digestion AAS .A-38
Dust Quality Assurance Plan. . . A-42
Paint Chip Sampling A-44
Paint Chip Analysis .. • A-45
Paint Chip Quality Assurance Plan .A-:48
Drinking Water Sampling -A-50
Drinking Water Analysis : •. -A-51
Drinking Water Quality Assurance Plan .A-53
Lead Paint Stabilization .A-58
Soil Abatement « 1 • • .A-61
-------
BLOOD LEAD, TIBC, AND FERRITIN
RT.OOn PnT.T.RCTION AND PROCESSING PROTOCOL
A. COLLECTION PROCEDURE
1. Materials needed per participant
- Gauze sponges, sterile, individually wrapped 2X2" (2)
- Alcohol wipe (2)
- Bandaid
- 3 mL lavender top vacutainer
- 5 mL red top vacutainer
- 21 or 23 gauge butterfly (used for children in place
of needle)
- 5 cc syringe
- Tourniquet
- 3 cc plastic screw top tube
- Pipette
- Pre-printed labels
- Refrigerator or cooler for holding blood
- Latex gloves
2. Venipuncture procedure
- Locate a suitable table for blood drawing and lay out
blood collection supplies. Put on gloves.
- Locate the puncture site. Hold with 2 fingers on one
side of the "alcohol wipe" so that only one side touches
the puncture site. Wipe the area in a circular motion
beginning with a narrow radius and moving outward so as
to not cross over the area already cleaned. Repeat with
a second alcohol wipe.
- Locate vein and cleanse in manner previously described,
then apply the tourniquet. If it is necessary to feel the
vein again, do so; but after you feel it, cleanse with
alcohol wipe again.
- Fix the vein by pressing down on the vein about 1/2
inch below the proposed point of entry into the skin.
- Approach the vein in the same direction the vein is
running, holding the needle so that it makes a 45 degree
angle with the examinee's arm.
- Push the needle, with bevel facing up, firmly and
deliberately into the vein. If the needle is in the vein,
blood will flow freely into the butterfly tubing. If .no
blood enters the tubing, probe for the vein until entry
is indicated by blood flowing freely into the tubing.
A - 1
-------
- For collection, loosen the tourniquet immediately after
blood flow is established and release entirely when the
syringe is filled.
- Completely fill the syringe and then withdraw the
needle with a slow but firm motion. When the needle is
out of the arm, press gauze firmly on the puncture. Heavy
pressure as the needle is being withdrawn -should be
avoided because it may cause the sharp point of the
needle to cut the vein.
- Have the examinee raise his arm (not bend it) and
continue to hold the gauze in place for three (3)
minutes. This will help prevent hematomas.
- Insert butterfly needle into purple top tube and allow
to fill. To ensure accurate results, a minimum of 1.75 mL
of blood must be drawn into the 3 mL lavender top tube to
provide the proper ratio of anticoagulant to blood.
Invert the lavender top tube several times to ensure
proper mixing.
- Insert needle into red top tube with examinee's ID
number written on it and allow rest of blood to flow in.
Allow red top tube to "cool" at room temperature 10
minutes. Spin red top tube 15 minutes in centrifuge.
Pipette sera to plastic screw top tube. Discard red top
tube, butterfly needle, and syringe in safety container.
- Label lavender top tubes with pre-printed labels
provided, and use a ballpoint pen to add the date
collected and your initials to the label. The lavender
top tube should be affixed with the label showing the
participant's ID number (e.g. 88002B1) identified "Lead
in Soil Blood Lead". The red top tube should be labeled
with the participants ID number. The plastic screw top
tube should have participant's ID number written in
indelible ink.
- Place a bandaid on the participant's arm.
B. PROCESSING=
- Place the lavender top tubes and the plastic screw top
tube upright in a rack in the cooler or refrigerator
within 30 minutes after being drawn. Log in the specimens
and keep cool until transported to lab. For the lavender
top tube, note on the log sheet if a full draw is not
obtained ( minimum blood volume is 1.75 mL) or if the
blood was not refrigerated within 30 minutes.
A - 2
-------
ANALYTICAL METHOD FOR BLOOD LEAD
GRAPHITE FURNACE AAS
A. GLASSWARE, SUPPLIES. AND EQUIPMENT:
1. Pipettes, 5 mL Mohr, graduated in 1/10 mL
2. Pipettes, Class A, Volumetric: 5.0 mL, 10.0 mL, 15.0 mL,
and 25.0 mL
3. Flasks.90Clm£s A, Volumetric: 100 mL, 200 mL, 500 mL,
4. Eppendorf Pipettes: 20 jiL and 500 nL
5. Eppendorf Pipette Tips: 20 jiL clear tips and 500 jiL blue
tips
6. Falcon Tubes (Falcon 2063): 12 x 75 mm, polypropylene,
round - bottom tube with cap
7. Disposable Sampling Cups: 1.5 mL polystyrene or
polyethylene
8. Analytical Balance
9. Pyrolytically Coated Graphite Tube (Perkin-Elmer, part
number B010-9322)
10. Pyrolytic Graphite L'vov Platform (Perkin-Elmer, part
number B010-9324)
11. Lead Hollow Cathode Lamp (Perkin-Elmer, part number
0303-6039)
12. Micromedic Automatic Pipette with 1 mL and 50 nL Sampling
Pumps and 1 mL and 200 jiL Dispensing Pumps
13. Rotator-Labindustries Labquake Shaker
14. Vbzrtex-Genie Mixer
15. Sonifier Cell Disrupter
16. Hydro-Ultrapure Water System
17. Perkin-Elmer Zeeman/5000 System (Model 5000 Atomic
Absorption Spectrometer, Zeeman Graphite Furnace, HGA-400
Furnace Programmer, AS-40 Autosampler). System is
interfaced to a PC with printer (IMS 286 Computer, CM
4531 EGA Monitor, Epson LX 810 Printer).
A - 3
-------
18. Argon Gas, 99.996% Purity (as per Perkih-Elmer
recommendation)
B. MATRIX MODIFIER
(0.2% NH4H2P04 and 0.05% Mg (N03) 2. 4H20 or 0.05% Mg(NO3)2. 6H2O in
1% HNO3 ULTREX:
1. Reagents Required;
a. Ammonium Phosphate, Monobasic
(NH4H2P04) ULTREX Ultrapure
Reagent, J.T.Baker Chemical Co.
b. Magnesium Nitrate Mg(NO3)2.4H20, or Mg(NO3)2.6H20,
Johnson Mathey Chemicals Limited (distributed by
Alfa Products) i
c. Nitric Acid (HN03), ULTREX, J.T. Baker Chemical Co.
2. Preparation of Matrix Modifier:
i-
a. 0.05% .Mg(NO3)2: weigh 0.3716 grams if using
Mg(N03)2.4H2O
or 0.4331 grams if using Mg(NO3)2.6H2O on an
analytical balance. Transfer to 500 raL volumetric
flask.
b. 0.2% NH4H2PO4: weigh 1.000 gram NH4H2PO4 on an
analytical balance. Transfer to same 500 mL
volumetric flask.
c. Add approx. 250 mL 1% HNO3 ULTREX to flask; stopper;
swirl contents to dissolve salts.
d.- After salts are in solution, qs to 500 mL with 1%
HNO3 ULTREX.
e,. Stopper and mix thoroughly.
C. PREPARATION OF LEAD STANDARDS:
CAUTION* All glassware must be "lead-free." Clean all
pipettes by soaking at least 24 hours in 30-50% HNO3, then
rinsing thoroughly with deionized water; oven dry. Fill
volumetric flasks with 30-50% HNO3. Let stand at least 24
hours. Rinse thoroughly with deionized water; air dry.
1. Aqueous Stock Standard (10 ppm Lead): prepare from 1000
ppm Lead Standard (Alfa Products, Ventron Division, or
Varian, Sunnyvale CA). Into a Class A volumetric flask,
pipette 5 mL of 1000 ppm Lead Standard, using a 5 mL Mohr
A - 4
-------
TD pipette or a 5 mL Class A volumetric pipette. QS to
500 mL with 1% HN03 ULTREX; stopper; mix.
2. Aqueous Working Standards (prepare weekly):
Working
jig/L
250
500
1000
1250
1500
3.
Standard
ppm or jig/mL
0.25
0.50
1.0
1.25
1.5
Stock
(mL of 10 ppm)
5
5
10
25
15
Total Volume (mL)
(qs with deionized water)
200
100
100
200
100
Label, date, and initial all solutions.
Blood Lead Standards:
a. Preparation of Base Blood (1/10 Dilution of a "Low
Lead" Blood): Using a Micromedic Automatic Pipette,
dispense 300 jiL of a child's non-hemolyzed, EDTA-
preserved blood, with a lead concentration of 4-6
jig Pb/dL) into a labeled Falcon tube. Add 2700 jiL
Triton X-100 in three 900 jiL portions. Mix by
swirling gently; cap tube; vortex (lowsetting) .
Prepare a week's supply of base bloods. Dilute
each patient's blood separately (DO NOT POOL
BLOOD); refrigerate. 3 mL base blood is needed for
a six point calibration curve. All standards in
one run must be prepared from the SAME base blood.
b. Preparation of Blood Lead Standards (method of
Standard Additions). Prepare fresh daily:
1) With an Eppendorf pipette transfer 20 jiL of
each aqueous standard (250, 500, 1000, 1250,
and 1500 U9/L) into one of five Falcon tubes,
labeled Stds 1,2,3,4,5.
2) With an Eppendorf pipette add 500 jjiL of base
blood (from Step 3a), to each tube. Mix by
swirling; cap tube. The remaining base blood
in the tube is the Zero Calibrator unspiked
blood).
The actual concentration of Pb (ng/L) in each Blood
Standard is: Std #1 = 9.615; Std #2 = 19.231;
Std #3 = 38.462; Std f4 = 48.077; Std #5.-= 57.692.
A - 5
-------
D. QUALITY CONTROL
1. Source of Controls:
a. CDC and CAP Proficiency Test Bloods. These
specimens are used as controls, after the test
values have been established by CDC or CAP.
b. Ciba-Corning. BLD Tox I, II (Bi-Level Whole Blood
Assayed Toxicology Controls), Product Code 9660.
The DHMH Laboratory establishes a mean and if ollows
CDC guidelines to determine the acceptable range (±
4 jig/dL, or 10%, whichever is greater) for these
controls.
2. Three Controls (low 10. ng/dL, middle 30 (ig/dL and high
40-60 |ig/dL) are analyzed in a run.
3. Maintaining QC Records: All QC results must be plotted
daily by the analyst on charts specifically reserved for
the instrument used in the analyses.
E. OPERATION OF ZEEMAN/5000 SYSTEM:
1. TURN ON ARGON TANK (tank pressure 350 kPa, 41 psi).
2. TURN ON H2O (faucet at sink - water flow not too fast;
optimum Flow Rate 2.5 L/min).
3. TURN ON AUTOSAMPLER (AS-40). Power Sequence takes 60
seconds; it then goes into STANDBY (sampling arm is above
overflow vessel) . ALWAYS TURN AS-40 AUTOSAMPLER ON BEFORE
THE HGA-400 PROGRAMMER AND
AA 5000.
4. TURN FURNACE GAS CONTROL ON (there are three positions:
ON, OFF, OPEN) . TURN POWER ON. E-45 error message
appears on digital display and a warning alarm sounds, if
the power is turned on
before the gas. Use CE (CLEAR ENTRY) to turn off alarm.
Make sure GAS CONTROL FLOW is ON.
5. TURN AA 5000 POWER ON.
TURN RUN SWITCH ON.
Turn optical interface (located top, right in lamp
compartment) to ZAA position.
6. TURN COMPUTER, MONITOR AND PRINTER ON (see Analytical
Instrument Software Operator's Manual).
A - 6
-------
SET AA 500.0 PARAMETERS:
a. Install lamp with lamp window facing the Zeeman
Furnace Module (active lamp position is in the
front).
b. Note the ZAA lamp position number and plug the lamp
into the receptacle in the turret hub,
corresponding to that number.
c. Turn Accessory Switch ON (on Zeeman Furnace
Module).
d. Select ZAA above Accessory Switch).
e. Press Lamp # (1-6), corresponding to position of
lamp location.
f. Select the proper current (Pb Lamp 10 ma). Never
use higher than the specified current. Press 10
and LAMP MA.
g. Select proper SLIT-LOW for the element (Pb 0.7 ma).
h. Press correct wavelength number (Pb 283.3) on
keyboard. Press \ PEAK.
ADJUST LAMP after at least 5 minutes warm-up time.
a. Press SET UP. The main display will show a value
of approximately 50. If upper limit of SET UP (99)
is displayed, press GAIN to bring display down to
50.
b. Adjust lamp with two adjustment screws (located
front-bottom of lead lamp on lamp mount) and slide
lamp forward or backward in the mount, until a
maximum reading is achieved. If an overrange (99)
is obtained, press GAIN and continue adjusting
lamp.
c. Press SET UP again to cancel SET UP Mode.
d. To check PM (photomultiplier) VOLTAGE press CHECK
and GAIN; record PM Voltage and Lamp Energy on Lab
Chart.
SET ADDITIONAL AA PARAMETERS AFTER LAMP ADJUSTMENT
(FOLLOW THIS ORDER):
a. PRESS 5.0 t (time must match Atomization Time).
A - 7
-------
b. CHOOSE PEAK AREA on AA 5000.
c. PRESS NUMBER AVG (omit this step if not averaging 2
or more readings).
d. PRESS PRINT.
e. PRESS 1 STO (to store AA Parameters).
10. PROGRAM THE HGA-400 PROGRAMMER:
Furnace Conditions for Pb Determination
1 2 34 5
100 130 650
5 5
10
STEP
TEMP(°C)
RAMP TIME *(s)
HOLD TIME (s)
RECORDER
READ
BASELINE
INT GAS (MINI FLOW) mL/min
EXT.ALT (STOP FLOW)**
10
5
45
20
1
4
48
6 7
1800 2600 20
0 11
5 5 10
REG REC
READ
0
STOP
FLOW
* When RAMP TIME is zero (Max. Power Heating), Temperature
Control requires calibration.
** If STOP FLOW does not work, press 0 MINI FLOW.
Note: Adjustment of Furnace Conditions must be made as
needed. Graphics and physical observations must
also yield favorable results.
11. 'CAliIBRATE OPTICAL TEMPERATURE SENSOR (the pptimal
temperature sensor is only used when operating in the
Maximum Power Heating Mode):
a. First purge line of air by entering 120 TEMP on
Furnace Programmer. Press MANUAL TEMP key and hold
for 5 seconds.
b. Set Range Selector on the Furnace Assembly to the
correct range for the required atomization
temperature for lead (1800°C); Range Selector
A - 8
-------
>1500°C. If unable to achievecalibration of the
optical sensor at this setting, use < 1500°C.
c. Press RECORDER MANUAL (to turn the electromagnet
on) . •
d. Enter the Atomization Temperature (1800°C) for lead
and press and hold MANUAL TEMP key. Hold entire
time while adjusting calibration control .(for 15-20
seconds when atomization temperature is <. 2000°C;
for 5 seconds when atomization temperature is >
2000°C). Wait briefly and then adjust the CAL
control fairly rapidly to obtain balance of the
indicator lights (red/green).
e. Release MANUAL TEMP key.
f. Press REG MAN (to turn electromagnet off).
It is important NOT TO OVERHEAT THE MAGNET.
Repeat Steps llc-llg several times to get correct
balance.
12. PRELIMINARIES FOR AS-40 AUTOSAMPLER:
a. The AS-40 has been in STANDBY (the arm is up in the
air) . Keep the flushing liquid reservoir filled
with 0.05% Triton X-100. Empty Waste Bottle
regularly.
b. While in STANDBY:
NOTE: The sampling arm can only be moved manually when
the Autosampler is in STANDBY Mode.
1) Check that the pipette tip is positioned
correctly, by checking the tip as it enters
the entry hole in the graphite tube (use
flashlight or lamp to look into the entry hole
of the graphite tube).
2) Check the tip in relation to the L'vov
Platform (with a dental mirror, positioned to
the right of the right window assembly, of the
furnace assembly).
3) Adjustments can be made to improve
positioning:
a) The whole sample carriage can be realigned to
center the tip as it enters furnace. Release Table
LOCK Control (on right front of table; turn
counterclockwise).
A - 9
-------
(1) To adjust left-right movement: turn LATERAL
Control Knob on left side of the table.
(2) To adjust forward-backward movement: turn
HORIZONTAL Control Knob on front of table, to
left of the LOCK/RELEASE Control Knob.. After
adjustment, tighten table LOCK/RELEASE Control
Knob.
b) Knob to right of sampling arm (closest to Furnace)
adjusts immersion depth of tip into the solution in
sample cup. Directly in front of this knob is the
knob to adjust penetration depth of tip into
graphite, tube.
13. PROGRAM AS-40 AUTOSAMPLER:
a. Press STANDBY to take it out of STANDBY.
b. If there are less than 35 cups being used, enter
the LAST SAMPLE location number.
c. Press 10 (jiL), SAMPLE VOLUME.
d. Press 5 (flL), ALT VOLUME.
e. Press 1, INST PROG (if 1 STO already programmed on
AA 5000 (Step 9e).
f. Press 1 HGA PROG {HGA 400 Programmer cannot store
program).
14. PROGRAM COMPUTER: see Analytical Instrument Software
Operator Manual, pp. 3-16; also, DHMH Lead Laboratory
Computer Guide for Zeeman 5000 AA, pp. 1-17.
15. PRELIMINARIES TO BLOOD LEAD ANALYSIS:
a. Blank Furnace
1) "Enter 120 TEMP on Furnace Programmer.
2) Press MANUAL TEMP key and hold for 5 seconds
to purge line of air.
3) Press .START on Furnace Programmer. When
operation is complete, the absorbance value
appears on the AA 5000 display.
4) Press AZ (Automatic Zero) on Model 5000 to
make Furnace read 0.
A - 10
-------
5) Repeat Steps 15a(3) and 15a(4) to ensure
Furnace is blanked (0.001 or better).
Check the deionized water to determine quality of
H20, the Triton X-100 with matrix modifier (referred
to as Sample Blank) and the highest Blood Pb
Standard (to check that the absorbance of this
standard will give an acceptable Characteristic
Mass of 12 ± 20%.
A good technique for checking the proper set-up of
the instrument is by running the highest Blood Pb
Standard, the calculation of its Characteristic.
Mass, and the comparison of this value with the
accepted value (12 ± 20%).
1) Place cup containing 1% HNO3, in position #1 of
AS-40 Method Tray.
2) Place deionized H2O in cup #2.
3) Place Triton'X-100 in cup #3.
4) Place highest Blood Pb Standard in cup #4.
5) Press 1, MANUAL, START on the Autosampler.
6) Note Absorbance of each sample on the Digital
Display. Record Absorbance of H20 on Lab
Chart.
7) If Absorbance reading of deionized H20 or
Triton X-100 is not acceptable (must be ±
0.002 or better), re-run using a fresh aliquot
in another cup.
a) Press STANDBY at Step 6 of Furnace
Program (to stop Autosampler), otherwise
the sample tray will rotate to the next
position in the Autosampler.
b) Press STANDBY to take the Autosampler out
of STANDBY.
c) Press the Cup Position # desired.
d) Press MANUAL, START.
e) Note Absorbance on Digital Display.
A - 11
-------
c. Calculate the CHARACTERISTIC'MASS (CM) pg/0.0044 A-s:
CM - Sample Volume X Std Cone X 0.0044 (1% Absorption)
Abs Blood Std - Abs Base Blood
Example:
CM « 10 'uL x 57.692 ua/L'x 0.0044 = 11.81 pg/0.0044 A-s
.215 (.235 - .020)
Enter CM value on Lab Chart. Ideal CM value for Pb is
12 ± 20%.
G. SPECIMEN PREPARATION:
1. Write up worksheets with patients' names, allowing two
cups for each patient. Computer generated worksheets may
be used. • Specimen position # on worksheet corresponds to
cup position # on Autosampler. Cups # 1-7 are for the
Blank (0.5% Triton X-100) and six Blood Lead Standards,
including Base Blood Blank. Three Blood Lead Standards
(run as check standards) and 3 Controls are included in
each run: one of each at the beginning, in the middle,
and at the end.
2. Mix specimen .by rotating. If specimen is clotted,
sonification is necessary.
3. Micromedic Automatic Pipette Parameters:
a. Set 50 jiL Sampling Pump at 0600.
b. Set 200 }iL Dispensing Pump (for 0.5% Triton X-100)
at 0450.
4. With the Micromedic Pipette, aspirate 30 (iL specimen into
the delivery tip. Dispense the sample, with 90 uL 0.5%
Triton X-100 diluent, into the bottom of a properly
labeled Falcon tube.
5. Aspirate air into the delivery tip and dispense 90 (iL
of 0.5% Triton X-100 into the same tube, as in S1iep 4.
6. Repeat "Step 5.
7. Cap tube; mix contents of tube by swirling, or, Vortex
(low setting).
8. Before pipetting next specimen, wipe tip with tissue and
rinse tip, by aspirating air and dispensing 0.5% Triton
X-100 into the waste beaker. Wipe off tip with tissue
and proceed to next specimen, repeating Steps G(4-7).
A - 12
-------
Pipette each patient's specimen in duplicate.
H. SPECIMEN ANALYSIS:
1. Zeeman/5000 System:
a. Screen on Computer Monitor displays "Data
Collection - Tray Position Pointer". Note: Refer
to Step E14 for Programming Computer.
THE CURSOR MUST BE ON "AUTO ZERO". .
b. Place cup containing Matrix Modifier into position
ZERO (to the left of the sampling tray).
c. Place 0.5% Triton X-100 in sampling cup in
positions AZ (Auto Zero) and 1.
d. Place standards in sampling cups, positions 2-7.
Replace cover.
e. Press STANDBY to take Autosampler out of STANDBY.
f. Press RESET and START.
g. While the Triton X-100, in the AZ position, is
being delivered into the graphite tube, check tip
position for proper delivery, using a dental mirror
placed to the right of the window assembly. If it
is delivering properly, press Fl on the computer
keyboard. If it is not, press STANDBY on the
Autosampler to stop it. Manually adjust the tip,
according to "Preliminaries for AS-40 Autosampler"
(Step E12). Wait until all steps in the Furnace
Program have been completed. Press STANDBY.
Repeat Steps G5-6.
h. While standards are running, start to load
specimens and controls onto sampling tray.
i. After all standards have been run, check slope,
intercept and correlation coefficient, to determine
acceptability of curve. R (correlation
coefficient) must be 0.999.
j. Continue to load samples onto sampling tray. The
last specimen should always be a freshly pipetted
control. Cover sampling tray to minimize
evaporation.
k. When the analysis of the last specimen is complete,
a buzzer in the AS-40 Autosampler Programmer sounds
A - 13
-------
briefly. This completes the first tray. At this
point, additional specimens may be run. One
standard curve is used for approximately 20-25
specimens, in duplicate. ;
1. Results are calculated by the computer.
m. Repeat the analysis, if the following occurs:
1) The Absorbance Reading of a specimen is higher
than the Absorbance Reading of the highest
standard: With an Eppendorf pipette, make a
1:2 dilution of patient's blood with 0.5%
Triton X-100 prior, to repeat analysis.
2) The initial Pb result is >35 jig/dL.
3) The initial Pb result is <2 jig/dL.
I. INSTRUMENT SHUT DOWN:
1. Zeeman/5000 AA:
a. Temporary Shut Down:
1) Gas Switch OFF on HGA-400.
2) E45 displays on HGA-400 and alarm sounds.
Press CLEAR ENTRY to turn alarm off.
b. Complete Shut Down:
1) Computer, Monitor, Printer: TURN OFF, as per
computer instructions in DHMH Lead Laboratory
Computer Guide, pp.19-20.
2) Turn Argon gas tank OFF. Dials on tank
regulator must be at ZERO.
3) Turn H2O OFF (faucet at sink).
4) HGA-400 Programmer:
c) Turn Gas Switch OFF.
d) E45 displays and buzzer alarm sounds; press CLEAR
ENTRY to turn alarm off.
e) Press OFF on Furnace Programmer..
5) Zeeraan 5000: Turn ACCESSORY Switch OFF.
* ' i
A - 14
-------
6) AA 5000:
a) Press POWER OFF.
b) Press STANDBY.
c) Close lamp compartment cover, if open.
7) AS-40 Autosamplers Press POWER OFF. This module must
ALWAYS BE TURNED OFF LAST.
J. REFERENCES:
Pruszkowska, E., Carnrick, G.R., Slavin, W.: "Blood Lead
Determination with the Platform Furnace Technique." Atomic
Spectroscopy 1983; Vol. 4, No. 1: 59-61.
Pascal, D.: "Calibration Procedure for Graphite Furnace Blood
Lead". Centers for Disease Control, Atlanta, GA 30333: Feb,
1986. Perkin-Elmer Model 5000 Atomic Absorption Spectrometer;
Zeeman/5000 System; AS-40 Autosampler Sequencer Version;
HGA-400 Graphite Furnace Instruction Manuals. Norwalk,
Connecticut, 1984.
Perkin-Elmer Zeeman/3030 Atomic Absorption Spectrometer Operator's
Manual (Publications B385). Norwalk, Connecticut.
Analytical Instrument Software, Inc.: "Auto-AA on Line QC
Software" Operator's Manual, 1989.
Micromedic Systems Automatic Pipette. Operating Manual.
Horsham, Pennsylvania.
DHMH Lead Laboratory Computer Guide for Zeeman/5000 AA, Sept. 1991.
DHMH Lead Laboratory Computer Guide for Zeeman ' 030 AA, Sept.
1991.
A - 15
-------
BALTIMORE SOIL LEAD ABATEMENT DEMONSTRATION PROJECT
HAND WIPE PROTOCOL FOR LEAD DETECTION
Purpose The purpose of the testing is to detect lead levels
on children's hands. Sampling occurs during the
clinic visit. During the summer sampling period,
samples are obtained from the hand area. During
the winter sampling, samples are obtained only from
the hand area. Elbow wipes were also obtained
during the initial summer sampling to eliminate
false negatives due to hand washing.
Materials
Procedure
- wipes
- disposable gloves
- plastic bags
- labels
At the opening of each box of gloves or Wash-a-bye
baby Wipes, a blank control is obtained. The
investigator wears disposable gloves. The
investigators wipe all surfaces of their own gloved
hands using three-wipes per hand. These six wipes
are placed in a plastic bag with the sample surface
folded inward and labeled as blank with the test
date.
For each child to be sampled, the investigator
identifies the child and obtains the ID number that
the child has been assigned and dons a new pair of
gloves.
For hand levels, the investigator wipes the child's
hand on all surfaces using three wipes per hand.
Each wipe is applied to all hand surfaces, up to
and including the wrist. A total of six wipes are
used per child per sampling. The six wipes are
then be placed in a single plastic bag with the
subject's ID number as follows:
88 C _
Sampling of the elbow area also requires three
wipes per elbow. Each wipe is be applied to the
entire posterior elbow and two inches up and down
the arm. All six wipes are placed in a single
plastic bag and labeled as follows using the
subject's ID number:
88 E _
At the end of the sampling period, all bags of
samples are collected and transported to the state
laboratory administration for analysis.
A - 16
-------
LEAD ANALYSIS OF HANDWIPES
Nitric Acid/Perchloric Acid Digestion
(Used in .Cleaning Children's Hands and Elbows)
A. GLASSWARE, SUPPLIES AND EQUIPMENT:
1. Beakers: Griffin, Pyrex brand/ graduated, 800 mL
2. Flasks, volumetric, Class A: 100 mL, 200 mL and 1 Liter
3. Test tubes: polypropylene, round bottom with caps, 17 x
100 mm (Falcon #2059)
4. Watch glasses, Pyrex brand, 75 mm diameter
5. Pipettes, volumetric, Class A: 2.0 mL, 5.0 mL, 10.0 mL,
15.0 mL and 20.0 mL
6. Pipette, graduated: 10 mL
7. Cylinders, Pyrex brand, graduated: 10 mL with stopper;
100 mL and 250 mL without stopper
8. Baby Wipes: commercial, alcohol free baby wipes (ex:
Wash a-Bye Baby)
9. Hotplates
10. Atomic Absorption Flame Spectrophotometer, Varian Model
AA5 with Background corrector and IM 6 Indicating Module
11. Atomic Absorption Flame Spectrometer, Perkin Elmer 3030B
12. Deionized water: Hydro Service Ultrapure Water System
13. Perchloric Acid Fume Hood
B. REAGENTS 8
1. Nitric Acid, 10.0% (prepared from 'Baker Analyzed'
Reagent)
2. Nitric Acid/Perchloric Acid (Ratio 5:4)
Nitric Acid: 'Baker Analyzed' Reagent
Perchloric Acid, 60%: G. Frederick Smith
A - 17
-------
c.
STANDARDS :
1. "Stock" Standard (100 ppm Lead):
20 ml/ of Varian Techtron Lead Standard (1000 ppm) diluted
to 200 mL with 10.0% HN03
2. "Working Standards":
Working Standard
ppm
2.0
5.0
. 10.0
15.0
20.0
mL of Stock
(100 ppm)
2.0
5.0
10.0
15.0
20.0
Total Volume i(mL)
qs with 10.0% HN03 |
100
100
100
100
100
D.
A 0.6 "working" standard is prepared with 6 mL of 10.0
ppm standard, diluted to 100 m: with 10.0% HNO3.
3. Label, date and initial all solutions.
SAMPLES;
Samples are received from the "Lead in Soil Project" for AAS
analyses. Twelve wipes per child are received; 3 wipes per
hand in one baggie and 3 wipes per elbow in a second baggie.
E.
CONTROLS and BLANKS:
Four Controls (2.0, 5.0, 10.0 and 20.0 ppm working standards
added to wipes) and two Total Reagent Blanks are included in
every group of 40 samples.
F. PROCEDURE!
1. Preparation of Samples for Acid Digestion: ,
a. Acid washing of glassware: All glassware must be
acid washed prior to use. Soak for 24 hours in 30%
v/v nitric acid/deionized water; rinse with
deionized water. Glassware must be oven-dried and
cooled to room temperature prior to use/.
b. Label each beaker with the sample number.
c. Transfer, with a minimum of contact, all handwipes
from the child's hands to a labeled, acid washed
800 mL beaker. To a second 800 mL beaker transfer
all wipes from the child's elbows. Partially cover
each beaker with a watch glass.
