v>EPA
United States
Environmental Protection Agency
:H| Office of
Research arid Development
National Human Exposure Assessment Survey
(NHEXAS)
Arizona Study
Quality Systems and Implementation Plan
for Human Exposure Assessment
Title: Quality Systems and Implementation Plan for Feasibility Studies
in Support of NHEXAS
Source: The University of Arizona
Notice: The U.S. Environmental Protection Agency (EPA), through its Office of Research and Development (ORD), partially funded
and collaborated in the research described here. This protocol is part of the Quality Systems Implementation Plan (QSIP)
that was reviewed by the EPA and approved for use in this demonstration/scoping study. Mention of trade names or
commercial products does not constitute endorsement or recommendation by EPA for use.
The University of Arizona
Tucson, Arizona 85721
Cooperative Agreement CR 821560
Standard Operating Procedure
UA-G-QSIP
U.S. Environmental Protection Agency
Office of Research and Development
Human Exposure & Atmospheric Sciences Division
Human Exposure Research Branch
-------
ENDNOTES
The QSIP was designed as a "living" document assuring the quality of the National
Human Exposure Assessment project in Arizona. Researchers knew that even the best
plan may need to change and agreed to document all necessary changes when
implemented. These changes are described in the "ENDNOTES." Each endnote is
designated as a number that appears as a superscript in the QSIP text to document each
change. All changes were made with the concurrence of our EPA Project Officers. The
endnotes (xx pages) can be found at the end of the document. Careful attention to the
endnotes should be made when reading the QSIP.
-------
Section No.TOC
Revision #0
June 1995
Page: 2 of 116
TABLE OF CONTENTS
1.0 PROJECT PLANNING AND ORGANIZATION 5
1.1 Identification of the Problem 5
1.2 Background of Specific Pollutants & Pollutant Classes
Selected for Study 5
1.3 Project Scope and Work Objectives 14
1.4 Project Description 29
1.5 Personnel Qualifications 39
1.6 Training Required 39
1.7 Experimental Design. . 40
2.0 MANAGEMENT ASSESSMENT 59
2.1 Assessment Responsibility 59
2.2 Assessment Types . 60
2.3 Assessment Usage 61
2.4 Assessment Criteria 61
2.5 Assessment Documentation 61
3.0 PROJECT IMPLEMENTATION PLAN 62
3.1 Project Design Criteria 62
3.2 Data Quality Indicators 84
3.3 Project Responsibilities 96
4.0 DATA ACQUISITION AND MANAGEMENT 99
4.1 Data Acquisition 101
4.2 Control and Calibration of
Measurements and Testing Equipment 102
4.3 Identification of Data 102
4.4 Control of Erroneous Data 102
4.5 Data Evaluation. 103
4.6 Procedures 103
5.0 RECORDS USAGE AND MANAGEMENT 103
5.1 Data Records 103
5.2 Records Management System 103
5.3 Record Validation 104
5.4 Records Identification, Indexing and Retention ...... .105
5.5 Records Distribution and Storage 105
6.0 ROUTINE CONTROLS AND PROCEDURES 106
6.1 Maintenance of Equipment 106
6.2 Quality of Consumables 107
6.3 Labeling • .108
6.4 Acceptance of Equipment and Materials 108
6.5 Storage of Equipment and Materials 108
-------
Section No. TOC
Revision #0
June 1995
Page: 3 of 116
7.0 TECHNICAL ASSESSMENT AND RESPONSE 108
7.1 Assessment Procedures 108
7.2 Assessment Evaluation 110
7.3 Assessment Response and Follow-up. 110
DISTRIBUTION LIST 112
References 113
TABLES
Table 1. CAS Numbers of Target Analytes 13
Table 2. Race and ethnicity by county based on the
1990 census data 26
Table 3. Sampling and Analysis Methods to be employed in Stage 2. 32
Table 4. Sample Collection methods of Stage 3 . .34
Table 5. Classes of Hypotheses and Statistical Approach taken. . .35
Table 6. Person-level sampling weights for NHEXAS Arizona
(HH = households) 48
Table 7. The likelihood (denoted by K%) of capturing at least
1 person with a significant exposure in the
of the distribution ( = p) given the sample size. . .50
Table 8. The sample size required to capture 1 person
in the exposure distribution 51
Table 9. Probability of high end capture for a selected
domain given a design effect of 1.5 52
Table 10. General Field Assays (HDL) . .55
Table 11. General Laboratory Procedures (MDL) 56
Table 12. Quality Assurance Objectives for Sample Characterization.57
Table 13. Siting Criteria 63
Table 14. Projected Volume of Air Collected by Active Samplers. . .72
Table 15. Sampling Media for Pollutant Classes. . .73
Table 16. Specific Equipment and operation parameters
used in Residential Sampling (Stages 2 through 5) . . .81
Table 17. Specific Equipment and operation parameters
used in Residential Sampling (Stages 3, 4, 5) 82
Table 18. Summary of samples collected for each medium
in each stage .83
-------
Section No. TQC
Revision #0
June 1995
Page: 4 of 116
Revision #0
June 1995
Page: 4 of 116
FIGURES
Figure 1. Major classes of Standard Operating Procedures in
support of NHSXAS Arizona 17
Figure 2. Administrative Procedures 18
Figure 3. Field Collection Procedures associated
with metals, pesticides and VOCs 19
Figure 4. Laboratory Procedures associated with tasks and
analyte classes 20
Figure 5. Data Management Procedures 21
Figure 6. Data Analysis Procedures . . 22
Figure 7. An overview of NHSXAS Arizona Implementation 30
Figure 8. Diagrammatic of the Survey Design . 45
Figure 9. Diagrammatic of the NHEXAS Arizona Project and
Data Tracking components 75
Figure 10. Study Stages and Questionnaire Collection burden ... 77
Figure 11. Study Stages and Technician Task Completion burden . . 78
Figure 12. Time lines and information collected in Stages 1 & 2 . 79
Figure 13. Time lines and information collected in Stages 3,
4 & 5 80
Figure 14. Handling the Photo-Ionization Detector for
Total VOC in air 85
Figure 15. Handling of Soil Composites for yard and foundation. . 86
Figure 16. Handling of dust samples from floors collected
using a vacuum cleaner ..... 87
Figure 17. Handling of particulate from pumped air 88
Figure 18. Handling of all VOC samples (active and passive) ... 89
Figure 19. Handling Wipe Samples 90
Figure 20. Handling Food and Beverage Samples ..... 91
Figure 21. Handling Water Samples ... 92
Figure 22. Handling Urine and Blood Samples ..... 93
Figure 23. Organizational Structure of NHEXAS Arizona 100
APPENDICES
Appendix A: Standard Operating Procedure List
Appendix B: Questionnaire (9/94) Evaluation by Arizona
Appendix C: QA Audit Documents
-------
Section No. j _ Q
Revision #0
June 1995
Page: 5 of 116
1.0 PROJECT PLANNING AND ORGANIZATION
1.1 Identification of the Problem
Some pollutants pose well-known risks to health, whereas for most
pollutants we have insufficient health information. Currently, we lack
enough information about the population distribution and the determinants
of exposure to perform risk management and protect public health. There is
an obvious public health need to try to answer important questions
concerning exposure to environmental pollutants that are likely to produce
harm and increase the risk of disease in humans. Unfortunately, there are
many such questions that are insufficiently addressed to properly enable
scientific decisions about management of these risks and protection of
public health [Sexton et al., 1992]. There is a major scientific need to
provide exposure data that are directly relevant to answering these
important questions. The NHEXAS concepts are very appropriate to such
goals.
Basically, there are several classes of pollutants that are potentially
great health risks for which we have little information on population
exposure. Further, these chemicals can lead to exposure from multiple
sources (air, water, soil, food, dust, etc.). Also, we have the sense that
certain populations, including low-income individuals, minorities and the
biologically susceptible, are at high risk and their exposures need to be
identified. Further, little is known about temporal and spatial
distributions, and trends in these distributions.
The pollutant classes that are of special concern, as identified by EPA
NHEXAS and the scientific community, are the volatile organic compounds
(VOC's), metals, polycyclic organic matter (POM's/PAH's), pesticides, and
dioxins. In general, these pollutants have been shown to cause such
concern based on limited epidemiological or occupational studies in humans
or in animal models utilized for this purpose. Some, considered
prototypic, are well-known to cause harm to humans and are considered of
major public health significance; the best examples are benzene and lead.
As will be seen, almost all have documented concerns that are of public
health significance, relating to their different risks - carcinogenic,
teratogenic, etc. (based on reports of EPA, ATSDR, NIOSH, WHO, WHO/EURO).
Some health risks and outcomes, like asthma, the clustering of birth
defects and some cancers, appear to be increasing [PHS, WHO]. These
diseases may be related to the NHEXAS target pollutants (prototypic,
priorities 1 & 2) [ATSDR, CDC]. The need to provide such information is
urgent.
1.2 Background of Specific Pollutants & Pollutant Classes Selected for Study
We selected pollutant classes to meet the criteria of i) public health
significance, ii) the potential for NHEXAS in terras of specific areas where
high exposures occur and in which high-risk groups (those biologically
susceptible, minorities, low income and/or inner city residents) are
especially exposed, iii) the cost and feasibility of determining relevant
exposures (and sources), and iv) the feasibility of assessing exposure-dose
-------
Section No. 1.2
Revision #0
June 1995
Page: 6 of 116
and dose-response relationships to provide the information directly
relevant for risk assessment, management and for making public health
decisions. The negotiated pollutant classes selected from the NH1XAS
Request for Proposal (RPP) include (in order of priority) metals,
pesticides and volatile organic compounds (VOCs). The list of target
compounds were separated into primary and secondary categories. We plan to
analyze each media for all target analytes. This ranking approach enables
us to consistently select the same analytes when confronted with choices
due to limitations in sample size or funding. For instance, if the amount
of sample is limited the primary targets (metals) only will be analyzed.
Further limitations will exclude secondary targets (secondary metals) from
analysis. Discussion continues about incorporating polycyclic aromatic
hydrocarbons (PAHs) in our project? If funding is forthcoming this
document will be appended to address QA issues unique to PAHs. The
following discussion outlines the importance of the specific pollutants
selected within each class.
1.2.1 Target Metals
Primary Metals: Lead (Pb), Arsenic (As), Cadmium (Cd), Nickel (Ni) and
Chromium (Cr)
Secondary Metals: Barium (Ba), Iron (Fej3, Manganese (Mn), Selenium (Se),
Vanadium (V), Copper (Cu) and Zinc (Zn)
4
Selection of metals to evaluate, and the rationale for these choices, are
based on the EPA Criteria Document for Lead, other EPA documents on metals,
and IPCS/WHO Criteria & Guidelines documents. All can produce a variety of
health effects, and many are considered definite or probable public health
problems. Lead is the obvious prototype metal as it is a NAAQS pollutant,
with clearly defined potential human harm. However, there is a lack of
sufficient data on population distributions of exposure, especially for all
high risk populations.
Arsenic in soil (and dust) appears to be ubiquitous (Arizona & rural
California sampling reports), and may be locally ubiquitous in air and
water. Given its known health risks for humans (op cit.), and the lack of
data on population distributions of exposure, it is a good metal to target.
Cadmium, chromium, and nickel are also of interest for their potential harm
to human health and the lack of data on population distributions of
exposure. Chromium is of special interest as it is associated with several
health outcomes and is a source marker (op cit., below). We are also
interested in selenium concentrations, as selenium (a natural ingredient in
food) appears to be protective of cancer in non-toxic amounts. Vast
quantities of sulfide minerals are deposited in the rock strata of Arizona.
Copper mining and smelting is a major industry in Arizona.
1.2.2 Target Pesticides
Primary Pesticides: Chlorpyrifos and Diazinon
Secondary Pesticides: Malathion and Carbaryl
-------
Section No. 1.2.2
Revision #0
June 1995
Pages 7 of 116
Pesticide exposure studies conducted in recent years in the U.S. include
the U.S. EPA's Non- Occupational (NOPES) [Lewis et al., 1988], Household
Infant (HIPES) [Lewis et al., 1991], and the Second National Health and
Nutrition Examination Survey (NHANES II) [Murphy et al., 1983]. On- going
pesticide exposure studies include the NCI-EPA Farm Occupational Exposure
Study (NEFOES), Lawn Application Pesticide Exposure Study (LAPES) [Nishioka
et al., 1992] and the Brownsville, TX study. On the basis of these studies
we have selected the following four compounds from the NHEXAS pesticides
category: chlorpyrifos, diazinon, malathion and carbaryl. Chlorpyrifos,
diazinon and malathion are thiophosphate (SP) insecticides; carbaryl is a
broad-spectrum insecticide used more than twice as often as any other
carbamate.
The criteria used here to select representative pesticides for the US
population include toxicity, frequency of use, frequency of detection, and
existence of highly exposed populations. One comparison of risk to
pesticides from indoor exposure has indicated that four chlorinated
insecticides (heptachlor, chlordane, aldrin, dieldrin) may pose the
greatest risk [Wallace, 1991]. However, this risk assessment used data
from only two cities and may have relied too heavily on data from
analytical methods that were better suited to detection of these relatively
non-polar pesticides rather than to more polar and acidic pesticides. In
addition, this risk assessment noted that inadequate information on
exposure and potency prevented a calculation of risk for chlorpyrifos and
diazinon, which, as noted there, are two of the most frequently detected
pesticides in indoor air. Evaluations by the Arizona Department of
Environmental Quality indicate that chlordane is found throughout the state
at low background levels. This pesticide is no longer sold. It was
replace by chlorpyrifos for treatment of termites. According to the
Uniform Building Code (employed throughout the state), the soil under the
foundation/slab of all homes must be treated with a termaticide. We
selected chlorpyrifos as it is the most common termaticide currently
employed. In our unpublished survey of home in Pima county, Arizona, aside
from the pyrethroids which are very unstable, diazinon is the most common
pesticide used inside homes. Malathion is used in the western part of the
state along the California border to prevent Mediterranean fruit fly
infestations of citrus. Carbaryl is also used heavily in agriculture in
the lower Colorado around Yuma. Current sales, frequency of use, and cost
effectiveness of analysis procedures dominated our selection criteria for
pesticides. The following discussion is provided as additional rational
chemical for the four pesticides that have been selected for the NHEXAS
program.
The four compounds selected here are known to have toxic endpoints and
suspected as having carcinogenic endpoints. Acute toxicity for
organophosphate insecticides (organophosphate including phosphate OP and
thiophosphate SP insecticides, e.g. diazinon, chlorpyrifos, malathion)
involves acetylcholinesterase inhibition, with nausea, vomiting,
brachycardia, tachycardia, ataxia, and paralysis [Salem and Olajos, 1988].
Chronic exposure can lead to a "dying-back" of the peripheral nervous
system. While both diazinon and chlorpyrifos are SP (thiophosphates),
rather than strict OP (organophosphates), both pesticides are oxidized to
-------
Section No. 1,2.2
Revision #0
June 1995
Pages 8 of 116
the corresponding OP (e.g. diazinon) in the environment [Glotfelty et al.,
1990] and these OP forms are highly toxic and exhibit about 10,000 fold
higher acetyl cholinesterase inhibition [Fujii and Asaka, 1982]. Diazinon
use in the garden or orchard has been associated with increased incidence
of brain cancer in children, relative to cancer controls and adjusted for
ETS exposure, income, and education [Davis et al., 1993]. Diazinon has
been banned for use on golf courses and sod farms because of high rates of
water-fowl kills related to toxic run-off [OS Senate, 1991].