A - 18
-------
d. Air dry overnight prior to addition of acid.
2. Hot Nitric Acid/Perchloric Acid Digestion:
a. To each beaker, add 100 mL of the acid mixture (5:4
ratio nitric acid/perchloric acid).
b. Cover the beakers with watch glasses and place the
beakers .on hot plate in a perchloric acid fumehood.
Adjust the setting to the hotplate to achieve low
boil or simmer. The handwipe material will
dissolve in the acid at this temperature. Swirl
the beakers frequently to prevent material from
sticking to the sides of the beakers.
c. After the material is dissolved, continue heating
until the sample is evaporated just .to dryness. DO
NOT BAKE.
d. Add 5 mL of 10.0% nitric acid to each beaker. Heat
the sample at a low boil or simmer on a hot plate
.to redissolve lead; swirl beaker.
e. Transfer the solution into a labeled, acid washed
10 mL graduated cylinder with stopper or to a new
labeled graduated Falcon tube. Rinse beaker and
watch glass with a very small amount (±1.5 mL) of
10.0% nitric acid and transfer to the same
graduated cylinder or Falcon tube. .
f. Repeat rinse procedure three times.
g. Allow solution to come to room temperature.
h. Dilute to 10 mL volume with 10.0% nitric acid.
Stopper cylinder or cap tube and mix well. Allow
contents to settle to avoid necessity of filtering.
The samples are ready for AA analyses.
YARIAN AA5 FLAME PARAMETERSs
AA settings used:
Resonance line 2833 A°
Slit Width 100+ microns
Lead Lamp: Source Current 4-5 mamps
Fuel acetylene 3.0
Oxidant air 7.0 -
Support Pressure 21-22
Recorder Varian Model 9176, Span 2 mv/FS, Speed 2mm/min
A - 19
-------
PERKIN ELMER 3030B FLAME PARAMETERS:
ELEMENT: Pb WAVELENGTH (NM): 283.3 SLIT (NM): 0.7
FLAME: AIR-ACETYLENE, OXIDIZING (LEAN, BLUE)
CHAR CONG: 0.45 SENS CHECK (MG/L) : 20.0 LINEAR TO (MG/L) :
20.0
1. TECHNIQUE: AA 2. LAMP CURRENT (MA): 10
3. SIGNAL PROCESSING: HOLD 4. CALIBRATION: LINEAR
5. NOMINAL WEIGHT: 1.0 6. STATISTICS: SINGLE READING
7. TIME (SECOND): 5.0 8. READ DELAY (SECONDS): 0.0
9. SCREEN FORMAT: BASIC DATA 10. PRINTER: OFF
11. RECORDER SIGNAL: 0.2 CONT ABS .12. RECORDER EXP: 1000
13. SI: 20.0 14. S2: 15.0 15. S3: 10.0
16. S4: 5.0 17. S5: 2.0 18. S6: 0.6
19. S7: 20. S8: 21. RSLP:
Computer IMS 286 used in conjunction with the Perkin Elmer 3030B.
Results are obtained in ug Pb.
REFERENCES
University of Cincinnati Medical Center, Lead Program Project, "Institute of
Environmental Health. "7.4.2. Digestion of Handwipes Samples, " p.
100 (received August, 1988).
Perkin Elmer Model 303B Atomic Absorption Spectrometer Instruction Manual.
Norwalk, Connecticut, J.987.
Varian Techtron Model AA-5 Atomic Absorption Spectrophotometer instruction
Manual. Melbourne, Victoria, Australia, January, 1971.
Analytical Instrument Software, Inc.: "Auto-AA on Line QC Software"
Operators Manual, 1989.
A - 20
-------
LEAD ANALYSIS OF HANDWIPES
Nitric Acid (1M) Extraction
(Used in Cleaning Children's Hands)
A. GLASSWARE, SUPPLIES AND EQUIPMENT:
1. Beakers: Griffin, Pyrex brand, graduated, 50 mL, 250 mL,
and 800 mL
2. Flasks, volumetric. Class A: 100 mL, 200 mL, and 2 liter.
3. Watch glasses, Pyrex brand, 75 mm diameter
4. Test tubes: polypropylene, round bottom with caps, 17 x
100 mm (Falcon #2059)
5. Filter paper: Whatman #40, 9.0 cm diameter
6. Pipettes, volumetric, Class A: 2.0 mL, 5.0 mL, 10.0 mL,
15.0 mL, and 20.0 mL
7. Pipette, graduated: 10 mL
8. Cylinders, Pyrex brand, graduated: 50 mL, 100 mL and 250
. mL
9. Hotplates
10. Atomic Absorption Flame Spectrometer, Perkin Elmer 3030B
11. Deionized water: Hydro Service Ultrapure Water System
12. Fume Hood
B. REAGENT:
1. Nitric Acid, 1M (prepared from Baker Analyzed Reagent)
C. STANDARDS:
1. "Stock" Standard (100 ppm Lead):
20 mL of Varian Techtron Lead Standard (1000 ppm) diluted
to 200 mL with 1.0 M HNO3
A - 21
-------
2.
"Working Standards":
Working Standard
ppm
2.0
5.0
10.0
15.0
20.0
mL of Stock
(100 ppm)
2.0
5.0
10.0
15.0
20.0
Total Volume (mL)
qs with 1.0 M HN03
100
100
100
100
100
A 0.6 "working" standard is prepared with 6 mL of 1
standard, diluted to 100 mL with 1.0 M HN03.
3. Label, date and initial all solutions'.
D. SAMPLES
Samples are received from the "Lead in Soil Project" for AAs
analyses. Six wipes per child are received; three wipes per
hand in one baggie.
E. CONTROLS AND BLANKS:
Four Controls (2.0, 5.0, 10.0 and 20.0 ppm working standards
added to wipes and two Total Reagent Blanks are included in
every group of 42 samples.
F. PROCEDURES:
1. Preparation of Samples for Acid Digestion:
a. Acid washing of glassware: All glassware must be
acid washed prior to use. soak for 24 hours in 30%
v/v nitric acid/deionized water; rinse with deionized water.
Glassware must be oven-dried and cooled to room temperature
prior to -use.
b. Label each beaker with the sample number.
c. Transfer, with a minimum of contact, all handwipes
from the child's hands to a labeled, acid washed
800 mL beaker. Partially cover the beaker with a
watch glass.
d. Air dry overnight prior to addition of acid.
A - 22
-------
Nitric Acid Digestion:
a. Add 100 mL 1 M HN03 to each beaker.
b. Swirl beaker:for 10 seconds.
c. Cover beaker with a watch glass and allow acid to
extract at room temperature for 2 hours.
d. Decant the acid solution from the beaker containing
handwipes into a labeled, acid washed 250.mL beaker.
e. Add 50 mL 1 M HN03 to the handwipes in the 800 mL beaker.
f. Swirl the beaker for 10 seconds.
g. Decant the acid solution into the same 250 mL beaker to
composite the acid rinses.
h. Repeat Steps e, f and g to achieve a total acid solution of
approximately 200 mL.
i. Cover each 250 mL beaker with a watch glass and place beakers
on a hotplate. Adjust setting so that contents of beakers
simmer (low boil) for two hours.
j. Evaporate the samples to dryness. DO NOT BAKE.
k. Add approximately 3-5 mL 1 M HN03, rinsing the watch
glass and the sides of the beaker.
1. Heat beakers on a hotplate to redissolve the lead.
Adjust the setting to achieve a low boil or simmer.
m. Filter to remove undissolved material into a 50 mL
labeled, acid washed beaker. Make .several rinsings
of the 250 mL beaker and the filter paper with ±1.5
mL 1 M HNO3.
n. Place the 50 mL beakers on a hotplate. Heat at a low boil or
simmer to reduce volume to approximately 5.0 mL.
o. Transfer the solution into a new labeled, graduated
Falcon tube.
p. Rinse the beaker with a very small amount (±1.5 mL)
of 1 M HNO3 and transfer to the same Falcon tube.
q. Repeat rinse procedure three times.
r. Allow solution to come to room temperature.
A - 23
-------
s. Dilute to 10 mL volume with 1.0 M HN03. Cap tube and mix
well. Allow contents to settle to avoid necessity of
filtering. The samples are ready for AA analyses.
PERKIN ELMER 30SOB FLAME PARAMETER:
ELEMENTS Pb WAVELENGTH (NM): 283.3 SLIT (NM): 0.7
FLAME: AIR-ACETYLENE, OXIDIZING (LEAN, BLUE)
CHAR CONG: 0.45 SENS CHECK (MG/L) : 20.0 LINEAR TO .(MG/L) : 20.0
1. TECHNIQUE: AA 2. LAMP CURRENT (MA): 10
3. SIGNAL PROCESSING: HOLD 4. CALIBRATION: LINEAR
5. NOMINAL WEIGHT: 1.0 6. STATISTICS: SINGLE READING
7. TIME (SECOND): 5.0 8.' READ DELAY (SECONDS): 0.0
9. SCREEN'FORMAT: BASIC DATA 10. PRINTER: OFF
11. RECORDER SIGNAL: 0.2 CONT ABS 12. RECORDER EXP: 1000
13. Sis 20.0 14. S2: 15.0 15. S3: 10.0
16. S4: 5.0 17. S5: 2.0 ' 18. S6: 0.6
19. S7: 20. S8: 21. RSLP:
Computer IMS 286 used in conjunction with the Perkin Elmer 3030B.
Results obtained are in u'g Pb. ,
REFERENCESs
University of Cincinnati Medical Center, Lead Program Project, Institute of
Environmental Health "Acid Digestion of Handwipe Samples, Method B: 1
M Nitric Acid Extraction" pp 41 & 42 (received May, 1990) .
Perkin Elmer Model 3030B Atomic Absorption Spectrometer Instruction Manual.
Norwalk, Connecticut, 1987.
Analytical Instrument Software, Inc.: "Auto-AA on Line QC Software"
Operators Manual, 1989.
A - 24
-------
BALTIMORE SOIL LEAD ABATEMENT DEMONSTRATION PROJECT
SOIL SAMPLING PROTOCOL
I. Site Description
For each location, a detailed drawing should be made that shows the
boundary of the lot, the position of the main building and any other
buildings such as storage sheds or garages, the position of the
sidewalks, driveways, and other paved areas, the position of the play
areas if obvious, and the position of the areas with exposed soil
(grassy or bare), also, showing roof rain spouts and general drainage
patterns.
In addition to the diagram, briefly describe the location, including the
following information:
Type of building construction
Condition of main building
Condition of property (debris, standing water, vegetation
cover)
Nature of adjacent property
Presence and type of fence
Animals on property
Apparent use of yard (toys, sandbox, children present)
Underground utilities
II. Soil Area Description
For each soil area (i.e. front patch, front yard, back yard, side yards)
identified on the general diagram, draw a full page diagram showing the
approximate dimensions and position relative .to the building foundation.
Indicate vegetation "and bare soil areas, as well as obvious traffic
patterns. Identify the category of land use, such as roadside, property
boundary, adjacent to foundation, play area. Mark the sample location
on the diagram.
III. Sampling Schemes
Measure the soil area to determine the sampling scheme. Select the
sample scheme for each soil area which adequately characterize the
potential exposure of children to lead in the dust from this soil.
Identify the suspected areas of high lead concentrations and the assumed
general distribution pattern of lead concentrations at the soil surface.
Small Area Pattern. Measure and mark off an area 20 inches from the
base of the foundation into the soil area. Repeat measuring and marking
at the boundaries. The area inside the marked pattern indicates the
sampling collection area. If the sampling collection area is less than
two meters in each dimension, a single composite sample may be taken if
it appears that such a sample would adequately represent the soil area.
(Collect two sample bags, mark one bag top and the other bag bottom.)
A - 25
-------
Large Area Pattern. Measure and mark off an area 20 inches from the
base of the foundation into the soil area. Repeat measuring and marking
at the boundaries. The area inside the marked pattern indicates the
sampling collection area. Collect one composite sample at the
foundation and one composite sample at the boundary of the yard if the
area is less than 10 feet wide. (Collect fqur sample bags, two bags
marked top and two bags marked bottom.) Collect an additional composite
sample at an imaginary sample line between the foundation and boundary
sample areas if the yard is larger than 16 feet wide. (Collect six
sample bags, three bags marked top and three bags marked bottom.)
Very Large Area Pattern. Measure and mark off an area 20 inches from
the base of the foundation into the soil area. Repeat measuring and
marking at the boundaries. If a yard is wider than 16 feet and more
than 20 feet long then divide the yard into a vertical half and a
horizontal half. Collect one composite sample at the from each section
of the yard. (Collect twelve sample bags, six bags marked top and six
bags marked bottom.)
Sample Collection
Clean and decontaminate the corer after each sample collection. Remove
vegetation and debris from the corer at the point of insertion into the
soil, but do not remove any soil or decayed litter. Drive the corer in
to the ground to a depth of 15 cm (6 in.). If this depth cannot be
reached, the corer should be extracted and cleaned, and another attempt
made nearby. If repeated attempts do not permit a 15 cm core, take the
sample as deep as possible, and record the maximum penetration depth on
the sample record sheet.
Collect ten randomly selected core samples from within the sampling
area. The cores make a composite sample identified as a single sample.
Record composite information on the sample sheet. 1
Combine the top two inch segment of each core into one composite sample
and combine the bottom two inch segment of each core into second
composite sample. Assemble composite soil core segments in clean
previously unused plastic bags suitable for prevention of contamination
and loss of the sample. Remove debris and leafy vegetation from the top
sample material. Do not remove soil or decomposed litter from the
sample material. This is the most critical part of the soil sample and
is likely to be the highest in lead concentration.
Record the sample identification number on the bag and the sample record
sheet. Store the composite soil sample at ambient temperature until
submitted to the laboratory for analysis.
Clean the corer after collecting each sample composite by reinsertion of
the corer into the soil of the next sampling area. Draw field blanks for
each soil area by inserting the core borer into randomly selected
locations within the sample area. These blanks are drawn prior to
sample collection and at the conclusion of sampling.
A - 26
-------
V. Sample Handling and Storage
Seal the sample bags to prevent loss or contamination of the sample and
storage samples in a cool, dry location.
Record-keeping and Sample Custody
Initiate soil sample records for each location which consists of a
location diagram and description, a plot diagram for each distinct soil
plot, and sample record sheet for each sample in a plot.
Sequentially number samples bags. Record sample numbers on location
diagram, soil area description, and sample record sheet.
Deliver the sample to the laboratory and release the sample to the
laboratory personnel for analysis.
A - 27
-------
BALTIMORE SOIL LEAD ABATEMENT DEMONSTRATION PROJECT
SOIL ANALYSIS (XSF) PROTOCOL
A. SAMPLE PREPARATION
1. Identify sample information to be logged in on the Lead in Soil
Processing Sheets. Record the contract number, sample information,
date, time and total sample number on the processing sheet.
Example:
Date received: 03/30/89
Time: 12:30 p.m.
Total Number of Samples Received from: 41 from Ms. Merrill Brophy
Sample Identification Number: #590312565
Site address: 2092 W. Preston Street
Area: Front yard
2. Record the soil sample information on the XRF Run Sheets. Assign
sequential analysis identification number to the sample.
Example:
The last Fine Soil Fraction sample number was 0436, then the next
sample would be a Total Soil Fraction sample numbered 0437.
3. Specimen containers and XRF sample cups are to be prepared before
soil samples can be processed.
j-
a. Label specimen containers. Include the date, the analysis
number, the sample's identification number, and the particular
soil fraction - Fine or Total. ;
b. Label XRF sample cups. Include only the analysis (cup) number.
4. Air dry samples overnight at room temperature. Use disposable
weigh boat or Kraft paper to air dry sample. Wear gloves during
this process.
a. Label weigh boat. Include the sample's identification number,
.and the sample's analysis number (cup number).
b. Place weigh paper (glassine) on disposable weigh boat.
c. Transfer sample onto weigh boat to air dry.
d. Return samples to corresponding bags after air drying.
5. Sieving process must be done under the hood. Gloves and dust mist
respirators must be worn throughout this process.
A - 28
-------
a. Sift pulverized sample in a 2 mm 9.0 mesh sieve by using the
back of a gloved hand to crush larger particles.
b. Place the sample that passed through the sieve into a specimen
container labeled "Total Soil Fraction".
6. Run samples through an Open Pan Riffle Sampler to obtain a
homogenous sample.
7. Place the homogenized Total Soil Fraction sample into the open pan
and riffle once. This will divide the sample into two parts.
8. Then take one part of the sample and put into the open' pan and
riffle to yield a quarter sample. The remaining three-fourths of
the sample .should be placed into a specimen container labeled Total
Soil Fraction.
9. Pass the quartered sample in a 250 urn 60.0 mesh sieve. This
represents the Fine Soil Fraction. Discard particles that cannot
pass through the 250 urn sieve.
10. If the quartered sample does not seem to be at least two grams,
then take the Total Soil Fraction from its specimen container and
repeat steps 5 --8. After enough "Fine Soil Fraction has been
collected, remember to take the soil that did not pass through the
250 um and replace it back into the specimen container labeled
Total Soil Fraction.
11. Clean sieves by tapping on a hard surface to remove residual
particles. This must be done between sample processing.
12. After steps 4-10 are completed, the Total Soil and Fine Soil
Fraction of a sample should be placed in XRF sample cups
respectively. Use a spoon or spatula to place the sample into a
labeled XRF sample cup.
13. Seal XRF sample cup with mylar film and a ring.
Samples are now ready to be analyzed by Kevex X-ray Fluorescence Spectroscopy
(XRF). List samples according to their analysis number that corresponds with
the sample's identification number.
B. XRF ANALYSIS
1. Approximately 2 g of Total Soil Fraction or Fine Soil Fraction
sample are poured into sample cups (Somar Labs, Inc., Cat. No.
340), fitted with windows of 1/4 mil thick X-ray polypropylene film
(Chemplex Industries, Inc., Cat. No. 425). The sample cup should
be at least half full.
2. The sample cup is sealed with a sheet of micropourous film (Spex
Industries, Inc., Cat. No. 352A) held in place by the snap-on
sample cup cap. The exact weight of the sample is not important,
but should be in the range of 2-6 g.
A - 29
-------
The instrument configuration for the Kevex Delta Analyst Energy
Dispersive X-ray spectrometer is:
a. Kevex Analyst 770/8000 Excitation/Detection Subsystem:
1) X-ray tube: Kevex high output rhodium anode
2) Power supply: Kevex 60 KV, 3.3 mA.
3) Detector/cryostat: Kevex Quantum - UTW lithium,
drifted silicon.
b. Kevex Delta Analyzer:
1) Computer mainframe: Digital Equipment Corp, PDF 11/73
2) Computer software: • Kevex XRF Toolbox II, Version 4.14
3) Disk drives: Iomega Bernoulli box, dual drives, 10 MB
4) Pulse processor: Kevex 4^60
5) Energy to digital converterr Kevex 5230
c. Operating conditions:
1) Excitation mode: Mo secondary target with 4 mil thick Mb
filter .
2) Excitation conditions: 30 kV, 0.4 MA
3) Acquisition timer 100 livetime seconds
4) Shaping time constant: 7.5 microseconds
5) Sample chamber atmosphere: air .
6) Detector collimator: TA
d. Analytical conditions:
1) Escape peaks, and background should be removed from all
spectra.
2) The intensity ratio, defined as the integral of counts in
the Pb (I»A) window divided by the integral of the counts in
the Mo (KA) Compton scatter window, should be determined for
each spectrum.
3) The intensity ratios for the standards should be used to
determine a linear least squares calibration curve.
A - 30
-------
At the time of the analysis, there was no established detection limit for
this type of analytical method. However, the laboratory employed a standard
calibration range from 78 to 4,000 ppm lead. The lab worked within this
linear range for all analysis. SRM 1645 (714 ± 28 ppm Pb) was used as part
of quality control.
A - 31
-------
QUALITY ASSURANCE PLAN
SOIL
MARYLAND STATE LABORATORIES ADMINISTRATION
I. INTRODUCTION
This quality assurance document sets forth the Division of
Environmental Chemistry Multi-Element Analysis Laboratory (MEAL)
policies and procedures that maximize the quality of laboratory
performance. The goal of the laboratory is to provide a quality
service of elemental analysis.
It is the policy of MEAL to maintain an active quality
assurance/quality control (QA/QC) program to provide analytical
data of known and supportable quality and ensure a high
professional standard in analytical data generated in support of
projects undertaken for the public by state and federal agencies.
II. SAMPLE COLLECTION . ' s
Soil samples are collected by the Maryland Department of the
Environment (MDE) personnel for a variety of programs and projects.
All collectors are trained in sampling procedures.
Soil samples that are to be analyzed for metals are collected and
stored in clean previously unused polystyrene bags.
III. SAMPLE ACCEPTANCE, PRESERVATION, AND STORAGE
All incoming samples are delivered to the Soil Laboratory. As the
samples are accepted, they are assigned a laboratory sample number
and the submission form is dated with the current date.
i
The quantity of sample submitted must be adequate for all analyses
requested.
IV. METHODOLOGY
Lead is quantified via the Kevex XRF analyzer.
Ten percent of the samples are replicated. Certified SRM are
included throughout each tray run.
V. A. Requirements
1. Perform routine preventive maintenance on the Kevex unit.
2. A NBS standard should be analyzed once per tray of samples for
lead. The measured value should be within the control limits
established by NBS.
3. At least one replicate sample should be run every 10 samples,
or with each set of samples to verify precision of the method.
A - 32
-------
VI. DATA REDUCTION, VALIDATION, AND REPORTING
An important element In the quality control program is the
validation of data by the use of accuracy and precision
determinations. .Precision describes the degree to which data
generated from replicate measurements differ from one another.
Accuracy refers to the correctness of the data. Replicate samples
are analyzed periodically. Analysis and replicated data is also
graphically illustrated by plotting the numerical difference
between replicates versus sample number. The mean and standard
deviation are calculated for sample data.
Blind samples of known values are inserted into the sample stream
for analysis by the sample collectors.
VII. INSTRUMENT RECORDS AND LOGBOOKS
Maintain instrument records and logbooks for each instrument
including the following:
.1. Operations manuals with updates as provided by the
manufacturers
2. Service manuals and schedules of recommended preventive
maintenance
3. Maintenance logbooks containing entries describing all
maintenance performed on the instrument both by the multi-
element laboratory personnel, as well as qualified service
engineers
4. Sample logbooks containing a record of all samples analyzed
listed by date of analysis. These logbooks contain pertinent
information, such as sample identification, instrument
conditions, and analyst. Any special modifications made to
either the instrument or to the analytical protocol are also
noted.
VIII. GENERAL LABORATORY PRACTICES
The purchase of standard (or reference) material must be
accompanied by a certification or assay of composition.
A - 33
-------
BALTIMORE SOIL LEAD ABATEMENT DEMONSTRATION PROJECT
HOUSEHOLD DUST PROTOCOL
Household dust sampling should be carried out at the time of the
environmental visit to the home of the study participant.
For this study, the household dust samples are defined as the
samples that represent dust most likely to impact on a child's hands
during indoor activity. This would include dust on window sills, and
furniture, as well as dust on toys and other objects likely to be
handled by children. A minimum of three areas should be sampled: at the
main entrance to the household, and two areas most frequently used for
play activities by the child or children. Additional areas may be
selected that represent: 1) secondary entrances to the household (back
or side doors); 2), sources or accumulation of dust within the household
(paint, rugs, upholstered furniture); 3) additional play areas or other
areas of activity frequented by the children.
The sample has two components that are important to interpreting
lead exposure, the concentration of lead in the dust and the amount of
dust, or loading, on the surface. The concentration of lead in dust
appears to be closely related to the amount of lead on children's hands,
whereas the amount of dust on surfaces is an indicator of the importance
of this route of human exposure. At least 10% of the samples should be
over a defined area to determine the household loading factor.
Sketch the approximate layout of the residence and select to
sampling. Bear in mind that some areas, such as foyer, may reflect
outdoor dust to a greater degree than others.
The, sampling apparatus is the Sirchee-Spittler Hand Held Dust
Vacuum unit which is attached to a 'Dustbuster' hand held type vacuum.
Prior to the sample collection the sample collection screen; must be
clean.
For some samples, both the weight of the dust and the lead
concentration of the dust will be measured. In this case, it is
necessary to sample a defined area, so that the results may be expressed
in ng Pb/m . Mark the 4' x 4' sample area with tape. The surface of
the sample area is vacuumed with back and forth strokes about 1-2 inches
in width. The vacuum is most efficient if the head is held parallel to
the ground at a 45 degree angle. A single pass across the surface of
the sample area is sufficient to collect adequate sample amounts. After
dust sampling, the vacuum unit is kept in an upright position until the
sample screen is ready to be removed. Turn the vacuum off and remove
the sample screen. Empty the contents of the sample screen into a
labeled-reinforced paper envelope. Seal the envelope with scotch tape.
The sample amount required for analysis is equal to 2 grams of dust. If
the sample amount from the area is not sufficient additional sample
material may be collected from another 4'x 4' sample area and ;added to
the initial sample. Record sample data on the appropriate chain of
custody form. Transport the sample to the laboratory in a manner to
ensure upright envelope delivery.
A - 34
-------
BALTIMORE SOIL LEAD ABATEMENT DEMONSTRATION PROJECT
DUST ANALYSIS (XRF) PROTOCOIi
Identify sample information to be logged in on the Lead in Dust
Processing Sheets. Record the contract number, sample information/
date, time and total sample number on the processing sheet.
Example:
Date received: 06/30/90
Time: 12:30 p.m.
Total Number of Samples Received from: 41 from Ms. Merrill
Brophy
Sample Identification Number: #590312565
Site address: 2092 W. Preston Street
Area: Front door area
Area Collected (square feet): 4
Record the dust sample information on the XRF Run Sheets. Assign
analysis identification number to the sample.
Example:
The last dust sample .number was 0436, then the next sample
would be numbered 0437.
XRF sample cups (Somar Labs, Inc., Cat. No. 340) are to be prepared
before dust samples can be processed.
A. Assemble XRF cup for weighing.
1) Place the 26 mm ring, with rounded surface down, on a
flat surface
2) Cover with a 3 X 3 inch piece of mylar film (Spex
Industries, Inc., Cat. No.352A)
3) Snap/fit the 24 mm ring over the mylar film and inner
ring
B. Label XRF sample cups. Include only the analysis (cup)
number.
Transfer dust from envelope, as quantitatively as possible, onto a
60.0 mesh, 250 um, 3 inch wide stainless steel sieve with a pan and
cover.
A; Discard particles that cannot pass through the 60 mesh
sieve.
B. Clean sieves by tapping on a hard surface to remove
residual particles. This must be done between sample
processing.
Balance scale to nearest mg. and weigh empty XRF sample cup. Record
the weight (e.g. 28 mg). '
A - 35
-------
5. Using a pyrex funnel tripod, transfer dust into center of sample
cup.
A. Weigh and record weight of dust samples. The minimum
acceptable sample is 20 mg.
B. Seal XRF cup, containing dust sample, with another piece of
mylar file. Snap/fit 21 mm ring over this assembly.
C. Record lab accession number on outer ring of cup and on
side rim at finger grip cover.
D. Clean glass funnel with compressed air.
6. Remove cover of cup containing dust sample and place cup in the
Kevex XRF 7700/8000 for analysis.
A. The instrument reading in ppm is obtained.
B. Calculation:
sample weight = mg/sq. ft.; ppm
no. sq. ft.
Example: if sample weight = 28 mg; XRF reading = 200 ppm;
area sampled = 4 sg. ft.
28 mg = 7 mg./sq. ft.; 200 ppm.
4 sq. ft.
XRF ANALYSIS
The instrument configuration for the Kevex Delta Analyst Energy
Dispersive X-ray spectrometer is: . '
A. Kevex Analyst 770/8000 Excitation/Detection Subsystem:
1) X-ray tube: Kevex high output rhodium anode
2) Power supply: Kevex 60 xC, 3.3 mA.
3) Detector/cryostat: Kevex Quantum - UTW lithium, .drifted
silicon.
B. Kevex Delta Analyzer:
1) Computer mainframe: Digital Equipment Corp, PDF 11/73
2) Computer software: Kevex XRF Toolbox II, Version 4.14
3) Disk drives: Iomega Bernoulli box, dual drives, 10 MB
4) Pulse processor: Kevex 4460
A - 36
-------
5) Energy to digital converter: Kevex 5230
Operating conditions:
1) Excitation mode: Mo secondary target with 4 mil thick Mo
filter.
2) Excitation conditions: 30 keV, 0.40 - 0.60 mA
3) Acquisition time: 100 livetime seconds
4) Shaping time constant: 7.5 microseconds
5) Sample chamber atmosphere: air
6) Detector collimator: TA
Analytical conditions:
1) Escape peaks, and background should be removed from all
spectra.
2) The intensity ratio, defined as the integral of counts in
the Pb (LA) window divided by the integral of the counts in
the Mo (KA) Compton scatter window, should be determined for
each spectrum.
3) The intensity ratios for the standards should be used to
determine a linear least squares calibration curve.
A - 37
-------
LEAD ANALYSIS OF SURFACE DUST
A. GLASSWARE, SUPPLIES AND EQUIPMENT:
1. Beakers: Griffin, Pyrex brand, graduated, 100 mL
2. Flasks, volumetric, Class A: 25 mL, 50 mL, 100 mL, 200 mL, 1
liter and 2 liter
I
3. Funnels, micro, polypropylene, 24 mm top I.D., 4.5 mm stem
diameter (Bel-Art #14685-0024)
4. Test tubes: polypropylene, round bottom with caps, 17 x 100
mm (Falcon #2059)
5. Watch glasses, Pyrex brand, 75 mm diameter
5. Filter paper, Whatman #42, 5.5 cm diameter
'. Pipettes, volumetric, Class A: 2.0 mL, 5.0 mL, 6:0 mL, 10,0
mL, 15.0 mL and 20.0 mL
8. Pipettes, graduated: 2mL and 5 mL
9, Cylinders, Pyrex brand, graduated: 25 mL, 100 mL and 500 mL
10. Analytical Balance, 4-place, Mettler AE240
11. Hotplates
12. Atomic Absorption Flame Spectrophotometer, Perkin Elmer 3030B
13. Deionized water: Hydro Service Ultrapure Water System
14. Fume Hood
B. REAGENTSs
1. Nitric Acid, 7 M (prepared from 'Baker Analyzed' Reagent)
2. Nitric Acid, 1 M (prepared from 'Baker Analyzed' Reagent)
C. STANDARDS:
1. "Stock" Standard (100 ppm Lead):
20 mL of Varian Techtron Lead Standard (1000 ppm) diluted to
200 mL with 1M HNO3
A - 38
-------
2. "Working" Standards:
Working Standard
ppm
2.0
5.0
10.0
15.0
20.0
—
mL of Stock
(100 ppm)
2.0
5.0
10.0
15.0
20.0
Total Volume (mL)
| qs with 1M HNO3
100
100
100
100
100
A 0.6 ppm "working" standard is prepared with 6 mL of 10.0 ppm
standard, diluted to 100 mL with 1 M HN03.