Not only are the toxicity and other adverse health effects of these
pesticides clearly established, but these specific pesticides are among the
most widely used and frequently detected pesticides in the U.S. The NOPES
study found the insecticides chlorpyrifos and diazinon in the indoor air of
over 80% of the 260 homes investigated [Lewis et al., 1988]. The HIPES
study, which focused on the house dust and indoor air of 9 homes, also
found chlorpyrifos to be the most frequently detected pesticide, with the
absence of diazinon from the HIPES samples attributed to analytical
difficulties [Fortmann et al., 1991]. The results of the NHANES II study,
where the urine of nearly 6000 O.S. residents was analyzed for the residues
of a variety of pesticides, confirm the exposure of the population to these
pesticides [Murphy et al., 1983]. Metabolites of chlorpyrifos and diazinon
were detected in 6-7% of the urine samples. Pesticides have also been
detected in food and municipal drinking water, indicating numerous routes
of exposure [Schattenberg and Hsu, 1992; Lin et al., 1981].
The NOPES study has suggested that outside of the environment defined by
farm and agricultural workers, exposure to pesticides is defined primarily
by use in and around the home. The pesticides selected here are used
extensively for residential and community applications. Because these
pesticides are also used for agricultural purposes, farm and non-farm
populations can be evaluated in NHEXAS. The NOPES and HIPES studies have
shown that for the residential environment, pesticides are present in
indoor air and dust at higher levels than are found in outdoor soil and
air. These studies suggest that for the non-farm environment, indoor
exposure to pesticides is more important than outdoor exposure and that
inadvertent dermal contact and ingestion of contaminated house duet may
account for significant pesticide exposure in young children.
The highly exposed populations for diazinon and chlorpyrifos are those
which routinely treat the home for insect control, either with bombs,
sprays, no-pest strips, crack-and-crevice treatments or commercials
applications. Those homes located near orchards have been identified as
potential highly exposed populations. Homeowners who make their own
application may be the more highly exposed population, both from
inadvertent misuse and contaminants that brought into the home on clothing.
Chlorpyrifos, Diazinon and Malathion:
Chlorpyrifos, diazinon and malathion are thiophosphate pesticides (SP) used
extensively throughout the U.S. in agriculture, on turf [Nishioka et al.,
1992], in community pest eradication programs [Weiskopf et al., 1988] and
indoors. Malathion attained notoriety when sprayed from aircraft in
-------
Section No. 1,2.2
Revision #0
June 1995
Pages 9 of 116
southern California as part of the Mediterranean fruit fly eradication
program. The potential exposure pathways for chlorpyrifos, diazinon and
malathion are inhalation, dermal contact and ingestion, particularly since
SP are frequently used indoors. The potential pathways of entry from the
outdoors into the indoor environment include airborne routes such as
spray-drift and resuspension from soil, and track-in of soil residues.
Exposure to SP in food may be expected as these insecticides have been
detected in diverse produce [Iverson et al., 1975]. Exposure to home-grown
produce may represent a dietary exposure route. Although dust-bound
residues may constitute a significant source of exposure for young
children, indoor use of SPs could result in exposure pathways such as
inhalation, dermal contact and ingestion of solid aerosols with higher
pesticide concentrations than those in dust- bound residues. These
compounds are semi-volatile organic compounds (SVOCs) with little
information currently available on the distribution between the vapor and
particle phases and the temporal variability of the phase distribution.
Biomarkers of exposure to chlorpyrifos and diazinon have been established
from the NHANES II study [Murphy et al., 1983]. Chlorpyrifos exposure is
indicated toy the metabolite 3,5,6-trichloro-2-pyridinol, while
diethylphosphorothionate (DETP) could result from both diazinon and
chlorpyrifos exposure. The phenolic metabolite in urine specific for
diazinon, 2-isopropyl-6-methylpyrimidin-4-ol, may be detected by a method
similar to that used for detecting the phenolic chlorpyrifos metabolite
[Grover et al., 1986].
Ca.rba.ryl:
5
Carbaryl (chemical name: 1-naphthol-N-methylcarbamate) is a wide-spectrum
carbamate insecticide that is used to control over 100 species of insects
on citrus, fruit, vegetables, lawns, and ornamentals, as well as on pets
and poultry [Extoxnet, 1989]. It is formulated commercially as baits,
dusts, granules, wetable powders, flowables, and aqueous dispersions
[Carbaryl, 1988]. The widespread use of this insecticide makes it a prime
candidate for evaluation.
Carbaryl is moderately to very toxic [Extoxnet, 1989]. Carbamate poisoning
is similar to organophosphate poisoning, since both Inhibit
acetylcholinesterase [Salem and Olajos, 1988]. It can produce adverse
effects in humans by skin contact, inhalation, or ingestion. Inhalation or
ingestion of very large amounts can be toxic to the nervous and respiratory
systems resulting in nausea, stomach cramps, diarrhea, and excessive
salivation. Other symptoms at high doses include sweating, blurring of
vision, loss of coordination, and convulsions.
Although carbaryl does cause skin and eye irritation, it does not appear to
be a significant chronic health risk at or below occupational levels
[Carbaryl, 1988]. It is a weak mutagen and has only low teratogenic
potential [Extoxnet, 1989; Carbaryl, 1988]. Carbaryl has shown no signs of
carcinogenicity in long-term and lifetime studies of mice and rats
[Extoxnet, 1989; NTI, 1968]. Most animals, including humans, readily break
down carbaryl and rapidly excrete it in the urine and feces. Workers
-------
Section No. 1,2.2
Revision #0
June 1995
Page: 10 of 116
occupationally exposed by inhalation to carbaryl dust were shown to excrete
74% of the inhaled dose in the urine in the form of a breakdown product
[Extoxnet, 1989].
Carbaryl has a half-life of 7 days in aerobic soil and 28 days in anaerobic
soil. Degradation in the soil is mostly due to sunlight and bacterial
action. The half-life of carbaryl in pond water is 1 - 32 days, and in air
the half-life is 1 - 4 months.
1.2.3 Target VOCs
Primary VOCs: Benzene, Formaldehyde and 1,3-butadiene
6
Secondary VOCs: Toluene, Trichloroethylene (TCI) and Total VOCs
The selection of our primary and secondary VOCs is based on data gathered
by EPA and others [Wallace, 1987; Lofroth et al, 1989; Sheldon and
Jenkins, 1990; Wallace, 1991; Cooke, 1991; Hodgson and Wooley, 1991;
Dann and Wang, 1992; Lofgren and Petersson, 1992; Wallace, 1992].
A number of studies have documented the important contribution of indoor
personal activities and consumer products to personal exposures to VOCs
[for example, Wallace, 1992; Clobes et al, 1992]. Benzene is one of the
VOCs that has been studied most extensively. Personal exposures (personal
air and breath) to benzene as well as its indoor and outdoor concentrations
have been well characterized [Wallace, 1987]. Although inhalation is the
dominant pathway of human exposure [Hattemer-Prey et al., 1990], benzene
and other toxic VOCs are also found in water. In certain situations, this
exposure can be appreciable. For example, it has been shown that
residential use of benzene-contaminated water, such as in a hot shower, may
result in significant inhalation and dermal exposures [Shehata, 1985;
Lindstrom et al, 1992]. A ubiquitous compound, benzene is considered a
human carcinogen and has been estimated to have one of the highest
upper-bound lifetime cancer risks of all VOCs evaluated [Wallace, 1991;
Dann and Wang, 1992]. One recent study concluded that benzene contributes
almost 20% to the computed overall VOC-related lifetime cancer risk [Dann
and Wang, 1992].
We have also selected formaldehyde and 1,3-butadiene as primary VOCs for
investigation because: (1) they are important as potential health hazards,
(2) they are prevalent in air (indoor and ambient) and house dust
[Kirchherr et al., 1992], and (3) we lack adequate exposure information
on either compound. The individual lifetime cancer risk estimates for
formaldehyde and 1,3-butadiene are of the same order as, or possibly even
higher than, the airborne risks calculated for benzene [Wallace, 1991]. A
more recent study suggests that 1,3-butadiene and formaldehyde, along with
benzene, make the greatest relative contribution to overall VOC-related
cancer risk, namely, 41% for 1,3-butadiene, 18% for benzene, and 15% for
formaldehyde. However, there still appears to be uncertainty about the
carcinogenic potency of formaldehyde. Also, the lack of personal exposure
or indoor concentration data for 1,3-butadiene make the risk estimates for
both of these compounds extremely speculative [Wallace, 1991].
-------
Section No. 1.2,3
Revision #0
June 1995
Pages 11 of 116
Arizona had an extensive aviation industry including major air force bases.
In the last 10 years TCE has been identified as a pollutant of ground
water, a major component of the domestic water supply. Further, recent
malfunctions in the water clean-up site have resulted in atmospheric
release of TCE. This is a VOC of major local concern and plays a role in
the selection of this locally significant VOC. We expect examination of
this VOC to enhance subject participation. The secondary VOCs (e.g., TCE)
occur extensively in drinking water in certain locations, and are currently
of great concern because of indications that they are potential teratogens
[ATSDR; Goldberg and Lebowitz]. Information on their occurrence and
concentrations may be obtained from city water department records in most
larger cities in the U.S. Toluene has also been included because it is
toxic and can be used to determine the sources of benzene (from the
benzene/toluene concentration ratio).
Most earlier personal exposure studies employed sampler tubes containing
Tenax sorbent through which a known volume of air was drawn with a personal
sampling pump. Because Tenax does not retain very volatile compounds, such
as 1,3-butadiene, and cannot be used to trap reactive compounds, such as
formaldehyde [WHO, 1987], no large-scale personal exposure studies have
been carried out to assess exposure to these target compounds.
Benzene:
Exposure to benzene occurs mainly as a result of active and passive
smoking, vehicle exhaust (including driving and other personal activities
associated with motor vehicles, use of attached garages for parking cars),
storing gasoline, the use of certain consumer products, and exposure to
aerosolized water-containing benzene. High doses affect the central
nervous system, while exposure levels as low as 1 ppm are reported to
result in a high risk of leukemia and anemia in children. Studies have
shown that children of smokers die of leukemia at more than twice the rate
of children of nonsmokers [Wallace, 1991]. Techniques to monitor benzene
in indoor and outdoor air are well- established [Wallace, 1987]; benzene
also provides clear-cut biomarkers for direct assessment of exposure and
risk. The unmetabo1ized parent compound in exhaled breath [Wallace, 1987;
Gordon et al., 1988; Gordon 1990; Gordon et al., 1992; Pellizzari et
al., 1992; Wallace et al., 1993] serves as a biomarker of exposure whereas
the urinary metabolite trans, trans-rauconic acid is a biomarker of dose
[Buckley et al., 1992; Ducos et al., 1992].
Formaldehyde:
Because of its widespread use, formaldehyde is one of the most frequently
measured compounds in indoor air studies [WHO, 1987; Otson and Fellin,
1992; Moschandreas and Gordon, 1991]. Potential sources of formaldehyde
exposure include building materials (urea-formaldehyde foam insulation,
pressed wood products, particle board, adhesives, carpeting, new
furniture), consumer products (cleaners, fabric softeners), permanent-press
fabrics, paper products, cosmetics, and incomplete combustion (vehicle
exhaust, emissions from gas stoves, burning cigarettes, wood smoke) [Cooke,
1991], Indoor air concentrations are generally significantly higher than
-------
Section No. 1.2.3
Revision #0
June 1995
Page: 12 of 116
outdoor concentrations. Recent emissions test data suggest that the
emissions of formaldehyde increase significantly (by 20-75% depending on
the emission control technology being used on the test vehicle) when either
methyl t-butyl ether (MTBE) or methanol/ethanol blended (alternative
oxygenated) fuels are used [Anderson et al., 1993]. These researchers
estimate that, in Denver, motor vehicle emissions are the major source of
formaldehyde during the winter (>1 ton of formaldehyde per day). Human
exposure to formaldehyde is principally through inhalation and skin
absorption or less frequently by ingestion [Sittig, 1985]. Two
subpopulations have been identified as having particularly high potential
for formaldehyde exposure, namely, residents of mobile homes containing
particle board and plywood, and persons living in conventional homes
insulated with urea-formaldehyde foam. In addition to its designation as a
suspect carcinogen and mutagen, formaldehyde in the gas phase is an eye,
nose, and throat irritant and a skin irritant in liquid form [WHO, 1989].
Formaldehyde is rapidly metabolized in humans to form formic acid.
Although it has been widely monitored in indoor environments, direct
measurements of personal exposure to formaldehyde are relatively sparse
[Chan et al., 1991].
1,3-butadiene:
Environmental tobacco smoke from burning cigarettes is an important indoor
source of 1,3- butadiene [Lofroth et al., 1989]. Vehicle emissions are a
major outdoor source of this compound [Lofgren and Petersson, 1992], and
vehicle exhaust and evaporative emissions of fuel from vehicles may enter
the living space of houses with attached garages [Hodgson and Wooley,
1991]. Low levels of 1,3-butadiene have been measured in ambient urban air
[Lofgren and Petersson, 1992; Cote and Bayard, 1990] and it is regarded as
a probable human carcinogen by U.S. EPA [Hallenbeck, 1992]. However, data
on indoor concentrations of 1,3-butadiene are very limited. A single pilot
study conducted recently to measure indoor concentrations and personal
exposures for several air toxics indicated that outdoor concentrations for
1,3-butadiene were generally higher than indoor levels [Sheldon and
Jenkins, 1990]. Like benzene, 1,3-butadiene is not very soluble in water.
Thus, unmetabolized parent compound may serve in much the same way as a
biomarker of exposure. Urinary metabolites of 1,3-butadiene that may be
used as biomarkers of dose are the mono- and diepoxides as well as the
recently identified 1,2-dihydroxy-4-(N-acetylcysteinyl)butane [Sabourin et
al., 1992],
Our literature evaluation of these pollutants (see Table 1) suggests they
will be good choices for evaluation within the NHEXAS framework. As a
result we formed a consortium in response to EPA's NHEXAS RFP consisting of
the University of Arizona, Battelle Memorial Institute and Illinois
Institute of Technology (aka NHEXAS Arizona). We propose evaluating
exposure to the afore-mentioned metals, pesticides and VOCs using a
non-stratified, nested sampling strategy. The consortium will evaluate a
subset of the State of Arizona's population determined statistically as
described in Section 1.7.
-------
Section No. 1,2.3
Revision #0
June 1995
Page: 13 of 116
Table 1. Target analytes and their associated Chemical Abstract Service
{CAS} Registry Number.
Target Chemicals
CAS Registry Number
Primary Metals
Lead
Arsenic
Cadmium
Nickel
Chromium
Secondary Metals 1
Barium
Manganese
Selenium
Vanadium
Copper
Zinc
Primary Pesticides
Chlorpyrifos
Diazinon
Secondary Pesticides
Malathion
Carbaryl
Primary VOCs
Benzene
1, 3-butadiene
Formaldehyde
Secondary VOCs
Total VOC
Toluene
Trichloroethylene
1»1-Dichloroethylene
D ichloromethane
1.1-Dichloroethane
cis-1,2-Dichloroethylene
Trichloromethane
1.2-Dichloroethane
1.1.1-Trichloroethane
Carbon tetrachloride
1,2-Dichloropropane
1.1.2-Trichloroethane
Tetrachloroethylene
Chlorobenzene
Ethylbenzene
m-Sp-Xylene
Styrene
1,1,2,2-Tetrachloroethane
o-Xylene
m-Dichlorobenzene
p-D ichlorobenzene
o-Dichlorobenzene
7439-92-1
7440-38-2
7440-43-9
7440-02-0
7440-47-3
7440-39-3
7439-96-5
7782-49-2
7440-62-2
7440-50-8
7440-66-6
2921-88-2
333-41-5
121-75-5
63-25-2
71-43-2
106-99-0
50-00-0
No CAS Number
108-88-3
79-01-6
75-35-4
75-09-2
75-34-3
156-59-2
67-66-3
107-06-2
71-55-6
56-23-5
78-87-5
79-00-5
127-18-4
108-90-7
100-41-4
108-38-3 & 106-42-3
100-42-5
79-34-5
95-47-6
541-73-1
106-46-7
95-50-1
-------
Section No. jj
Revision #0
June 1995
Page: 14 of 116
3 Project Scope and Work Objectives
Exposure assessment studies are generally limited by the number of media
and pathways evaluated for any given pollutant. Our consortium will
demonstrate the gains in accuracy and precision of exposure assessment
models when multiple media (air, soil, dust, skin, water, food, blood and
urine) that accrue and multiple pathways (inhalation, ingestion,
absorption) are evaluated for the target contaminants described above.
Our NHEXAS project will evaluate methods and hypotheses of importance for
exposure assessment in the State of Arizona.
The evaluated population will be selected using a population-based
probability design for recruitment (see section 1.7.3). A considerable
literature base exists about exposures in the work-place. Our emphasis
will be placed on residential exposures; information about occupational
exposures will be limited to questionnaire responses by respondents. The
exception is occupational exposure by a small subset of subjects where
personal air is collected for metals and pesticides. Our overall project
goals are to:
A. Obtain data for exposure and risk assessment.
B. Determine total residential exposure from multiple exposure pathways.
C. Determine the high end (upper 10%) of the residential exposure from a
population-based probability distribution for metals (Pb, As, Cd, Ni,
Cr, Ba, Mn, Se, V, Cu and Zn) in the state of Arizona.
D. Determine the high end (upper 10%) of the residential exposure from a
population-based probability distribution for 4 pesticides
(chlorpyrifos, diazinon, malathion and carbaryl).
E. Determine the high end (upper 10%) of the residential exposure from a
population based probability distribution for total VOC concentrations
and evaluate compound specific VOC exposures (benzene, formaldehyde,
toluene, TCI and 1,3-butadiene) in a small population subset.
F. Use available methods and technologies to obtain technically valid
distributors of the selected pollutants.
3.1 Project Objectives related to scope include:
A. The most accurate way to fulfill the broad-based project objectives is
to evaluate all people, at all times, throughout the country. This is
not possible in terms of cost and practicality. As a result we must
select a representative subset of our target population. Our sampling
strategy to meet this goal is outlined in section 1.7.3.
B. Documenting the occurrence, distribution & determinants of total
exposures to selected metals, pesticides and VOCs (pollutants) in the
general population.
-------
Section No. | ^ j
Revision #0
June 1995
Page: 15 of 116
Specific pollutants include:
metals (Pb, As, Cd, Ni, Cr, Ba, Mn, Se, V, Cu and Zn),
pesticides (chlorpyrifos, diazinon, malathion and carbaryl), and
VOCs (benzene, formaldehyde, toluene, TCE and 1-3 butadiene and total
VOCs).
C. Characterizing the 90th percentiles of total exposures to each of the
pollutants in the general population.
g
D. Monitoring geographic and temporal trends of the multi-media exposures.
1. Evaluating exposures in a proportionate-based population sample for all
media, personal characteristics, time-activity, and geographic factors
contributing to total exposure.
P. Evaluating exposures in a proportionate-based sample as reflected in
blood and urine samples (biomarkers) by target pollutant concentrations.
G. Exploring the use of intensive/precise techniques in a small, nested
subset of the population with the goal of generalizing results to the
entire population.
Results from the intensive monitoring phase of this study will be
projected to the larger but less precise data of the earlier stages and
then to the general population of the state of Arizona. From this
project we will determine the feasibility of using the methodologies
employed in this feasibility study in larger population surveys (NHEXAS
Stage 2).
H. Objectively measuring time-space locations of subjects in small subsets
of the population to determine whether these methods add significant
information for the exposure assessment.9
I. Constructing predictive models of total exposure (at each survey level).
These models will help formulate initial exposure assessments based on
information currently available. They will be compared with successive
models generated by using the data that becomes available at each study
stage. Iterative model comparisons will enable investigators to assess
the cost/benefit of progressively more intensive and expensive analysis
for accurate exposure assessment.
J. Determining whether exposure can be predicted by questionnaire-based
predictive models which can be used to validly formulate relationships
between "explanatory variables" of exposure measured by questionnaire
or measured in media and exposure as evidenced by biomarker samples. We
will evaluate relationships between exposure reports, environmental
measurements, and biomarkers of target pollutants.
K. Evaluating current and needed information on pharmaco- & toxico-kinetics
for dosimetry based on our predictive models and information available
in the literature.
-------
Section No. 1.3,1
Revision #0
June 1995
Page: 16 of 116
L. Collaborating and coordinating the exposure assessment modeling with
EPA, the other NHEXAS phase 1 field studies, and the other federal
agencies involved.
1.3.2 The following Work Objectives and implementation steps must be met to
complete the projects
All procedures supporting the work objectives are divided into the
following project implementation areas, administrative, field, laboratory,
data management and data analysis. Figure 1 provides an overview and
Figures 2-6 report specific protocols and their relationship to the
project.
A. Preparation—Work Objectives
1. Completion of the contributory documents for the Office of Management
and Budget (OMB); review of the ICR including finalization of the
questionnaires. (Az PI & Co-PI will interact with EPA project
officers and RTI/EOHSI Consortium).
2. Completion of the QSIP and associated Standard Operating Procedures
(SOPs? UA-G-1.0). These documents will define responsibility for all
work done in the project and the QA measures required as part of task
completion. (The PI and c-PIs of the Arizona Consortium interact with
each other and the EPA project officers to develop an appropriate
QSIP).
3. The selection of the study population occurs. We will use a three
phase sampling approach to identify the housing units to be sampled
(see section 1.7.3). Although the sample design is nested, none of
the subsamples (stages) will be stratified. Each stage will be a
subset of the preceding stage.10
4. Assembly of the necessary equipment, sampling and analytical
methodologies.
5. In-house, pre-field evaluation of all equipment and supplies (field,
lab & data). Development of any trial in-house laboratory forms.
(Consultation with EPA personnel, State DEQ, discussions within the
consortium; PI, Co-PIs).
6. Hiring and training additional personnel to fulfill project needs in
the field, laboratory, data and data analysis components of the
project. Certifying all project personnel for performance of
specific tasks. Training and documentation11 of the field staff
including interviewers and technical support is discussed in SOP
UA-T-1.0 through UA-T-4.0. Data assistant and lab training plans
are described as UA-T-5.0 and UA-T-6.0 respectively. (Consultation
within the Consortium re: personnel availability; Co-PIs Arizona &
Battelle will cooperate in training field personnel. Each consortium
-------
Section No. 1.
Revision #0
June 1995
Page: 17 of 116
Figure 1. Major classes of Standard operating Procedures in support of
NH1XAS Arizona.
Overall Procedures Supporting Work Objectives
Custody
Archive
Dust
Soil
Dermal
Biological
General
Sample Archive
Instrument
Operation &
Calibration
System Definition
Form Coding
Data Entry
System Definition
& Maintenance
Administrative
Procedures
Field
Collection
Procedures
Laboratory
Procedures
Data
Management
Procedures
Data
Analysis
Procedures
General
Training
Air
Water
Food/Beverage
Analysis
Data Cleaning
& Verification
Data Validation
Data Appendage
Data Transfer
& Receipt
Statistical
Procedures
& Analysis
-------
Figure 2. SOPs included in Administrative Procedures.
internal QC requirements for that procedure.11
Section No. 1.3.2
Revision #0
June 1995
Pages 18 of 116
Each SOP addresses
UA-C-1.0
UA-C-2.0
UA-C-3.0
UA-C-4.0
UA-C-5.0
UA-C-6.0
UA-C-7.0
UA-C-8.0
UA-D-28.0
A - *
> Custody
Administrative
Procedures
General
e>Z0-Q-ho
G ~2*t>
Archive
UA-G-4.0
~B€0=€-'1 0-
mifiL
6 CO-£7 -3-0
Training4
q 11
UA-G-1,0
UA-G-2.0
UA-G-3.0
UA-G-5.0
UA-T-1.0
UA-T-2.0
UA-T-3.0
UA-T-4.0
UA-T-5.0
UA-T-6.0
BCO-T-l.Q
Uft -b Hi. O
-------
Section No. 1.3,2
Revision #0
June 1995
Page: 19 of 116
Figure 3. SOPa included in Field Collection Procedures associated with
metals, pesticides and VOCs. lach sop addresses internal QC
requirements for that procedure.13
metals
pesticides
* VOCs
UA-F-7.0
UA-F-8.0
UA-F-21.0
UA-F-7.0
UA-F-8.0
N/A Dust
UA-F-S.Q
UA-F-6.0
UA-F-21.0
UA-F-5.0
UA-F-6.0
N/A Soil
UA-F-9.0
UA-F-9.0
N/A Dermal
UA-F-19.0
UA-F-20.0
UA-F-19.0
UA-F-20.0
UA-F-19.0 . . ,
UA-F-20.0 Blol°glcal
Field
Collection
Procedures
AIR
metals
UA-F-3.0
UA-F-4.0
UA-F-10.0
UA-F-14.0
Water UA-F-16.0
Food UA-F-15.0
pesticides
UA-F-3.0
UA-F-4.0
UA-F-10.0
UA-F-14.0
UA-F-17.0
UA-F-15.0
VOCs
UA-F-11.0
UA-F-12.0
UA-F-13.0
UA-F-18.0
aA'f-23'0 UA'f-23,0
icc^on o( - Ufl-F-2z<0
9//5yfs""
-------
Section No. 1.3.2
Revision #0
June 1995
Page: 20 of 116
Figure 4. SOPs included in Laboratory Procedures associated with tasks and
analyte classes. Each SOP addresses internal QC requirements
associated with the Procedure.'
12
UA-L-2.0
UA-L-3.0
UA-L-4.0
UA-L-5.0
UA-L-11.0
UA-G-4.0
UA-G-5.0
UA-L-12.0
UA-L-13.0
BCO-G-2.0
BCO-G-3,0
-8CO-G-4©-
General
1 Sample
J Archive
>
ua-l-6.0
UA-L-7.0
ua-l-8.0
UA-L-10.0
bco-l-2,0
Bco-L-5.0
BCO-L-6.0
BCO-L-7.0
Bco-L-8.0
BCO-L-9.0
BCO-L-IO.O
BCO-L-18.0
5CO-L-23.0
BCO-L-24.0
it/8 jqT
Instrument
Operation &
Calibration
Laboratory
Procedures
metals
Analysis
AIR
SOIL
DUST
DERMAL
FOOD
WATER
( EPA
methods)
UA-L-9,0
BCO-L-3.0
BCO-L-l.O
BCO-L-3.0
BCO-L-l.O
BCO-L-3.0
BCO-L-3.0
BCO-L-4,0
BCO-L-19.0
fOI
102.
TM »°3
200,8
pesticides
BCO-L-ll.O
bco-l-kt2^,
BCO-L-14.0
BCO-L-13.0
BCO-L-14.0
6 CO-L-24.
BCO-L-12.0
BCO-L-2Q.O
PDA 202
FD* 203
Fb# aexf
525$%
VOCs
BCO-L-16.0
BCO-L-17.0
BCO-L-18,0
BCO-L-22.0
N/A
N/A
s~
N/A
Biological
524.2
-------
Section No. 1.3.2
Revision #0
June 1995
Page: 21 of 116
Figure 5. SOPs included in Data Management Procedures. Each SOP addresses
internal QC requirements for that procedure.14
UA-D-1.0
UA-D-2.0
UA-D-3.0 } System
r t a a n J Definition
UA-D-4.0
l/(rD i UA-D-28.0
it U A * D * 3o 10
/3 UA-D-15.0 Data Entry
UA-D-5.0
UA-D-6.0
UA-D-7.0
t r > n ci a
UA:D-s.u—
UA-D-9.0 1 F Coding
UA-D-10.0 / s
UA-D-11.0 a , .
«A=©=tt^rVW3PU
UA-D-13.0
UA-D-14.0
uft-b-31.0 —
H A-0-3T-0
im -v yi'O
Data
Procedures
Data Cleaning
& Verification
\ UA-D-16.0
/ UA-D-17.0
r UA-D-18.0
{ UA-D-20.0
1 UA-D-21.0
UA-D-22.0
VA-©-J3t9
UA-D-24.0
ua-o-M-v
y A -b -
dA:e> -58.o
Ho. o)
Data Validation
UA-D-25,0
UA-D-26,0
Data Appendage UA-D-27.0
Data Transfer
& Receipt
UA-D-29.0
-------
Section No. 1,3.2
Revision #0
June 1995
Page: 22 of 116
Figure 6. Data Analysis Procedures. Each SOP addresses QC internal
requirements for that procedure.15
IIT-A-LG
IIT-A-2.0
IIT-A-3.0
System Definition
& Maintenance
Data
Analysis
Procedures
Statistical Approaches
Univariate Distributions
Bivariate Distributions
Multivariate Distributions
Univariate Analyses
Bivariate Analyses
Multivariate Analyses
Multivariate Distributions by Domain
Multivariate Distributions over Domains
Multivariate Analyses by Correlate Predictors
TEA Predictive Models
Multivariate Analyses of Predictive Models
-------
Section No. 1.3.2
Revision #0
June 1995
Page: 23 of 116
member is responsible for training and certifying its additional
support personnel.)
Begin the Project Documentation Form Notebooks specified in the SOPs
for each area (Packet prep, Field, Lab(s), Data and Analysis). Begin
a Master Project Form Notebook (in the Data Section) that contains a
copy of all forms to be used in the project. These will be
supplemented with any new form from field, lab and data sections.
Ideally, there should be no form changes within the study. As a fail
safe mechanism, a copy of every form (and every version) used in the
project should be kept in these notebooks maintained by the Project
Field and Data Coordinators, and laboratory Supervisors. (University
of Arizona & Battelle, data, field & lab Coordinators cooperation).
Complete Dictionaries, Coding and Cleaning SOPs for Submitted
Questionnaire Version (UA—D-3-0 through UA-D—24.0; University of
Arizona, Co-PI & Data section, Coding input from EPA through
* 13
conference calls & personal interaction).
Build Data Entry Screens for Submitted Questionnaire Version
13
(University of Arizona, Data Section; UA-D-3.0).