3. Label, date and initial all solutions.
SAMPLES:
Dust samples received from the "Lead in Soil Project" for AAS analysis
are in XRF Cups (Somar), covered with mylar film. They were previously
analyzed by XRF 700/8000. For the most part, sample weight is well
below 100 mg.
CONTROLS AND BLANKS:
Duplicate NBS #1579 (11.87% Pb) Controls and two Total Reagent Blanks
are included in every group of 25 samples.
PROCEDURE:
1. Preparation of Samples for Acid Digestion:
a. Acid washing of glassware: All glassware must be acid washed
prior to use. Soak for 24 hours in 30% v/v nitric
acid/deionized water; rinse with deionized water.
Glassware must be oven-dried and cooled to room temperature
prior to use.
b. Label a 100 mL beaker with sample number. Tare beaker on a
calibrated 4-place analytical balance.
c. Transfer dust sample from XRF cup into tared beaker.
1) Small quantity dust sample (25 mg or less): Transfer
dust as completely as possible. Some dust adhers to the
mylar film and cannot be transferred.
A - 39
-------
2) Large quantity dust sample (50 mg or more): Transfer a
minimum of 50 mg dust. If quantity is 100 mg or more,
analyze in duplicate for precision check.
d. Weigh sample and record weight on form provided by
Environmental Chemistry Division.
One run constitutes approximately 25 samples.
Hot Nitric Acid Digestion:
a. To each sample in the beaker, add 25 mL 7 M nitric acid,
washing down the dust from the sides of the beaker.
b. Cover each beaker with a watch glass and place beakers on a
hotplate. Adjust setting so that contents of beakers simmer
(low boil) for two hours.-
c. Remove beakers from hotplate and allow to cool.
d. For samples weighing 25 mg or less:
1) Transfer digested sample to a 25 mL Class A volumetric
flask.
2) Rinse beaker and watch glass with 5 mL of 1M nitric acid
and transfer to same flask.
Repeat rinse procedure at least three times.
3) Dilute to volume with 1M nitric acid.
4) Stopper flask and mix well.
5) Filter through Whatman #42 filter paper, using a
polypropylene micro funnel, into a new labeled.Falcon
tube. Cap the tube. The sample is ready for AA
analysis. '
e. For samples weighing 50 mg or greater:
1) Filter the digested sample into a 50 mL Class A
volumetric flask, rinsing the lOOmL beaker, watch glass
and filter paper with 5 mL of 1M nitric acid.
2) Repeat rinse procedure at least three times.
3) Dilute to volume with 1M nitric acid.
4) Stopper flask and mix well.
5) Pour contents of volumetric flask into a new labeled
Falcon tube. Cap the tube. The sample is ready for AA
analysis.
A - 40
-------
PERKIN ELMER 3030B FLAME PARAMETERS:
ELEMENT: PB WAVELENGTH (NM): 283.3 SLIT (NM): 0.7
FLAME: AIR-ACETYLENE, OXIDIZING (LEAN, BLUE)
CHAR CONG: 0.45 SENS CHECK (MG/L) : 20.0 LINEAR TO (MG/L): 20.0
1.
3.
5.
7.
9.
11.
13.
16.
19.
TECHNIQUE: AA
SIGNAL PROCESSING: HOLD
NOMINAL WEIGHT 1.0
TIME (SECOND): 5.0
SCREEN FORMAT: BASIC DATA
RECORDER SIGNAL: 0.2 CONT ABS
SI: 20.0 14. S2: 15.0
S4: 5.0 17. S5: 2.0
S7: 20. S8:
2.
4.
6.
8 .
10.
12.
LAMP CURRENT (MA) r
10
CALIBRATION: LINEAR
STATISTICS : SINGLE
READ DELAY (SECONDS)
PRINTER: OFF
RECORDER EXP: 1000
15. S3: 10.0
18. S6: 0.6
21. RSLP:
READING
: 0.0
Computer IMS 286 used in conjunction with the Perkin Elmer 3030B. Results
obtained in percent; then converted to ppm. Results are reported in both
units.
REFERENCES;
University of Cincinnati Medical Center, Lead Program Project,
Institute of Environmental Health. "Surface Dust Analysis Protcol"
pp 11-13 (received 12/90).
Perkin Elmer Model 3030B Atomic Absorption Spectrophotometer Instruction
Manual. Norwalk, Connecticut, 1987.
Analytical Instrument Software, Inc.: "Auto-AA on Line QC
Software"Operators Manual,1989.
National Bureau of Standards Report 10674, "Experimental Evaluation
of Analytical Methods for ' Determining Lead in Paint and Building
Materials". U.S. Department of Commerce National Bureau of Standards,
Jan. 6, 1972.
A - 41
-------
QUALITY ASSURANCE PLAN
DUST
STATE LABORATORIES ADMINISTRATION
INTRODUCTION
This quality assurance document sets forth the Division of
Environmental Chemistry Multi-Element Analysis Laboratory (MEAL)
policies and procedures that maximize the quality of laboratory
performance. The goal of the laboratory is to provide a quality
service of elemental analysis.
It is the policy of MEAL to maintain an active quality
assurance/quality control (QA/QC)•program to provide analytical
data of known and supportable quality and ensure a high
professional standard in analytical data generated in support of
projects undertaken for the public by state and federal agencies.
SAMPLE COLLECTION
Dust samples are collected by the Maryland Department of the
Environment (MDE) personnel for a variety of programs ancl projects.
All collectors are trained in sampling procedures.
Dust samples that are to be analyzed for metals are collected and
stored in clean previously unused paper envelopes.
SAMPLE ACCEPTANCE, PRESERVATION, AND STORAGE
All incoming samples are delivered to the Multi-Elements Analysis
Laboratory. As the samples are accepted, they are assigned a
laboratory sample number and the submission form is dated with the
current date.
The quantity of sample submitted must be adequate for all .analyses
requested.
METHODOLOGY
j
Lead is quantified via the Kevex XRF analyzer. Ten percent of the
samples are replicated. Certified SRM is included each tray run.
QUALITY CONTROL
A. Requirements
1. Perform routine preventive maintenance on the Kevex unit.
2. A NBS standard should be analyzed once per tray of samples for
lead. The measured value should be within the control limits
established by NBS.
A - 42
-------
samples Kofverify . of the method.
VI. DATA REDUCTION, VALIDATION, AND REPORTING
An important element in the quality control program is the
validation of data by the use of accuracy and precision
determinations. Precision describes the degree to which data
generated from replicate measurements differ from one another.
Accuracy refers to the correctness of the data. Replicate samples
are analyzed periodically. Analysis and replicated data is also
graphically illustrated by plotting the numerical difference
between replicates versus sample number. The mean and standard
deviation are calculated for sample data.
VII. INSTRUMENT RECORDS AND LOGBOOKS
Maintain instrument records and logbooks for each instrument
including the following:
1. Operations manuals with updates as provided by the
manufacturers
2. Service manuals and schedules of recommended preventive
maintenance
3, Maintenance logbooks containing entries describing all
maintenance performed on the instrument both by the multi-
element laboratory personnel, as well as qualified service
engineers
4. Sample logbooks containing a record of all samples analyzed
listed by date of analysis. These logbooks contain pertinent
information, such as sample identification, instrument
conditions, and analyst. Any special modifications made to
either the instrument or to the analytical protocol are also
noted.
VIII. GENERAL LABORATORY PRACTICES
The purchase of standard (or reference) material must be
accompanied by a certification or assay of composition.
A - 43
-------
BMiTIMORE SOIL LEAD ABATEMENT DEMONSTRATION PROJECT
PAINT CHIP SAMPLING PROTOCOL
1, Visually evaluate the residence for evidence of peeled, chipped, cracked
paint on all surfaces.
2. Identify sample locations of painted surfaces that are peeling/ chipped
or cracked.
3. Collect paint samples using the sharp edge of a small knife blade to
scrape all layers of the suspect material down to the substrate. The
area sampled will equal a diameter of 2.0 inches.
4. Place the sampled material in a previously unused sampling paper
envelope and seal all edges of the envelope.
5. Mark the sample envelope with the property identification number, sample
code and sample number.
6. Return samples to the office.
7 * Record sample informa'tion on index card file.
8. Deliver samples to lab for analysis.
9. Report sample results to Lead In Soil personnel using the modem.
10. Samples which contain 0.06% lead will be positive for lead in this
study.
11. Lead In Soil personnel record sample results on main property file.
12. Residences which do not reflect sample results of 0.06% lead in the
paint chips sampled will not be scheduled for paint stabilization.
A - 44
-------
BALTIMORE SOIL LEAD ABATEMENT DEMONSTRATION PROJECT
PAINT CHIP ANALYSIS PROTOCOL
A. Using the Mortar & Pestle Method;
1. Paint chips delivered to the Lab.
2. Samples must be logged in on the Lead in Soil Processing Sheets.
The date, the time, the total number of samples brought by the
collector, and all the information listed on the sample bag should
be written on this sheet. The information listed on the sample bag
will include the sample identification number, the address and
particular area from which the sample was taken needs to be written
on the Lead in Soil Processing Sheet. Example:
Date received: 3/22/89
Time: 12:30 p.m.
Total Number of Samples Received From: 135 from Ms. Merrill
Brophy
Sample Identification Number: #590316535
Address: 2092 W. Preston Street
Area: Side of Front Door
3. The paint samples then need to be written up on the XRF Run Sheets.
Identification number is assigned. The sample is then given an
analysis number by the analyst. The number given to the sample is
used only as a means to identify a particular sample for analysis.
The samples should be written in consecutive sequence. Example:
The last sample analyzed was number 0439, then the next paint
chips sample should be numbered 0440.
4. Specimen Containers and XRF Sample Cups are to be prepared before
samples can be processed.
a. Label Specimen Containers - Include the date, the analysis
number, and the Samples's Identification Number.
b. Label XRF Sample Cups - Include only the analysis (i.e. cup)
number only.
5. Mortar & Pestle should always be clean.
6. Place paint chips into mortar and use the pestle to crush the
sample. Continue to crush the sample until a homogeneous mixture
is attained. Gloves and respirators must be worn.
7. Use a spoon or spatula to place the sample into a corresponding XRF
sample cup, then seal the cup with mylar film and a ring.
A - 45
-------
8. Before next sample is crushed, the mortar and pestle should be
wiped clean. Wipe the mortar and pestle with a clean paper towel,
then wash them with distilled water and dry them with a clean paper
towel. This process should be done after each sample.
9. Once all samples have completed steps 1-7, the samples are now
ready for analysis.
10. Analyzed sample results are recorded onto XRF Run Sheets in ppm's..
I?s ing Electric Mill Method
1. Paint chips delivered to the lab.
2. Samples must be logged in on the Lead in Soil Processing Sheets.
The date, the time, the total number of samples brought by the
collector, and all the information listed on the sample bag should
be written on this sheet. The information listed on the sample bag
will include the sample's identification number, the site address
and particular are from which the sample was taken needs to be
written on the Lead in Soil Processing Sheet. Example:
Date received: 3/30/89
Time: 12:30 p.m.
Total number of Samples Received From: 135 from Ms. Merrill
Brophy
Sample Identification Number: #590316521
Address: 2092 W. Preston Street
Area: Side of Front Door
3. The paint chip samples identification numbers are recorded on the
XRF Run Sheets. The sample is then assigned an analysis number by
the analyst. The number given to the sample by the analyst is used
only as a means to identify a particular sample for analysis. The
samples should be written in consecutive sequence.
Example:
The last sample analyzed was number 0439, then the paint chip
sample should be numbered 0440.
4. Specimen containers and XRF sample cups are to be prepared before
sample can be processed.
a. Label Specimen containers - Include the date, the .analysis
number, and the sample's identification number.
b. Label XRF Sample Cups - Include analysis number only.
5. Electric Mill should always be clean.
A - 46
-------
6. Electrical grinding must always be done under the hood. Gloves and
respirators must be worn.
a. Place paint chip samples into the Electric Mill.
b. Turn Electric Mill oh for approximately 3 minutes.
c. Turn grinder off after 3 minutes, wait for the dust to settle,
remove lid and check to see if a homogeneous mixture was
attained.
7. Use a spoon or spatula to place the sample into a corresponding XRF
sample cup, then seal the. cup with mylar film and a ring.
8. Before the next sample can be processed, the Electric Mill should
be cleaned. Wipe the Electric Mill with a clean paper towel inside
and out, dampen another paper towel and clean the mill very well,
and then dry the Electric Mill with another clean, dry paper towel.
This process should be done between each sample.
9. Once all samples have completed steps 1-7, the samples are now
ready for analysis.
10. Analyzed sample results are recorded onto XRF Run Sheets in ppm's.
A - 47
-------
QUALITY ASSURANCE PLAN
PAINT CHIPS
MARYLAND STATE LABORATORIES ADMINISTRATION
I. INTRODUCTION
This quality assurance document: sets forth the Division of
Environmental Chemistry Multi-Element Analysis Laboratory (MEAL)
policies and procedures that maximize the quality of laboratory
performance. The goal of the laboratory is to provide a quality
service of elemental' analysis.
It is the policy of MEAL to maintain an active quality
assurance/quality control (QA/QC) program to provide analytical
data of known and supportable quality and ensure a high
professional standard in analytical data generated in support of
projects undertaken for the public by state and federal agencies.
II. SAMPLE COLLECTION
Paint chip samples are collected by the Maryland Department of the
Environment (MDE) personnel for a variety of programs and projects.
All collectors are trained in sampling procedures.
Paint chip samples that are to be analyzed for metals are collected
and stored in clean previously unused paper envelopes.
III. SAMPLE ACCEPTANCE, PRESERVATION, AND STORAGE
All incoming samples are delivered to the Multi-Elements Analysis
Laboratory. As the samples are accepted, they are assigned a
laboratory sample number and the submission form is dated with the
current date.
The quantity of sample submitted must be adequate for all analyses
requested.
IV. METHODOLOGY
Lead is quantified via the Kevex XRF analyzer.
V. QUALITY CONTROL
A. Requirements
1. Perform routine preventive maintenance on the Kevex unit.
2. By NBS standard should be analyzed once per tray for the lead
measured. The measured value should be within the control
limits established by NBS.
3. At least one replicate sample should be run every 10 samples,
or with each set of samples to verify precision of the method.
A - 48
-------
VI. DATA REDUCTION, VALIDATION, AND REPORTING
An important element in the quality control program is the
validation of data by the use of accuracy and precision
determinations. Precision describes the degree to which data
generated from replicate measurements differ from one another.
Accuracy refers to the correctness of the data. Replicate samples
are analyzed periodically. . Analysis and replicated data is also
graphically illustrated by plotting the numerical difference
between replicates versus sample number. The mean and standard
deviation are calculated for sample data.
VII. INSTRUMENT RECORDS AND LOGBOOKS
Maintain instrument records and logbooks for each instrument
including the following:
1. Operations manuals with updates as provided by the
manufacturers
2. Service manuals and schedules of recommended preventive
maintenance
3. Maintenance, logbooks containing entries describing all
maintenance performed on the instrument both by the multi-
element laboratory personnel, as well as qualified service
engineers
4. Sample logbooks containing a record of all samples analyzed
listed by date of analysis. These logbooks contain pertinent
information, such as sample identification, instrument
conditions, and analyst. Any special modifications made to
either the instrument or to the analytical protocol are also
noted.
VIII. GENERAL LABORATORY PRACTICES
The purchase of standard (or reference) material must be
accompanied by a certification or assay of composition.
A -.49
-------
BALTIMORE SOIL LEAD ABATEMENT DEMONSTRATION PROJECT
DRINKING WATER SAMPLING PROTOCOL
Residents are notified that water must not be turned on prior to the
Environmental Health Aide sampling the system on the sampling day.
Do not shut off water flow valve to the sink fixture (which would
prevent use of the system prior to first draw) as this may introduce
lead corrosion products into the sample.
Morning first draw is collected from a cold water tap which had not been
used for 8-18 hours. Determine . if water was used prior to sample
collection. If water was used, state the use in the 'remarks on the
sample collection form.
Water samples are collected from each household faucet in 250 ml
cubitainers.
Water samples are preserved on site with 5 ml of nitric "acid per liter.
Water tap is closed after filling each sample container to prevent loss
of product and to ensure representative collections.
Keep samples cool (4 degrees C) after collection prior to analysis.
A - 50
-------
BALTIMORE LEAD IN SOIL
WATER ANALYSIS PROTOCOL
LEAD
Reference; Method 239.2 (Atomic Absorption, furnace technique) EPA - 600/4-
79-020
Optimum Concentration Range: 5-100 fig/L .
Detection Limit: 1 jig/L
Application;
Tests for lead are carried out using the graphite furnace atomic absorption
technique as described herein. Samples, blanks, quality control, replicate,
and spike test solutions are prepared as described and placed in trays for
automatic sampling (see Quality Assurance Plan). This instrument setup and
analysis steps are performed using the parameters defined.
Preparation of Standard Solution
1. Stock lead solution: ,Commercially available containing 1000 mg/L (1000
ppm) of lead.
2. Matrix modifier - Ammonium monobasic Phosphate + Magnesium Nitrate
Solution: Transfer 4 grams of NH4H2PO* monobasic Ultrex reagent and 0.2
grams of Mg (NO3)2, Suprapure, to a 100-mL volumetric flask and makeup
to mark with deionized distilled water (DW) containing 0.5% (v.v) HNO3.
3. Working lead solution: Dilute the stock solution to the ratios needed
as calibration standards .at the time of analysis. The calibration
standards and reagent blank must be prepared with the same acid, i.e.,
0.5% (v/v) HNO3. The reagent blank (RB) used in all subsequent dilutions
is prepared by diluting 5 mL cone. HNO3 to 1 L with DW. A 1-ppm solution
is prepared by dilution of the 1000-ppm stock solution with RB. This 1-
ppm solution is used to obtain calibration standards of 0, 5, 10, 25,
50, and 100 ppb lead. To obtain the calibration standards, withdraw
appropriate aliquots of the 1-ppm solution and dilute to 100 mL with RB.
Sample Preparation
All samples solutions for analysis are acidified in the field or in the
laboratory and contain 0.5% (v/v) cone. HNO3.
Instrument Parameters for Lead Analysis
1. Drying Time and Temp: 40 sec. - 120 degrees C
2. Charring Time and Temp: 40 sec - 1000 degrees C
3. Atomizing Time and Temp: 5 sec - 1800 degrees C
A - 51
-------
4. Cleaning Time and Temp: 5 sec - 2600 degrees C
5. Cooling Time and Temp: 20 sec - 25 degrees C
6. Purge Gas Atmosphere: Argon
7. Wavelength: 283.3 nm
8. Slit: 0.7 nm
9. Tub/site: Pyro coated tube with L'vbv platform
10. Matrix Modifier Setting: 5 fil
11. Sample and Standard Quantity Setting: 20 jiL
12. Max power: 30
13. Background correction mode: on
14. Lamp: electrodeless discharge lamp (EDL)
Notes Parameters 1, 2, 4, and 5 use 1 sec. ramp time. Parameter 3 uses
0 sec. ramp time and gas stop flow,
Instruments Used
1* Perkin-Elmer model 3030 Atomic Absorption Spectrophotometer with p2 Arc
Background Corrector
2. Perkin-Elmer PR-100 printer
3. HGA-300 graphite furnace
4. AS-40 Auto Sampler
A - 52
-------
QUALITY ASSURANCE PLAN
WATER
1YLAND
I. INTRODUCTION
This quality assurance document sets forth the Division of
Environmental Chemistry Multi-Element Analysis Laboratory (MEAL)
policies and procedures that maximize the quality of laboratory
performance. The goal of the laboratory is to provide a quality
service of elemental analysis.
It is the policy of MEAL to maintain an active quality
assurance/quality control (QA/QC) program to provide analytical
data of known and supportable quality and ensure a high
professional standard in analytical data generated in support of
projects undertaken for the public by state and federal agencies.
II. SAMPLE COLLECTION
Water samples are collected by the Maryland Department of the
Environment (MDE) personnel for a variety of programs and projects.
All collectors are trained in sampling procedures and approved by
the Division of Water Supply of the MDE.
Water samples that are to be analyzed for metals are collected and
stored in clean polyethylene or polypropylene containers with
teflon-lined lids.
III. SAMPLE ACCEPTANCE, PRESERVATION, AND STORAGE
All incoming samples are delivered to the Water Laboratory. As the
samples are accepted, they are assigned a laboratory sample number
and the submission form is dated with the current date.
To ensure that samples are not degraded and that their integrity is
maintained, all samples for metal analysis must be kept at 4
degrees C until receipt, and must be received by the laboratory no
later than one day after collection. Water samples for total
metals analysis should be preserved with analytical grade nitric
acid at a pH of 2 or less (typically 0.5% v/v) . The quantity of
sample submitted must be adequate for all analyses requested.
IV. METHODOLOGY
The following elements (arsenic, cadmium, chromium, lead, silver
and selenium) are quantified via graphite furnace atomic absorption
spectrophotometer.
Samples are analyzed after a blank and three different standard
calibration concentrations are completed. The characteristic mass
or sensitivity of the analyte is calculated for any of three
standards using the following equation:
A - 53
-------
(mL sample > x (cone, of standard in ppb) x (0.0044)
peak area (abs.)
The first sample • in each tray of 35 positions is always an EPA
water supply quality control sample and is followed by a standard
equivalent to one half of the maximum contaminant level (MCL) . Ten
percent of the samples are replicated. One hundred percent are
spiked. Blanks and different concentrations of standards are
included throughout each tray run.
V. QUALITY CONTROL
A. Minimum Requirements
1. All quality control data should be maintained and
available for easy reference or inspection.
2. An unknown performance evaluation sample must be analyzed
once per year for the metals measured. Results must be
within the control limit established by EPA. If problems
arise, they should be corrected, and a follow-up
performance sample should be analyzed.
3. Minimum Daily Control
a. After a calibration curve composed of a minimum of
a reagent blank and three standards has been
prepared, subsequent calibration curves must be
verified by use of at least a reagent blank and one
standard at or near the MCL. Daily checks must be
within + 10 percent of original curve.
b. If 20 or more samples per day are analyzed, the
working standard curve must be verified by running
an additional standard at or near-the MCL every 20
samples'. Checks must be within + 10 percent of the
original curve.
B. Optional Requirements
1. Routine preventive maintenance on balances and the atomic
absorption spectrophotometer.
2. Class S weights should be available to make periodic
checks on balances.
3. Chemicals should be dated upon receipt of shipment and
replaced as needed or before shelf life has been exceed.
4. A known reference sample (NBS) should be analyzed once
per quarter for the metals measured. The measured value
should be within the control limits established by NBS.
A - 54
-------
5. At least one duplicate sample should be run every 10
samples, or with each set of samples, to verify precision
of the method. Checks should be within the control limit
established by EPA.
6 . Standard deviation should be obtained and documented for
all measurements being conducted.
7 . Quality control charts or a tabulation of mean and
standard deviation should be used to document validity of
data on a daily basis. See Attachments 2 and 3 for
accuracy Quality Control chart sheets and precision
Quality Control chart sheets, respectively.
VI. DATA REDUCTION, VALIDATION, AND REPORTING
An important element in the quality control program is the
validation of data by the use of accuracy and precision
determinations . Precision describes the degree to which data
generated from replicate measurements differ from one another.
Accuracy refers to the correctness of the data. In an analysis
run, a replicate and spike are run periodically. Percent
recoveries are calculated on spike sample data and accepted when
recoveries are between 85% and 115%. If recoveries are outside
this range, samples are re-poured and re-spiked for additional
determinations. Percent recovery data is transferred onto graphs.
Replicated data is also graphically illustrated by plotting the
numerical difference between replicates versus sample number. An
upper warning limit and upper control limit is calculated by
multiplying the mean by 2.51 and 3.27 respectively. These quality
control charts are very useful in determining if a system is in a
state of statistical control. These charts are also used to
visualize and monitor the relative variability of repetitive data.
( See Attachment 2). the formula to calculate the mean of the
precision data is as follows:
R = ( d± .
di = the difference btween replicate results
N = number of samples
The mean and standard deviation are calculated for the spiked
sample data.
The percent recovery is calculated using the following formula:
% R = (SSR - SR^ x 100
SA
SSR = spiked sample result
SR = sample result
SA = spike added
In addition to replicates and spikes, the analytical sample runs
include numerous calibration checks. A reagent blank and standard
are run periodically. EPA water supply quality control samples and
NBS trace elements standards, and various other commercially
prepared standards are also analyzed during each run.
A - 55
-------
VII. . INSTRUMENT RECORDS AND LOGBOOKS
Instrument records and logbooks are maintained for each instrument.
These records include the following:
1. Operations manuals with updates as provided by the
manufacturers.
2. Service manuals and schedules of recommended preventive
maintenance.
3. Maintenance logbooks containing entries describing all
maintenance performed on the instrument both by the multi-
element laboratory, as well as by qualified service engineers.
4. Sample logbooks containing a record of all samples analyzed
listed by date of analysis. These logbooks contain pertinent
information, such as sample identification, instrument
conditions, and analyst. Any special modifications made to
either the instrument or to the analytical protocol are also
noted.
VIII. GENERAL LABORATORY PRACTICES
A. Laboratory Water
Laboratory pure water is supplied by a reverse osmosis, mixed
bed ion exchange system. Effluent water passes through a 0.5
um filter and the resistance of the outlet water is monitored
with an in-line conductivity probe (18 megohms).
B. Analytical Reagent
Analytical reagent grade chemicals are purchased for all
analyses and the following requirements are maintained:
1. All chemicals and standards are dated upon receipt and
the expiration date is also posted on the container.
2. Stock and working standards are labeled with
concentration, date prepared and'expiration date and with
the initials of the preparer.
C. Analytical Glassware
All volumetric glassware used in chemical analysis is
certified to be Class A Grade. ,
Disposable plastic tubes are used to minimize contamination.
D. Preparation of Standard Solutions
1. All standard solutions are made by one of the following
techniques:
A - 56
-------
(a) Dilution of a primary standard grade reagent to
volume using Class A volumetric glassware.
(b) Dilution of a known standard solution to volume
using Class A volumetric glassware (serial standard
method).
2. Shelf life of standard solution is dependent upon the
stability of reagent used and the frequency of use.
Standard solutions are labeled with date of preparation
and expiration, and the initials of the person who made
them.
3. The purchase of any standard (or reference) solution must
be accompanied by a certification or assay of
composition. Without such certification, said standard
will not be used.
E. Standardization Procedures
Any solution that will be used as a standard is checked
against a primary standard unless otherwise certified.
F. Hollow Cathode Lamp (HCL) and Electrodeless Discharge Lamp
(EDL) Documentation
1. All HCL and EDL lamps are dated upon arrival.
2. The intensity of each lamp is check upon arrival and
recorded with each use. The lamp is replaced if the
intensity goes below 75% of its original value.
A - 57
-------
BALTIMORE SOIL IN LEAD ABATEMENT DEMONSTRATION PROJECT
LEAD PAINT STABILIZATION PROTOCOL
The work included in this portion of the contract includes:
A. Before-preparation practices
1. Prior to preparation, occupants shall be notified of
starting date and expected date of completion. They
shall be instructed to remove all movable objects from
the work area and be informed about the proper method of
entrance to and egress from the property. Signs of heavy
cardboard shall be posted at each property, in a location
clearly visible to passersby, at least seven days prior
to the start of work.
2. Windows and doors in the work area shall be taped using
duct tape or equivalent water proof tape to seal out dust
for the duration of the work. Six mil thick plastic
should be installed on vertical surfaces where wet
scraping occurs.
3. All workers will be required to change into appropriate
work clothes, including shoes, upon arrival at the work
site. Remove work clothing before leaving work site.
Each worker will be required to wear a half-mask air
purifying respirator equipped with high efficiency
filters while in the work area. Smoking, eating, and
drinking will not be permitted in the work area. The
contractor will provide water, a dressing room, washroom
and toilet facilities for use of his employees.
4. Blood will be taken from the workers and tested prior to
starting the project, at two months and at the conclusion
of the project.
B. Complete preparation of exterior surfaces containing cracking,
chipping, peeling or chalking lead based paint includes
removal of deteriorated paint, High Efficiency Particulate Air
(HEPA) vacuum cleaning', washing and rinsing.
1. Methods of removal are limited to wet scraping in which
the surface being worked on is kept constantly wet using
a water spray.
2. Removal includes complete collection and disposal of all
resulting debris and dust.
3. Cleanup shall include HEPA vacuuming of all surfaces to
remove dust; if this is not feasible, wet methods may be
used, including wet sweeping, or shovelling.
A - 58
-------
4. Debris, including used sealing tape, used drop cloths,
filters, and disposable clothing shall be disposed of
according to hazardous waste regulations. Waste shall be
recorded by type, quantity, and disposal site.
C. Minimum preparation, washing and rinsing to remove dust and
dirt, of surfaces adjacent to those treated above shall be
done as necessary, to match those surfaces.
D. Repainting can begin after the site has been inspected and
approved and shall begin within 48 hours of completion of
surface preparation. The contractor shall provide all labor,
materials, equipment, and services necessary for satisfactory
application of field painting.
1. All caulking shall be done as directed by the Project
Manager. Caulk shall be a one part 100% liquid polymer,
acrylic base compound, non-sagging, non-staining, and of
gum consistency.
2. All paint shall be applied using a brush or roller. All
surfaces being repainted shall receive one coat of primer
and two finish coats. Paint shall be unscarred and
completely . integral. Sufficient drying time must .be
allowed between coats to satisfy the manufacturer's
requirements. Paint shall be a high quality latex based
system composed of a primer and an exterior finish, and
shall have a lead content of not more than 0.06%.
3. Upon completion of the work, all paint spots shall be
removed from walls, glass and other surfaces. All
rubbish and accumulated materials shall be removed and
area must be left in a clean, orderly and acceptable
condition.
E. The contractor may, with approval of the project manager,
choose to cover such items as window frames, porch eaves and
door frames with 0.032 inch thick, alloy 3004 - H 134 aluminum
sheets as an alternate to scraping and repainting them.
1. Prior to starting work, all windows and doors of the
affected structure shall be taped to seal out any dust.
2. Covering shall be formed on site to produce a close
fitting cover with smooth bends, close joints, and a
generally neat appearance.
3. Fasteners shall be spaced in accordance with good
practice, hidden where possible.
4. Joints at corners and at edges where aluminum abuts
masonry shall be neatly made and neatly and fully
caulked. Caulk shall be one part 100% liquid polymer,
acrylic base compound, non-sagging, non-staining, gum
consistency.
A - 59
-------
5. After completion of work and at the end of the day, all
resulting debris and dust shall be removed using a HEPA
vacuum cleaner and disposed of according to regulations.
A - 60
-------
BALTIMORE SOIL IN LEAD ABATEMENT DEMOSTRATION PROJECT
SOIL ABATEMENT PROTOCOL
Background: Removal of soil to a depth of 6 inches and replacement
of the soil with topsoil and sod is based on the results of
analysis for lead levels in the soil. If the lead levels in the
soil were equal to or greater than 500 ppm; the study property will.
be scheduled for removal of the soil. If the lead levels in the
soil are less than 500 ppm; bare spots of the study property were
reseeded to reduce dust levels.
Work included in this part of the contract includes:
A. Before preparation practices
1.. Prior to preparation residents shall be notified of
starting date and estimated date of completion. They
shall be instructed to remove all movable objects from
the work area and the contractor will control access to
the work area. Signs of heavy cardboard shall be posted
at each property in a location clearly visible to
passersby at least seven days before work begins.
2. Blood will be taken from the workers and tested prior to
starting the project, at two months, and at the
conclusion of the project.
\
3. Windows and doors adjacent to the work area shall be
taped with duct tape or equivalent waterproof tape. The
contractor shall use a light water spray.to eliminate or
capture any dust produced by the abatement procedures.
4. Workers will be required to change into work clothes,
including shoes, upon arrival at the site. Remove work
clothes before leaving the work site. Each worker will
be required to wear a half-mask air purifying respirator
equipped with high efficiency filters while working.
Eating, drinking, and smoking will not be permitted in
the work area. The contractor will provide water, a
dressing room, washroom and toilet facilities for the use
of his employees.
B. Preparation of the designated area includes physically
locating and marking the limits of the area using stakes and
tape or some other approved method, removing and disposing of
any trash within the work area limits, and carefully removing
any existing shrubs, plants, or ground cover other than grass
to an adequate storage place. Fencing shall also be removed
and stored.
C. Excavation of the soil to a depth of six inches. Mist the
area to be excavated with water to control dust levels.
Dispose of the excavated soil in an appropriate manner
depending on the toxicity or lack thereof.
A - 61
-------
D. Placing and compaction of "clean" soil (earth material
obtained off site, which shall have been previously tested (at
least 5 days prior) for lead and found to have less than 50
ppra).
E. After the refill has been acceptably compacted, the area shall
be covered with two inches of clean topsoil and sodded.
Shrubs, plants, and/or ground cover shall be replanted. In an
area close to a work area where there is an established stand
of grass, the contractor may be required to seed existing bare
areas. Seeding shall consist of loosening the existing soil
to a depth of two inches ; removal of all clods, stone and
other foreign materials larger than three inches in any
dimension, application of a 5-10-10 fertilizer at a rate of 5
pounds per 100 square feet and seed mix No. 2 at a rate of
0.25 pounds per 100 square feet, raking fertilizer and seed
into the prepared bed to a depth of not more than 1/4 inch,
compaction using an approved lawn roller, and application of
an approved mulch. All existing fencing removed during the
course of soil lead abatement work shall be re-erected.
F. At the end of each workday cleanup shall include vacuuming,
with a High Efficiency Particulate Air (HEPA) vacuum cleaner,
all surfaces adjacent to the work area to remove all dust.
After complete removal and cleanup, the site is ready for
inspection.
G. All lead contaminated debris including soil, filters and
disposable clothing shall be disposed of in accordance with
hazardous or solid waste regulations.
A - 62
-------
Appendix B
Quality Assurance Plans
PAGE
Program .Quality Assurance Plan -.- B-l
CDC Quality Assurance for Blood Lead Analyses. ....... B-83
QA/QC for Soil, Dust, and Handwipes B-92
-------
-------
QUALITY ASSURANCE PLAN
PROJECT TITLE: BALTIMORE SOIL LEAD ABATEMENT DEMONSTRATION PROJECT
(LIS)
EPA PROJECT MANAGER: RICHARD BRUNKER, Ph.D.
LIS PROJECT MANAGER: MERRILL BROPHY, R.N., M.S.N.
PERFORMING ORGANIZATION: MARYLAND DEPARTMENT OF THE ENVIRONMENT,
TOXICS, ENVIRONMENTAL SCIENCE & HEALTH
APPROVALS:
PROJECT MANAGER:
(MERRILL BROPHY, R.N., M.S.N.)
QA MONITOR:
PROGRAM ADMINISTRATOR:.
PRIMARY INVESTIGATOR:
EPA PROJECT MANAGER:
(ELI REINHARZ)
(BARBARA CONRAD, R.N., M.P.H. )
(KATHERINE FARRELL, M.D., M.P.H)
(RICHARD BRUNKER, Ph. D.)
B - 1
-------
Section 1.0
Revision 1.0
Date: March 90
Page: 1 of 37
QUALITY ASSURANCE PLAN
Section No.
1.0 Table of Contents
2.0 Project General Overview
2.1 Statement of Decision
2.2. Purpose of the Study
2.3 Description of the Project
2.4 Anticipated Results
2.5 Consequences of Incorrect Decisions or Conclusions
2.6 Project Measurements .
2.7 Application of Environmental Findings
2.8 Sampling Summary Table
2.9 Project Time Line
2.10 Project Flow Chart
2.11 Organizational Chart
2.12 Contractor and Subcontractor Geographical Locations
2.13 Procedure for Monitoring Contractors and Subcontractors
2.14 Description of Contractor and Subcontractor
Communications with MDE and EPA
3.0 Quality Assurance Plan Description
3.1 Introduction
3.2 Quality Assurance Plan Summary
3.2.1 Quality Assurance Plan Background
3.2.2 Demonstration Project
3.3 Quality Assurance Plan Objectives
3.3.1 Major Task Summary
4.0 Quality Assurance Objectives
4.1 General
4.2 Representativeness
4.3 Precision and Accuracy
4.4 Completeness
4.5 Comparatibility
5.0 Responsibility for Quality Assurance Plan
5.1 LIS Project Manager
5.2 LIS Environmental Coordinator
5.3 LIS Environmental Health Aides
6.0 Sampling Procedures
6.1 General
6.2 Equipment List
6.3 Record Keeping
B - 2
-------
Section 1.0
Revision 1.0
Date: March 90
Page: 2 of 37
7.0 Soil Sampling
7.1 Detailed Soil Sampling
7.1.1 Sampling Schemes
7.1.2 Handling of Samples
7.2 TOXICITY Soil Sampling
7.2.1 EP TOX Sampling Schemes
7.2.2 TCLP Sampling Schemes
7.3 Post Stabilization. Soil Sampling
7i3.1 Sampling Schemes
7.3.2 Handling of Samples
7.4 QA objectives for soil sampling
8.0 Paint Sampling
8.1 Exterior Paint Sampling
8.1..1 Sampling Schemes
8.1.2 Handling of Samples
8.2 Interior Paint Sampling
8.2.1 Sampling Schemes
8.2.2 Handling of Samples
8.2.3 Equipment Calibration
8.3 QA for paint sampling
9.0 Dust Sampling
9.1 Pre-Stabilization
9.1.1 Sampling Schemes
9.1.2 Handling of Samples
9.2 Post-Stabilization
9.2.1 Sampling Schemes
9.2.2 Handling of Samples
9.3 QA for dust sampling
10.0 Water Sampling
10.1 Sampling Schemes
10.2 Handling of Samples
10.3 QA for water sampling
11.0 Chain-of-Custody
11.1 General
11.2 Sample Receipt
11.3 Sample Storage
12.0 Quality .Assurance Plan - Soil Analysis
12.1 Introduction
12.2 Sample Collection
B - 3
-------
Section
Revision
Date:
Page:
12.3 Sample Acceptance, Preservation and Storage
12.4 Methodology
12.5 Quality Control Minimum Requirements
12.6 Data Reduction, Validation and Reporting
12.7 Instrument Records and Logbooks
12.8 General Laboratory Practices
13.0 Quality Assurance Plan - Water Analysis
13.1 Introduction
13.2 Sample Collection
13.3 Sample Acceptance, Preservation and Storage
13.4 Methodology
13.5 Quality Control Minimum Requirements
13.6 Data Reduction, Validation and Reporting
13.7 Instrument Records and Logbooks
13.8 General Laboratory Practices
1.0
1.0
March 90
3 of 37
14.0
Quality Assurance Plan - Dust Analysis
14.1 Introduction
14.2- Sample Collection
Sample Acceptance, Preservation and Storage
Methodology
Quality Control Minimum Requirements
Data Reduction, Validation and Reporting
Instrument Records and Logbooks
General Laboratory Practices
Atomic Absorption Spectrometry
14.3
14.4
14.5
14.6
14.7
14.8
14.9
15.0
Quality Assurance Plan - Paint Chips
15.1 Introduction
15.2 Sample Collection
15.3 Sample Acceptance, Preservation and Storage
15.4 Methodology
15.5 • Quality Control Minimum Requirements
15.6 Data Reduction, Validation and Reporting
15.7 Instrument Records and Logbooks
15.8. General Laboratory Practices
16.0 Data Assessment
16.1 General
16.2 Precision and Accuracy
16.3 Completeness
16.4 Corrective Action
17.0 Appendices
B - 4
-------
Section 2.0
Revision 1.0
Date: March 90
Page: 4 of 37
2.0 General Project Overview
2.1 Project Details
Project Title: Baltimore Lead In Soil Demonstration Project
EPA Project .Managers Richard Drunker, Ph.D.
LIS Investigators: Dr. Katherine Farrell, M.P.H.
Dr. J. Julian Chisolm, Jr.
Dr. Boon Lim, M.P.H.
*Charles Rodhe, Ph.D.
MDE Project Manager:
Project QA Monitor:
Project Category:
Project Duration:
Type of Project:
Project Address:
Project Phone No:
Merrill Brophy, R.N., M.S.N.
Alice Zeiger, Toxicologist 89-90
Eli Reinharz, Ecological Assessment 90-92
Type II
3 years
Superfund Urban Soil Abatement
Demonstration Project
Maryland Department of the Environment
TESH/LIS
2500 Broening Hwy
Baltimore, MD 21224'
410-631-3820
B - 5
-------
Section 2.0
Revision 1.0
Date: March 90
Page: 5 of 37
2.2. Statement of Decision
Study the correlation between removal of soil lead hazards
from child's environment at home and blood lead levels.
2.3 Purpose of the Study
Observe effects of soil abatement on children's blood lead
levels.
2.4 Description of the Project
Project will study two areas in Baltimore City which are of
pre-1978 construction and typical of lead containing type
houses in the urban area. A control group will be established
and will receive all testing and treatment as study group
except the soil abatement. Blood lead levels will be studied
over the course of the project beginning with baseline levels
and at regularly scheduled times. Soil, dust, water and paint
samples will be collected at all project houses. Paint
stabilization will be conducted at all project houses to
reduce likelihood of recqntamination of house soils. Soil
abatement for areas of the houses with lead results of >500
ppm will be conducted on study area houses.
2.5 Anticipated Results
Hypothesis stated a reduction of 1000 ppm of soil lead would
result in a reduction of 3 - 6 ng/dl blood lead level.
2.6 Consequences of Incorrect Decisions
The consequences of incorrect decisions would be the wrongful
assumption of soil lead levels influencing children's blood
lead levels.
2.7 List of Project Measurements
Table 2.7.1
B - 6
-------
Section 2.0
Revision 1.0
Date: March 90
Page: 6 of 37
2.8 Application of Environmental Findings
These findings will be used by EPA in determining future
actions/protocols for management of soil lead.
2.9 Sampling Summary Table
Table 2.9.1
2.10 Project Time Line
Figure 2.10.1
2.11 Project Flow Chart
Figure 2.11.1
*
2.12 Organizational Chart
Figure 2.11,1
2.13 Contractor and Subcontractor Geographical Locations
Contractor: .
Maryland Department of the Environment
Toxics, Environmental Science and Health
2500 Broening Hwy - .
Baltimore, MD 21224
Subcontractor:
State of Maryland Department of Health and Mental Hygiene
Laboratories Administration
301 W. Preston St.
Baltimore, MD 21204
2.14 Procedure for Monitoring Contractors and Subcontractors
The procedure for EPA monitoring of MDE will be through
quarterly reports and site visits.
The procedure for MDE monitoring of subcontractors of the
project will include blind audits and site visits.
2.15 Description of Contractor and Subcontractor
Communications with EPA or MDE
Quarterly reports to be submitted to Region III EPA by
MDE. Interim report will be submitted to EPA. Financial
Status Reports will be submitted annually to EPA.
Regularly scheduled meetings to be conducted by EPA
Region III Oversite Coordinator.
Laboratories Administration communicates with MDE through
laboratory analysis results. Regularly scheduled
meetings to be conducted by MDE with Laboratory
Administration personnel. A Memorandum of Understanding
is to be developed between MBE and the Laboratories
Administration.
B -
-------
Section 3.0
Revision 1.0
Date: March 90
Page: 7 of 37
3.0 Quality Assurance Plan Description
3.1 Introduction
The purpose of the Quality Assurance Plan (QAP) for the
Baltimore Lead-in-Soil Demonstration Project is to indicate
prime responsibilities and prescribe requirements for assuring
that the project is planned and executed in a manner
consistent with defined quality assurance objectives. This
QAP provides guidance and specifications to assure that:
1. All field sampling, methodologies and documentation,
sample preparation, handling and transportation are
conducted consistently according to established
procedures;
2. All laboratory determinations and analytical results
are valid through preventative maintenance, instrument
calibration, and analytical protocols;
i
3. Samples are identified and controlled through sample
tracking systems and chain-of-custody protocols;
4. Records are retained as documentary evidence of the
sample integrity, applied processes, equipment used, and
analytical results; and
5. Generated data is validated and its use in
calculations documented.
3.2 Quality Assurance Plan Summary
Information provided within this document summarizes the
specific tasks required for the project as well as other
pertinent information. .
3.2.1 Quality Assurance Plan Background
Data collected from Baltimore's Childhood Lead Poisoning
Prevention Program, coupled with the CDC report, led to the
following conclusion:
a. Children playing in the area of exposed, lead
contaminated soil may ingest lead in the course of their
normal hand-to-mouth activities.
b. Direct contact with lead contaminated soil may result
in increased body burden of lead.
B - 8
-------
Section 3.0
Revision 1.0
Date: March 90
Page: 8 of 37
c. Exposure of humans to lead through ingestion or
inhalation can result in toxic effects in the brain,
central and peripheral nervous system, kidney, and
hematopoietic system. Anemia is an early manifestation
of lead poisoning. Peripheral neuropathy also results
from lead poisoning. Young children under the age of six
are especially prone to the most profound and deleterious
effects of lead exposure. Chronic exposure to low levels
of lead can cause permanent learning disabilities in
children.
3.2.2 Demonstration Project
The Baltimore Soil Lead Abatement Demonstration Project
shall involve sampling approximately 400 selected
children from two urban neighborhoods (Park Heights -
study area, Walbrook Junction - control area) for blood
lead levels to determine base line data, sampling their
yards to establish soil lead levels, and their residences
to determine dust lead levels, removing contaminated soil
in the study area, re-sampling the children during the
following year to observe the effects of the soil
removal. Property owner consent is required for the
houses to be enrolled in the project. This QAP addresses
the soil, dust, water and paint collection and analysis.
The environmental staff will consist of one industrial
hygienist, as the Environmental Coordinator, and six
Environmental Health Aides. The Environmental Health
Aides will conduct environmental sampling at project
properties according to the attached protocols. The
Environmental Coordinator schedules the sampling and
supervises the sampling and documentation of samples
Detailed environmental sampling will be conducted, during
1988 and 1989, as necessary, at the selected children's
properties. Environmental sampling includes: soil,
interior dust, water, and exterior paint. Common
environmental characteristics of the study and control
area properties include:
a. exterior paint positive for lead
b. soil areas included in the property
B - 9
-------
Section 4.0
Revision 1.0
Date: March 90
Page: 9 of 37
4.0 Quality Assurance Objectives
4.1 General
The quality of environmental sampling made during this study
will be determined by the following characteristics:
accuracy, precision, representativeness, completeness, and
comparability.
4.2 Represenativeness
Sampling procedures will be used to assure that samples
collected are representative of the media. Sample handling
protocols protect the representativeness of the collected
sample. Proper documentation will ensure that protocols have
been followed and that sample identification and integrity are
assured.
B - 10
-------
Section 5.0
Revision 1.0
Date: March 90
Page: 10 of 37
5.0 Responsibility for Quality Assurance
5.1 Lead-in-Soil Project Manager
The project manager is responsible for:
-approving, maintaining and implementing the Quality
Assurance Plan for this project,
-indicating the types of quality assurance records to be
maintained for this project and,
-approving sampling procedures and operating systems.
5.2 Lead-in-Soil Environmental Coordinator
The environmental coordinator is responsible for:
-developing sampling protocols for soil, dust, water and
paint,
-training other LIS personnel in the implementation of
environmental protocols,
-scheduling environmental sample collection,
-reviewing sample collection information, results and
data,
-periodically checking the sampling equipment for
cleanliness and condition
-informing the project manager of sampling issues and,
-randomly checking the data entry forms of sample
analysis results for accuracy against the original sample
results form submitted by the laboratory.
5.3 Lead-in-Soil Environmental Health Aides
The environmental health aides are responsible for:
-collecting environmental samples from project properties
as scheduled.
-collecting environmental samples according to project
protocols.
-submitting samples and appropriate paperwork to the lab
for sample analysis.
-transcribing sample analysis results to the data entry
forms.
B - 11
-------
-checking data entry forms for accuracy before submitting
them to the environmental coordinator.
Section 6.0
, Revision 1.0
Date: March 90
Page: 11 of 37
6.0 Sampling Procedures
6.1 General
Sampling protocols will be followed as closely as possible
allowing however in the inherent differences between each
project property. Deviation from sampling protocols will be
clearly documented on diagrams or forms.
6.2 Equipment List
Standard sampling equipment is provided for the project. A
complete listing of equipment is located in Appendix M.
6.3 Record Keeping
All environmental materials, forms, equipment and supplies
which have been used for this project are maintained by the
Environmental Coordinator.
B - 12
-------
Section 7.0
Revision 1.0
Date: March 90
Page: 12 of 37
7.0 Soil Sampling
7.1 Detailed Soil Sampling
7.1.1 Sampling Schemes
After properties are selected and enrolled in the project,
they undergo detailed soil sampling. Top soil samples will be
collected by soil corer at the surface to a depth of 2 inches.
Bottom soil samples will be collected from the soil corer at
a depth of 4-6 inches by the project staff according to the
attached protocols. The soil sample will be collected using
one of the defined pattern: line source, targeted area, small
area, or grid pattern. Pattern selection will be based upon
the layout of the subject property at the discretion of the
Environmental Coordinator. Sketches indicating property
details and sample locations will be made by Environmental
Health Aides. .
7.1.2 Handling of Samples
Soil samples are collected in polyproplyene plastic bags and
labeled at the time of collection. Samples are recorded on a
chain-of-custody form and submitted to the laboratory by the
Environmental Coordinator. Laboratory personnel receive
samples and issue a sample receipt for the project file.
Analysis results are recorded on Soil Processing Sheets and
returned to the Environmental Coordinator.
7.2 Toxicity Soil Testing
7.2.1 EP Toxicity Soil Testing Sampling Scheme (EPTOX)
Soil samples from study area properties will be analyzed by
EPTOX prior to the soil stabilization phase of the project to
determine disposal requirements of the soil removed during
stabilization. Top soil samples will be collected by soil
corer at the surface to a depth of 2 inches from locations
within each area to be stabilized. Randomly selected sites
within the area to be stabilized will be sampled in the
following manner:
One sample from an area 20 inches from foundation, one sample
from the middle of the area and one sample from 20 inches
inside the area will be composited for EPTOX testing.
Soil samples are collected in polyproplyene plastic bags and
labeled at the time of collection. Samples are recorded on a
chain-of-custody form and submitted to the laboratory by the
Environmental Coordinator. Laboratory personnel receive
B - 13
-------
Section 7.0
Revision 1.0
Date: March 90
Page: 13 of 37
samples and issue a sample receipt for the project file.
Analysis results are recorded on Sample Processing Sheets and
submitted to the -Environmental Coordinator.
7.2.2 Toxicity Characteristics Leaching Procedure (TCLP)
Soil samples from control area properties will be analyzed by
TCLP to determine disposal requirements of the soil. Top soil
samples will be collected by soil corer at the surface to a
depth of 2 inches from locations within each area to be
stabilized. Randomly selected sites within the area to be
stabilized will be sampled in the following manner:
One sample from an area 20 inches from foundation, one sample
from the middle of the area and one sample from 20 inches
inside the area will be composited for EP Toxicity testing.
Soil samples are collected in polyproplyene plastic bags and
labeled at the time of collection. Samples are recorded on a
chain-of-custody form and submitted to the laboratory by the
Environmental Coordinator. Laboratory personnel receive
samples and issue a sample receipt for the project file.
Analysis results are recorded on Sample Processing Sheets and
returned to the Environmental Coordinator.
7.3 Handling of Samples
7 .3 Post Stabilization Soil Sampling
7.3.1 Sampling Schemes
Post stabilization soil sampling is conducted after paint
stabilization and soil abatement. The primary purpose of this
sampling is . to document the effectiveness of stabilization
activities. Soil abatement contracts require replacement soil
to contain less than SOppm lead. Top soil samples will be
collected by soil corer at the surface to a depth of 2 inches.
Bottom soil samples will be collected from the soil corer at
a depth of 4-6 inches by the project staff according to the
attached protocols. The soil sample will be collected using
one of the defined pattern: line source, targeted area, small
area, or grid pattern. Pattern selection will be based upon
the layout of the subject property at the discretion of the
environmental coordinator. Sketches indicating property
details and sample locations will be made by environmental
health aides.
B - 14
-------
Section 7.0
Revision 1.0
Date: March 90
Page: 14 of 37
7.3.2 Handling of Samples
«
Soil samples are collected in polyproplyene plastic bags and
labeled at the time of collection. Samples are recorded on a
chain-of-custody form and submitted to the laboratory by the
Environmental Coordinator. Laboratory personnel receive
samples and issue a sample receipt for the project file.
Analysis results are recorded on Sample Processing Sheets and
submitted to the Environmental Coordinator.
B - 15
-------
Section 8.0
Revision 1.0
Date: March 90
Page: 15 of 37
8.0 Paint Sampling
8.1 Exterior Paint Sampling
8.1.1 Sampling Schemes
Properties selected for the project will be sampled for lead
content of exterior painted -surfaces of all structures.
Representative samples of chipped, cracked, flaking or peeling
paint chips will be collected for analysis.
will be stored in reinforced
at the time of collection.
a chain-of-custody form and
the Environmental Coordinator.
samples and issue a sample
Analysis results will be
Sheets and submitted to the
8.1.2 Handling of Samples
Paint chip samples collected
paper envelopes and labeled
Samples will be recorded on
submitted to the laboratory by
Laboratory personnel receive
receipt for the project file
recorded on Paint Processing
Environmental Coordinator.
8.2 Interior Paint Sampling
8.2.1 Sampling Schemes
Properties selected for the project will be sampled for lead
content of two interior surfaces of three rooms within the
house. Record sample identification numbers, codes, condition
of surface tested and results on Lead-in-Soil LBP Inspection
Forms. A sample copy of the LBP Inspection Form is located in
Appendix F.
PGT XRF Calibration
Calibration of the PGT XRF Lead Based Paint Analyzer is
required at the beginning and end of each interior survey.
Calibration checks against three paint standards is required.
Each check is recorded in the calibration record and the page
where it is recorded is indicated on the LBP Inspection Form.
Each calibration check must fall within the acceptable range
as indicated on the standards.
B - 16
-------
Section 9.0
Revision 1.0
Date: March 90
Page: 16 of 37
9.0 Dust Sampling
9.1 Pre-Stabilization - •
9.1.1 Sampling Schemes
Properties selected for the project will be sampled for lead
content of dust on interior surfaces within the house.
9.1.2 Handling of Samples
Dust samples collected will be stored in reinforced paper
envelopes and labeled at the time of collection. Samples will
be recorded on a chain-of-custody form and submitted to the
laboratory by the Environmental Coordinator. Laboratory
personnel receive samples and issue a sample receipt for the
project file. Analysis results will be recorded on Paint
Processing Sheets and submitted to the Environmental
Coordinator.
9.2 Post-Stabilization
9.2.1 Sampling Schemes
Properties selected for the project will be sampled for lead
content of dust on interior surfaces within the house.
9.2.2 Handling of Samples
Dust samples collected will be stored in reinforced paper
envelopes and labeled at the time of collection. Samples will
be recorded on a chain-of-custody form and submitted to the
laboratory by the Environmental Coordinator. Laboratory
personnel receive samples and issue a sample receipt for the
project file. Analysis results will be recorded on Paint
Processing Sheets and submitted to the Environmental
Coordinator.
B - 17
-------
10.0 Water Sampling
10.1 Sampling Schemes
Section 10.0
Revision 1.6
Date: March 90
Page: 17 of 37
Properties selected for the project will be sampled for lead
content of drinking water sources within the house. Annotate
on the Water Analysis Form those samples .collected that will
not be first draw (from pipes not used for 8-18 hours
previously) .
10.2 Handling of Samples
A vial of 1M nitric acid was added to labeled sample
collection containers prior to collection of the water sample.
Record water samples on a Water Analysis Form, store on ice
and submit to the laboratory at the end of the day.
B - 18
-------
Section 11.0
Revision 1.0
Date: March 90
Page: 18 of 37
11.0 Chain-of-Custody
, 11.1 General
EPA has established a program of sample chain—of-custody that
is followed during sample handling activities in both field
and laboratory operations.
Chain-of-custody procedures document the sample history and
constitute a crucial part of sampling and analysis programs.
Chain-of-custody documentation verifies the identification and
history of a sample from collection through the time of
analysis.
The objective of sample custody identification and control is
to ensure that:
-all samples scheduled for collection, as appropriate for
the data required, will be uniquely identified;
-the collection samples will be analyzed and traceable to
specific analysis records,
-important sample characteristics will be preserved;
-samples will be protected from loss or damage;
-an alteration of samples (e.g., filtration,
preservation) is documented;
-a record of sample integrity is established for legal
and technical purposes; and
The chain-of-custody record is used to:
-document sample handling procedures, including sample
location, and sample number; and
-describe the chain-of-custody process.
The chain-of-custody description section requires:
-the sample number;
-the name(s) of the sampler(s) and the person shipping
the samples;
-the date and time that the samples were delivered for
shipping; and
Section 11.0
B - 19
-------
Revision 1.0
Date: March 90
Page: 19 of 37
-the names of those responsible for receiving the samples
at the laboratory.
Samples of a chain-of-custody record for soil, paint, and dust
samples are shown in Appendix F. A sample of the water sample
analysis form is shown in Appendix F.
As samples are collected, entries will be made on the chain-
of-custody form. Data to be noted includes:
-date and time,
-sampler(s) name,
-type of sample,
-sample identification number,
-project name,
-name of person to receive results,
-property identification number and street address.
Soil, dust, water and paint chip sample containers will be
labelled by an indelible marker with the appropriate
information necessary to match the sample container to the
chain-of-custody record.
When samples are received at the laboratory, the laboratory
technician will verify each and every sample against the
chain-of-custody, note any discrepancies or losses of samples,
and then sign for receipt of the samples. The laboratory
technician may also contact filed personnel to resolve
deficiencies, irregularities, discrepancies, etc., prior to
accepting the samples. Samples will remain under the control
of the laboratory technician until samples are ultimately
disposed.
A sample is considered to be in custody if it:
-is in the physical possession of the responsible party;
-is in view of the responsible party;
-is secured by the responsible party to prevent
tampering; or
-is secured by the responsible party in a restricted
area.
B - 20
-------
Section 11.0
Revision 1.0
Date: March 90
Page: 20 of 37
Chain-of-Custody is initiated at the time of sample collection
and follows the sample through to the laboratory where it is
replaced by a Laboratory Processing Sheet. Sample information
and results will be recorded on the Laboratory Processing
Sheet and returned to the Environmental Coordinator.
11.2 Sample,Receipt
All samples will be delivered to the laboratory by a member of
the LIS field sampling team. Upon receipt chain-of-custody
and sample integrity will be checked and any problems
recorded. Samples will then be logged in by laboratory
personnel who will accept and sign the chain-of-custody
record.
Each sample received by the laboratory is assigned a unique
sequential Laboratory Identification number which will
identify the sample in the laboratory's internal tracking
system.
11.3 Sample Storage
Samples not destroyed by the analysis process will be returned
to the Environmental Coordinator for inventory and storage at
the secured facility for the duration of the project.
B - 21
-------
Section 12.0
Revision 1.0
Date: March 90
Page: 21 of 37
12.0 Quality Assurance Plan Soil Analysis
12.1. Introduction
This quality assurance document sets forth laboratory policies
and procedures that maximize the quality of laboratory
performance. The laboratory goal is to provide a quality
service of elemental analysis.
Soil samples submitted for analysis for metals will be
collected and stored in clean previously unused polystyrene
bags.
12.2 Sample Acceptance, Preservation, and Storage
All incoming samples will be delivered to the Soil Laboratory.