Pre-assessment Studies:
a. Pre-field evaluation of the methodology in terms of equipment use,
timing and practicality, streamline the field process to minimize
the impact on project respondents and maximize the amount,
quality, precision, accuracy, and integrity of data and samples
gathered in the field. (University of Arizona, Co-PI & Field
Coordinator, Interaction and discussion within the Consortium and
with EPA consultation).
b. Storage and Shipping trials of any samples collected during the
pre-field evaluation (UA-G-1.0 through UA-G-4.0 plus handling
procedures described in Field and Lab SOPs; University of Arizona
Co-PI & Field Coordinator; Interaction within the consortium and
with cooperating labs analyzing food, water and tissue).
c. In-house lab trials to evaluate custody transfer, sample storage
and pre-coded laboratory forms (UA—C—1.0 through UA-C-8.0;
University of Arizona-—Field Coordinator and Lab Supervisor &
Battelle; Interaction and discussion within the consortium).
d. Revision or further development of the QSIP or any associated
SOPs including revised field forms. (PI & Co-PIs University of
Arizona, Battelle & IIT; Interaction within the consortium and
with cooperating labs analyzing food, water and tissue).
e. Review and finalize field and lab forms and determine those that
must be maintained as databases (as opposed to filed field
records).(Internal University of Arizona).
-------
Section No. 1.3.2
Revision #0
June 1995
Page: 24 of 116
11. Build necessary data entry screens for field forms. Complete
Dictionaries, Coding and Cleaning SOPs for any field forms.
(UA-D-3.0; University of Arizona, Interaction and discussion within
the Consortium and with EPA consultation).
12. Build data Entry Screens for any in-house laboratory data forms.
Complete Coding, Dictionaries and Cleaning SOPs for any in-house lab
data. (UA-D-3.0 through UA-D-26.0; University of Arizona,
Interaction and discussion within the Consortium and with EPA
consultation) .13
13. Retrieve and manipulate data collected by other sources that may
complement or supplement NHEXAS data and be used in modeling. This
includes databases like weather, County and State Department of
Environmental Quality regional pollutant databases. Build
dictionaries, complete appropriate documentation and build Master
Databases. (UA-D-4.0 through UA-D-27.0; Data Section, University of
, 13
Arizona, Interaction and discussion within the Consortium).
14. Prepare "Adult" and "Minor Child" written consent forms describing
the respondent risks and benefits. Outline precisely respondent
burden in terms of time, environmental and biological samples. State
in broad terms the analyses to be performed on the samples. If the
respondent cannot read, read the consent form to him/her. Be sure
the respondents understand what they are agreeing to do. Describe
the specifics of subject confidentiality. Obtain the signature of
the adult or the minor child and legal guardian prior to collecting
the Baseline Questionnaire. All communications, consents, assents
and questionnaires will be available in English and Spanish.
~Notes Completion of the preliminary work should coincide with the
approval of the OMB Package.
15. Complete the application for the Internal Review Board (Human
Subjects) and attach the questionnaires and the consent forms.
16. After IRB approval, immediately format and print the approved
questionnaires (Field Coordinator, University of Arizona).
17. If the final approved questionnaire has been modified by OMB, make
the appropriate changes in the data dictionaries, the data entry
screens, the coding and cleaning instructions and SOPs (UA-D-3.0
through UA-D-25.0; Data Coordinator, University of Arizona).13
B. Implementation of Field Survey—Work Objectives:
1. Field Sampling
a. Recruitment (Arizona, PI Co-PI & Field Coordinator); Statewide,
1200 households16will be asked to participate in the NHEXAS Arizona
project. They will be mailed a letter describing the study and
informing them that an interview team will visit their home. The
-------
Section No. 1.3.2
Revision #0
June 1995
Page: 25 of 116
interview team will visit the home and a short Descriptive
Questionnaire will be completed at these homes (10% will be
readministered using a telephone follow-up). A minimum of 75%
response is mandatory (900 homes). Interview techniques and
questionnaires completion are covered in UA-T-1.0 through
UA-T-2.0. Recruitment will be representative of the population
groups living in the State of Arizona (Table 2).
b. Baseline Questionnaire Administration (Arizona, PI Co-PI & Field
Coordinator); Of the 1200 homes invited to participate, Baseline
Questionnaires will be administered to the Primary Respondent in a
minimum of 900 homes.17
c. Initial Sampling (UA-F-1.0 through UA-F-7.0; Arizona, PI Co-PI &
Field Coordinator): Households will be selected for sampling to
minimize the sampl^8 design effect (section 1.7.3). Using this
strategy, 450 homes will be approached for environmental sampling
using less intensive sampling procedures and additional
questionnaire administration.
d. Intensive & Temporal Sampling8 (UA-F-1.0 through UA-F-20.0;
Arizona, PI Co-PI & Field Coordinator): 175 homes1 will be
selected to minimize the sample design effect (section 1.7.3)
from the 450 homes of stage 2 and asked to participate in the more
intensive standardized environmental sampling (see Table 3). They
will be subject to environmental monitoring and asked to complete
all questionnaires. A randomly selected subset of 50 stage 3
homes will be asked to participate in the same stage of sampling
during two additional seasons to determine variation in pollutant
concentrations associated with seasonal differences.
20
2. Analysis of Samples—Lab Processing (UA) and Sample Shipment to
Battelle & Other Labs (Arizona). These procedures run concurrently
with Field sampling (Section 1.3.2, B 1)—Work Objectives (p. 24).
a. Sample Custody & Integrity (UA-C-1.0 through UA-C-8.0; Arizona,
Battelle and other Labs): All samples are accompanied by custody
sheets and signed by the relinquisher and the recipient at each
stage after collection. Custody of samples collected in the field
transfers to the Project Field Coordinator. He is responsible for
transfer of the samples and custody forms to the appropriate labs
and for sample integrity evaluation and maintenance prior to
shipment or transfer.
b. Sample Analysis (Arizona): Samples to be analyzed weighed and/or
aliquoted in Arizona (soil and dust) are transferred to the Lab
Supervisor who assigns analysis and QC checks the analysis and
completed data forms (UA-L-9.0 through UA-L-10.0). Sample residue
is appropriately stored or shipped to Battelle (BCO-L-1.0 through
BCO-L-22 .0) by the laboratory supervisor for further analysis.
The lab Supervisor provides the data coordinator with a copy of
-------
Section No. 1,3.2
Revision #0
June 1995
Page: 26 of 116
Table 2. Race and ethnicity by county based on the 1990 census data.
Census Race Response by County
Spanish Descent
County
White
% %Native %Asian
Black American or Isl.
Apache
20.3
0.2
77.6
<0.1
Cochise
81.5
5.2
1.2
2.2
Coconino
64.5
1.3
29.3
0.7
Gila
76.6
0.2
13.1
0.3
Graham
77.6
1.6
14.7
0.6
Greenlee
86.4
0.3
1.9
0.6
La Paz
74.7
0.3
17.7
0.6
Maricopa
84.9
3.5
1.8
1.7
Mohave
95.3
0.1
2.3
0.7
Navajo
44.0
1.0
52.2
0.3
Pima
78.9
3.1
3.0
1.8
Pinal
74.9
3.1
9.6
0.6
Santa Cruz
74.9
0.2
0.2
0.4
Yavapai
95.8
0.2
1.6
0.5
Yuma
75.7
2.8
1.5
1.3
%
Other
1.9
10.0
4.2
9.8
5.3
10.7
6.7
8.1
1.6
2.5
13.1
11.8
24.3
1.9
18.8
% % Non
Hispanic Hispanic
3.9
28.4
10.1
18.4
24.6
42.8
24.7
16.0
5.0
7.1
12.4
29.4
77.1
6.4
41.8
96.1
71.6
89.9
81.6
75.4
57.2
75.3
84.0
95.0
92
87
70
22
93
58
-------
Section No. 1.3.2
Revision #0
June 1995
Page: 27 of 116
the custody form indicating sample shipment or transfer to
storage.
c. Sample Analysis (elsewhere, e.g. Battelle (BCO-L-1.0 through
BCO-L-IO.0) or the Federal Labs): Interim sample storage will
conform with requirements specified in the various SOPs.
Samples will be batched and shipped in a timely fashion according
to specifications in the SOPs by the Project field Coordinator at
Arizona. Samples to be shipped to Battelle will include: soils,
dust, air filters (PM10, PUP, Carbo traps, OVM badges,
Formaldehyde badges, personal air samples).
d. Lab Data Processing (UA-D-25.0; Arizona): Laboratory data sheets
are copied by the Lab Supervisor and copies are placed in the lab
notebook. Original data sheets are assembled by type and
transferred to the Project Data Coordinator at weekly intervals.
e. Lab Data Processing (UA-D-4.0 through UA-D-27; from elsewhere):
Laboratory results will be sent to the Principal Investigator at
the University of Arizona from other labs. Data can be sent in
two forms. (1) Copies of the laboratory data sheets are given to
the Project Data Coordinator, assembled into packets and batches
and processed as the Arizona Lab data. (2) Electronic data and
data format definition are provided to the Project Data Manager
for creation of appropriate master data bases. EPA-Cincinnati
(Maurice Berry) will serve as a Clearing House for data analyzed
by other agencies (CDC, FDA and water contract labs) and transfer
data to the consortia under EPA ORDER 2180.2).
3. Reduction and Preparation of Data—Questionnaires, Laboratory Data
and Field forms [These procedures run concurrently with work
described above (Section 1.3.2, B 1 & B 2)]—Work Objectives (p. 24).
a. QC & QA checks of forms: Questionnaires, Laboratory Data and
Field forms will be QC checked in the field by field staff for
accuracy and completions. A copy of the field form is retained in
the field office. All Questionnaires and Field forms will be QA
checked by the Project Field Coordinator within 24 hours of return
to the Field Office. These checks will ensure Field QC was adhered
to and the forms are complete. Laboratory Data will be QA checked
by the Project Laboratory Supervisor at the end of each day of
laboratory analysis for adherence to all QC/QA procedures
specified in the individual SOPs (see all Field and Laboratory
SOPs, Arizona & Battelle).
b. Data Packet Assembly: Field Forms and Questionnaires will be
collected into household packets and passed to the Project Data
Coordinator at the end of each week (UA-C-3.0, Arizona).
c. Data Processing: The Project Data Coordinator will batch data
Field and Arizona lab data (UA-C-4.0 through UA-C-8.0), assign
data coding, entry, 100% verification and validation tasks.
-------
Section No. J . 3 . 2
Revision #0
June 1995
Page: 28 of 116
Processed data batches are assembled into master databases by the
Project Data Manager (UA-D-27.0; Arizona).
At Battelle the project Lab supervisor will assign the task of
data coding, entry, verification, validation and creation of a
master data base of laboratory measurements. Battelle (and
National Labs) will calculate concentrations of pollutants and
transfer the completed data set, plus formatting and all key
variables to Principal Investigator at the University of Arizona.
These data will be handled as a supplemental data base by the
Project Data Manager (D).
21
d. Data Processing of Supplemental Data Bases: Public record data
will be sought for various analysis. This will mean re-formatting
databases created by others to make them compatible with our
NHEXAS databases. Databases currently identified include: weather
data (NOAA), data bases from Air Quality Districts in the State of
Arizona, data from FDA reporting contaminants in commonly consumed
food using a mini-market basket approach and complete analysis
databases provided by cooperating laboratories. These data may be
used in modeling (UA-D-4.0 through UA-D-27.0; Arizona—Project
Data Manager).
e. Final Data Base QA Check: Logged, batched data and appropriate
custody forms are filed in the project data offices. Each
processed data batch is examined for errors. Ten percent of the
forms are randomly selected for final 10% data checks. The batch
is deemed to be accurate if no more than 5% of the data is
erroneous. If systematic errors are found they can be documented
and corrected. More than 5% error rate requires examination of
50% of the data batch. If the 50% check has more than a 5% error
rate, the entire batch is re-processed as per step c (above)
(UA-D-26.0; Arizona— Project Data Coordinator).
f. Data Delivery: The project Data Manager will deliver data bases
to EPA (and others) in the specified format as instructed by the
Principal Investigator.
4. Analysis of Prepared Data—Questionnaires, Laboratory Data and Field
forms [These procedures run concurrently with work described above
(Section 1.3.2, Bl; B2 & B3)]—Work Objectives (p. 24).
a. Sample Population Representativeness (Arizona): Several forms of
analysis will be run throughout the project to ensure proportional
representation of all groups evaluated. Recruited population
within census blocks and tracts will be compared with the census
data. Demographics of nonparticipants will be compared with
census data to determine any bias in sampling. Participation in
each of the environmental sampling stages will be evaluated
relative to census statistics (Table 2) to evaluate potential bias
due to non response.
-------
Section No. 1 • 3•2
Revision #0
June 1995
Pages 29 of 116
21
b. Appropriateness of Methods (including internal Questionnaire
Evaluation) (Arizona & IIT): Intensive and less intensive methods
will be used in different study stages. Exposure distributions
generated using these different techniques will be compared and
the increment of "improvement" in the exposure assessment will be
evaluated.
21
c. Exposure Assessment (Arizona and IIT): "Body burden" measures of
target pollutants, as measured by the biomarkers, integrate
cumulative exposures over long periods of time. We will compare
the biomarker values with the models derived for individual
multimedia exposure using different levels of data collection
intensity (including the inter and intra-individual variability).
There will be many iterations of this process. One will consider
the entire population. Others will examine possible "explanatory"
relationships. These models will be particularly important in
examining the high end of the distribution. These analyses will
integrate all the data collected.
d. A project implementation overview is presented in Figure 7.
1.4 Project Description
This effort is a feasibility study in support of the National Human
Exposure Assessment Survey (NHEXAS) funded by the United States
Environmental Protection Agency. This component will be called NHEXAS
ARIZONA.
NHEXAS Arizona will assess exposure to selected metals, pesticides and
VOCs. In Arizona, exposure information will be gathered directly from
subjects, from environments frequented by subjects (primarily subject home
environments) and from public records?^Questionnaires will be employed to
characterize the study population, evaluate common practices believed to
contribute to exposures, and evaluate potential bias in the study due to
non participation. Blood and urine samples will be collected directly from
the subjects and concentrations of target pollutants will be measured.
Additional concentrations of target pollutants will be measured from the
air, dust, soil and water of the home environments of subjects. Duplicate
diets (regardless of food source) will also be collected. Public records
containing usable information on target pollutants (soil, air, and water
could be employed. Exposure assessment models will be generated using
direct and surrogate measures varying in the intensity of detail.
NHEXAS Arizona is broken into a series of field stages reflecting the work
objectives. For cost effectiveness and efficiency of sampling, all 5
stages will be operational at the same time in the field.10 Forms (or
questionnaire covers) and sample buckets will be color coded to help keep
protocols straight within each household. Concurrent stage sampling will be
essential to obtain the temporal variability gathered in field stages 4-5.®
-------
Section No. ] 4
Revision #0
June 1995
Pages 30 of 116
Figure 7. An overview of NHEXAS Arizona Implementation.
Materials
r echnician
PREPARED
MATERIALS
RECEIVED
QA
Field
Coordinator
Team
Leader
FIELD ASSIGNMENT:
HH3D
HELD PERSONEL
I
QA
I
FIELD
COLLECTION
QA
Field
Coordinator
FIELD FORM AND SAMPLE LOGIN
TEMPORARY SAMPLE STORAGE
Data
Coordinator
QA
Laboratory
Supervisor
PHYSICAL
DATA
FORMS
VALUES BATCHED
SENT TO KEYPUNCH
ENTERED
VERIFIED
VALUES RETURNED
CLEANING
VALIDATION
Material
Technician
ARIZONA
laboratory;
-QA
L
Data
Manager
APPENDED TO
MASTER DATABASE
QA
QA
DATA
READIED
FOR USE
DIRECT
SAMPLE
SHIPMENT
SAMPLE
ALIQUOT
QA
QA
BATTELLE
ANALYSIS
QA
ARIZONA
ALIQUOT
ANALYSIS
SAMPLE
ANALYSIS
T
QA
-$A
ARCHIVE
-------
Section No. 1.4
Revision #0
June 1995
Page: 31 of 116
STUDY PHASE 0 (pre-study population recruitment): Identification of sample
design components to sample proportionately, subject geographic
locations, and demographic characteristics.