As the samples are accepted, they will be assigned a
laboratory sample number and the submission form is dated with
the current date.
The quantity of sample submitted must be adequate for all
analyses requested.
12.3 Methodology
Lead will be quantified via the Kevex XRF analyzer. Ten
percent of the samples will be replicated. Certified NBS
Standards will be included each tray run.
12.4 Quality Assurance
All quality assurance data will be maintained and available
for easy reference or inspection. An unknown performance
evaluation sample must be analyzed once per year for the
metals measured. If problems arise, they should be corrected,
and a follow-up performance sample should be analyzed.
B - 22
-------
Section 12.0
Revision 1.0
Date: March 90
Page: 22 of 37
Perform routine preventive maintenance on the Kevex unit.
A NBS standard should be analyzed once every tray run for lead
measured. The measured value should be within the control
limits established by NBS.
At least one replicate sample should be run every 10 samples,
or with each set of samples to verify precision of the method.
12.5 Data Reduction, Validation, and Reporting
An important element in the quality control program is the
validation of data by the use of accuracy and precision
determinations. Precision describes the degree to which data
generated from replicate measurements differ from one another.
Accuracy refers to the correctness of the data. Replicate
samples will be analyzed periodically. Analysis and
replicated data is also graphically illustrated by plotting
the numerical difference between replicates versus sample
number. The mean and standard deviation will be calculated
for sample data. Blind samples of known values will be
inserted into the sample* stream for analysis by the sample
collectors.
12.6 Instrument Records and Logbooks
Maintain instrument records and logbooks for each instrument
including the following:
Operations manuals with updates as provided by the
manufacturers. Service manuals and schedules of recommended
preventive maintenance, maintenance logbooks containing
entries describing all maintenance performed on the instrument
both by the multi-element laboratory personnel, as well as
qualified service engineers, and Sample logbooks containing a
record of all samples analyzed listed by date of analysis.
These logbooks contain pertinent information, such as sample
identification, instrument conditions, and analyst. Any
special modifications made to either the instrument or to the
analytical protocol will also noted.
12.7 General Laboratory Practices
The purchase of standard (or reference) material must be
accompanied by a certification or assay of composition.
B - 23
-------
Section 13.0
Revision 1.0
Date: March 90
Page: 23 of 37
13.0 Quality Assurance Plan Water Analysis
13.1. Introduction
This quality assurance document sets forth laboratory policies
and procedures that maximize the quality of laboratory
performance. The laboratory goal is to provide a quality
service of elemental analysis for the project.
It is the laboratories policy to maintain an active quality
assurance program to provide analytical data of known and
supportable quality and ensure a high professional standard in
analytical data generated in support of the project undertaken
for the public by state and federal agencies.
13.2 Sample Collection
Water samples for this project will be collected by trained
sampling collectors who are approved by the Division of Water
Supply of the MDE.
Water samples to be analyzed for metals will be collected and
stored in clean polyethylene or polypropylene containers with
teflon-lined lids.
13.3 Sample Acceptance, Preservation and Storage
All incoming samples will be delivered to the Water
Laboratory. As the samples are accepted, they are assigned a
laboratory sample number and the submission form is dated with
the current date.
To avoid sample degradation, all samples for metal analysis •
must be kept at 4 degrees C until receipt, and must be
received by the laboratory no later than one day after
collection. Water samples for total metals analysis should be
preserved with analytical grade nitric acid at a pH of 2 or
less (typically 0.5% v/v). The quantity of sample submitted
must be adequate for all analyses requested.
13.4 Methodology
Arsenic, cadmium, chromium, lead, silver and selenium will be
quantified via graphite furnace atomic absorption
spectrophotometer. Samples will be analyzed after a blank and
three different standard calibration concentrations will be
completed. The first sample in each tray of 35 positions is
always an EPA water supply quality control sample and is
followed by a standard equivalent to one half of the maximum
B - 24
-------
Section 13.0
Revision 1.0
Date: March 90
Page: 24 of 37
contaminant level (MCL). Ten percent of the samples are
replicated. One hundred percent will be spiked. Blanks and
different concentrations of standards will be included
throughout each tray run. .
13.5 Quality Control Minimum Requirements
All quality control data will be maintained and available for
easy reference or inspection. An unknown performance
evaluation sample must be analyzed once per year for the
metals measured. Results must be within the control limit
established by EPA. If problems arise, they should be
corrected, and a follow-up -performance sample should be
analyzed.
Prepare a calibration curve composed of a minimum of a reagent
blank and three standards, verify subsequent calibration
curves by use of at least a reagent blank and one standard at
or near the MCL. Daily checks must be within ± 10 percent .of
original curve. If 20 or more samples per day are analyzed,
the working standard curve must be verified by running an
additional standard at or near the MCL every 20 samples.
Checks must be within + 10 percent of the original curve.
Routine preventive maintenance on balances and the atomic
absorption spectrophotometer. Class S weights should be
available to make periodic checks on balances.
Chemicals should- be dated upon receipt of shipment and
replaced as needed or before shelf life has been exceed. A
known reference sample (NBS) should be analyzed once per
quarter for the metals measured. The measured value should be
within the control limits established by NBS. At least one
duplicate sample should be run every 10 samples, or with each
set of samples to verify precision of the method. Checks
should be within the control limit established by EPA.
Standard deviation should be obtained and documented for all
measurements being conducted. Quality control charts or a
tabulation of mean and standard deviation should be used to
document validity of data on a daily basis.
13.6 Data Reduction, Validation and Reporting
An important element in the quality control program is the
validation of data by the use of accuracy and precision
determinations. Precision describes the degree to which data
generated from replicate measurements differ from one another.
B - 25
-------
Section 13.0
Revision 1.0
Date: March 90
Page: 25 of 37
Accuracy refers to the correctness of the data. In an
analysis run, a replicate and spike will be run periodically.
Percent recoveries will be calculated on.spike sample data and
accepted when recoveries are between 85% and 115%. if
recoveries are outside this range, samples will be re-poured
and re-spiked for additional determinations. Percent recovery
data is transferred onto graphs. Replicated data is also
graphically illustrated by plotting the numerical difference
between replicates versus sample number. An upper warning
limit and upper control limit is calculated by multiplying the
mean by 2.51 and 3.27 respectively. Quality control charts
will be very useful in determining if a system is in a state
of statistical control and will be used to visually monitor
the relative variability of repetitive data.
13.7 Instrument Records and Logbooks
Instrument records and logbooks will be maintained for
each instrument. These records include the following:
1. Operations manuals with updates as provided by the
manufacturers. Service manuals and schedules of
recommended preventive maintenance.
i
2. Maintenance logbooks containing entries describing
all maintenance performed on the instrument both by
the multi-element laboratory, as well as by
qualified service engineers.
3. Sample logbooks containing a record of all samples
analyzed listed by date of analysis. These
logbooks contain pertinent information, such as
sample identification, instrument conditions, and
analyst. Any special modifications made to either
the instrument or to the analytical protocol will
be also noted.
13.8 General Laboratory Practices
13.8.1 Laboratory Water
Laboratory pure water is supplied by a reverse osmosis,
mixed bed ion exchange system. Effluent water passes
through filter and the resistance of the outlet water is
monitored with an in-line conductivity probe (18
megohms).
B - 26
-------
Section 13.0
Revision 1.0
Date: March 90
Page: 26 of 37
13.8.2 Analytical Reagent
Analytical reagent grade chemicals will be purchased for
all analyses and the following requirements will be
maintained. All chemicals and standards will be dated
upon receipt and the expiration date is also posted on
the container. Stock and working standards will be
labeled with concentration, date prepared and expiration
date and with the initials of the preparer.
13.8.3 Analytical Glassware
All volumetric glassware used in chemical analysis is
certified to be Class A Grade. Disposable plastic tubes
will be used to minimize contamination.
13.8.4 Preparation of Standard Solutions
All standard solutions will be made by diluting primary
standard grade reagent to volume using Class A volumetric
glassware, or diluting a known standard solution to
volume using Class A volumetric glassware (serial
standard method).
Shelf life of standard solution is dependent upon the
stability of reagent used and the frequency of use.
Standard solutions will be labeled with date of
preparation and expiration, and the initials of the
person who made them.
The purchase of any standard (or reference) solution must
be accompanied by a certification or assay of
composition. Without such certification, said standard
will not be used.
13.8.5 Standardization Procedures
Any solution that will be used as a standard is checked
against a primary standard unless otherwise certified.
13.9 Hollow Cathode Lamp (HCL) and Electrodeless Discharge
Lamp (EDL) Documentation
1. All HCL and EDL lamps will be dated upon arrival.
2. The intensity of each lamp is check upon arrival
and recorded with each use. The lamp is replaced
if the intensity goes below 75% of its original
value.
B - 27
-------
14.0 Quality Assurance Plan Dust Analysis
14.1 Introduction
Section 14,0
Revision 1.0
Date: March 90
Page: 27 of 37
This quality assurance document sets forth laboratory policies
and procedures that maximize the quality of laboratory
performance. The goal of the laboratory is to provide a
quality service of elemental analysis for the project.
The quality assurance program uses analytical data of known
and supportable quality to ensure a high professional standard
in analytical data generated in support of projects undertaken
for the public by state and federal agencies.
14.2 Sample Collection
Dust samples will be collected by trained .sampling collectors
for this project. All collectors will be trained in sampling
procedures. Dust samples to be analyzed for metals will be
collected and stored in clean previously unused paper
envelopes and recorded on the chain-of-custody.
14.3 Sample Acceptance, Preservation and Storage
All incoming samples will be delivered to the laboratory. As
the samples are accepted, they will be assigned a laboratory
sample number and the submission form is dated with the
current date. The quantity of sample submitted must be
adequate for all analyses requested.
14.4 Methodology
Lead will be quantified via the Kevex XRF analyzer. Ten
percent of the samples will be replicated. Certified NBS
standards will be included each tray run.
B - 28
-------
Section 14.0
Revision 1.0
Date: March 90
Page: 28 of 37
14.5 Quality Control
Perform routine preventive maintenance on the Kevex unit.
A NBS standard should be analyzed once per tray for lead. The
measured value should be within the control limits established
by NBS. At least one replicate sample should be run every 10
samples, or with each set of samples to verify precision of
the method.
14.6 Data Reduction, Validation and Reporting
An important element in the quality control program is the
validation of data by the use of accuracy and precision
determinations. Precision describes the degree to which data
generated from replicate measurements differ from one another.
Accuracy refers to the correctness of the data. Replicate
samples will be analyzed periodically. Analysis and
replicated data is also graphically illustrated by plotting
the. numerical difference between replicates versus sample
number. The mean and standard deviation will be calculated
for sample data.
B - 29
-------
14. 7 Instrument Records and Logbooks
Section 14.0
Revision 1.0
Date: March 90
Page: 29 of 37
Maintain instrument records and logbooks for each instrument
including the following:
1. Operations manuals with updates as provided by the
manufacturers. Service manuals and schedules of
recommended preventive maintenance
2. Maintenance logbooks containing entries describing
all maintenance performed on the instrument both by
the multi-element laboratory personnel, as well as
qualified service engineers
3. Sample logbooks containing a record of all samples
analyzed listed by date of analysis. These
logbooks contain pertinent information, such as
sample identification, instrument conditions, and
analyst. Any special modifications made to either
the instrument or to .the analytical protocol will
be also noted.
14.8 General Laboratory Practices
The purchase of standard (or reference) material must be
accompanied by a certification or assay of composition.
B - 30
-------
Section 15.0
Revision 1.0
Date: March 90
Page: 30 of 37
15.0 Quality Assurance Plan - Paint Chips
15.1 Introduction
This quality assurance document sets forth the laboratory
policies and procedures which maximize the quality of
laboratory performance. The goal of the laboratory is to
provide a quality service of elemental analysis.
It is the policy of the laboratory to maintain an active
quality assurance program to provide analytical data of known
and supportable quality and ensure a high professional
standard in analytical data generated in support of the
project.
15.2 Sample Collection
Paint chip samples will be collected by trained sampling
collectors and stored in clean previously unused paper
envelopes.
15.3 Sample Acceptance, Preservation and Storage
All incoming samples will be delivered to the laboratory,
assigned a laboratory sample number and the submission form is
dated with the current date.
The quantity of sample submitted must be adequate for all
analyses requested. - • -
15.4 Methodology
Lead will be quantified via the Kevex XRF analyzer. Ten
percent of the samples will be replicated. Blanks will be
included each tray run.
B - 31
-------
Section 15.0
Revision .1.0
Date: March 90
Page: 31 of 37
15.5 Quality Control
Perform routine preventive maintenance on the Kevex unit.
A NBS standard should be analyzed once per tray for lead. The
measured value should be within the control limits established
by NBS. At least one replicate sample should be run every 10
samples, or with each set of samples to verify precision of
the method.
15.6 Data Reduction, Validation and Reporting
An important element in the quality control program is the
validation of data by the use of accuracy and precision
determinations. Precision describes the degree to which data
generated from replicate measurements differ from one another.
Accuracy refers to the correctness of the data. Replicate
samples will be analyzed periodically. Analysis and
replicated data is also graphically illustrated by plotting
the numerical difference between replicates versus sample
number. The mean and standard deviation will be calculated
for sample data.
15.7 Instrument Records and Logbooks
Maintain instrument records and logbooks for each instrument
including the following:
Operations manuals with updates as provided by the
manufacturers
Service manuals and schedules of recommended preventive
maintenance
B - 32
-------
Section 15.0
Revision 1.0
Date: March 90
Page: 32 of 37
Maintenance logbooks containing entries describing all
maintenance performed on the instrument both by the multi-
element laboratory personnel, as well as qualified service
engineers
Sample logbooks containing a record of all samples analyzed
listed by date of analysis. These logbooks contain pertinent
information, such as sample identification, instrument
conditions, and analyst. Any special modifications made to
either the instrument or to the analytical protocol will be
also noted.
15.8 General Laboratory Practices
The purchase of standard (or reference) material must be
accompanied by a certification or assay of composition.
B - 33
-------
Section 16.0
Revision 1.0
Date: March 90
Page: 33 of 37
16.0 Data Assessment
16.1 General
The purpose of data quality assessment is to assure that data
generated under the program will be accurate and consistent
with project objectives. The quality of data will be
assessed based on the precision, accuracy, consistency, and
completeness of the data that area measured or generated.
Data quality assessment will be conducted in three phases:
Phase I
Prior to data collection, sampling, and analysis, procedures
will be evaluated in regard to their ability to generate the
appropriate, technically acceptable information required to
achieve project objectives. This QA plan meets this
requirement by establishing project objectives defined in the
terms of required sampling analysis protocols.
Phase 2
During data collection, results will be assessed to assure
that the selected procedures are efficient and effective and
the data generated provided sufficient information to achieve
project objectives. Precision and accuracy of measurement
systems will also be evaluated. In general, evaluation of
data will be based on performance audits, results of duplicate
and reference sample analyses, and review of completeness
objectives.
Phase 3
Throughout the data collection activities, an assessment of
the adequacy of the data base generated in regard to
completing project objectives will be undertaken.
Recommendations for improved quality control will be
developed, if appropriate. In the event that data gaps are
identified, the Project Manager may recommend the collection
of additional raw data to fully support the project's findings
and recommendations.
Documentation may include:
-number of duplicate and reference samples analyzed;
-identification of statistical techniques, if used, to measure
central tendency, dispersion, or testing for outlier;
-use of historical data and its reference; and
-i'dentification of analytical method.
B - 34
-------
Section 16.0
Revision 1.0
Date: March 90
Page: 34 of 37
16.2 Precision and Accuracy
Assessment of precision and accuracy of analytical data is
accomplished via review of duplicate analyses (precision) and
reference standard (accuracy) in soil. Precision is generally
expressed as the relative percent difference (RPD). Accuracy
is expressed as percent recovery. Precision must be assessed
for each matrix, since distribution of contaminants may be
non-homogeneous, especially in soil. Precision in samples
must be reviews with knowledge of the matrix and level of
analyte present. Corrective action or documentation of
substandard precision is the laboratory's responsibility.
Accuracy is also impacted by matrix interferences. Each
method which provides quality control requirements and
acceptance criteria also specifies the method of generating
the data to be reviewed. It is the laboratory's
responsibility to attempt to identify the source of
substandard recoveries and either take corrective action or
document the cause.
Precision control requirements and acceptance criteria
also specify the method of generating the data to be
reviewed. It is the laboratory's responsibility to
attempt to identify the source of substandard recoveries
and either take corrective action or document the cause.
16.3 Completeness
Completeness is generally assessed as a percentage of data
intended to be generated, and is most often utilized in Phase
3 of the data assessment process. Assessment of completeness
will be undertaken by the Project Manager in cooperation with
the LIS staff.
B - 35
-------
Section 16.0
Revision 1.0
Date: March 90
Page: 35 of 37
16.4 Corrective Action
Corrective or preventative action is required when potential
or existing conditions are identified that may have an adverse
impact on data quantity or quality. Corrective action could
be immediate or long-term. In general, any member of the
program staff who identifies a condition adversely affecting
quality can initiate corrective action by notifying in writing
his or her supervisor and the Project Manager. The written
communication will identify the condition and explain how it
may affect data quality or quantity.
An analysis or analytical system is considered to be out-of-
control when it does not conform to the conditions specified
by the method or standard operating procedures which apply.
To confirm that an analysis or analytical system is in
control, the laboratory routinely performs instrument
calibration checks, analysis of method blanks and method blank
spikes. These results will be compared to the results of
quality control samples to laboratory control charts or
analytical protocol criteria (e. g., U. S. EPA-CLP).
A Corrective Action Documentation Form, Appendix F is to be
completed for each out-of-control situation. The analyst,
working with his or her supervisor or task leader, will
attempt to determine the cause of the problem and take
appropriate corrective action. Analysis may not resume until
the problem has been corrected and it is determined that the
analysis is back in control. Demonstration of the restoration
of analytical control will normally be accomplished by
generating satisfactory calibration and or quality assurance
sample data. This documentation will be attached to the
corrective action documentation form to be placed in the
project files.
16.5 Immediate Corrective Action
Immediate corrective action is applied to spontaneous, non-
,recurring problems, such as an instrument malfunction. The
individual who detects or suspects non-conformace will
immediately notify his or her supervisor. . The supervisor and
the appropriate task leader will then investigate the extent
of the problem and take the necessary corrective steps. If a
large quantity of data is affected, the task leader must
prepare a memorandum to the Project Manager. The individual
will collectively decide how to proceed. If the problem is
limited in scope, the task leader will decide on the
corrective action measure, document the solution in the
appropriate workbook, and notify the Project Manager.
B - 36
-------
Section 16.0
Revision 1.0
Date: March 90
Page: 36 of 37
16.5 Long-Term Corrective Action
Long term corrective action procedures will be devised and
implemented to prevent the recurrence of a potentially serious
problem. The Project Manager will be notified of the problem
and will conduct an investigation to determine the severity
and extent of the problem. The Project Manager will then file
a corrective action request with the appropriate supervisory
personnel.
B - 37
-------
List of Appendices
Appendix A
Appendix B
Appendix C
Appendix D
Appendix E
Appendix F
Appendix G
Appendix H
Appendix- I
Appendix J
Appendix K
Appendix M
Section 17.0
Revision 1.0
Date: March 90
Page: 37 of 37
XRF Analyzer Calibration Procedure
Demonstration Project Flow Chart
Contract Abatement Diagram Example
Baltimore Project Vicinity Map
Protocols
Sample Forms
Glossaries
Method Detection Limits
Flow Chart of Data Handling
Environmental Sampling Statistics
LIS Project Personnel Flow Chart
Equipment List
B - 38
-------
Appendix A
XRF Analyzer Calibration Procedure
-------
-------
Baltimore Lead In Soil Demonstration Project
GUIDANCE DOCUMENT FOR X-RF
Definition: The field X-Ray fluorescence (XRF) analyzer is a
direct reading insturment which determines lead concentration on
painted surfaces. The advantages of the XRF analyzer are the
integrity of the surface is not disturbed and results are available
immediately. XRF reading represents the concentration of lead in
mg/cm2.
Analysis of paint chips is also used to identify lead in
painted surfaces. Obtaining paint chips samples breaks the
integrity of the surface and the chips must be submitted to a
laboratory for wet chemistry analysis. Readings represent the
percentage of lead in the paint by volume, or percentage.
Although both XRF and paint chip analysis should be fairly
consistent with each other, the results of one testing method
cannot be converted to the other.
Standard Procedures: Follow standard procedures to obtain reliable
results when using these analyzers. Factors that may interfer with
the XRF analyzers, include:
* Substrate material
* Temperature extremes (eg. below 35F or above 95F)
* Zero drift .
* Zinc
* Radio waves
* Vibration
Three XRF readings within a 1.7 range are recorded and the mean is
' reported as the lead concentration of a surface. If the readings
are not within a 1.7 range repeat the process of collecting
readings within the acceptable range. If the readings remain
greater than the 1.7 range, do not use this analyzer on this
substrate. Possible solution: Recalibrate the analyzer and
resample.
B - 39
-------
The basic technique for reducing the variability of XRF readings is
to take repeated measurements at the same point. Statistical
theory shows the variablity of the average of a set of repeated
measurements is less than the variabiltiy of individual
measurements. The greater the number of repeated measurements, the
greater the reduction in variability. Use paint chip samples
analysis to back up XRF readings as explained further in this
document.
Analyst Qualifications: The operator must have adequate training
and experience using the equipment. All XRF operators in TESH's
Lead Programs have the following qualifications:
* Trained in radiation health and safety, including
knowledge of Federal, state and local laws and
regulations governing the licensing and use of
radioactive devices.
* Listed by name on the license for the XRF equipment used,
issued by MDE Center for Radiological Health.
* Assigned ring badge for monitoring exposure.
* Attended classroom training in use of XRF analyzer,
principles of operation and calibration. Content
considered to be essential includes:
1~ Required number of readings per surface tested.
2. Factors that affect XRF analysis.;
3. Need for back up paint scrapings (when, where).
4. Knowledge of how to take paint scrapings.
5. On-the-Job training with other experienced
inspector/tester.
6. .Knowledge of building construction.
Field work Guideliness Inspectors using XRF analyzer should use a
logbook assigned fo the analyzer. All reading taken must be
recorded as well as operation, maintenance and repair information.
* Charge the batteries continually when the instunnent is
not in use and 12 hours before use in the field.
B - 4O
-------
* Calibrate the insturment on site before a survey is done.
Follow the manufactures instructions for calibration as
summarized below:
• Turn on the analyzer, allow the analyzer to warmed
up for at least ten minutes then take seven to
eight calibration readings using both the lead and
non-lead standards. Record the calibrations in the
notebook and indicate the calibration page on the
survey form.
* Compare warm up reading to post-warm up readings. The
differences between these readings may indicated low
battery problems, etc.
* Take a minimum of three readings per surface tested. If
the first two readings are very high, e.g., over 6.0
mg/cm2 it is not required to take the third reading.
* At levels where the XRF readings are questionable,
(0.5 - 2.5 mg/cm2) results must be verified by
paint scrapings. A representative sample of the
surface in question should be taken. For example,
if all wood trim in the living room appears similar
and all readings are 1.5 mg/cm2, one paint sample
would be adequate to verify these readings.
* Check for zero standard after a reading of 10.0mg/cm2 or
after. a series of 5.0 mg/cm2 or higher and record all
these readings on the test form.
* Use the analyzers only on surfaces that are flat and as
wide as the face of the analyzer. Paint scrapings should
be taken if surface is narrow or irregular.
* Use back-up paint scrapings on metal, concrete or brick
surfaces and components that contain air spaces, such as
hollow core doors.
* If the analyzer moves while taking a reading start
over with that reading.
Collection of Paint Scrapings: Collect samples in an uniform and
consistent manner. Samples should contain all paint layers but not
the substrate material.
B - 41
-------
Surfaces that should be tested: All painted interior and exterior
surfaces should be tested, including:
*all walls within each room,
*all parts of windows including sash, frames, wells, and
sills,
*all parts of stairs including risers, treads, balusters,
baseboards, and newel posts.
*all parts of porches including railings, balusters, columns,
ceilings * and floors.
Make a sketch of the dwelling and indicate the north direction.
Identify all rooms by a code number. Identify each sample by using
the sample identification number and appropriate code number.
Sample diagram attached.
Interpretation ,of results: XRF readings in excess of 2.5 mg/cm2
can be considered positive without additional testing. XRF
readings of 0.5 - 2.5 mg/cm2 should be confirmed with paint chip
analysis. Results of paint scrapings at 0.5 or higher are
considered to be positive. Some results fall into Vgrey areas"
that require additional professional assessment to make a
determination. In cases such as this, contact the Environmental
Coordinator for assistance.
B - 42
-------
Appendix B
Demonstration Project Flow Chart
-------
Figure 2.11.1
Demonstration Project Row Chart
Preliminary surface soil sampling
I
Analyze soil with XRF
Data to LIS
[Pb] > 500 ppm in soil areas
where children reside and LBP on exterior of residence
NO
1. Property scheduled for
paint stabilization.
2. Property not scheduled for
soil stabilization.
3. Analyze soil, dust, paint,
and water for lead.
4. Data to LIS
YES
1. Property scheduled for
paint stabilization.
2. Property scheduled for
soil stabilization.
3. Excavate soil to a depth
of 6 inches and remove soil.
Fill excavated area with clean
fill; re-sample abated areas
to determine post abatement
soil lead levels.
4. Data to LIS
B - 43
-------
Appendix C
Management Plans
PAGE
Data Management Plan C-l
Safety Guidelines C-39
Public Relations Plan. .C-42
-------
-------
Appendix C
Baltimore Contract Abatement Diagram Example
-------
, I _ M
5s-
r-
9
.»
'0'
Xs
!
1
1
1
OAR AC
<^~ COK
•v*- "•."* *"* •' .•".• : '.- • * ' .
** * . , •- * .« • • * «
B.
•7-0"
..*..,.•.. * • *••*.•??,
••-..••.«•••. c ••"-
"."• ". -. .*•'.*. "f--. •-" ".
COI
*.*• • - •**« *.•«
ae
JC
.'[
4
» ." "
u<
• • •
*
t
RETE —
b*"~'-*"'i'-: •* ~
--*• — ^-oc
(2 S
FROUT
•:.•':••'::':••;•:
1RETE
."A"-',-:/;"^.
\V-.* i"-j •" '
• •*•" "• . "•."•.'.'
DKJCRCTE
- .
TORY RC
.
PORCH
':.:• .i'-V/: •'•'.'•
"*. .^»**\* •"
WAl_Kv
•Q
JO
"~!
• «•— *
>w t
•
1
/
y
.
4out>e;
*" *""»*""• * \
•/•*••*'*'%
fev^
• • * " . • * *
* . • . '•* *
* * •" \ ** "."• * •
* * i* * • *
".VeVr-j
•_ 3 -b"
(
Q
s
p
1
ABATEMENT METHOD: RE NAOVE AMD D\SPOaS OF TOP
<& IWCHSS OF SOIU, REFILL. WITH
'CL.EANJ* MATERUAL.AS SPECIFIED.
WO A6ATEMEJJT
SPECIFIED.
*D* - /OS
E » 60
SF
COVHR: COVER AREAS "c"
THRU "F'WITH SOD
AS DIRECTED BY THE
ARCHITECT.
•EAVTN COMPANY
STATE OF MARYLAND
DEPARTMENT OF THE ENVIRONMENT
LEAD PAINT STABILIZATION PROJECT
IN BALTIMORE CITY
LEAD IN SOIL ABATEMENT
CONTRACT WO. MOE-88-OOI-TESH-l
CONT R ACT * 4,._
-------
Appendix D
Baltimore Project Viciinity Map
-------
BALTIMORE CITY
Bark Heights Area
Wall brook Junction Area
Patapsco
River
VICINITY MAP
No Scale
-------
Appendix E
Protocols
Sampling Collection
Sample Analysis QA
-------
BALTIMORE SOIL LEAD ABATEMENT DEMONSTRATION PROJECT
SOIL SAMPLING PROTOCOL
I. Site Description . -
For each location, a detailed drawing should be made that
shows the boundary of the lot, the position of the main
building and any other buildings such as storage sheds or
garages, the position of the sidewalks, driveways, and other
paved 'areas, the position of the play areas if obvious, and
the position of the areas with exposed soil (grassy or bare),
also, showing roof rain spouts and general drainage patterns.
In addition to the diagram, briefly describe the location,
including the following information:
Type of building construction
Condition of main building
Condition of property (debris, standing water, vegetation
cover)
Nature of adjacent property
Presence and type of fence .
Animals on property
Apparent use of yard (toys, sandbox, children present)
Underground utilities
XI. Soil Area Description
For each soil area (i.e. front patch, front yard, back yard,
side yards) identified on the general diagram, draw a full
page diagram showing the approximate dimensions and position
relative to the building foundation. Indicate vegetation and
bare soil areas, as well as obvious traffic patterns.
Identify the category of land use, such as roadside,, property
boundary, adjacent to foundation, play area. Mark the sample
location" on the diagram.
III. Sampling Schemes
Measure, the soil .area to determine the sampling scheme.
Select ;-the sample scheme for each soil area which adequately
characterize the potential exposure of children to lead in the
dust from this soil. Identify the suspected areas of high
lead concentrations and the assumed general distribution
pattern of lead concentrations at the soil surface.
Small Area Pattern. Measure and mark off an area 20 inches
from the base of the foundation into the soil area. Repeat
measuring and marking at the boundaries. The area inside the
marked pattern indicates the sampling collection area. If the
sampling collection area is less than two meters in each
dimension, a single composite sample may be taken if it
appears that such a sample would adequately represent the soil
B - 46
-------
area. (Collect two sample bags, one bag marked top and the
other bag marked bottom.)
Large Area Pattern. Measure and mark off an area 20 inches
from the base of the foundation into the soil area. Repeat
measuring and marking at the boundaries. The area inside the
marked pattern indicates the sampling collection area.
Collect one composite sample at the foundation and one
composite sample at the boundary of the yard if the .area is
less than 10 feet wide. (Collect four sample bags, two bags
marked top and two bags marked bottom.) Collect an additional
composite sample at an imaginary sample line between the
foundation and boundary sample areas if the yard is larger
than 16 feet wide. (Collect six sample bags, three bags
marked top and three bags marked bottom.)