STUDY PHASE 1/FIELD Stage 1: 1200 households*6will be selected as described
in section 1.7.3 Any adult household resident will be interviewed to
complete the Descriptive Questionnaire. The primary respondent will
be interviewed utilizing the NHEXAS Baseline Questionnaire in English or
Spanish at the discretion of the respondents. All other household
residents will be asked to self-complete the Baseline Questionnaire.
This is a survey of the permanent population of Arizona; part time
residents will be excluded. A 90% response rate for the first Q would
yield 1080; the requirement of a 75% (or better) response rate for the
second level, the Baseline Q, would yield a total of 900 plus houses.2
STUDY PHASE 2 (INITIAL SAMPLING)/FIELD Stage 2: For this stage, the target
population is the 900 (plus) households who completed Baseline Q's, and
are asked if they would like to participate further. Incentives offered
include free information about themselves and their micro-environment,
in relationship to the study as a whole. Additional questionnaires would
be completed and environmental sampling would take place in 50% of this
target population, i.e., about 450 households. There will be a primary
respondent in each household. Work-ups include (see Table 3):
Questionnaires (Q): These Stage 2 questionnaires are self-completed
or completion-assisted as requested by the subject. They include the
Activity Follow-up Q; The Time-activity Log; the Diet Diary (for
one week); and the Technician Questionnaire. All questionnaires are
available in English or Spanish at the choice of the respondent(s).
Environmental Sampling (inside & outside of residences): Metals
in dust, soil, & outside air; Pesticides in dust and soil; Total VOCs
in air. (Tap water for metals, pesticides, & VOCs, if resources
available). The sampling is described in greater detail in Section 3.
(Other Residents will be asked if they wish to continue being part of
the study, and they could complete the other NHEXAS Questionnaires
used in this Stage; they might also be eligible for environmental
sampling if further resources were made available. As other members
of the household have different exposures, related mostly to
different time-activities, secondary/tertiary individuals will be
identified among those in the household with different time-activity
patterns to be asked to also complete all the Q's, the Time-activity
Log, and the Diet Diary. Participation of multiple household members
will optimize participation of the primaries of the households (as
shown in our other studies and by other investigators), and will
maximize the number of individuals on whom information is obtained
without major cost increases. (Nevertheless, secondary and tertiary
participants will be identified as such and analyzed distinctly.)
-------
Section No. j 4
Revision #0
June 1995
Pages 32 of 116
Table 3, Sampling and Analysis Methods to be employed in Stage 2 (n=450).
These sampling methods will be employed side by side in Stage 3.
STAGE 2 SAMPLING & ANALYSIS METHODS
Media
Analyta
Collection
Analysis Method
Air
Indoor
Air
VOCs
n*
450
in situ-PID
PID
Outdoor
Outdoor
Metals
voc
125
450
Hi Vol/filter
in situ-PID
XRF
PID
Dust
Carpet/floor
Carpet/floor
Metals
Pesticides
450
450
Vacuum and filter
Vacuum and filter
XRF
ELISA
Soil
Foundation
Foundation
Composite (Yard)
Composite (Yard)
Metals
Pesticides
Metals
Pesticide
450
450
450
450
Mixed Aliquots to 2.5
Mixed Aliquots to 2.5
Mixed Aliquots to 2.5
Mixed Aliquots to 2.5
cm depth
cm depth
cm depth
cm depth
XRF
ELISA
XRF
ELISA
Water* *
Tap, DW
Tap, DW
Tap, DW
Metals
Pesticides
VOCs
125
125
125
Collect in plastic
Collect in glass
Collect in glass .
EPA
EPA
EPA
200. 8 -g.13
52 5 .J* -W
524.2 S3/.I
Background Information
Time activity
Diet
Follow-up
Technician
450+
450 +
450+
450+
Questionnaires
Questionnaires
Questionnaires
Questionnaire
This is the number of households, information will be obtained from all residents
** Contract analysis through EPA-EMSL
-------
Section No. 1.4
Revision #0
June 1995
Page: 33 of 116
STUDY PHASE 3/FIELD Stage 3 - INTENSIVE SAMPLING: A randomly selected
subset of households (n = 175)19 will be re-evaluated for metals,
pesticides and VOCs using methods (Table 4) with greater resolution and
reliability. Concurrently, the less intensive measurements used in
stage 2 will be repeated as an internal side by side calibration
approach. Use of these comparative technologies will be analyzed for
recommendations about equipment to be employed in Phase II of NHEXAS.
All follow-up NHEXAS questionnaires (with ancillary questions on
pesticides) (essentially the same as those obtained in Stage 2) will be
completed by at least the target respondent in the household;
secondary/tertiary household respondents will be asked to complete the
questionnaires and provide biological samples. Questionnaires are
available in English and Spanish at the choice of the respondent(s).
23 24
Media sampled will include air, dust, soil, water, food and beverages,
skin (dermal), blood, and urine.
Air monitoring for metals and pesticides1 will be done for residential
indoor and outdoor air, a subset will be done for VOCs, and a subset of
personal air samples for metals and pesticides.
Dust will be obtained from carpets/rugs, window sills,1 and from dermal
wipes, for metals and pesticides.
Soil will be obtained from the residential yard (as a composite) and
from the foundation, for metals & pesticides. Surface thin films
collection will also be made outside the home in one of the following
locations. Locations for selection are prioritized in order of
preference: (1) concrete curb, (2) concrete sidewalk, (3) concrete
driveway, (4) another concrete surface near the roadway, or (5) a
roadside mailbox at the home.
Water monitoring will be of both tap (all homes) and separate source
drinking water (estimated as 50% of homes) for metals, pesticides & VOCs
(specific on sample collection methods are outlined in SOP numbers
UA-F-16.0 - UA-F-17.0)1.3
- Food and beverages will be collected as a one-day duplicate diet. Food
and beverages24will be collected in separate containers. Specifics of
food and beverage collection and storage are outlined in SOP #
UA-F-15.0. Specifics for collection and storage are outlined as
UA-F—15.0. FDA wtXl conduct/a mini-market basket analysis ii» Arizona.
a. /
qJLo-V\£j
-------
Section No. 1.4
Revision #0
¦Tune 1995
Page: 34 of 116
Table 4. Sample Collection methods of Stage 3.
STAGE 3* ENVIRONMENTAL SAMPLE COLLECTION METHODS
Media
Contaminant
Class
Environment or
Type
Collection Method
Sample
Description
Air
Metals, Pesticides
Metals, Pesticides
VOC
Indoor/Outdoor
Personal
Indoor/Outdoor
Indoor/Outdoor
Indoor/Outdoor
Filter (Teflon) & Filter/PUF
Filter (Teflon) & Filter/PUF
OVM 3500 badge (VOC)
Pumped C-based tube (VOC)
PF-t tube (CHs0)
3-day integrated
1 -day integrated
7-day integrated
1-day integrated
7-day integrated
Dust
Metals, Pesticides
Carpet/floor
Other surfaces
Dermal
Vacuum surface sampler
Surface wipes, surface blots
Wipes
Defined area
Defined area
Hands to wrist
Soil
Metals, Pesticides
Foundation
Yard composite
Surface (curb)
Composite aliquot
Composite aliquot
Blot
Defined area
Defined area
Defined area
Water#
Met, Pest, VOC
Kitchen tap
Drinking water (if different)
Cubitainer, glass bottle/via Is
Cubitalner, glass bottle/vials
After flush
As stored
Biological#
Metals, VOC
Metals, Pesticides
Blood
Urine
Vacutainer
Plastic bottle
Venous blood
Overnight, morning cat,
Food#
Metals, Pesticides
Duplicate diet
Ziploc bags/plastic containers
Composite (1 day)
Beverage#
Metals, Pesticides
Duplicate beverage
Plastic bottle or cubitalner
Composite (1 day)
Background
Information:
Time activity
Diet Diary
Diet follow-up
Follow-up
Technician
Met, Pest, VOC
Met, Pest, VOC
Met, Pest, VOC
Met, Pest, VOC
Met, Pest, VOC
Household and work
Household and work
Household and work
Household and work
Household and work
Questionnaire
Questionnaire
Questionnaire
Questionnaire
Questionnaire
Daily for 1 week
Daily for 1 week
1 day
1-week integrated
1-week integrated
# Collected for analysis through IAG
* TTie same collections (excluding VOCs) are made for Stages 4 & 5
-------
Section No. 1<4
Revision #0
June 1995
Page: 35 of 116
Biological samples (blood and urine) will be obtained for metals VOCs
and pesticides in the primary and a subset of secondary respondents.
Urine will be collected as the first morning void not a 24-hour sample.
s
STUDY PHASE 3 (Intensive Sampling—continued)/FIELD Stages 4-5: Two
follow-up evaluations (Stages 4-5) will be performed in a randomly
selected subset of 50 homes evaluated in depth during Stage 3. Target
metals, pesticides (and VOCs/PAHs, given resources) will be evaluated in
environmental and biological samples (in the primary respondent) for
temporal (i.e., seasonal) variability.
8
STUDY PHASE 4 - Data Analysis, including Exposure Assessment: All analyses
will be hypothesis specific. General classes of analyses and the
statistical approaches that will be taken are listed in Table 5. All
analyses (except prediction models) are standard statistical techniques
found in textbooks and computer software package manuals.
Table 5. classes of Hypotheses and Statistical Approach taken.
Types of Statistical
Hypotheses Analyses Chemical Media
Char./Compar Popultns
1st Distributions
Domain Distributions
Correlates
Chem.-for all media
Chem.Sub-Classes
Methods Eval./Compar.
Compare Domains
Total Exposures(TEA)
Corr./Predn of TEA
Predictive Models
Corr./Pred.-pred.TEA'
Univar.fi Bivar. Analyses
Univariate Distributions
Bivariate Distributions
II M
Bivar./Multivar. Distributions
Multivar. Analyses
Bivar./Mult ivar. Analyses
Multivar. Distributions by Domain
Multivar. Distributions over Domains
Multivar. Analyses by Corr/Preds.
TEA Predictive Models
Multivar. Analyses of Pred.Models
None
None
Each/every
Each
Each/every
Each
Each/every
Each
Each
All
Ea. Class
All
Each
All
Each
All
i Each
All
Each
All
Each
All
Each
All
In general alpha will equal 0.05, except when corrected to accommodate
multiple comparisons. (However, since initial univariate evaluations are
not part of the hypothesis testing, they do not require correction for
multiple comparison.) The power (1-beta; type 2 error) will be deemed
satisfactory between 0.8 and 0.9 for the hypothesis testing. Power
calculations are generally made prior to designing a study to determine
the number of measurements needed. Our initial power calculations were
based on needed sample sizes for detecting exposures above the 90th
percentile, starting at the last stage then expanding to each earlier
stage. Calculations were based on the delineation of population
subgroups [within the design chosen] with a power of 0.80 and a type-I
error (p-value) of 0.05 for the detection of specific groups and those
at the 90th percentile of exposure - p^Q.50 & p2=0.10 for the final
stage [Fleiss, 1981]). Such calculations are valuable when the
distribution of the target pollutant is known. When unknown,
-------
Section No. 1.4
Revision #0
June 1995
Pages 36 of 116
assumptions are made about the variance and covariance of the
distribution and the resulting power determination may or may not be
valid. So, if the null hypothesis is not rejected, a power calculation
will be repeated to test the validity of the sample size.
Specific Statistical Approaches
The analyses are dictated by the design features related to
exposure-response determinations. All analyses, after data QA/QC (QA
specifics are provided in section 4 and SOPs UA-D-15.0, UA-D-16.0,
UA-D-25.0, UA-D-26.0), are reviewed to evaluate the distributions of the
variables (and presence of possible outliers). Depending on the demands
of the analysis, the purpose being descriptive or to test specific
hypotheses and the method/model chosen to perform the tests, we
determine if removal of outliers and/or transformations of variables are
necessary. These analyses will be performed at the University of
Arizona on the Local Area Network (LAN) consisting of several PCs linked
to two UNIX based SUN-System work stations (with a combined disk of 5.5
gig). This LAN and associated programming are described in UA-D-1.0.
Exposure assessment will be performed on the computer system at Illinois
Institute of Technology; its system is described in SOP # IIT-A-1.0.
Univariate analyses: Use of these methods assumes that measurements are
above the detection limits for each pollutant. First, frequency
distributions will be determined and histograms plotted (with
descriptive distribution statistics). Continuous distributions will be
examined; if they are non-Gaussian with skew around the mean (third
moment); then we will use transform functions (e.g., logarithmic) to
normalize the data.
If the data can not be normalized, then we will use a variety of
non-parametric statistics including rank correlation, nonparametric
linear models for regression and ANOVA (1 & 2 factor). If the variables
to be evaluated are categorical, we will use binomial or multinomial
2
approaches (e.g. contingency tables, X ).
Bivariate analyses include bivariate frequency distributions and
histograms for sub-classes of key variables, scatter diagrams,
contingency (e.g., chi-square) tables, and analysis of variance, with
appropriate simple statistics (as above, also t & P tests, chi-square
tests, correlation coefficients and regression slopes with intercepts).
Common indices of variable expression, known (or expected) bivariate
distributions, cell sizes, and other, similar factors, influence our
choice of approaches. Our choices are influenced by the nature of the
alternate hypotheses, whether we are testing primary or secondary
hypotheses, or performing exploratory (hypothesis-generating) analyses.
The next step is to examine the relationships between intervening/
confounder variables and the dependent variables (e.g., relations with
chronic occupational exposure reports, daily exposure events, SES
-------
Section No. 1.4
Revision #0
June 1995
Pages 37 of 116
factors), and to evaluate the relationships between confounder and
intervening variables with independent indoor and outdoor pollution
variables (e.g. relations with time-activities, meteorological
conditions, etc.), using similar univariate and bivariate approaches.
The univariate and bivariate relationships of dependent and independent
variables are analyzed during this phase, without and then with
appropriate co-, intervening, and confounding variables. Often within
these stages of analyses we will evaluate such concerns as homogeneity
of sub-group variances (or homoscedasticity in regressions), linearity,
and trends. Results therefrom determine the approaches to the next
phases of analysis. [Such analyses precede all multivariate approaches,
and dictate what approaches and what variables are to be considered for
each specific aim addressed.]
As for the multivariate analyses, our first choice is linear {or
nonlinear) parametric regression. For interactions we will employ ANOVA
and ANCOVA as tools. Discriminant analysis will include principal
components analysis, and possibly analysis of variance (with tabulated
coefficients and adjusted variables, and significance results). For
certain outcome variables, we will use multiple logistic regression and
regressive logistics; the latter has been used recently and very
effectively in analyzing time-dependent relationships. (Odds ratios
from multiple logistics are useful.)
Prediction models for total exposure can take various forms. Estimates
of the average exposure experienced in each location can be derived from
the time budget information and the integrated samplers in that
location. An average partial exposure component for the i-th location
can be calculated, E^ = [tf x c-] / T, where T is the total elapsed
time. The distribution of this average partial exposure component as a
proportion (P) of the measured total integrated personal exposure (E),
p. =1./1 can be used to summarize (mean or median percentage) the
relative importance of each component (location) to total personal
exposure in this population. Since this average partial exposure does
not use the actual concentration average specific to the time(s) an
individual was present for locations sampled with integrated samplers,
these proportions may over- or under-estimate the individual's actual
exposure component.
Differences between the integrated average concentration in a location
and the actual concentration to which an individual is exposed may be
due to the uneven spatial distribution of the pollutant within the
compartment, building/room or geographic area and topography (for
outdoor samples). Further, responses to questions on the Time/Activity
diary and the Follow-up questionnaire provide information facilitating
evaluation of sources of variation and the times of source use recorded
by the participants, so that the temporal relationship between the
presence of the sampled individual and source being used can be
assessed.
-------
Section No. 1.4
Revision #0
June 1995
Page: 38 of 116
Time Weighted Average (TWA) estimates of personal exposure will be
calculated based on the time and location information derived from the
daily diary. This contains a discrete sequence of time periods (j) that
are spent in the limited number (7) of location categories; each period
having a unique duration, t-. For each of these time periods a
concentration, Cj, can be estimated from the passive sampler or from
continuous data (if available) for that location and time period. The
TWA is calculated as:
TWA = S Pj * Cj, where p. = tj / T and T = 2 tj for
j = 1,..., number of time periods.
Information from questionnaires and diaries influencing exposures will
be included as covariates in relating concentrations to exposures. The
calculated TWA will be compared with the integrated personal exposure
measurement using an analysis of covariance (ANCOVA) procedure to assess
the agreement between the estimated and measured exposure, and to
estimate the average pollutant concentrations in non-measured locations
and their importance from the value and relative significance of the
regression coefficients (Quackenboss, 1982; Tosteson, 1981; Spengler,
1985; Quackenboss, 1987).
Assessment of exposure from multiple media can be evaluated by adapting
a general PEA model which states thats
where Cs is the contaminant concentration in the exposure media i, Ck is
the concentration in environmental media k; 10. is the intake/uptake
factor (per body size [BW]) for exposure media i; EF is the exposure
frequency (day/year) for this population, ED is the exposure
duration(years), and AT is the averaging time for population exposure
(days). These models help define the process linking sources to routes
of exposure. Human activity patterns associated with exposure are
critical to these single and multiple media models.
The specific exposure assessment modeling approach will be decided
through collaboration with EPA. In addition, we would like to
explore exposure assessment modeling using: (a) mass balance models,
(b) other non-mass balance models previously attempted in exposure
assessment, (c) beta and gamma distributions with multinomials
approximating the alpha level, and compound pdf's involving different
Pearson-type characteristics. These attempts will be time and resource
dependent.
Avg. daily exposure = x x EF x ED x Ck
AT
-------
Section No. 1.5
Revision #0
June 1995
Pages 39 of 116
1.5 Personnel Qualifications
The Principal Investigator and Co-Principal Investigators (Co-PI) all have
doctoral degrees and additional experience in their field(s). Each Co-PI
will be responsible for determining the education and experience level
required for employees performing tasks overseen by the consortium member.
Each Co-PI will be responsible for determining and documenting the
qualifications of local project employees. Overall 30 to 35 people will
work on the project. This includes Pi's; staff coordinators; field
personnel; lab personnel; data personnel; and analysis personnel. Specific
of training and job responsibilities are outlined in Sections 1.6 and 3.3.
1.6 Training Required
Training requirements will be task dependent and determined by each Co-PI.
Training requirements will be appropriately defined by the job description
and job responsibilities. Qualification and supplemental training records
will be maintained by each Co-PI for those serving on the project in
accordance with the Training SOPs (UA-T-1.0 through UA-T-6.0).
As part of the quality assurance program, the project has an independent
Quality Assurance Officer who will, in conjunction with Co-PIs and
appropriate Project Coordinator(s)/Supervisor(s), define the need for
formal project-specific training and indoctrination programs and will
supervise their conduct.
Training and indoctrination programs will involve familiarization of key
personnel with:
* Technical objectives of the project
* Project communication chain
* Project control documents
* Project QA requirements
* Project QA responsibilities
* Project documentation requirements
Coordinator/Supervisor Training: Each of these individuals will have
appropriate degree qualifications and a minimum of two years experience in
his/her area of expertise. Additional training will be provided by self
study, individual instruction as needed, and monthly area specific and
joint staff meetings (one each per month).
Field Interviewer(s)/Technician(s): Each person will have a degree in an
area related to his/her task assignments. Additional training will be
provided by self study, individual instruction, and monthly area specific
and joint staff meetings (one each per month). Supplemental training
specific to this project will be provided by the Field Coordinator as
outlined in UA-T-1.0 through UA-T-4.0.
Data Assistants: Each person will have course work in a data intensive
field leading to a degree. Additional training will be provided by self
study, individual instruction, and monthly area specific and joint staff
-------
Section No. 1-6
Revision #0
June 1995
Pages 40 of 116
meetings (one each per month). Supplemental training specific to this
project will be provided by the Data Coordinator as outlined in UA-T-5.0.
Laboratory Assistants: Each person will have course work in a laboratory
based discipline leading to a degree. Additional training will be
provided by self study, individual instruction, and monthly area specific
and joint staff meetings (one each per month). Supplemental training
specific to this project will be provided by the Laboratory Supervisor as
outlined in UA-T-6.0.
The QA Officer will inspect documented evidence (see UA-T-1.0 through
UA-T-6) of all project-specific training and indoctrination activities
maintained by the Co-PIs and the various Project Coordinator(s)/
Supervisors.
1.7 Experimental Design
1.7.1 Fundamental Objectives and specific Hypotheses to be tested:
A. To determine the distributions of total human exposures to selected
multi—media pollutants and classes. The pollutant classes chosen (in
order of priority) are: metals, pesticides, and volatile organic
compounds (VOCs).
Specific aims & Hypotheses:
1. To document the occurrence, distribution & determinants of total
exposures in the general population.
2. To characterize the 90th percentiles of total exposures to each of
the pollutants.
3. To monitor geographic and temporal trends of the multi-media
exposures .8
The null hypotheses are that
i) significant exposures do not occur for these pollutants,
ii) temporal trends are stationary,
iii) there is a representativeness (similarity) of geographic areas
and populations exposed for the different pollutants. (The null
hypothesis is that which will be tested statistically, rejection
indicating there are significant differences in total exposures.)
Rejection of the null hypotheses will lead to determining amounts and
types of variation in exposures (by media, sources, and time-activities
& characteristics of subjects). The current approach will allow
evaluation of the hypothesis that the distribution and determinants of
total exposure can best be evaluated using a population selected by a
proportionate-based approach, and that the results of sampling such a
population will provide an accurate estimate of the "high end" (90th
percentile) of total exposure to specific pollutants.
B. To survey a proportionate—based sample of the total population to
determine current real-time exposures to target pollutants.
-------
Section No. 1.7 . 1
Revision #0
June 1995
Page: 41 of 116
Specific aims include:
8
1. Evaluation of media, personal, time-activity, and geographical
factors that contribute to current total exposure.
2. Evaluation of biomarkers (in blood & urine) for the target
pollutants.
The null hypotheses include:
i) there are no differences in total exposure related to ethnicity,
socio-economic status (SIS), occupational, or urban/suburban/rural
residences,
ii) there are no differences of exposure in relation to time-activity
and source-usage patterns of individuals,
iii) there are no increased sensitivities or specificities, precision
or accuracy, from the use of biomarkers.
3. Evaluations of relationships between exposure reports, environmental
measurements, and biomarkers of target pollutants.
Short term (person/week) and chronic (occupational and long term)
exposure to pollutants will be reported by subjects. We will test
the hypothesis that these exposures will be evident in the tissue and
body fluids of subjects.
We will also evaluate the importance of questionnaires and the
identified factors, including subject characteristics, therefrom in
predicting the 90th percentiles of measurements. (Or, test the
hypothesis that the 90% percentile of exposure can be predicted from
questionnaire data alone.)
C. To use smaller nested samples for more intensive surveys, using more
precise pollution monitors and analytical methods.
Specific aims:
To further test the hypotheses mentioned (above in 1 & 2), to evaluate
the use of such monitors, and to test specific hypotheses concerning the
contributions of different media (air, water, soil, food, dust).
Further hypotheses could be generated regarding specific sources of
exposure. One could initially test hypotheses about the "causes" (if
any) of specific exposures for those exposed at the high end (90th
percentile) of the exposure concentration spectra.
D. To use standard instruments in smaller nested survey populations in
specific geographic areas.
Specific aims:
1. To test the feasibility of such approaches to exposure assessment.
-------
Section No. 1.7.1
Revision #0
June 1995
Page: 42 of 116
Specific null hypotheses concern the accuracy, precision, reliability
and external validity of the questionnaires, monitoring and survey
instruments, and the cost-benefit ratios.
2. To use instruments that could include an objective device for
measuring time—space locations of individuals, and determine
feasibility in population surveys.
The null hypotheses are that
i) the standard, more expensive instruments are not improvements over
the other instruments,
ii) additional instruments don't add much additional information
about exposure and any added information is not justifable in
terms of the cost,
iii) the use of biological markers of exposure
a) fail to more rigorously define total exposure,
b) do not have more utility (including reliability, sensitivity
& specificity, cost-benefit), and
c) are not more valid indicators of dose.
E. To construct prediction models of total exposure (at each survey level).
These prediction models will help formulate initial exposure assessments
based on information derived from exploring prior specific aims and
hypotheses.
Specific aims include:
1. Evaluation of the prediction models.
Hypotheses to be tested related to these models
i) do models add anything to previous objectives?
ii) do models and results help design the next survey stage?
iii) are models validated by the next survey stage monitoring?
iv) are models refined by the further survey stage monitoring, use of
specific instruments, and/or use of biological markers? (This
approach will analyze and demonstrate the amount of information
required to achieve the NHEXAS objectives.)
2. To determine if these models can be used to validly formulate
relationships between "explanatory variables" (discussed above) and
provide information that might be useful for risk assessment and
management. To iteratively refine these in each of the following
aims and retest the hypotheses.
3. To collaborate and coordinate the exposure assessment modeling with
EPA and the other cooperative agreement(s), and the other federal
agenc ies involved.
Specific Aims:
i) To insure the optimal use of the data.
ii) To improve compatibility and comparability of models.
-------
Section No. 1.7 . 1
Revision #0
June 1995
Page: 43 of 116
4. To evaluate the exposure assessments in regard to current and needed
information on pharmaco- & toxico-kinetics for dosimetry.
F. Relevance of Findings to an Improved Decision Process
As implied in the objectives, improved information of a direct nature on
relevant exposures will significantly improve risk assessment. Further,
such information and assessment should significantly improve risk
management and policy decisions. (For instance, we have already seen
this for lead and other NAAQS pollutants, to the extent that existing
information was used to make pro-active decisions for abatement and
exposure prevention.) Decision-making based on improved exposure
assessment directly aid human studies, promote stronger environmental
surveillance, promote better models of exposure - response
relationships, and lead to better risk assessment.
We think that the improved approaches for key pollutants and the
investigative approaches provided by NHEXAS would improve and facilitate
decision making even more. NHEXAS provides a logical, more focused and
directed approach to the steps necessary for obtaining information
directly relevant for such decision making on important environmental
problems in the U.S. Continued investigations beyond the stage, with
continued collaboration of EPA and other investigators, would likely
insure such benefit to public health.
1.7.2 Survey Design
Overview: Throughout this document NHEXAS Arizona "FIELD STAGES" refer
to specific sampling protocols that will be performed at the selected
household subsets. To minimize confusion, we refer to each component of
the survey design as a design phase. All of this work will precede our
entry into the field.
Arizona is the fifth largest state in the United States and is composed
of only 15 counties. The recommended NHEXAS design will be employed, but
the terminology (states, counties, census blocks) is suboptimal for the
NHEXAS Arizona design. NHEXAS Arizona will employ a multi-phase
probability sampling design. Arizona will be divided into 5-6 "regions"
analogous to states?7 Each region will contain a number of "combined
census block groups" that will be analogous to counties and serve as the
primary sampling unit (PSU) for the first design phase of the study.
There will be about 400-600 PSUs in the state; about 48 PSUs16will be
selected. The PSU will then be divided into area segments containing
20-30 houses each based on 1990 census "blocks"; 5 area segments will be
selected from each of the subdivided PSUs; these will become secondary
sampling units (SSUs) making up the second design phase. The area
segments will contain about 20-30 housing units each. Each house in
the selected SSUs will be listed; houses will be randomized and
sequentially selected until 5 participating households are obtained for
each area segment, thus completing the third phase selection.
-------
Section No. 1.7.2
Revision #0
June 1995
Page: 44 of 116
Finally, after each sample household has completed the Descriptive
Questionnaire, including a roster of all people living in the household,
a single "individual respondent" will be selected at the fourth design
phase of sampling for monitoring within the sample household. One- and
two-persons households will be sub-sampled to achieve approximately
equal person-level probabilities of selection in spite of unequal number
of persons in sample households. Sample households will be selected in
random waves until the appropriate number of participants have been
recruited. Figure 8 diagramatically represents this design.
Our study design has a longitudinal component facilitating evaluation of
the effect of temporal variability on sample concentrations and
individual respondent autocorrelation. To accommodate potential
temporal effects on samples, each PSU will be sampled in two different
seasons based on the possible return schedule of the field team.
Temporal evaluation of individual respondent autocorrelations indicate
the potential of short term evaluations for model-based extrapolation to
longer-term exposure distributions (e.g., annual or lifetime). Details
of how this design will be implemented follow in this section.
A. First Design Phase Sample of Combined Census Block Groups
Sampling frame - The sampling frame for the first design phase sample
will be constructed from the 1990 Census Summary Tape File 1A {STF 1A).
The sampling units will consist of combined census block groups derived
from census block-level records.
Sample size - A NHEXAS survey design study by Clickner [1993 J recommends
a relatively large sample of PSUs because of the high cost of collecting
and analyzing environmental samples and because of the possibility of
high intracluster correlations (i.e., tendency for measurements to be
more alike within counties). Therefore, NHEXAS Arizona will use a field
data collection protocol intended to maximize the number of PSUs
(combined census block groups) while maintaining a relatively efficient
field data collection protocol within the PSUs.
In keeping with the NHEXAS study survey design, environmental data will
not be used for stratification in the NHEXAS Arizona study survey
design. Although we anticipate no stratification, only stratification
based on general population characteristics (age, gender, and
ethnicity) potentially related to analysis domains and/or general
population differences in exposures to toxicants with proportionate
sampling rates will be undertaken in the study design phase.
Sampling method - We will select sample PSUs with probabilities
proportional to size (pps) using a sequential probability minimum
replacement (pmr) sampling algorithm [Chromy 1979]. Use of pps sampling
facilitates an approximately equal probability sample of housing units
with equal numbers of sample housing units per combined census block
group (PSU). The frame units will be sorted within the State in a
geographical order in a serpentine manner (i.e., small to large in
region 1, large to small in region 2, etc.). This will result in
-------
Section No. ^ j 2
Revision #0
June 1995
Page: 45 of 116
Figure 8. Diagrammatic of the Survey Design Phases.