Very Large Area Pattern. Measure and mark off an area 20
inches from the base of the foundation into the soil area.
Repeat measuring and marking at the boundaries. If a yard is
wider than 16 feet and more than 20 feet long then divide the
yard into a vertical half and a horizontal-half. Collect one
composite sample at the from each section of the yard.
(Collect twelve sample bags, six bags marked top and six bags
marked bottom.)
IV. Sample Collection . . •• -
Collect ten randomly selected core samples from within the
sampling area. The cores make a composite sample identified
as a single sample. Record composite .information on the
sample sheet.
Clean and decontaminate the corer after each sample
collection. Remove vegetation and debris from the corer at
the point of insertion into the soil, but do not remove any
soil or decayed litter. Drive the corer in to the ground to
a depth of 15 cm (6 in.). If this depth cannot be reached,
the corer should be extracted and cleaned, and another attempt
made nearby. If repeated attempts do not permit a 15 cm core,
take the sample as deep as possible, and record the maximum
penetration depth on the sample record sheet.
Combine the top two inch segment of each core into one
composite sample and combine the bottom two inch segment of
each core into second composite sample. Remove debris and
leafy vegetation from the top sample material. Do not remove
soil or decomposed litter from the sample material. This is
the most critical part of the soil sample and is likely to be
the highest in lead concentration.
Assemble composite soil core segments in clean previously
unused plastic bags suitable for prevention of contamination
of the sample. Record the sample identification number on the
B - 47
-------
bag and the sample record sheet. Store the composite soil
sample at ambient temperature until submitted to the
laboratory for analysis.
Clean the corer after collecting each sample composite by
reinsertion of the corer into the soil of the next sampling
area.
Draw field blanks for each soil area by inserting the core
borer into randomly selected locations within the sample area.
These blanks are drawn prior to sample collection and at the
conclusion of sampling.
V. Sample Handling and Storage
Seal the sample bags to prevent loss or contamination of the
sample and storage samples in a cool, dry location.
Record-keeping and Sample Custody
Initiate s.oil sample records for each location which consists
of a location diagram and description, a plot diagram for each
distinct soil plot, and sample record sheet for each sample in
a plot.
Sequentially number samples bags. Record sample numbers on
location diagram, soil area description, and sample record
sheet.
Deliver the sample to the laboratory and release the sample to
the laboratory personnel for analysis.
B - 48
-------
QUALITY ASSURANCE PLAN
FOR SOIL SCORER SAMPLE >
I. INTRODUCTION
This quality assurance document sets forth the Baltimore
Lead-in-Soil Project's (LIS) policies and procedures that
maximize the quality of sample collection and laboratory
performance. The goal of the sample collector is to
provide a representative sample of the surface to be
tested according to the appropriate protocol. The goal
of the laboratory is to provide a quality service of
elemental analysis.
It is the policy of LIS to maintain an active quality
assurance (QA) program to provide analytical data of
known and supportable quality and ensure a high
professional standard in analytical data generated in
support of projects undertaken for the public by state
and federal agencies.
II. SAMPLE COLLECTION
Soil samples are collected by LIS personnel for this
project. All collectors are trained in sampling
procedures.
Soil samples for metals analysis are collected and stored
in clean previously unused polystyrene bags. Sample bags
are labeled with a unique sample identification number
and sample code which reflects the location of the sample
site. A corresponding chain-of-custody form is completed
at the time of sample collection.
III. SAMPLE ACCEPTANCE, PRESERVATION, AND STORAGE
All incoming samples are delivered to Maryland's
Department of Health and Mental Hygiene " (DHMH) for
analysis. As the samples are accepted, they are assigned
a laboratory number and the chain-of-custody is dated
wit,h the current date.
'••"'.
The quantity of sample submitted must be adequate for all
analyses requested.
B - 49
-------
IV.
SAMPLE RESULTS
Following DHMH analysis, the results are reported to the
IiIS Environmental Coordinator (EC). Excess sample
material and DHMH sample cups are returned to the LIS
project for storage at a secured facility. The LIS EC
reviews the data results and assigns preliminary data
processing tasks to the Environmental Health Aides I
(EHA), who transfers the sample number and results to
data entry forms. Each set of data entry forms are
double checked by level II EHA personnel.
Upon completion of the data information transfer the data
entry sheets are surrendered to data entry personnel for
double entry into the data base.
V.
DATA REVIEW
A review of all raw data and data base soil files is to
be conducted prior to the end of the study.
B - 5O
-------
BALTIMORE SOIL LEAD ABATEMENT DEMONSTRATION PROJECT
PAINT CHIP ANALYSIS PROTOCOL
A. Using the Mortar & Pestle Methods
1. Paint chips delivered to the Lab.
2. Samples must be logged in on the Lead in Soil Processing
Sheets. The date, the time, the total number of samples
brought by the collector, and all the information listed
on the sample bag should be written on this sheet. The
information listed on the sample bag will include the
sample identification number, the address and particular
area from which the sample was taken needs to be written
on the Lead in Soil Processing Sheet. Example:
Date received: 3/22/89
Time: 12:30 p.m.
Total Number of Samples Received From: 135 from Ms.
Merrill Brophy
Sample Identification Number: #590316535
Address: 2092 W. Preston Street
Area: Side of Front Door
3. The paint samples then need to be written up on the XRF
Run Sheets. Identification number is assigned. The
sample is then given an analysis number by the analyst.
The number given to the sample is used only as a means to
identify a particular sample for analysis. The samples
should be written in consecutive sequence. Example:
The last sample analyzed was number 0439, then the
next paint chips sample should be numbered 0440.
4. SjISjcimen Containers and XRF Sample Cups are to be
prepared before samples can be processed.
a. Label Specimen Containers - Include the date, the
analysis number, and the Samples's Identification
Number.
b. Label XRF Sample Cups - Include, only the analysis
(i.e. cup) number only.
5.• Mortar & Pestle should always be clean.
B - 51
-------
6.
7.
8.
9. ,
10.
Place paint chips into mortar and use the pestle to crush
the sample. Continue to crush the sample until a
homogeneous mixture is attained. Gloves and respirators
must be worn.
Use a spoon or spatula to place the sample into a
corresponding XRF sample cup, then seal the cup with
mylar film and a ring.
Before next sample is crushed, the mortar and pestle
should be wiped clean. Wipe the mortar and pestle with
a clean paper towel, then wash them with distilled water
and dry them with a clean paper towel. This process
should be done after each sample-
Once all samples have completed steps 1 - 7, the samples
are now ready for analysis.
Analyzed sample results are recorded onto XRF Run Sheets
in ppm's.. '
B. UsingElectric Mill Method
1. Paint chips delivered to the lab:.
2. Samples must be logged in on the Lead in Soil Processing;
Sheets. The date, the time, the total number of samples
brought by the collector, and all the information listed
on the sample bag should be written "on this sheet. The
information listed on the sample bag will include the
sample's identification number, the site address and
particular are from which the sample was taken needs to
be written" on the Lead in Soil Processing Sheet.
Example: * -
Date received: 3/30/89
Time: 12:30 p.m..
Total number of Samples Received From: 135 from
Ms. Merrill Brophy
Sample Identification Number: #590316521
Address: 2092 W. Preston Street
Area: Side of Front Door
B - 52
-------
3. The paint chip samples identification numbers are
recorded on. the XRF Run Sheets. The sample is then
assigned an analysis number by the analyst. The number
given to the sample by the analyst is used only as a
means to identify a particular sample for analysis. The
samples should be written in consecutive sequence.
Example:
»
The last sample analyzed was number 0439, then the
paint chip sample should be numbered 0440.
4. Specimen containers and XRF sample cups are to .be
prepared before sample can be processed.
a. Label Specimen containers - Include the date, the
analysis number, and the sample's identification
number.
b. Label XRF Sample Cups - Include analysis number
only.
5. Electric Mill should always be clean.
6. Electrical grinding must always be done under the hood.
Gloves and respirators must be worn.
a. Place paint chip samples into the Electric Mill.
b. Turn Electric Mill on for approximately 3 minutes.
c. Turn grinder off after 3 minutes, wait for the dust
to settle, remove lid and check to see if a
homogeneous mixture was attained.
7. Use a spoon or spatula to place the sample into a
corresponding XRF sample cup, then seal the cup with
mylar film and a ring.
8. Before the next sample can be processed, the Electric
Mill should be cleaned. Wipe the Electric Mill with a
clean paper towel inside and out, dampen another paper
towel and clean the mill very well, and then dry the
Electric Mill with another clean, dry paper towel. This
process should be done between each sample.
9. Once all samples have completed steps 1 - 7, the samples
are now ready for analysis.
10. Analyzed sample results are recorded onto XRF Run Sheets
in ppm' s.
B - 53
-------
QUALITY ASSURANCE PLAN
FOR PAINT CHIP SAMPLES
I. INTRODUCTION
This quality assurance document sets forth the Baltimore
Lead-in-Soil Demonstration Project's (LIS) policies and
procedures that maximize the quality of sample.collection
and laboratory performance. The goal of the sample
collector is to provide a representative sample of the
surface to be tested according to the appropriate
protocol. The goal of the laboratory is to provide a
quality service of elemental analysis.
It is the policy of LIS to maintain an active quality
assurance (QA) program to provide analytical data of
known and supportable quality and ensure a high
professional standard in analytical data generated in
support of projects undertaken for the public by state
and federal agencies.
II. SAMPLE COLLECTION
Paint chip samples are collected by LIS personnel for
this project. All collectors are trained in sampling
procedures.
Paint chip samples for metals analysis are collected and
stored in clean previously unused paper envelopes.
Sample envelopes are labeled with a unique sample
identification number and sample code which reflects the
location of the sample site. A corresponding chain-of-
custody form -is completed at the time of sample
collection.
III. SAMPLE ACCEPTANCE, PRESERVATION, AND STORAGE
All incoming samples are delivered to Maryland's
Department of Health and Mental Hygiene (DHMH) for
analysis. As the samples are accepted, they are assigned
arj^aboratory sample number and the chain-of-custody form
is*dated with the current date.
.v'.,r"
The quantity of sample submitted must be adequate for all
analyses requested.
B - 54
-------
IV. SAMPLE RESULTS
Following DHMH analysis, the results are reported to the
LIS Environmental Coordinator (EC). Excess sample
material and DHMH sample cups are returned to the LIS
project for storage at a secured facility. The LIS EC
reviews the data results and assigns preliminary data
processing work to the environmental health aides I
(EHA), who transfer the sample number and results to data
entry forms. Each set of data entry forms are double
checked by level II EHA personnel.
Upon completion of the data information transfer the data
entry sheets are surrendered to data entry personnel for
double entry into the data base.
V. DATA REVIEW
A review of all raw data and data base paint chip files
is to be conducted at the prior to the end of the study.
B - 55
-------
BALTIMORE SOU. TVRAD ABATEMENT DEMONSTRATION PROJECT
DRINKING WATER SAMPLING PROTOCOL
1. Residents are notified that water must not be turned on prior
to the Environmental Health Aide sampling the system on the
sampling day.
2. Do not shut off water flow valve to the sink fixture (which
would prevent use of the system prior to first draw) as this
may introduce lead corrosion products into the sample.
3. " . Morning first draw is collected from a cold water tap which
had not been used for 8-18 hours. Determine if water was used
prior to sample collection. If water was used, state the use
in the remarks on the sample collection form.
4. Water samples are collected from each household faucet in 250
ml cubitainers.
5. Water samples are preserved on site with 5 ml of nitric acid
per liter.
6. Water tap is closed after filling each sample container to
prevent loss of product and to ensure representative
collections.
7. Keep samples cool (4 degrees C) after collection prior to
analysis.
B - 56
-------
QUALITY ASSURANCE PLAN
FOR DRINKING WATER SAMPLE
I. INTRODUCTION
This quality.assurance document sets forth the Baltimore
Lead-in-Soil Demonstration Project's (LIS) policies and
procedures that maximize the quality of sample collection
and laboratory performance. The goal of the sample
collector is to provide a represenative sample of the
surface to be tested according to the appropriate
protocol. The goal of the laboratory is to provide
quality service of elemental analysis.
It is the policy of LIS to maintain an active quality
assurance (QA) program to provide analytical data of
known and supportable quality and ensure a high
professional standard in analytical . data generated in
support of projects undertaken for the public by state
and federal agencies.
II. SAMPLE COLLECTION
Water samples are collected by LIS for this project. All
collectors are trained in sampling procedures and
approved by the Maryland Department of the Environment' s
Division of Water Supply.
Water samples analyzed for metals are collected and
stored in clean polyethylene or polypropylene cubitainers
with teflon-lined lids. A premeasured vial of nitric
acid is added to the container prior to the water sample
collection. The cubitainers are labeled with a unique
sample identification number and .sample code which
reflects the location of the sample site. A cor-
resposnding chain-of-custody form is completed at the
time of sample collection. Cubitainers are stored in a
small cooler partially filled with ice. Sample con-
taining cubitainers should not be allowed to freeze.
tfe"
III. SAMPLE ACCEPTANCE, PRESERVATION, AND STORAGE
All incoming samples are delivered to Maryland's
Department of Health and Mental Hygiene (DHMH) Water
Laboratory. As the samples are accepted, they are
assigned a laboratory sample number and the chain-of-
custody form is dated with the current date.
B - 57
-------
VI.
To ensure that samples are not degraded and that their
integrity is maintained, all samples for metal analysis
must be received by the laboratory no later than one day
after collection.
The quantity of sample submitted must be adequate for all
analyses requested.
SAMPLE RESULTS
Following DHMH analysis, the results are reported to the
LIS Environmental Coordinator (EC) who reviews trhe data
results and assigns preliminary data processing work to
the Environmental Health Aides I (EHA) who transfers the
sample number and results to data entry forms. Each set
of data entry forms are double checked by level II EHA
personnel. .
Upon completion of the data information transfer the data
entry sheets are surrendered to data processing personnel
for double entry into the data base.
DATA REVIEW
A review of all raw data and data base water files is to
be conducted prior to the end of the study.
B - 58
-------
BALTIMORE SOIL LEAD ABATEMENT DEMONSTRATION PROJECT
HOUSEHOLD DUST PROTOCOL
Household dust sampling should be carried out at the time of
the environmental visit to the home of the study participant.
For this study, the household dust samples are defined as the
samples that represent dust most likely to impact on a child's
hands during indoor activity. This would include dust on window
sills, and furniture, as well as dust on toys and other objects
likely to be handled by children. A minimum of three areas should
be sampled: at the main entrance to the household, and two areas
most frequently used for play activities by the child or children.
Additional areas may be selected that represent: 1) secondary
entrances to the household (back or side doors); 2) sources or
accumulation of dust within the household (paint, rugs, upholstered
furniture); 3) additional play areas or other areas of activity
frequented by the children. • .
The sample has two components that are important to
interpreting lead exposure, the jconcentration of lead in the dust
and the amount of dust, or loading, on the surface. The
concentration of lead in dust appears to be closely related to the
amount of lead on children's hands, whereas the amount of dust on
surfaces is an indicator of the importance of this route of human
exposure. At least 10% of the samples should be over a defined
area to determine the household loading factor.
Sketch the approximate layout of the residence and select to
sampling. Bear in mi'nd that some areas, such as entryway, may
reflect outdoor dust to a greater degree than others.
The sampling apparatus is the Sirchee-Spittler Hand Held Dust
Vacuum unit which is attached to a 'Dustbuster' hand held type
vacuum. Prior to the sample collection the sample collection screen
must be clean.
For some samples, both the weight of the dust and the lead
concentration of the dust will be measured. In this case, it is
necessary to sample a defined area, so that the results may be
expressed in ug Pb/m . Mark the 4' x 4' sample area with tape.
The surface of the sample area is vacuumed with back and forth
strokes about 1—2 inches in width. The vacuum is most efficient if
the head is held parallel to the ground at a 45 degree angle. A
single pass across the surface of the sample area is sufficient to
collect adequate sample amounts. After dust sampling, the vacuum
unit is kept in an upright position until the sample screen is
ready to be removed. Turn the vacuum off and remove the sample
screen. Empty the contents of the sample screen into a labeled-
reinforced paper envelope. Seal the envelope with scotch tape.
B - 59
-------
The sample amount required for analysis is equal to 2 grams of
dust. If the sample amount from the area is not sufficient
additional sample material may be collected from another 4'x 4'
sample area and added to the initial sample.
Record sample data on the appropriate chain of custody form.
Transport the sample to the laboratory in a manner to ensure
upright envelope delivery.
B - 60
-------
QUALITY ASSURANCE PLAN
FOR DUST (VACUUM SAMPLED^
I. INTRODUCTION
This quality assurance document sets forth the Baltimore
Lead-in-Soil Demonstration Project's (LIS) policies and
procedures that maximize the quality of sample collection
and laboratory, performance. The goal of the sample
collector is to provide a representative sample of the
surface to be tested according to the appropriate
protocol. The goal of the laboratory is to provide a
quality service of elemental analysis.
It is the policy of LIS to maintain an active quality
assurance (QA) program to provide analytical data of
known and supportable quality and ensure a high
professional standard in analytical data generated in
support of projects undertaken for the public by state
and federal agencies.
II. SAMPLE COLLECTION
Dust samples are collected by LIS personnel for this
project. All collectors are trained in sampling
procedures.
Dust samples for metal analysis are collected and stored
in clean previously unused paper envelopes. Sample
envelopes are labeled with a unique sample identification
number and sample code which reflects the location of the
sample site. A corresponding chain-of-custody form is
completed at the time of sample collection.
III. SAMPLE ACCEPTANCE, PRESERVATION, AND STORAGE
All incoming samples are delivered to -Maryland's
Department of Health and Mental Hygiene (DHMH) for
analysis. As the samples are accepted, they are assigned
a laboratory sample number and the chain-of-custody form
is? dated with the current date.
The quantity of sample submitted must be adequate for all
analyses requested.
B -61
-------
IV.
V.
SAMPLE RESULTS
Following DHMH analysis, the results are reported to the
LIS Environmental Coordinator (EC) who reviews the data
results and assigns preliminary data processing tasks to
the Environmental Health Aides I (EHA), who transfer to
sample number and results to data entry forms. jEach set
of data entry forms are double checked by level II EHA
personnel.
Upon completion of the data information transfer the data
entry sheets are surrendered to the data entry personnel
for double entry into the data.base.
DATA REVIEW
A review of all raw data and data base dust files is to
be conducted prior to the end of the study.
B - 62
-------
Appendix F
Sample Forms
-------
PRIORITY.
Collector _
STATE OF MARYLAND
DEPARTMENT OF HEALTH AND MENTAL HYGIENE
Laboratoriaa Administration
201 W. Preston St.
P.O. Box 2355, Baltimore, Maryland 21203
J. Mahaan Joaaph. Ph.D., Director
HAZARDOUS WASTE LABORATORY
Metals Analysis Report Form
LAB NO.
Nama/Time/Date
Sample ID No..
Sample Alert _
Specify Program:
Circle Type of Analysis:
„ 1. EP Toxicity
Sample Source,
Preservative Used.
NPDES:.
OTHER:
Chain of Custody Sample Possession:
Prom: To:
Narm/Time/Oata
Nama/Tima/Data
Name/Time/Data
Name/Tima/Date
2. Priority Pollutant
a Total Metals
4. Dissolved Metals
Indicate Type of Sample:
Liquid
Solid.
Percent Solids.
Element
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cooper
Iron
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Zinc
Chromium Cr+6
Metals in ppm
EP
Total
Element
Aluminum
Calcium
Cobalt
Magnesium
Manganese
Potassium
Sodium
Vanadium
EP
Total
Section Chief:,
OHMHJKO" «i
SELECT OTHER ELEMENTS FROM REVERSE SIDE OF THIS FORM
Date: Verified By: Authorized By:
-------
97494
PRESS HARD—BALL POINT PEN ONLY
STATE OF MARYLAND
Bottle
Number:
Source of Sample:
Sample Drinking Wat
Types Landfill
(Circle): Stream
Other
Remarks:
County Plant No.
Field Data:
DEPARTMENT OF HEALTH AND MENTAL HYGIENE
Laboratories Administration
201 W. Preston St.
P.O. Box 2355, Baltimore. Maryland 21203
J. Mehsen Joseph. Ph.D.. Director
Name:
WATER ANALYSIS
Lao NO. uaie neceivea
Do not write above this line.
Data Category Code j j
Countv:
fVilloctnr
Street Town or City (include telephone Number
er Community (Public Treated) Source (Raw Water) Emergency
Non-Community (Pub. Untreated) Distribution (Treated) Routine
Private MCL ' Recheck
Other
Sampling
Station
Chlorine
Residual
are Required
niype or
. . Ac,d:
Date Collected • Time Iced Acid
. ,_. [___
i i !
PH*
Free
Total
Specific Conductance
**
ANALYSIS
PH-
Alkalinity (Total)
• pH*. Ca CO, SAT.
Alkalinity, Ca CO, SAT.
Hardness
Ammonia-N
Nitrate-Nitrate N
Nitrite N
MB AS
Chloride
Fluoride
Color*
Turbidity
Conductance*. SPEC
Sulfate
Total Solids
Dissolved Solids
CODE
00403
00410
70311
74023
00900
00608
00630
00615
38260
00940
00951
00081
00076
00095
00945
00500
70300
RES
3U
t
-TJ
J
J
V*
-
ANALYSIS
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
Aluminum
Calcium
Copper
Iron
Magnesium
. Manganese
Nickel
Potassium
Sodium
Zinc
CODE
01002
01007
01027
01034
01051
71900
01147
01077
01105
00916
01042
01045
00927
01055
01067
00937
00929
01092
"
RESULTS ;
i j i i i
: - i '
! ' • .
I i J : ' :
! .j | .. ' !
!'- ! :
i i ! :
4 ' '
i ! * '•
i j ,
i -: '
! f „
i : .
i
ii,-
i i :
I j "- : •
i ' ! ;
•I . ! . '
!
•Results reported in units, all others in rriilligrams per liter ippm>
-------
LEAD IN SOIL SAMPLE PROCESSING/XMET RUN SHEET
RECEIVED BY: , ANALYSIS DATE:
RECEIVED FROM:
SHEET #
ANALYSIST:
PROPERTY ADDRESS j SAMP ID NUMBER i PPM -1
! ; ! '
1 ' • • i ;
! » i '
• i
3' i
i ' !
i ' ' ;
i ^ '
i i
! • ! !
1 J :
! 3 i
i i
:! i
i :| i
i J 1 .
1 1 !
i i :
! ! !
! ' 1 !
i j
i ' • j i
i '
J i
i! :
1 1 i
• . . • 1
.. • l| '!
• ' 1 ;
j - .
ii i
1 :
i
l - ;
PPM -2 ]
i
li
1
•I
>\
1
a
ii
i
i
•-— j
M
1
•
•
B - 65
-------
COUNTY CODE
STATE OF MARYLAND
DEPARTMENT OF HEALTH AND MENTAL HYGIENE
LABORATORIES ADMINISTRATION
201 W. Preston Street
Baltimore, MD 21203
DUST EXAMINATION FOR LEAD •
• DATE SUBMITTED
SAMPLES COLLECTED BY:
SCHOOL/DAY CARE IDENTIFICATION NUMBER:
SAMPLE TYPE: DAYCARE DUST WIPE SURVEY SCHOOL DUST WIPE SURVEY
PROGRAM: MULTI MEDIA LEAD STUDY MDE STATE CODE; (DC ONLY)
REPORT RESULTS TO:
(NAME) (PHONE NUMBER) .
MDE/TESH 2500 BROENING HWY. BALTIMORE, MD 21224
NO. Sample
Code
Location Area in
(sill, well, Inches
floor) (Ixw)
Laboratory
Results
DO NOT COMPLETE - FOR LAB USE ONLY
SAMPLE RECE|VED BY:
DATE RECEIVED:
REMARKS:
REPORTED
ANALYST
INTERPRETATION OF RESULTS
(Results reported in Micrograms Lead per Square Foot (ug/ft2)
Threshold Limit: Floor 200 Window Sill 500 Window Well 800
B - 66
-------
ENVIRONMENTAL DATA ENTRY FORM
PROP
ID
SAMPLE
NUMBER
SAMPLE
CODE
RESULTS
WGT
MGM
XRF
PPM
WGT
AAS
PPM
FLAG
B - 67
-------
DEPARTMENT OF THE ENVIRONMENT
TESH LEAD LAB
MULTI-MEDIA LEAD STUDY
XRF SAMPLE RECORD
FIELD STAFF SUBMITTING:,
SAMPLE DATE:
PRIORITY: ^_____^
SAMPLE LOCATION:.
SAMPLE TYPE:
SAMPLE ID NUMBER
WET WT
DRY WT
-
SAMP WT
RESULTS
SAMPLE RECEIVED BY:.
DATE RECEIVED:
ANALYST ASSIGNED:
ANALYSIS DATE:;
REVIEWED BY:
RESULTS REPORTED:
B - 68
-------
MARYLAND DEPARTMENT OF THE ENVIRONMENT
TESH/LIS
2500 BROENING HWY
BALTIMORE, MD 21224
CHAIN OF CUSTODY
SAMPLE TYPE
* *
SAMPLE LOCATION
SAMPLE NUMBER
PERSON RELEASING SAMPLES:
PERSON RECEIVING SAMPLES:.
PERSON RELEASING SAMPLES:.
PERSON RECEIVING: SAMPLES*
DATE:.
DATE:
DATE:
DATE:
B - 69
-------
1
E-t
0
H
0
P;
P<
as
O -o
H J
EH •§
»5 2
* » *
CQ 8 "S
o °- fa
33
Q |
H I
o
CQ
S
H
| ,
L.
V
L.
4) u
0) U
4 1
S 3
j Is-1
§<* o g, a
to O C
O ° I o '5>
IV. t» CD t- 0)
IM V £ A 0£
55 5 or £ S
O z x o o
H
O
Pi H
CQ C9
HPJ
EH PS
55 H
S O
Pi O
0
W o
CQ •"
< t;
W tu
*o*
a rt "•
s "
i
H
W
«
fe
O
EH
§
EH
Pi
Pi
Q
Q
^|
^M
x
g
o
0.
0) M
S -
" "T? ^*
8)
4)
0
CO
•o B.C. 7.
1-1 < 4-> 0 0
•^ 4-< 01 <-• *->
r* o ** o o
Q> V 01 . 0) U
••{ — . a> X Q. S.
< 0 U « « g
S a. «/> CJ — —
1
z
S
tn
•B
3
|
«
S
H-
o
0)
01
3
a
at
1
§
V
X
^
^
1
V
a>
2
U
o
"o
1
.. 1
I *
**• fc
o o
o
ft.
—
O) O)
O -Q
a _j
-------
g
3
eg
d*
55
O
H
B
g
E<
CO
25
0 «
•2 B
Q
^
H
O
CO
§ *
§
ta
»3
1
Q.
?5
S
11
P\
CJ ci|
J*4 jp£
^« ^^
5 §
JTJ ST*
M^
P4
*9
1
1
H
i;
ia
I'M
1
n
ft
o
B
H
g4
s
Pi
a
a
*
a
71
en
u.
oe
X
Inspector
g
i
I
0
I
E
g
^M
a
, e
u
u.
oc
X
o
I
t—
fe
•u
5
1
f—
8
(n
O
QL
36 e» N
•n ; °
JE^ 0, og t/> o in
CO
V
ae
L.
2
t.
1
_j
^»
C OJ3
CO
_i
«
AJ
1
itenance
a
X
«J
»—
OT
«
"8-
u
1
•
u -t
t in
< v>
U —1
_1 III
< tn
1
4J
C
•*•
1
CO
V)
V
«^
X
o
g
.
M
c
u
Inspector
g
4-1
a
-j
01
1
f/9
1
C_
0»
CO
Q
o
ac
0)
D>
ID
a.
§
ty>
(U
>-
0
z
4J
to
£
4J
o:
*-«.
g)
tu
>> "5
b o
c
uj Jr
!_i t
X | ~
1 tB
«t Q*
X **
\ s
-------
u
h>
o
OS
§
H
(H
S
£4
O
w
Q
H
O
to
H
O
§
J
cu
t_
•
4J o
eg 4-1
° I
3
0
IV. *+.
" O
a H
o o
E< ft
CO
_c
g
a.
I
i
ri «>
W^t O)
|Cu CD
ta rt
55 Q
H
ft
s
^4
H
§
•8
H
1
H
H
«
H
IV.
Ft
0
EH
S5
H
«
<
ft
H
a
Q
2
1
^->
c.
"B
•^
•g
a
i
o
O)
OJ
a.
o>
.0
.^
c
o
4^'
I
a
u
u.
s
1
^~
o>
H 4J CO C
7* o z *< •—
f^ aj 01 ^
5 •" e S a
rt! o S t- e
3 c. o 4-1
c
1*
c
a
X
J
u
cou
UCJ
«
c
Q>
CO
0
u
S.
CO
ID
X
u
S
^
f
«
•5
I
J
K>
IM
«
CJ
"i-
a
CO
O -J
—1 Ul
< CO
*
CJ — 1
—1 Ul
< CO
%
CJ -1
*
CJ — 1
_1 Ul
< CO
*
.
CJ -1
— 1 Ul
< CO
*
CJ — 1
-J Ul
< CO
It
CJ _l
—1 Ul
< CO
*
CJ _l
— 1 Ul
< CO
«
U _l
-1 Ul
< CO
»
CJ _l
-J Ul
< CO
%
CJ —I
—1 IU
<
-------
CORRECTIVE ACTION DOCUMENTATION FORM
?TQN QF PROBLEM and when identified: ______
State cause of problem if known or suspected:
SEQUENCE OF CORRECTIVE ACTION: (If no responsible person is
identified, bring this form directly to the QA Coordinator)
State date, person, and action planned:
CA Initially Approved By:
Date: • ^
Follow-up dates:
Description of follow-up:
Final CA Approved By:
Date:
-------
Appendix G
Glossaries
Units of Measure and Abbreviations
-------
GLOSSARY OF UNITS OF MEASURE
C — Celsius
M - Mole, a chemical unit of measure
megaohms - Ten6 ohms
mg/1 - milligrams per liter
ml.- milliliter
ppm - Parts Per Million
B - 74
-------
GLOSSARY OF ABBREVIATIONS
AA - Atomic Adsorption
CDC - Centers for Disease Control
EC - Environmental Coordinator
EDL - Electrodeless Discharge Lamp
EHA - Environmental Health Aid
EPA - Environmental Protection Agency
EPTOX - EP Toxicity Testing
HCL - Hollow Cathode Lamp
LBP - Lead Based Paint
LIS - Lead-in-Soil Program
MCL - Maximum Contaminant Level
MDE - Maryland Department of the Environment
MEAL - Multi-Element Analysis Laboratory
NBS - National Bureau of Standards
Pb - Lead
QA - Quality Assurance
QAP - Quality Assurance Plan
QC - Quality Control
RPD - Relative Percent Differance
S - Classification of Weights
TCLP - Toxicity Characteristics Leaching Procedure
XRF - X-Ray Fluorescence
B - 75
-------
-------
Appendix H
Method Detection Limits
-------
DETECTION LIMITS
Blood Lead
FEP
Ferritin
Soil - XRF
Dust - XRF
Dust - AAS
Dust - Handwipes
Paint - XRF
1.6 jug/dl
6 /ng/dl
1.2 ng/dl
Calibrated
4,000 ppm
Calibrated
for 78
for 78
to
to
12,000 ppm
1,600 ppm (if sample
quantity is adequate)
1,8 /xg
Calibrated for 1,000 to
18,000 ppm.
B - 76
-------
Appendix I
Flow Chart of Data Handling
-------
FLOW CHART OF DATA HANDLING
1AB REPORT RECEIVED BY
ENVIRONMENTAL
COORDINATOR
ERRORS BACK
TOLAS
TRANSCRIBED BY
ENVIRONMENTAL HEALTH
, • AIDEI
ERRORS BACK
TO LAB
REVIEWED BY
ENVIRONMENTAL HEALTH
AIDE ITS
ERRORS BACK
TO HEALTH AIDETS
RANDOM SPOT CHECKS BY
ENVIRONMENTAL
COORDINATOR
ERRORS BACK
TOLAS
DATA ENTERED BY
JOANNE SMITH
MARKSCHERER
ERRORS BACK TO
ENVIRONMENTAL
COORDINATOR
DATA CHECKED BY
URSULA PARKER
ERRORS CORRECTED
FROM ORIGINAL
SAMPLING RESULTS
DATA PROVIDED TO
PRO JECT MANAGER
B - 77
-------
Appendix J
Environmental Sampling Statistics
-------
Table 2.9.1
Sampling Summary Table
ENVIRONMENTAL STATISTICS
LOCATION: AREA 1
Number of Observations
Sample Mean
Maximum Value
Minimum Value
Sample Median
Upper Quartile
Lower Quartile
Sample Unit of Measure
SOIL
1339
550
6800
22
348
607
206
ppra
WATER
294
9.9
420
0
2.2
6.1
.9
ppb
'PAINT
485
4.9
37.1
0
3.0
6.88
.91
ppm
DUST
333
1068
22600
2
418
935
180
ppra
POST
SOIL
153
60
619
1
22
54
13
ppra
POST
DUST
118
754
7300
13
430
929
235
ppm
ENVIRONMENTAL STATISTICS
LOCATION: AREA 2
•
Number of Observations
Sample Mean
Maximum Value
Minimum Value
Sample Median
Upper Quartile
Lower Quartile
Sample Unit of Measure
SOIL
826
596
7500
39
409
693
243
ppm
WATER
252
6.6
103
0
1.9
4.9
.4
ppb
PAINT
373
5.4
70
.02
2.4
7.11
.76
ppm
DUST
297
1077
21200
1
436
1100
187
ppm
POST
SOIL
N/A
N/A
N/A
N/A
N/A
N/A
N/A
ppm
POST
DUST
59
954
7000
5
438
926
159
ppm
-------
Appendix K
LIS Demonstration Project Personnel Flow Chart
-------
I
z
UJ
I
m
Z
o
ec
>
HI
III
I'-
LL
O
Z
LU
1 > > Z
i - , t
d? '-i
d 21 g. ^
< S 2 *^
": x i a
0. O 0 j
in -, ui -
Z 7 L*^
0 z
= <
m ?
uj o
2 E
Q.
I
LLJ
z
s
-3
|
Uj
5
UJ
CO
__
sii
2 Z Q
Sl8
Q 5
§ si
!K§|
z2w O
SS|£
Isil
03 D
LLJ Z
Q ^
si
D O
5Q
m
/
h
n 4i
-DATAPRI
1-DATA
.T-
J
^ O
Zui
"i
a
!O O c
j||J Z f
•a: ui
i
-------
Appendix L
LIS Demonstration Project Time Line
-------
U!
3
O>
3
en
CTv
U
u
u-
o
u
to
2
HI
o
z
O
U
c:
o
CTl
OS
cn
CO
CO
cr»
(0
0)
C
o
u
•o
0)
JJ
'10
II
I
O
M '
J
O <
CO CO
> J
CO
*&
CO £
3 <
Q CO
M
OS ,-q
u a,
EH S
U C5
l-l Z
CD CO
< EH
Q< U)
z
u
M <
o «
en «;
-------
Appendix M
Equipment List
-------
ENVIRONMENTAL SURVEY CHECKLLIST
EQUIPMENT REQUIRED
CODE MANUAL
747-XRF ANALYZER
RING BADGE
TAPE MEASURE
ARCHITECT'S TEMPLATE
CLIP BOARD
LBP-FORMS
PERMANENT LAB MARKER
VACUUM
SIRCHEE ATTACHMENT
FILTERS
SAMPLE COLLECTION TUBES
LAB POLICEMAN
4'X 4' TEMPLATE
PAINT SCRAPPER
SAMPLE COLLECTION ENVELOPES
Y/N
COMMENTS/ CONDITION
-
B - 81
-------
EQUIPMENT REQUIRED I Y/N 1 COMMENTS/CONDITION
SOIL PROBE
PROBE CLEANING TOOL
SAMPLE COLLECTION BAGS
B - 82
-------
LEAD IN SOIL PKOJECT
EXTERNAL QUALITY ASSURANCE/QUALITY CONTROL
QUALITY ASSURANCE FOR BLOOD LEAD ANALYSIS (CENTERS FOR DISEASE CONTROL)
A. INTRODUCTION
The lead in soil demonstration project of its nature requires blood lead data of the highest quality.
Expected differences in blood lead levels from successful abatement are of the order of 2-4 lAg/dL, thus
placing unusually stringent requirements on long term laboratory precision. The quality control issues
including establishment and maintenance of a high degree of precision over the entire duration of the
project. The key. function of-the quality assurance system is to ensure the absence of any "drift"
(downward or upward) with analytical values with time, such that any difference in blood lead values
over time cannot be attributed to with time, such that any difference in blood lead values over time
cannot be attributed to changes in the analytical system. Simply stated, this will help insure that
statistically speaking, observed changes in blood lead are real- that is, due to intervention and not
attributable to changes in the laboratory method over time. Since the CDC has extensive experience
in such activities from the National Health and Nutrition Examination Survey (NHANES) and other
long term studies, we were asked by the USEPA to provide assistance. The material following is a
summary of laboratory related issues that were included in the overall QC program.
B. ELEMENTS OF A QUALITY CONTROL SYSTEM
In order for any analytical measurements to be valid and interpretable, the sources of error for each
unique measurement system must be identified and minimized. This, then, is the major function of
quality control. In the specific example of blood lead measurements, the following have been shown
from experience to be the major sources of error:
1) contamination of the specimen during collection, storage, or analysis
2) deterioration of the specimen by clotting, denaturation, or other processes
3) instability of the measurement system, either over a short (within run/day) or
long time span
4) improper calibration for the measurement system
5} errors in data handling, storage, or reporting
Quality control therefore must include a number of components, both within and external to the
laboratory: 1) collection of an uncontaminated specimen; 2) preservation and shipping (if needed) of
the specimen under conditions that assure integrity, 3) monitoring of analytical method performance, to
include instrumental stability,-maintenance, and performance of the analyst(s); and 4) accuracy and
completeness of all data, to include specimen identification, data reduction, and data interpretation.
Some critical components of each of these areas include:
B-83
-------
1. Specimen Collection
Proper screening of all specimen collection equipment to define any detectable levels of the analyte,
and estimate variability of this contamination.
Written protocols for specimen collection which describe in detail all sampling equipment and its use,
precautions to avoid contamination, and other requirements (time of day, fasting/non-fasting state of
subject) which might- affect specimen integrity.
2. Specimen Preservation and Shipping
Proper packing, storage and shipping temperatures, suggested means of conveyance for timely receipt
of specimens.
Detailed shipping and specimen log forms to allow description of each specimen to record any variances
from collection or shipping protocols.
3. A"»lytical method Performance
Method selected must demonstrate precision and accuracy in the appropriate analytical range and
should be simple, rugged, rapid and cost-effective. Ideally, the detection limit should be ca. 2 ng/dL
with precision about 5 % at the 10 mg/dL level for the proposed study.
Instrumental stability, and by inference "method" stability, should be documented by analysis of control
materials, both "bench" and "blind". It is desirable that materials with certified values of the analyte of
interest be an analyzed regularly to demonstrate method accuracy. It is suggested that at least 1 10%
of the specimens be quality control pools.
4. Bench and Bwnd Quality Control
Blind quality control pools should be inserted at a rate of 5% by a source external to the laboratory.
These specimens should be in the same container type and labelled with pseudopatient numbers such
that they are indistinguishable from patient samples. It is suggested that the blind (and bench) pools
have two concentrations- one in the "expected" range of values for the majority of patient samples and
one at or near the "decision level" for undue exposure. It is important that the blind materials be truly
blind to the analyst for ma-mmnm effectiveness in the detection of analytical system error. The
"pseudopatient" numbers used in labelling of the blinds will be decoded by the supervisor only, and that
analytical run evaluated on the basis of pre-established control limits.
Use of quality control charts for means (X bar) and ranges (R) is essential; it is suggested that 20 runs
be made for characterization of all quality, control materials, and that these data be analyzed by two-
way analysis of variance (ANOVA) to produce these charts. These charts should be in use by the
analyst for each run for the evaluation of "bench" or known blood controls (and by the supervisor for
blinds) by use of mean and range control limits, such that corrective actions needed may be made in a
timely way.
Criteria for repeat analytical runs (due to "out of control" condition as indicated by results from quality
control samples) are dependent on the number of pools in the quality control system.
Inclusion of blind splits (duplicate samples within run, with different identification number such that
identification by the analyst is prevented) is suggested at a 5% rate; some split specimens may be
submitted to an external laboratory for verification of accuracy or comparability. If specimen collection
B-84
-------
constraints allow, it is recommended that at least 10% of the specimens be split with an external
laboratory.
Criteria should be established as to "acceptable" agreement with the external laboratory.
5. Accuracy and Blanks
Blanks, consisting of samples in which ultrapure water is processed through the entire analytical
procedure, are a useful part of quality assurance. The data from these determinations can be used to
evaluate potential contamination in the laboratory environment as well as estimate the limit of
detection to the analytical method.
Establishment of accuracy through the regular analysis of reference materials or proficiency testing
pools is an essential part of good laboratory practice, and .will help establish the accuracy of the
method. The pools used for this accuracy assessment should be as close to identical to the survey
samples as possible.
6, Data Integrity
Data logging should be performed for each run in approved notebooks or other data forms as soon as
possible following each run. Electronic data entry may be desirable either as an adjunct to or
replacement for "hard copy". It is recommended, however, that instrumental data be collected on hard
copy in such a way that all data can be independently verified or reconstructed.
Data reduction should be standardized; all records of calculations should be secured and available for
review.
C. DESCKgTION OF QUALITY CONTROL SYSTEM USED
From previous experience in "long-term" quality control, a system was established that is similar to that
used in the NHANES surveys. The cardinal features of such a system include written protocols for
specimen collection, shipping, and analysis, a systematic screening of all specimen collection equipment
and containers, establishment of statistical control limits by each individual laboratory, and supervision
of all QC activities by a local laboratory supervisor. Since the three laboratories already had QC
systems in place, there was a need to establish a common set of protocols and procedures for the entire
project.
1. Initial Activities
Each laboratory was provided with a description of the sample collection and shipping protocols
developed at CDC (1), as well as a reprint of our analytical method for blood lead (Appendix A).
Summary descriptions of the QC system used in NHANES, as well as general descriptions of the
NHANES quality control system were distributed (Appendix B,C).
Four whole bovine blood pools were collected at CDC, evaluated for lead content, and aliquoted into 2
mTf Vacutainer brand whole blood collection containers (blind pools) or plastic screw-capped vials
(bench pools). The Vacutainer specimen containers (as well as the plastic vials for the bench controls)
were screened by established protocol (1), and had been purchased in sufficient quantity to allow all
thee projects to use them as standard specimen containers. Pools such as these (whole bovine blood,
stabilized with 1.5 mg/mL disodium EDTA) have been shown to be stable at least two years at 4 C, the
recommended storage temperature. Data from this screening are presented in Table 1. Aliquots of
these four-pools were distributed to the laboratories, and duplicate analysis of the four pools was
B-85
-------
performed over a series of twenty analytical runs. The data generated from these analyses were used
to calculate the QC limits for both means (X bar) and ranges of duplicate measurements of these pools.
The method of calculation is presented in Appendix D, using POOL "A" from Standard Reference
Materials (SRM) 955 from the National Institutes of Standards and Technology (NIST). The
calculations are based on two-way analysis of variance (ANOVA) as described by Shewhart (2,3).
Results of the calculated limits for these four pools were sent to the three laboratories to be used as
part of the laboratory quality control program. Results of the calculation for the three laboratories, as
well as GDC, are presented in Table 2. The quality control limits could then be used in two ways:
1) the limits for the "blind" pools were used to evaluate the blind quality control pools,
which were inserted into each analytical run by the supervisor; and
2) the limits for the "bench" pools could be used by the analyst (along with those for any
additional pools) to evaluate the degree, of statistical control for the analysis.
Insertion of the "blind" pools was random, using a random numbering table numbering scheme
presented by Taylor (4), with identical labels as study subject specimens and identical Vacutainers (2
mL liquid EDTA, lot # 8E014 EXP 5/90). An example of the labelling system is given in Table 3. If
names were provided on the sample labels, then fictitious names were provided for the "blinds" by the
supervisor. The source if banes could be random names from a metro phone book, or.any other
appropriate source.
2. Calibration
Since three different analytical methods were used hi the study, the issue of calibration of the
analytical systems was very important. The CDC recommendation to all three laboratories was that
either SRM 3128 (from NIST) or equivalent aqueous standards for lead be used. In the case of the
graphite furnace AAS methods (Boston and Baltimore), a version of the CDC published method was
used for analysis, which includes "matrix matched" standards and lead nitrate aqueous standards. The
DPASV method used by Cincinnati. (5) includes standards analyzed by isotope dilution mass-
spectroscopy (IDMS). In all three laboratories, the ultimate test of the accuracy of calibrations
generation of accurate values for reference materials. As can be seen from Table 2, all three
laboratories agreed well (within 5%) with each other, and generated comparable results on the four
pools provided by CDC (Figure 1).
3. Interpretation of Data
The quality control system outlined here has multiple uses:
1) evaluation of "day-today" statistical control of the analytical system;
2) verification of analytical performance on "blinds" - known samples inserted in each
analytical run to verify precision
3) evaluation of any "trends" hi the analytical performance of the method over time- either
short term (days/week) or long term (months^years)
With the use of common rules for the verification of statistical control (4), all the laboratories would
follow a statistically valid and proven method for data evaluation. Any problems not resolved at the
local level were presented to CDC for resolution.
B-86
-------
D. RESULTS OF QUALITY CONTROL SYSTEMS
Data from the initial characterization for the four whole blood pools used in this project are presented
in Table 2. Each laboratory can be individually compared as to within-run precision, among runs
precision, and total precision. Using the definition of the limit of detection as 3 SD(wr) developed by
Winefordner (6), the laboratory detection limits may also be compared.
Of equal importance are the long-term quality control data, especially in terms of time trends. The .
Shewhart plots for the three laboratories are presented in Figure 2. As can can be readily seen, no
long-term trends in analytical values with time are evident. Statistical tests of the null hypothesis
(that is, a "0" slope of X bar versus time) reveled no statistically significant trends with time.
The conclusions that can be drawn from these three systems are as follows:
1) comparable values were obtained on common quality control materials, which covered
the analytical concentration range of interest;
2) laboratory data for blood lead were produced from analytical systems in statistical
control (as defined by Shewhart); and
3) no statistically significant time trends were observed in the data- that is , the
difference in pre- and post abatement blood lead values are real and not the product of
unstable analytical systems.
B-87
-------
Table 1 Data from Lead Screening
2 mL Vacutainers (B D lot #8E016 Exp. 5/90) Catalog # 6384
Analytical result N = 42 tubes;
soaked overnight (12 hr in 1% v/v nitric acid)
X = 0.0964 ug/dL lead (SD = 0596 ug/dL CV = 62%)
3 mL plastic vials (linear polyethylene) Falcon Catalog #
Analytical result N = 42 tubes;
soaked overnight (12 hr in 1% v/v nitric acid)
X = 0.51 ng/mL Equivalent to 0.025 mg/dL (SD = 0.36 ng/mL CV = 71%)
Capillary Collection Butterflies BD Catalog # 7251: 7253)
Analytical results; One mL 1% v/v nitric acid passed through each collector)
N = 5 results collectors each size
X = <0.1 ng/mL (cat 7251)
X = <0.1 ng/mL (cat 7253)
B-88
-------
Table 2 quality Contol Limits- Means and Ranges
LAB
CDC
MD
CN
BOS
POOL
BLIND1
BENCH2
BENCH1
BLIND2
BLIND1
BENCH2
BBNCH1
BLMD2
BLEND1
BENCH2
BENCH1
BLIND2
BLEND1
BENCH2
BENCH1
BLIND2
MEAN
4.6
43.5
1.8
10.7
5.1
45.7.
2.0
11.1
3.5
43.3
2.4
8.9
4.0
47.0
0.2
10.6
95%
MEAN
3.0-6.2
38.2-48.8
1.0-2.5
8.5-12.9
4.2-5.9
43.9-47.6
1.45-2.63
9.6-12.6
1.9-5.1
40.5-46.1
0.9-4.0
7.1-10.7
2.4-5.6
42.9-51.2
-1.2-1.5
8.6-12.5
Conf T limits
RANGE
1.6
2.2
1.4
1.4
0.9
1.1
-0.6
1.0
3.2
2.2
2.0
3.1
0.8
2.9
0.8
1.3
99%
MEAN
2.5-6.7
36.5-50.5
0.8-2.7
7.8-13.6
4.0-6.2
43.3-48.2
1.27-2.8
9.2-13.1
1.4-5.6
39.6-46.9
0.4-4.5
6.5-11.2
1.9-6.1
41.6-52.6
-1.6-1.9
8.0-13.1
Conf Limits
RA*
2.1
2.9
1.9
1.8
1.1
1.5
0.8
1.4
4.2
2.9
2.6
4.1
1.0
3.8
1.1
1.7
B-89
-------
Table 3 Labelling System for Blinds
BUN NUMBER
1
2
3
4
5
6
' 7
8
9
10
11
12
13
14
15
16
17
18 .
SPECIMEN NUMBERS
1-25
26-50
51-75
76-100
101-125 .
126-150
151-175
176-200
201-225
226-250
251-275
276-300
301-325
326-350
351-375
376-400
401-425
426-450
BLIND SPECIMENS (L)
OR(H)
10 (L) 15(H)
26(L) 50(L)
51(H) 52(L)
84(L) 96(H)
107(H) 118(L)
136(L) 137(H)
158(L) 159(H)
185(H) 195CH)
204(L) 214(L)
232(L) 239(11)
264(L) 266(L)
286(L) 298(L)
301(H) 317(H)
328(L) 348(L)
374(L) 359(H)
394(L) 399(H)
404CL) 417(H)
427(H) 431CH)
B-90
-------
E, REFERENCES:
1) "Lake Couer d' Alene Idaho Cadmium and Lead Study-Specimen Collection and
Shipping Protocol" Division of Environmental Health Laboratory Sciences, Center for
Environmental Health, Centers for Disease Control, Atlanta, GA 30333 8/6/86.
2) "Determination of Lead in Blood Using Electrothermal Atomization Atomic Absorption
Spectrophotometry with a L'vov Platform and Matrix Modifier", D.T. Miller, D.C.
Paschal, E.W. Gunter, P.E. Stroud, and J. D' Angelo, Analyst, 112, pp 1701-4 (1987).
3) "A Multi-Rule Shewhart Chart for Quality Control in Clinical Chemistry", J.O
Westgaard, P.L. Barry, and M. R. Hunt, Clinical Cbpmiat.ryr 27, pp. 493-501 (1981).
4) "A Quality Assurance Program for Health and Environmental Chemistry", M.A.
Gaultier and E.S. Gladney, Amproan Laboratory, pp. 17-22, July 1987.
5) "Elements of Sequential Analysis'' Chapter in Biostatistics, A.E. Lewis, ed. Reinhold
Publishing Corp., New York, 1966.
6) "Quality Assurance of Chemical Measurements^, John Taylor, Lewis Publishers,
Cheslea, MI 1987.
7) "Anodic Stripping Voltammetry Procedure Modified for Improved Accuracy of Blood
Lead Analysis", S.M. Roda, RJX Greenland, ILL. Bornschein, and P.B. Hammond,
Clinical Chemistry. 34. pp. 563-7 (1988).
8) "Limit of Detection- A Closer Look at the IDPAC Definition", G.L. Long and J.D.
Winefordner, Analytical Chemistry, 55, pp 712A.-724A, (1983)
B-91
-------
LEAD IN SOIL PROJECT
QA/QC for Seal, Dust, and Handwipes (US BPA/EMSL/LV)
Section 1
Preparation Laboratory Operations
1.1 Sample Receipt . .
Three cities are involved in the superfund Lead Abatement Program; Baltimore, Boston, and
Cincinnati. EMSL-LV supplies the field samplers in each city with 30-gallon plastic barrels for soil
samples and 1-gallon metal containers for interior dust samples. A minimum of two soil and two dust
samples are collected in each city and shipped to EMSL-LV. 'The preparation laboratory manager
records the arrival date of all samples received.
1.2 Sample Labeling
1.2.1 Soil and Dust
Each soil sample is labeled and identified by a unique sample code as described below.
A EOS H 01 001 (example)
digits 1 234 5 67 8-10
Digits Representation
1 Sample type - "A" = audit "C" = calibration
2-4 City code - "BOS", "BAL", "GIN"
5 Concentration - "H" = high, "M" = medium, "L" = low
6-7 2 kg sample - represents number of the 2 kg container in which soil was subsampled. If sample
is dust the number would represent the lOOg container.
8-10 20 g aliquot - numbered aliquot from soil 2 kg container or 2g aliquot from dust 100 g container.
Analytical laboratories at each city provide sample labels and containers to be used for that city. Prior
to shipping, the EMSL labels are removed and the city labels are affixed to the sample containers.
Also, the EMSL-LV codes and corresponding city codes are recorded in a log book for each sample.
1.2.2 Handwipes
Each handwipe sample is labeled and identified by a unique sample code as described below.
A EOS H QQ1 (example)
digits 123456-8
Digits Representation
1 Sample type - "A" = audit "C" = calibration
2-4 City code - "BOS", "BAL", "CIN"
5 Concentration - "H" = high, "M" = medium, "L" = low
6-8 Internal ID - the last three numbers of the internal LESC ID.
B-92
-------
Analytical laboratories at each city provide sample labels and containers to be used for that city. Prior
to shipping, the EMSL labels are removed and the city labels are affixed to the sample containers.
Also, the EMSL-LV codes and corresponding city codes are recorded in a log book for each sample.
1.8 Sample Tracking
1.3.1 Soil and Dust
The preparation laboratory manager tracks each sample as it progresses through the preparation
procedures and records progress in a logbook.
The following information is recorded on a daily basis.
Sample Type - soil, interior dust
City - Boston, Baltimore, Cincinnati
Concentration - high, medium, low
Dried - whether sample has been dried (yes/no)
Crushed - whether sample has been crushed (yes/no)
Bulk homogenization - Whether bulk sample has been homogenized (yes/no)
Pulverized * whether sample baa been pulverized (yes/no)
2 kg split - Whether bulk sample has been split into 2 Kg samples. If this step is partially
complete, the number of aliquots prepared will be recorded.
100 g split - Whether 2 kg soil aliquots have been split into 100 g aliquots or whether the bulk
dust sample have been split into 100 g aliquots. If this step is partially complete, the number
of aliquots prepared will be recorded.
20 g split - Whether 100 g soil aliquots have been split into 20 g aliquots or 100 g dust aliquots
have been split into 2 g aliquots. If this step is partially complete, the number of aliquots
prepared will be recorded. •
The appropriate types of information will be made available for dust and handwipe samples. As
aliquots are sent to analytical laboratories, this information will also be recorded (see sample
shipment).
1.3.2 Handwipes
The appropriate types of information will be made available for dust and handwipe samples. As
aliquots are sent to analytical laboratories, this information will also be recorded (see sample
shipment).
. Sample Type - handwipe
. City - Boston, Baltimore, Cincinnati
B-93
-------
. Concentration - high, medium, low
. Spiked - whether sample has been spiked (yes/no)
The appropriate types of information will be made available for dust and handwipe samples. As
^aliquots are sent to analytical laboratories, this information will also be recorded (see sample
shipment) .-
1.4 Sample Custody
Custody is transferred from the field samplers to the preparation laboratory manager when the
samples are received. The sample remain in the custody of the preparation laboratory manager until
they are shipped to the analytical laboratories.
1.5 Sample Storage
All samples are placed in cold storage upon receipt until there is room for them in the drying room.
After air drying, the samples are returned to cold storage until processing.
1.6 Sample Shipment
As samples are shipped a shipping form (Figure 1.1) is sent to both the laboratory manager and QA
manager. The form sent to the laboratory manager contains only the types and numbers of samples
sent and the city sample code information for each sample. The form sent to the QA manager contains
information as well as the EMSL sample code, which identifies the concentrations of each sample.
B-94
-------
LEAD ABATEMENT QA SAMPLE SHIPMENT FORM
LAB SAMPLE TYPE
BATCH
DATE SHIPPED
NO OF SAMPLES
Sample
Number
1
2
3
4 •
6
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20 '
21
22
23
24
Ciiy Sample Code
•
-
EMSL Sample Code
-
.
FIGURE 1.1 LEAD ABATEMENT SAMPLE SHIPMENT
FORM
B-95
-------
Section 2
Soil Audit Sample Preparation Procedures
2.1 Overview
Specific areas of the preparation laboratory are designated for sample processing. Sample integrity
during processing is ensured by: (1) the use of detailed sample labels, (2) documenting the status of
each sample during the processing, (3) following the preparation documenting the status of each sample
during each processing step.
Bulk soil samples are processed as outlined in Figure 2.1 Each step is detailed in sections 2.2 -2.8.
SOIL SAMPLE
DRY
SIEVE . ; 1 DISCARD \
<20MM i > 20MM '.
FRACTION
-
CRUSH
<20MM
FRACTION
•
PULVERIZE SAMPLE
TO 0.25MM
HOMOGENIZED &
SUBSAMPLE 2000
GRAM AUQUOTS
HOMOGENIZE &
SUBSAMPLE100
GRAM AUQUOTS
HOMOGENIZE &
SUBSAMPLE 20
GRAM AUQUOTS
BATCHING &
SHIPMENT
FIGURE 2.1 SOIL AUDIT SAMPLE PREPARATION FUJW
-------
2.2 Sample drying
2.2.1 Summary
Samples tables constructed of PVC and heavy nylon mesh are used to air dry the samples. Use of the
mesh enhances air circulation and increases the rate of sample drying. These tables are located in a
dust free drying room.
Chemicals as well as food, drinks and smoking are prohibited in the drying area. A separate pair of
gloves is worn when handling each sample. Care is exercised during the cleaning operation to avoid
contamination of samples. Only one sample at a time is dried to avoid cross contamination. Weekly
vacuuming or sweeping is performed to clean the floors of the drying room. Sweep EZ, a sweeping
compound) is used at least once a week to control dust accumulation in the drying area.,
2.2.2 Equipment:
Drying tables with nylon mesh surface
Exalt paper, 86 inch wide rolls
Rubber gloves, unpowdered
25.3 Procedure
Label a bulk sample processing data form for each sample to be air dried. Place two fresh sheets of
kraft paper, approximately 1 square meter in area, on the drying table. Wearing gloves, slowly spread
the sample on top of the paper, firing care not to lose any soil off the paper or contaminate any
adjacent samples. Disaggregate any large peds. Soils high in clay may harden nearly irreversibly if
allowed to dry without a preliminary disaggregation of medium and coarse peds. Place an additional
sheet of kraft paper loosely over the sample. Daily stir the soil sample to facilitate drying. During the
first few days replace the bottom sheet of paper in order to alleviate excessive moisture accumulation.
Note any observations of fungal or algal growth on the data form.
Allow the sample to air dry for a minimum of four days. Prior experience indicates that samples dry to
a constant moisture content (1-2.5%) within three days at the EMSL-LV preparation laboratory.
25.4 Quality Control
When samples are received, labels are checked and recorded. Wearing gloves, the samples are spread
out on kraft paper, which is an effective barrier separating the samples from the PVC mesh tables. A
cover sheet of kraft paper is used to reduce potential contamination. -When handling the samples,
gloves are always worn.
2.3 Tnitial Disaggregation and Sieving
2.3.1 Summary
When a bulk soil sample is air dry, it is disaggregated and sieved in order to remove large rock
fragments and to prepare the sample for crushing, pulverization, homogenization and subsampling.
This procedure is accomplished in two steps: (1) disaggregation and sieving through a 20-mm sieve
and, (2) onioning, pulverizing, and sieving through a 2-mm sieve.