16 27
IDEALIZED NHEXAS
ARIZONA SAMPLING DESIGN
PHASE 1
Primary Sampling Units
Select 48 Primary Sampling Units from
400-600 combined block groups?6
PHASE 2
PHASE 3
inn*
~ ¦ ~
~ ~~ ¦
PHASE 4
Area Segments
Select S area segments from the
division of the Primary Sampling
units into Secondary Sampling units.
Each area segment is constructed to
contain 20-30 Housing Units.
Housing Units
Select 5 Housing Units from each
selected Area Segment.
Primary Respondents
A designated primary respondent from
each Housing Unit. Other Housing Unit
residents will be asked to participate.
-------
Section No. 1.7.2
Revision #0
June 1995
Page: 46 of 116
implicit stratification with exactly proportionate sampling rates based
on the frame ordering because of the sequential nature of the sampling
method.
Second Design Phase Sample of Area Segments
16
Sampling frame - We anticipate a selection of 48 PSUs in Design Phase 1.
The sampling frame for the second phase sample will be constructed from
the 1990 Census Summary Tape File IB (STF IB) block-level data records
within each of these PSUs. The sampling units will be constructed by
combining blocks to yield secondary sampling units of 20-30 housing
units each from the 1990 census data. In no case will the combination
of blocks forming an SSU cross census block boundaries. Also, the
occurrence of SSUs that cross block group or block numbering area
(tract) boundaries will be minimized. Prior to proceeding to the third
decision phase, each SSU will be checked against census data to insure
that all possible housing units are considered. Bach missed housing
unit identified in this manner will be included in the sample. Each
missed housing unit that produces a study participant will count toward
the requisite number of participants in the sample segment. This
procedure ensures maximum coverage of occupied housing units in the
target population.
Sample size - The recommended NHEXAS survey design study suggests
minimizing sample clustering because of the high cost of collecting and
analyzing environmental samples and because of the possibility of high
intracluster correlations (i.e., tendency for the measurements of some
chemicals in some media to be more alike within PSUs and within area
segments). Based on this criteria and logistical considerations^ the
NHEXAS Arizona field studies will select five area segments per PSU.
Demographic variables from the Census STF IB data base will be used to
evaluate the secondary sampling units. They are (1) percent urban
population; (2) percent Hispanic population; (3) average housing unit
value; and (4) percent single family residences. The distributions of
these variables will be examined within each of the PSUs to assure that
the sample segment reflects the characteristics of the census block
group. This analysis will define the extent of representativeness with
regard to (1) homes in rural and urban areas, (2) representation of
Hispanics and Native Americans, and (3) representation of households
from all socio-economic strata relative to the 1990 and incremental
census data. Deviations will be evaluated and reported relative to
representativeness of the population.
Sampling method - The same sequential pmr sampling algorithm used to
select sample PSUs will be used to select the five sample SSUs (area
segments) within each sample PSU with probabilities proportional to size
(pps) (occupied housing units in the 1990 Census). Use of pps sampling
facilitates an approximately equal probability sample of housing units
with equal numbers of sample housing units per segment. The second
phase sampling frame will be sorted by the demographic variables and by
size within sampling strata in a serpentine manner. This sampling
-------
Section No. 1.7.2
Revision #0
June 1995
Page: 47 of 116
method ensures proportionate representation of the strata, as well as
proportionate representation of SSUs containing relatively large and
small numbers of housing units.
C. Third Design Phase Sample of Households
Sampling frame - The third phase sampling frame consists of a list of
all potential housing units in each selected area segment. This list is
compiled by the field interviewers assigned to the PSU.
Sample size - A multinomial sampling procedure will be implemented in
each sample segment to achieve a pre-assigned number of participants in
each segment. Hence, the sample size will be a random variable, the
number of sample housing units needed to produce "m" successes
(participants) {see below).
Five housing units will be selected from each of the five area segments
located in a SSU. This will yield a total of 1200 houses to be surveyed^
using the Demographic Questionnaire (5 houses x 5 area segments x 48 PSU
= 1200 housing units). No more than 2 housing units per area segment
will be evaluated for environmental monitoring [Stage 2 environmental
monitoring will total (450) housing units].
Sampling method - Use of the multinomial sampling procedure requires
that all housing units listed for each sample segment first be placed in
a random order so that the first "x" sample housing units constitute a
simple random sample of size "x," for every integral value "x." To
achieve "m" participants in a segment, we will begin by fielding the
first "m" housing units on the randomly ordered list (a simple random
sample of size "m"). When the final result for one of these sample
housing units is a "refusal," an ineligibility, or an incapacity, the
field coordinator will increase the size of the simple random sample by
one housing unit. The increase is accomplished by accessing the next
housing unit on the randomly ordered list (a supplemental sample of size
one).
The third design phase sample size for each segment will be the number
of sample housing units activated for data collection. Each refusal or
other non-interview will count against the survey response rate.
Therefore, sample housing units beyond the initial "m" housing units
will be activated only by the field supervisor who has verified that all
possible attempts to obtain a response have been completed. Inability
to contact the members of an occupied housing unit after repeated
attempts will not be sufficient reason to activate an additional sample
housing unit.
Whenever a sample housing unit is released for contact, the interviewer
will check for missed respondents within the sample housing.
D. Fourth Design Phase Sample of Persons for Monitoring
-------
Section No. 1.7.2
Revision #0
June 1995
Pages 48 of 116
Sampling frame - The fourth phase sample frame will be derived from the
the roster of household members compiled in the Descriptive
Questionnaire discussed above?0 The interviewer will use appropriate
probes to check for completeness of the roster and will remove any
household members not eligible for the study (see section 3.1.1).
Sample size - Housing units will be recruited in each sample segment
until an appropriate number of persons willing to participate in the
environmental and biological monitoring stages (Stages 2 & 3) have been
recruited (1-2 people per SSU). One person in each selected house will
be designated as the primary respondent.0 NHSXAS Arizona will enhance
participation by asking all household residents to participate in the
survey. We will evaluate the differences among exposures detected from
those sharing a common residence. Since the greatest cost of the survey
is encountered in the environmental sampling and "getting into the
house," the potential for evaluating exposures among family members
should not be overlooked. Further, if autocorrelation among residents
is low, then the power available for some analyses is enhanced. Data
obtained from the Primary Respondent can be readily identified and kept
separate for transfer to SPA.
Sampling method - If one person were selected at random from every
sample household, the survey design effect, or variance inflation
factor, due to unequal probabilities of selection within households of
different sizes would be approximately 1.5. Therefore, the design calls
for selection of a person from about 25% of the one—person households,
from about 32% of the two-person households, and the remainder (43%)
from all other households. This reduces the survey design effect for
unequal person-level probabilities of selection to less than 1.1.
Within those households chosen to have a sample member recruited, one
eligible household member will be selected at random using the
appropriate (unit size dependent) random selection table. Accounting
for the survey design effect will improve the capture chances a small
amount. This proposed design is a compromise between equal
probabilities of selection for monitored households and for monitored
persons with an unequal weighting design effect of about 1.1 expected in
each case (see Table 6 below).
Table 6. person-level sampling weights for NHSXAS Arizona (HH = households)
Housing Unit % Arizona
Size J Households
1-person
2-people
>3-people
Total
25%
32%
43%
100%
HH Completing
Descriptive QX
300
384
Asked to
Participate
100
516
1200 I
256
344
700
Expected to Partic-
ipate in Activity
65
164
221
+-
450
-------
Section No. 1.7,2
Revision #0
June 1995
Page: 49 of 116
Even if there were a comparatively high survey design effect (1.5),
NHEXAS Arizona could expect to capture at least one respondent with an
exposure in the upper 90th % for sample sizes greater than 50 (see Table
8).
1.7.3 Longitudinal Monitoring Design
The health effects of many of the NHEXAS target toxicants (e.g., metals
and VOCs) are long-term chronic effects. Therefore, the estimated
distributions of long-term average exposures (e.g., annual or lifetime)
are needed to estimate risks. Hence, the monitoring periods need to be
as long as possible to support risk assessment. For the toxicants of
interest, week-long monitoring is currently about the longest feasible
monitoring period. Since one week is still short relative to a year or
a lifetime, the distribution of long-term exposures must be estimated
using models of the relationship between the long-term and short-term
. 31
measurements.
NHEXAS AZ will investigate similarity among samples collected during^
different one-week exposures for the original sampling and two lag
intervals. Lag times between one and eight months will be investigated.
This will enable sufficient sample sizes for each lag interval (in
months) and produce reasonably stable estimates of the autocorrelations
for these lag intervals. Autocorrelations will not be directly
estimated for longer lag times because of the need to complete field
data collection within about 15 months and because the autocorrelations
for these longer-term lags should be small. Moreover, they can be
modeled by extrapolating from a plot of autocorrelation as a function of
lag times. The function should be relatively flat and stable for longer
lag times.
Each participant in about the first 60% of PSUs will be asked (after
completing the primary data collection activities) to participate in the
longitudinal follow-up for two different lag times (e.g., 1 month and 6
months)8. The follow—up visits will be conducted as the initial Stage 3
environmental monitoring visit (excluding the VOC analysis? our tertiary
priority) and will consist of the activities listed in Table 4 and
Figure 13 for the second and third visits.
The assignment to lag times will be stratified so that participants
during the first 5 to 7 months are assigned to the full range of
potential lag times from 1 to 8 months and participants in the last 5 to
7 months are assigned only to shorter lag times. This will enable all
the follow-up data collection activities to be completed approximately
at the same time that primary data collection activities are being
completed in the final sample PSU.
1.7.4 Precision Requirements
A. Survey errors are dichotomous; some result from sampling, other errors
result from nonsampling. Sampling errors are observation errors
-------
Section No. 1,7.4
Revision #0
June 1995
Page: 50 of 116
resulting from an observation of the survey population only. These
would not occur if the entire population was evaluated. All other
errors in the survey measurement process are nonsampling errors.
The expected precision for the NHEXAS Arizona survey population is
discussed below (Table 7) in terms of noncoverage, and survey
nonresponse errors. We will assess these using a survey design effect
(Table 8) of 1.5 to ensure that the sample size precision estimates are
adequately rigorous. Measurement errors associated with the
questionnaires are being evaluated by the RTI/EOHSI consortium for the
entire SPA sponsored NHEXAS project. The evaluation is planned, but
not yet complete. We will incorporate the questionnaire evaluation as
Appendix B when completed.
A clear view of necessary sample size is presented in Table 7. It
illustrates that with a design effect of 1.5 and for the sample sizes we
propose (see section 3.1), at least one individual from the total
population should be captured at the 90th percentile and an N (number
of houses monitored) of 12 5 housing units or better with a probability
of 95%. The probability drops to the 90% level with an N of 50 housing
units. A sample size of 25 is not desirable, although necessary,
because of limited equipment and funds. Results will be used to
identify trends, not for detailed analysis. Inferences can not be
statistically supported for analyses with an N of 50 or 25.
Table 7. The likelihood (denoted by K%) of
capturing at least 1 person with an
elevated exposure (upper 10%; = p) in the
distribution function given the sample size (N)
N
= 0.01
K%
p= 0
K%
,02
P= 0.
K%
05
p= 0.10
K%
450
95.1
99.9
99.9
100
175
69.0
90.0
99.3
100
125
56.7
82.1
98.8
99.9
50
28.5
68.8
81.9
97.0
25
15.4
39.9
57.5
82.8
+
Precision Targets
The precision of statistics based on a given sampling design is often
discussed in terms of estimates of population proportions (e.g.,
frequency of occurrence of elevated levels of VOCs in air or arsenic in
water) because the standard errors of proportions depend only on (1) the
population proportion being estimated, (2) the sample size available for
the estimate, and (3) the survey design effect for the estimate. In
particular, no independent sources of information regarding population
variability are needed for estimates of proportions.
-------
Section No. 1,7.4
Revision #0
June 1995
Pages 51 of 116
The survey design effect for a statistic is the ratio of the sampling
variance for the statistic under the actual sampling design divided by
the variance that would have been achieved with a simple random sample
of the same size. It can be factored into components associated with
the effects of (1) stratification, (2) multistage (phase) sampling
(clustering), and (3) unequal weighting (unequal probabilities of
selection and weight adjustments for unit nonresponse). Stratification
tends to decrease the design effect (increase precision), whereas
multistage sampling and unequal weighting increase the design effect.
Multistage sampling effects usually predominate.
NHEXAS Arizona has selected sample sizes that are sufficiently large to
be almost certain that at least one sample observation will be from the
upper 10 percent of the population distribution. Table 7 presents the
sample sizes needed to be from 98 to 99.9 percent confident that the
sample contains at least one observation from the upper 1 to 10 percent
of the population distribution for estimates with survey design effects
of 1.5. Sample sizes corresponding approximately to those being
proposed for various stages of NHEXAS Arizona (n = 450, 175, 50) are
shown in bold italics (Table 8). Because of unequal probabilities of
selection and intracluster correlations, the survey design effects are
expected to be on the order of 1.5 or greater for most estimates
produced by NHEXAS Arizona. We see from Table 7 that we can be about
99.9 percent certain that the sample will include at least one
observation from the upper 5—6 percent of the population distribution
(where n = 175) for those statistics for which the survey design effect
is 1.5.
Table 8. The sample size required (N) to achieve the likelihood (K;
98%-99.9%) of capturing at least 1 person with an exposure in the upper
pth percentile of the distribution function.
Sample Size
Sample Size
(N) Needed
(N) Needed
With Survey
With Survey
Design
Design
Survey Design
p
Effect = 1.5
K
P
Effect =1.5
K
P
Effect = 1.5|
+ ++ + + ++ + + ++
1%
1,031
99%
1%
687
98%
1%
584
2%
513
2%
342
2%
290
3%
340
3%
227
3%
193
6%
167
4%
169
5%
114
9%
110
6%
110
10%
56
10*
98
10%
66
One can also compute the probability that specific upper population
percentiles will be covered by the range of sample observations based on
the expected number of participants for the different types of data to
be collected in the NHEXAS field studies.
-------
Section No. 1.7.4
Revision #0
June 1995
Page: 52 of 116
Table 9 indicates the probability that various sample ranges cover the
particular percentile (p) of the distribution for selected domains.
These probabilities assume a survey design effect of 1.5, which is
expected to be approximately correct for most of the environmental media
measured. The table presents the probabilities of coverage of the
specified percentile of the population distribution for samples from the
full population, for samples from domains containing 25 percent of the
population members (such as the household residents of Arizona under 18
years of age), and for samples from domains containing 10 percent of the
population members (e.g., children under 7 years or minority members).
In each case, we see that the expected numbers of respondents are
generally sufficient to be almost certain that the sample size for the
full population will cover the 95th percentile of the population
distribution. However, for a domain containing 25 percent of the
population members, the sample size within this domain will be
sufficient for the environmental sampling of Stage 2 to the 95th
percentile but for the more intensive sampling of Stage 3 only to the
90th percentile. For population domains in the 10% range, the
environmental sampling of Stage 2 will be reliable to the 90th
percentile, but the number of houses in the intensive Stage 3 monitoring
will be inadequate.