2.35 Equipment:
Pumehood
Kraft paper
Plastic bags
Respirator
rolling pin
Rubber stopper
Tyvek suit
B - 97
-------
2mm sieve
20mm sieve
2.3.3 Procedure
Place a 1 m2 sheet of kraft paper on the sieving table under a vented fumehood. Place a 60 cm2 sheet
of kraft papers on the larger piece of paper and spread a portion of one bulk sample within the
confines of the 60 cm2 sheet. Carefully examine the nature of the rock fragments within the sample
and determine the amount of pressure necessary in order to disaggregate the soil peds without
fracturing or crushing the fragments. Place another 60 cm2 sheet of kraft paper over the sample and
gently roll the rolling pin across the sample. Enough force should be applied to break up the peds, but
not so much that weathered rock fragments are crushed. Place this crushed sample in the 20-mm
mesh sieve and push the soil through the sieve with a rubber stopper onto the kraft paper. Attempt
to include any soil adhering to rock fragments. Place the sieved material in a clean container and
repeat the process until all of the soil of from one bulk sample is sieved. All rock fragments and other
material larger than 20-mm is placed in a plastic bag and properly discarded.
Crush the minus 20-mm fraction (The Crushing procedure is described in section 2.4) then passed
through a 2-mm sieve using .the procedure described above.
The sieves are cleaned after each sample by tapping the sieve on a hard surface and brushing out the
sieve to expunge any remaining soil particles.
2.3.4 Quality Control
The disaggregation and sieving areas should be covered with kraft paper and cleaned after each sample
has been sieved. When sieving, gloves must be worn, as well as an appropriate mask and protective
clothing. The laboratory manager will frequently check the sieving operation for proper equipment and
for adherence to protocol A member of the EMSL-LV QA staff will visit the preparation laboratory to
ensure adherence to protocol
2.4 Crushing
2.4.1 Summary
After soils are sieved through the 20-mm sieve, the < 20-mm material is passed through a rock
crusher. The intent of .crushing is to further reduce the particle size to 2mm.
2.4.2 Equipment
Brush , .
Compressed air
Crusher .
Gloves
Mask
Protective Clothing
Plastic bags
Scoop
2-mm sieve
2.4.3 Procedure
With a scoop, place a portion of the mimiam soil fraction to the crusher opening. Turn the crusher on.
The crusher deposits the resulting crushed material into a collection bin at the bottom of the machine.
After the first scoop is crushed, shut the machine off and sieve the crushed material through the 2-mm
sieve (described in Section 2.4). If all the material passes through this sieve, the crushing plates are
sufficiently close enough to continue processing. If not, adjust the plates and repeat the procedure on
the same sample until all the material passes through the 2-mm sieve. Once the collection bin is full
turn the machine off and deposit the material into a clean labeled plastic bag. Repeat the operation
' .. B-98
-------
until all soil from one bulk sample is crushed. Thoroughly clean the machine with compressed air and
a brush between samples.
2.4.4 Quality Control
When crushing, gloves must be worn, as well as a mask and protective clothing. The machine opening
should be tightly fastened to minimize dust. The laboratory manager will frequently check processing
equipment for proper operations, for adherence to protocol including proper maintenance. A member
of the EMSL QA staff will visit the preparation laboratory to ensure adherence to protocol
2.5 Pulverizing
2.5.1 Summary
The routine soil samples that are analyzed by the cities are ground to a particle size of less than
0.25mm. Therefore, it is necessary to provide audit materials with the same particle size fraction. The
preparation laboratory pulverizes the minus 2-mm soil fraction to a particle size of less than 0.25mm.
2.5.2 Equipment
Brush
Compressed air
Gloves
Mask
Plastic bags
Protective Clothing
Pulvenizer
Scoop
0.25mm sieve
2.6.3 Procedure
With a scoop, place a portion of the minus 2-mm soil fraction material into the pulverizer opening.
Turn the power on. The pulverizer grinds the soil and deposits it into a collection bin at the bottom of
the machine. After the first scoop is pulverized, shut the machine off and sieve the material through
the 0.25-mm sieve. If all the material passes through this sieve, the grinding plates are sufficiently
close enough to continue pulverization. If not, adjust the plates and repeat the procedure on the same
sample until all the material passes through the 0.25-mm sieve (described in Section 3.6). Once the
collection bin is full, turn the machine off and deposit the pulverized material into a clean labeled
container. Repeat the operation until all soil is pulverized. Thoroughly clean the machine with
compressed air and a brush.
2.5.4 Quality Control
When pulverizing, gloves must be worn, as well as a mask and protective clothing. The machine
opening should be tightly fastened to minimize dust, the laboratory manager will frequently check the
processing equipment for proper operation, for adherence to protocol including proper maintenance. A
member of the EMSL QA staff will visit the preparation laboratory to ensure adherence to protocol.
2.6 Final Sieving
2.6.1 Summary
To ensure that the pulverized audit sample has a particle size < 0.25mm it is resieved through
0.25mm sieve.
2.6.2 Equipment
Fumehood'
B-99
-------
Kraft paper
Paint Brush
Plastic bags
0.25-mm sieve
3x5 card
2.6.3 Procedure
- Place aim2 sheet of kraft paper on the sieving table under a vented fumehood. Place a 60 cm2 sheet
of kraft paper on the larger piece of paper. Place a portion of the soil material in the 0.25-mm sieve
and screen the material using a rocking motion. Use a paint brush or 3 x 5 card to gently push the
material through. Place any material > 0.25mm into a separate pile. Continue this procedure until
the complete sample is sieved. Save the material not passing through the .25mm sieve for further
pulverization.
2.6.4 Quality Control
When sieving, gloves must be worn, as well as a mask and protective clothing. The laboratory manager
will frequently check the sieving processing equipment for proper operation and for adherence to
protocol A member of the EMSL-LV QA staff will visit the preparation laboratory to ensure
adherence to protocol.
2.7 Homogenization and Subsampling to 2-kg Aliquots
2.7.1 Summary
Prior to splitting the 2 kg aliquots into 20 g aliquots, the bulk soil (minus 0.25mm fraction) is
homogenized using a combination of three techniques; drum-rolling, cone and quartering, and riffle-
splitting. After homogenizing, the bulk sample is split into 2 kg aliquots using a riffle splitter.
2.7.2 Equipment
Drum homogenizer
Gloves
Kraft paper
Labels
Large riffle splitter
Mask
Protective clothing
Shovel
Top loading balance
2-L sample bottles
2.7.3 Procedure
2.7.3.1 Drum homogenization/Cone and Quartering
Place all of the < 0.25mm fraction from one soil sample into the drum homogenizer. Slower rotate the
drum for five minutes. Pour the entire sample onto a large piece of kraft paper so that the sample
takes on the shape of a cone. Homogenize the cone by dividing the cone into four equal quarters by
lines going clockwise from 1 to 4. Using a shovel, remove the first quarter to form anew cone. The
third, second and fourth quarters are piled sequentially over the first quarter. The procedure is
performed seven times in succession. Figure 2.2 illustrates the technique.
Figure 2.2 Top and side views of the soil cone
B- 100
-------
2.7.3.2 Riffle splitting
Position the two collecting bins under the large riffle splitter (Figure 2.3). Pour the entire sample
evenly across the baffles of the riffle splitter. Transfer the soil from each collecting bin into the
distribution pan and replace the receiving pans under the riffle splitter. Repeat this procedure five
times in succession.
2.7.8.8 Subsampling
After the homogenization, 2 kg aliquots are obtained. If the cone and quartering technique is used,
place a clean 2-L sample bottle at the bottom of the cone and, with an upward movement, collect a
sample weighing approximately 2000 grams (+/-20 grams). If the riffle splitting technique is used,
place a clean 2-L sample bottle at one end of the collecting bin and moved to the other end to fill the
bottle. The sample is labeled using the procedure described in Section 2.2. The first 2 kg aliquots for
each audit concentration's identified with "01" and subsequent aliquots numbered consecutively. The
other information within the sample code will ensure a unique sample identity. Repeat this procedure
for the entire amount of homogenized audit sample. Store the audit samples in cold storage until
further processing. .
B-101
-------
Figure 2.3 Large Biffle Splitter
2.7.4 Quality Control
When homogenizing, gloves must be worn, as well as a mask and protective clothing: Prepare labels
for the 2 kg samples prior to the processing step in order to avoid mislabeling. the laboratory manager
will frequently check the homogenization operation for proper processing equipment and for adherence
to protocol A member of the EMSL-LV QA staff will visit the preparation laboratory to ensure
adherence to protocol.
2.8 Homogenization and Subsampling to 100 g and 20g Aliquots
2.8.1 Summary , ~
Each 2 kg aliquot prepared in section 2.7 is further homogenized in a medium sized riffle splitter and
split into 100 g aliquot, the 100 g aliquots are then homogenized in a small riffle splitter and split into
20 g aliquots. These two procedures are done simultaneously in order to avoid the use of intermediate
sample containers and the possibility of mislabeling.
2.8.2 Equipment
Gloves
Fumehood
Laboratory containers (20 g samples)
Open pan balance
Plastic bags
Riffle splitters, medium (24 chute 13-1/2" x 15-3/8") and small (32 chutes 6-W x 9")
Scoop
2.8.3 Procedure
2.8.3.1 Homogenization and Subsampling to 100 grams '
2.8.3.1.1 Initial Homogenization— Position the two receiving pans under the medium riffle splitter.
Pour the entire 2 kg sample evenly across the baffles of the riffle splitter. Transfer the soil from each
receiving pan into the distribution pans and replace the receiving pans under the riffle splitter. Repeat
this procedure five times in succession.
2.8.3.1.2 Splitting to 500 g Aliquots-- Pour the sample evenly across the baffles «nri place the soil from
one receiving pan aside. Transfer the soil in the other receiving pan to the distribution pan and split
once more. This should produce approximately a 500 g samples in each receiving pan. Place these
samples on separate sheets of kraft paper. Split the soil from the other receiving pan similarly. This
produces a total of four 500 g aliquots from each 2kg aliquot.
2.8.3.1.3 Splitting to 100 g aliquots- Pour the 500 g sample evenly across the baffles and place the soil
from one receiving pan into a plastic bag. Transfer the soil in the other receiving pan to the
distribution pan and continue splitting as necessary until approximately 100-g- of sol occupies one of
the receiving pans. Place the entire contents of this pan into the distribution paa of the small riffle
splitter (see section below). Repeat the procedure until all of the 2 kg aliquot is spit into 100 g
aliquots.
2.8.3.2 Homogenization and Subsampling to 20 grams
2.8.3.2.1 Initial Homogenization—Position the two receiving pans under tile small riffle splitter. Pour
the entire 100 g aliquot from the distribution pan evenly across the baffles of the riffle splitter.
Transfer the soil from each receiving pan into the distribution pan and replace the receiving pans
under the riffle splitter. Repeat this step five times in succession.
B- 102
-------
2.8.3.2.2 Splitting into 20 g Aliquots-Pour a 100 g aliquots evenly across the baffles of the small riffle
splitter. Place the soil from one receiving pan into a plastic bag. Transfer the soil from other receiving
pans under the riffle splitter. Repeat this step five times in succession.
2.8.3.2.2 Splitting into 20 g Aliquots- Pour a 100 g aliquots evenly across the baffles of the small riffle
splitter. Place the soil from one receiving pan into a plastic bag. Transfer the soil from other receiving
pan to the distribution pan and continue splitting as necessary until approximately 20 g of soil occupies
one of the receiving pans. Place the entire contents of the pan into the pre-labeled sample container
provided by the analytical laboratories. Repeat the procedure until the entire 100 g sample is split into
five 20 g aliquots.
2.8.4 Quality Control . .
When homogenizing and subsampling, gloves must be worn, as well as a mask and protective clothing.
The laboratory manager will frequently check the operation for proper use of equipment and for
adherence to protocol. A member of the EMSL QA staff will visit the preparation laboratory to ensure
adherence to protocol As samples are characterized, precision estimates for each audit sample type will
be developed. If the pooled relative precision estimate (RSD) for an audit sample whose concentration
is above 10 times the detection limit (~100ppm) is greater than ten percent, the preparation
laboratory will combine all 20 g aliquots, rehomogenize, then resplit the sample.
B-103
-------
3.1 Overview
Section 3
Dust Audit Sample Preparation Procedures
Dust samples of different concentration will be supplied to EMSL-LV from each city. Prom these
samples, EMSL-LV will provide three audit samples with Pb at low, mid, and high concentration ranges
and three calibration standards at similar concentrations. The bulk samples are air dried, sieved,
homogenized and split into 2-gram aliquots as outlined in Figure 4.1. Participating laboratories supply
EMSL-LV with sample containers, labels, and the appropriate labeling techniques for the samples
A random subsample of the audit samples will be characterized by EMSL-LV. Fifty samples at each
concentration range will be analyzed for Pb by XRF. A subset of these samples will be analyzed by
ICPES after nitric acid extraction. Characterization data will be supplied to the Lead Abatement QA
manager.
DUST SAMPLE
AIR DRY
SIEVE
TO .25MM
HOMOGENIZE &
SUBSAMPLE 100
GRAM AUQUOTS
• HOMOGENIZE & >
SUBSAMPLE 2 i
GRAM ALIQUOTS j
BATCHING & j
SHIPMENT i
FIGURE 3J. DUST AUDIT SAMPLE PREPARATION FLOW
B-104
-------
3.2 Sample drying
Interior dust audit samples will be sent from the participating cities of Boston, Baltimore and
Cincinnati in one gallon containers. Upon arrival, remove the lid of each container and allowed it to air
dry before further preparation. The samples are kept in the shipping container during air-drying to
prevent loss of sample.
3.3 Sieving to 0.25mm
3.8.1 Equipment
Fumehood
Kraft paper
Paint brush
0.25mm Sieve
3x5 index cards
3.3.2 Procedure
Place a 1 m2 sheet of kraft paper onto the preparation table. On top of this sheet place a 60 cm2 sheet
of kraft paper. Set a 0.25mm mesh sieve on top of the smaller sheet of kraft paper. Portions of the
dust sample are placed into the sieve and gently, pushed through with either a paint brush or a 3 x 5
card. Material greater than 0.25m is placed in a plastic bag for proper disposal
3.4 Homogenization and Subsampling to 2 Gram Aliquots
3.4.1 Equipment
Fumehood
Gloves
Laboratory containers (2 gram samples)
Open, pan balance
Plastic bags
Eiffle spUtter, medium (24 chute 13-1/2" X125-3/8"
Eiffle spUtter, mini (14 chutes, 2-1/16" X 3-3/4")"
Scoop
8.4.2 Procedure
4.4.2.1 Homogenization and Subsampling to 100 grams
Position the two receiving pans under the small riffle spUtter. Pour the entire contents of the minus
0.25mm dust fraction evenly across the baffles of the riffle spUtter. Transfer the dust from each
receiving pan into the distribution pan and replace the receiving pans under the riffle spUtter. Repeat
this step five tune* in succession with the material in each receiving pan.
Pour the sample evenly across the baffles and place the dust from one receiving pan into a plastic bag.
Transfer the soil in the other receiving pan to the distribution pan and continue splitting as necessary
until approximately 100 g of dust occupies one of the receiving pans. Place the entire contents of this
pan into the distribution pan of the mini riffle spUtter (see section below). Repeat the procedure until
all of the dust sample is split into 100 g aliquots.
3.4.2.2 Homogenization and Subsampling to 2 Grams
Position the two receiving pans under the mini riffle splitter. Pour the 100 g aUquot evenly across the
baffles of the riffle spUtter. Transfer the dust from each receiving pan into the distribution pan and
replace the receiving pans under the riffle spUtter. Repeat this step five times in succession with the
material in each receiving pan.
B-105
-------
Splitting to 25 g Aliquots—
Pour the 100 g aliquot evenly across the baffles and place the dust from one receiving pan aside.
Transfer the dust in the other receiving pan to the distribution pan and split once more. This
produces a 25 g aliquot in each receiving pan. Place the 25 g aliquots on separate sheets of kraft
paper. Similarly split the remaining dust to produce an additional a total of two 5 g aliquots.
Splitting to 2 g Subsamples-
Pour the 25 g aliquot evenly across the baffles of the mini riffle splitter and place the soil from one
receiving pan into a plastic bag. Transfer the soil in the other receiving pan to the distribution pan
and continue splitting as necessary until approximately 2 g of soil occupies one of the receiving pans.
Place the entire contents of the this pan into the pre-labeled sample container provided by the
analytical laboratories. Similarfy split the dust set aside in the plastic bag. Repeat the procedure until
all of the 25g aliquots are split into 2 g samples.
3.4.3 Quality Control
When homogenizing and subsampling, gloves must be worn, as well as a mask and protective clothing.
The laboratory manager will frequently check the sieving operation for proper equipment and for
adherence to protocol A member of the EMSL QA staff will visit the preparation laboratory to ensure
adherence to protocol As samples are characterized, precision estimates at each concentration will be
developed. If the pooled precision estimate for an audit sample whose concentration is above 10 times
the detection limit (~100ppm) is greater then ten percent relative standard deviation, the preparation
laboratory will and resplit rehomogenize the sample.
3.4.2 Procedure
4.4.2.1 Homogenization and Subsampling to 100 grams
Position the two receiving pans under the small riffle splitter. Pour the entire contents of the minus
0.25mm dust fraction evenly across the baffles of the riffle splitter. Transfer the dust from each
receiving pan into the distribution pan and replace the receiving pans under the riffle splitter. Repeat
this step five times in succession with the material in each receiving pan.
Pour the sample evenly across the baffles and place the dust from one receiving pan into a plastic bag.
Transfer the soil int the other receiving pan to the distribution p[an and continue splitting as necessary
until approximately 100 g of dust occupies one of the receiving pans. Place the entire contents of this
pan into the distribution pan of the mini riffle splitter (see section below). Repeat the procedure until
all of the dust sample is split into 100 g aliquots.
3.4.2.2 Homogenization and Subsampling to 2 Grams
Position the two receiving pans under the mini riffle splitter. Pour the 100 g dust aliquot evenfy across
the baffles of the riffle splitter. Transfer the dust from each receiving pan into the distribution pan
and replace the receiving pans under the riffle splitter. Repeat this step five times in succession with
the material in each receiving pan.
Splitting to 25 g Aliquots-
Pour the 100 g aliquot evenly across the baffles and place the dust from one receiving pan aside.
Transfer the dust in the other receiving pan to the distribution pan and split once more. This
produces a 25 g aliquot in each receiving pan. Place the 25 g aliquots on separate sheets of kraft
paper. Similarly split the remaining dust to produce an additional a total of two 25 g aliquots.
Splitting to 2 g Subsamples--
B-106
-------
Pour the 25 g aliquot evenly across the baffles of the mini riffle splitter and place the soil from one
receiving pan into a plastic bag. Transfer the soil in the other receiving pan to the distribution pan
and continue splitting as necessary until approximately 2 g of soil occupies one of the receiving pans.
Place the entire contents of the this pan into the pre-labeled sample container provided by the
analytical laboratories. Similarly split the dust set aside in the plastic bag. Repeat the procedure until
all of the 25g aliquots are split into 2 g samples.
3.4.3 Quality Control
When homogenizing and subsampling, gloves must be worn, as well as a mask and protective clothing.
The laboratory manager will frequently check the sieving operation for proper equipment and for
adherence to protocol. A member of the EMSL QA staff will visit the preparation laboratory to ensure
adherence to protocol As samples are characterized, precision estimates at each concentration will be
developed. If the pooled precision estimate for an audit sample whose concentration is above 10 times
the detection limit (~100ppm) is greater than ten percent relative standard deviation, the preparation
laboratory will and resplit rehomogenize the sample.
Section 4
Handwipe Audit Sample Preparation
4.1 Summary
As part of the Superfund Lead Abatement program, children's hands are swabbed with handwipes
which are then analyzed for lead. As part of the quality control, handwipes audit samples are included
with the unknown handwipe samples for analysis. Handwipe audit samples are spiked with lead at
three different levels; 5ug, 20ug, and 40ug lead.
4.2 Equipment
Box of wet handwipes
200 mg/L and 1000 mg/L solutions
ml pipette
Ziploc type plastic bags
Plastic gloves
4.8 Procedure .
4.3.1 Regents
. 1000 mg/L Pb - Certified standard obtained commerciaDy.
. 200 mg/L Pb - Dilute 1000 mg/L Pb solution 1:5 with reagent water.
4.8.2 Spiking Procedure
. Unopened containers of wet-wipes are provided by the participating cities,
. Working in laminar flow dean hood, wearing clean gloves, Pull out 6 wet wipes from the
same container ^n^ place into a stack (Le., one on top of the other). Using a micropipet, add the spike
to the center of the wet wipe stack between the third and forth wipe). The spike volumes are given
below;
- 5 ug spike - 25 uL of 200 mg/L Pb standard
- 20 ug spike - 20 uL of 1000 mg/Pb standard
- 40 ug spike - 40 uL of 1000 mg/L Pb standard
. Fold and crumple the wet wipe stack and place into a zip-lock bag. Seal and label the bag
with lab ID number. Record the lab ID and spike level into a lab notebook.
B-107
-------
Section 5
Urban Soil, Urban Dust, And Wet-Wipe
Audit Sample Characterization
5.1 Sample Preparation
5.1.1 Reagents
Concentrated nitric acid (ACS Reagent grade)
Concentrated nitric acid (Double deionized)
Hydrofluoric acid (48% high purity)
Reagent water (ASTM type ID
5.1.2 Hot nitric Arid (HNO3) Extraction
Place 1 g sample (weighed to nearest 0.1 mg) or packet of wet wipes into a clean 100 mL beaker. Add
50 mL 7N HNO3 to soil or dust samples. Add 50 mL IN HNO3 to wet wipe samples. Push wet wipes
down with glass stirring rod to ensure complete coverage. Cover with a watch glass and heat gently at
95°C for 2 hours. Maintain at least 25 mL volume in the beaker by adding 7N HNO3 (IN for wet wipe
samples) as necessary. After digesting, cool and add 10 mL of water. Filter through Whatman No. 1
filter paper into a 100 mL volumetric flask. Rinse beaker and filter with additional water. Dilute to
volume with water. '
5.1.3 Total Digestion of Urban Soil and Dust Samples
. Add 0.5g (weighed to nearest 0.1 mg) sample into a clean teflon microwave digestion vessel
Add 9 mL of concentrated HNO3, and 4 mL of 48% HF. Cap and seal the vessels. Weigh capped
vessel to the nearest .Olg and place in microwave oven. A total of 12 vessels must be placed in oven.
Use blanks if extra spaces must be filled. Heat at 520 Watts for 30 minutes. Let the samples cool and
irradiate again at the same setting.
. CooL Weigh capped vessels. Rinse condensate from cap and vessel walls into vessel
Transfer quantitatively to a 100 mL polypropylene volumetric flask. Dilute with reagent water to the
Tnark.
. If not determined previously, determine percent solids as in Section 6.2.
5.1.4 Preparation of Loose Powder Samples for XRF Analysis
. Pour a 5g soil sample or 2 g dust sample into a powder cup and seal with 3.6 um mylar film.
5.2 Percent Solid Determination .
Determine the percent solids in the soil or dust samples by drying a 5g aliquot at 105°C for 24 hours.
Place a 5g sample (weighed to the nearest mg) in a tared aluminum weighing dish. Dry at 105°C for
24 hours. Cool in a desiccator. Reweigh to the nearest mg. .
Percent solids = [100 (wet wt. - dry wt.) /wet wtJ. .
5.3 Sample Analysis
5.3.1 Summary
Samples were analyzed by XRF to determine Pb concentrations and homogeneity. The XRF soil audit
concentrations were verified by ICP or GFFAAS. From the fifty aliquots of each soil analyzed by XRF,
a subset of 7 aliquots were analyzed by ICP or GFFAAS.
B-108
-------
5.8.2 ICPAES and GFAAS Analysis
The acid digests are analyzed by either ICPAES or GFFAAS depending on the lead concentration in
the digestate. Solutions containing Pb concentrations greater than 10 times the ICPAES IDL are
analyzed by ICPAES ODL is about 50 ppb). Lower concentrations are measured by GFAAS. The
instruments are calibrated and the digestates analyzed. HF resistant components are used for the
total digest solutions. Qualify control is described in Section 5.5.
5.3.3 XEF Analysis •
Loose powder samples are analyzed by XRF. The analysis conditions for lead are; Ag secondary target,
X-ray tube voltage = 35 Kev, X-ray tube current = 3 mA, atmosphere = air, counting time = 200
sec. live time. The lead L-beta peak/ Ag compton peak ratio is calculated. The lead concentration is
determined from the ratio and the calibration curve (Ratio vs. Concentration). Qualify control is
described in Section 6.5.
5.4 Instrument Calibration
5.4.1 ICPAES and GFFAAS Analysis
The instruments are calibrated following the manufacturer's guidelines. A series of calibration
standards are analyzed and a calibration line calculated using linear regression of intensity vs. standard
concentration.
5.4 3. XRF Analysis
The XRF is calibrated by acquiring spectra from a series of urban soil standards with known lead
concentrations. Acquisition conditions are given in Section 5.3.3. The Pb L-beta peak and Ag compton
peak are measured from the spectra and the Pb LB peak/Ag Compton peak ratios are calculated. A
calibration line is calculated using linear regression of ratio vs. standard concentration.
5.5 Quality Control
5.5.1 Sample Related Quality Control
The following QC are prepared for ICPEAS and GFAAS analysis
. Matrix Spike Sample - one sample per 20 will be spiked with lead prior to digestion.
. Reagent Blank Sample - One reagent blank win be prepared per group of 20 samples.
group o£ l&barrqbfczy Control Sample (LCS) - One LCS sample will be prepared and analyzed per
5.5.2 Analysis Related Quality Control
The following QC q*™p1»< are analyzed along with routine samples:
5.5.2.1 ICPAES and GFAAS Analyses
Initial Calibration Verification (ICV) Standard - After calibration, the ICV is analyzed. The
percent recovery must be 90-110%. The ICV solution is a standard from a different source
than the calibration standards.
Initial Calibration Blank (ICB) - After = analysis of the ICV, the ICB is analyzed. The
measured concentration must be less than 2 times the IDL.
Interference Check Solution (ICS) - An ICS solution is analyzed after the ICV and ICB are
analyzed. The ICS contains 500 ppm of major interferents (Mg, Ca, Fe, Al) and a known Pb
concentration. The % recovery of Pb must be 75-125%.
B-109
-------
Detection. Limit Sample (DL) - A DL sample is analyzed after the ICS solution. The
concentration of the DL solution is twice the IDL.
. ^Continuing Calibration Verification Standard (CCV) - A CCV is analyzed every 10 sample and
after the last sample. The CCV concentration is in the mid-calibration range. The % recovery
must be 90-110%. If not, the instrument must be recalibrates and all samples up to the last
acceptable CCV must be reanalyzed.
Continuing Calibration Blank Sample (CCB) - A CCB is analyzed after every CCV. The
concentration must be less than twice the IDL.
5.5.2.2 XKF Analyses
Reference Monitor (KM) - Prior to analysis, a reference monitor sample is measured. The
reference monitor intensity provides a standard measure of the x-ray flux that irradiates the
samples being analyzed. The reference monitor provides a method of standardizing and/or
compensating for changes in the x-ray tube flux. •
High Initial Calibration Verification Standard (ICVH) - An ICVH sample is analyzed after the
KM and after the last sample in a run. The concentration of Pb is at the high end of the range
of interest.
Low Initial Calibration Verification Sample (ICVL) - an ICVL is analyzed after the ICVH. The
concentration of Pb is at low end of the range of interest.
Section 6
Audit Sample Window Generation
6.1 Soil, Dust, and Handwipe Audit Samples
At least 50 aliquots from each soil and dust are analyzed by XRF, wet wipes are analyzed by ICPES.
"A biweight statistical procedure is used to calculate audit windows. The biweight approach has an
advantage over the classical approach in that it identifies outliers and weights them in a manner that
gives them less influence on the accuracy window.
After analysis, enter the data into the program, which then generates three estimates of prediction
intervals for single future observation from a univariate normal, population (Figure 7.1).
(1) Classical - Based on all data Reference: Whitmore, G.A. "Prediction Limits for a
Univariate Normal Observation", The American Statistician, VOL. 40, NO. 2, may
1986, PP 141-143.
(2) W/O Outliers - Outliers Removed by Grubbs' Test Reference: Barnett, V. and
Lewis, T. "Outliers in Statistical Data", 2ND ED., John Wiley and Sons, New York,
1984, P. 167.
(8) Biweight - Robust Estimation Using Biweight Procedure Reference: Kafadar, K "A
Biweight Approach to the One Sample Problem", Journal of the American Statistical
Association, VOL. 77, NO. 378, PP. 416-424.
B-110
-------
PREDICTION INTERVAL SUMMARY REPORT
DATA FILE:
or
gSTIHATOR
I or
DATA
CLASSICAL 50
M/0 OUTLIERS SO
BIHBIGHT 50
SAMPLE SAMPLE 95% INTERVAL 99% INTERVAL
HEAH STD DEV LOWER UPPER LOWER UPPER
927.1480 41.4193' 843.0907 1011.2050 814.9757 1039.3200
927.1480 41.4193 843.0907 1011.2050 814.9757 1039.3200
923.4212 43.1311 835.8742 1010.9680 806.5920 1040.2500
ii 7.1 Example of Audit Sample .Prediction Interval Summary Report _
The program also performs the following;
1) Tests for normality using the Kolmogorov-Smirnov and the Anderson-Darling statistic
2) Presents a histogram of the data
8) Lists the data and the biweight weighting factors
The information is sent to the project QA manger for review before audit samples are sent to *
laboratories for inclusion in sample batches.
Section 7
Safely
7.1 Laboratory Safety
Environmental samples invariabfy involve undesirable if not hazardous materials and must be handled
with respect. Special equipment and facilities are provided to prevent cross contamination of space and
other samples. Special training hi the use of the above may be needed (Section 1.3.3).
Personnel engaged in handling hazardous samples undergo initial and periodic medical examinations to
insure that they have not contracted medical problems related to the materials with which they are
involved, *
7.1.1 Equipment and Supplies .
Dust
Full face respirator
Laboratory coat
PVC gloves
Tyveksuits
7.1J2 Preparation Laboratory
Dedicated equipment and special facilities are used during sample preparation. The LESC warehouse
has two rooms dedicated to sample drying, sieving, homogenization, riffle splitting, and sample
aliquoting. During each of the above procedures the following equipment is required: full face
respirator, tyvek suit, and PVC gloves.
7.1.3 Characterization Laboratory
The analytical laboratory requires personnel to: 1) work in a laminar hood and wear a dust mask while
splitting samples, 2) wear PVC gloves while handling samples.
B - 111
•>?-U.S. GOVERNMENT PRINTING OFFICE: 1993 - 750-068/60017
------- |