Table 9. The probability that the NHEXAS Arizona sample range covers the pth
percentile of the measured concentrations for selected domains where the
sample design effect is 1.5.
| Domain = 100* || Domain = 25% || Domain = 10% ||
Sample
Size
p=.90
p=.95
p=. 98
p=.90
p=. 95
p=« 98
p=.90
p=. 95
p=.98
—h 4-
-++-
-++
450
1.00
O
o
o
o
1.00
0.98
0.78
0.96
0.79
0.45
175
1.00
1.00
0.90
0.95
0.78
0.44
0.71
0.45
0.21
125
1.00
0.99
0.84
0.89
0.66
0.34
0.58
0.35
0.15
-++-
-++-
+-
-++
C. Non-sampling Errors
Noncoverage Error. Noncoverage occurs when the sampling frame(s) fail
to include some members of the target population. The use of area
sampling frames for the first two phases of the sampling design provide
theoretically complete coverage at those phases of sampling. Some
noncoverage can occur in the third-phase sample of housing units if the
lists of current housing units compiled by the household interviewers
are imperfect. Errors may occur because of problems accurately locating
the boundaries of the sampled census blocks in the field, especially
from redevelopment since the 1990 census. Errors may also occur because
some housing units are not immediately apparent (e.g., garage apartments
or apartments off a common entry hall). Project staff will be available
to assist interviewers having difficulty locating the boundaries of
-------
Section No. 1.7,4
Revision #0
June 1995
Page: 53 of 116
sample segments by extracting additional information from the Census
TIGER files and/or interpreting the information provided. The third
phase sampling procedure includes techniques to locate and include any
missed housing units within, or between, the sample housing units and
the next listed housing unit. Given these field procedures, nearly
complete coverage of the targeted households and residents should be
obtained.
Another potential source of noncoverage occurs in the fourth phase. If
the household respondent fails to report all household eligible members,
then a nonsampling error occurs. The respondent may prefer not to
acknowledge that a particular individual is living at the residence or
may not consider an unrelated member of the household as belonging to
the household. Therefore, the interviewer will be trained to probe for
completeness of the household roster, including unrelated persons
residing in the same housing unit. Given this training, coverage of the
household population should be as good as practically possible.
Nonresponse Error. Nonresponse can occur at several levels in the
NHEXAS Arizona study design. The first opportunity for nonresponse is
to the Descriptive Questionnaire, which we expect to complete for about
90 percent of the occupied sample housing units. We also expect to
complete a few "key" items on the Descriptive Questionnaire (e.g., the
household roster with gender, ethnicity and approximate age of each
household member) for nearly all (say, 98 percent) of the occupied
sample housing units by using information from neighbors and other
sources. We will test for significant differences between respondents
and nonrespondents for all characteristics known for both, such as the
"key" Descriptive Questionnaire items.
Adjustments to the sampling weights (reciprocals of the probabilities of
selection) will also be computed to reduce the potential for nonresponse
bias for each phase of nonresponse to NHEXAS. The ability to compensate
by making meaningful weight adjustments for the potential nonresponse
bias depends heavily on the availability of sufficient data for the
nonrespondent to adequately model the response propensity (probability
of responding). This must be balanced against the essential need for
informed consent from each respondent. Preceding collection of
questionnaire information from any potential housing unit, we must
secure the respondent's signature on an "Informed Consent Form." This
form must be approved by the University of Arizona's Human Subjects
Review Board (our IRB). Pinal IRB approval is predicated on review of
the final NHEXAS Questionnaire. When in the field, we will acquire as
much descriptive information as possible to characterize all refusals.
We hope to have the Descriptive Questionnaire data to model nonresponse
against the Baseline Questionnaire. We expect that standard
weighting-class nonresponse adjustments will be sufficient to compensate
for the lower levels of nonresponse to the Descriptive Questionnaire and
the Baseline Questionnaire. However, logistic regression models for
the probability of responding will also be explored for nonresponse in
the monitoring phase because these models have the potential to make
-------
Section No. 1.7.4
Revision #0
June 1995
Page: 54 of 116
greater use of the data available in the Baseline Questionnaire and
produce greater reduction of nonresponse bias.
Questionnaire Measurement Error. Problems associated with questionnaire
construction will be evaluated for the entire NHEXAS project by the
RTI/EOHSI Consortium. Specific problems identified by Arizona are
identified in Appendix B. Further, errors will be limited by attempting
to hire people with some training or experience in interviewing
technique. We will provide additional training for new and
inexperienced interviewers, who will be trained. Interviewers will ask
the questions exactly as written and not use leading probes. These
procedures should reduce questionnaire measurement errors to the extent
practicable. All questionnaires will be available in English and
Spanish at the choice of the respondent.
Field Sampling Error. NHEXAS Arizona will hire the most qualified and
experienced monitoring personnel possible. They will all be cross-
trained in all sampling techniques. The Field Team Leader has the
authority to make field based decisions. Each field procedure has a
complete SOP written to cover appropriate sampling methods and ensure
consistent quality sampling. Field personnel are certified as being
competent to perform these procedures through training protocols
(UA-T-1.0 through UA-T-4.0) before they independently collect field
samples. If a protocol does not cover the situation confronted in the
field, the technician will consult with the Field Team Leader. The
Field Team Leader will determine whether a sample should be taken (and
how). If the need for consultation arises, the Field Team Leader can
call the project office. Any non-conforming collections will be noted
on the sample and field form.
Analytical Chemistry Measurement Error. National Institute of Science
and Technology (NIST) performance calibrations and standards, field
blanks, co-located field duplicate samples, replicate sample extract
analyses, etc. will be used to quantify measurement error variance and
bias for the collection and analysis methods employed in the study.
Overall, this will amount to a 10% collection of field and lab
blanks/spikes/duplicates/replicates. We expect to find that each field
and laboratory measurement procedure has no significant bias. Standard
errors of survey statistics will be based on the estimated total survey
variance, which includes all sources of variation without separately
estimating each variance component. If any significant measurement
biases are detected, survey estimates will be statistically adjusted to
produce unbiased population estimates.
1.7.5 Dependent and Independent Variables
All analyses will be hypothesis specific. In general:
Dependent Variables = all values measured from samples. This includes
measured amounts of each metal, pesticide, and VOC from each of the media
sampled. Detection of target pollutants in blood or urine {"biomarkers")
are independent variables.
-------
Section No. 1.7.5
Revision #0
June 1995
Page; 55 of 116
Independent Variables = all data obtained from questionnaires, the
geographic location (including elevation), and time of sampling.
After the exposure assessment model is complete, a dose model will be
created. At that point the biomarkers are the independent variable and
the exposure model results are the dependent variables.
1.7.6 Discriminating Power of Tests
A. statistical: Sample Size Determinations: Sample sizes for target
populations and nested sub-samples were calculated based on needed
sample sizes for "high exposure" (90th percentiles). Sample sizes are
extensively characterized in section 1.7.5. These are the maximum
attainable sizes based on resource availability.
Choice of Statistical Models: Based on the hypothesis being tested and
the study design, there will be a limited choice of statistical models
that can be used. We will choose the most robust, most discriminating
model amongst them, such as a general linear model. In general, the
relationships will be viewed as significant if p < .05 . Multiple
comparisons will be taken into account. More detail has been provided
in Section 1.4 Stage 4.
B. Instrument, Laboratory & Chemical Assays:
Equipment and analyses performed in the laboratory must meet some
minimal standards for each of the procedures performed. Outlined below
(Tables 10-12) are the methods and the expected detection levels
attainable for procedures and target pollutants. This section is in
terms of General Techniques followed by Detection Limits (DL) of the
target pollutants. When available, accuracy is established based on
standard reference material.
Table 10. General Field Assays
Medium/Material
Purpose
Method
Quantitative/
Qualitative
Toler-
ance
Air/PM10 +
other pumps
Air- Temp/RH
VOC Monitor
Location
Distance
Dust
(Pumps)
Field calibration
(Pumps)
Pump + Sensor
Site location Lat SLong
Distance from Reference
points
Collection
PsychrometerJ Quantitative!+10% RH
Sentex PID
GPS Meter
Tape Measure
Quantitative
Quantitative
Quant itat ive
Vacuum Cleaner Quant/Qual
± 2° F
± 1 ml/
min
+ 200 m
+.25 cm
+ .2 L/
+ min
-------
Section No. 1.7.6
Revision #0
June 1995
Pages 56 of 116
Table 11. General Laboratory Procedures
Medium/Material
Mr/PM10:
Soil
Dust (Vacuum)
Pood
Beverage
Purpose
Weight Concentration
Method
Balance
Texture Characterization Sieve
3 size fractions
gravel-silt (-1 to >4 o
pH
Electro-conductivity
Color Characterization
Weight/ Unit area
PH
Electro-conductivity
Veracity Assessment
comparison of food
sample with Diary
Report
Lab assay of Food Mass
pH meter
ohm meter
Munsell Cht
Balance
pH meter
ohm meter
Volumetric
Measures
GC/MS or
GC/ECD
Veracity Assessment
comparison of Fluid
sample with Diary
Report
Lab assay of Beverages
Volumetric
Measures
GC/MS or
GC/ECD
Quantitative/
Qualitative
Toler-
ance
Quant/estimate.01 mg
Quantitative
Quantitative
Quantitative
Qualitative
Quantitative
Quantitative
Quantitative
Qualitative
Quantitative
+ 5%
+ .2
+ .02
mho
.Olmg
+ .2
+ .02
mho
001-
025
mg/kg
metals
5-2.5
ng/g
pesticides
Qualitative
Quantitative
.001-
.025
g/L
metals
.01-.03
g/L
pesticides
-------
Section No. 1.7.6
Revision #0
June 1995
Page: 57 of 116
Table 12. Quality Assurance Objectives for Sample Characterization.
Medium Sample Type
Sample
Location/Form
Indoor/Filter"
Analytical
Laboratory
Analytical
Method
Precision
%RSD
Recovery
%
MDL
Air
Metals
Pesticides
VOCs
Outdoor/Filter
Personal/Filter1
Indoor/Filter/PUF
Outdoor/Filter/PUF
Personal/Filter/PUF
Indoor/whole air
Indoor/PSD (OVM3500)
Indoor/multisorbent tube
Outdoor/whole air
Outdoor/PSD (OVM35O0)
Outdoor/multisorbent tube
Formaldehyde Indoor/PSD (PF-1)
OutdoocflPSD (PF-1)
Battel le
UA
Battelle
UA
Battelle
UA
Battelle
Battelle
Battelle
UA
Battelle
Battelle
UA
Battelle
Battelle
Battelle
Battelle
ICP-AES
XRF (Stage II)
ICP-AES
XRF (Stage II)
ICP-AES
XRF (Stage II)
GC/MSD
GC/MSD
GC/MSD
PID* (Stage II)
GC/MS*
TD-GC/MSD*
pro (Stage II)
GC/MS
TD-GC/MSD
Colorimetric
Colorimetric
±30%
±30%
±30%
±30%
±30%
±30%
±15%
±15%
±15%
±20%
±15%
±20%
±20%
±15%
±20%
±15%
±15%
70-130%
70-130%
70-130%
70-130%
70-130%
70-130%
70-130%
70-130%
70-130%
75-125%
95-100%
80-120%
75-125%
95-100%
80-120%
75-125%
17.4-1,400 tig/m"1
0.1-3.5 ng/m3
17.4-1,400 itg/m3
0.1-3.5 ng/m3
52.1-4,170 ng/m3
0.1-2.7 ng/m3
3-20 ng/m3
3-20 ng/m1
6-40 ng/m3
EVOCs —500 ppb
0.5 ng/m3
0.1 ppbv
SVOCs -500 ppb
0.5 ng/m3
0.1 ppbv
10 ppbv
Dust&
Dermal
Metals
Carpet Dust"
Battelle
UA
ICP-AES
XRF (Stage II)
±30%
±30%
70-130%
70-130%
0.1-8 jtg/g
0.1-2.5 ng/g
Surface Wipe®
Battelle
ICP-AES*
±30%
70-130%
2-160 ng/m2
Hand Wipe'
Battelle
ICP-AES
±30%
70-130%
0.1-8 ng/sampie
Pesticides
Carpet Dust1-"
Carpet Dust"
Carpet Dust1-4
Surface Wipe*
Hand Wipe
Battelle
Battelle
Battelle
Battelle
Battelle
ELISA (Stage II
optional)
GC/ECD (Stage II)
GC/MSD
GC/MSD
GC/MSD
±25%
±15%
±15%
±15%
±15%
50-150%
70-130%
70-130%
70-130%
70-130%
0.25 ngfe
(0.03 ng/m )
o.i ng/g(0.oi ng/m2)
0.02-0.1 ng/g
(0.003-0.0! ng/m2)
0.4-2 ng/m2
0.02-0.1 ng/sample
-------
Section No. j_ j g
Revision #0
•Tune 1995
Page: 58 of 116
Table 12 (cont.) Quality Assurance Objectives for Sample Characterization.
Medium
Sample Type
Sample
Location/Form
Analytical
Laboratory
Analytical
Method
Precision
%RSD
Recovery
%
MDL
Soil
Metals
Yard/Foundation'
Battelle
UA
ICP-AES
XRF (Stage II)
±30%
±20%
70-130%
70-130%
0.1-8 fig/g
0.1-2.5 (ig/g
Pesticides
Yard/Foundation
Battelle
GC/MSD
±15%
70-130%
0.02-0.1 jig/g
Water
Metals
Tap; drinking
EPA-Ci
ICP-AES
±10%
85-115%
0.2-1 fig/L
Pesticides
Tap; drinking
EPA-Ci
GC/MS
±15%
85-115%
0.02-0.1 ng/L
VOCs
Tap; drinking
EPA-Ci
GC/MSD
±25%
75-125%
0.05 ng/L
Food
Metals
24-Hour diet33
FDA
ICP-AES
±20%
(M002-0.05 mg/kg
Pesticides
24-Hour diet0"
Pre /post ooofetd33
>h=p ttpLO ec. t/K»
FDA QC/M3 F^D, ±25%
70-130%
in-tiim
1-5 ng/g
t-i ng/g
Blood
Metals
Venous blood
CDC
ICP-AES
±10%
70-130%
6-10 fxg/L
Pesticides
(persistent)
Venous blood
CDC
GC/MS
±25%
50-120%
0.1-1 ng/L
VOCs
Venous blood
CDC
GC/HRMS
±25%
75-125%
0.03 ng/L
Urine
Metals
1st Morning void
CDC
ICP-AES
±10%
70-130%
2-6 ng/L
Pesticides
1st Morning void
CDC
GC/MS
±15%
50-130%
1-5 Hg/L
1 Assumes 1 g of dust and 8 m2 area vacuumed.
b Chlorpyrifos analysis only.
c Chlorpyrifos and diazinon analysis.
d Analysis of all four target pesticides.
* Variable surface area (calculated here in units of ng/m2,
assuming 0.05 m' sill area wiped.
Assumes 25 g aliquot of food extracted.
* PID « photoionization detector, GC/MS = gas chromatograph/
mass spectrometer; ID - thermal desorption-GC/MSD (mass selective
detector).
" Assumes air sample volume of 5,760 L over 24-hour sampling period.
' Assumes air sample volume of 1,920 L over 8-hour sampling period.
' Assumes 1 g sample.
-------
Section No. 1.7 . 6
Revision #0
June 1995
Page: 59 of 116
The detection limits and the ability of any evaluation to measure the
presence of a contaminant are also outlined in the Laboratory SOPs.
1.7.7 Acceptable limits on false positives and false negatives
Acceptable limits on false positive and false negative data depend on data
type. Our project has two major types of data and different approaches are
used to determine false positive and false negative values.
A. Questionnaire data: In general, a question will be regarded as not
valid and not usable if more than 5% of the responses are found to be
false. This determination will be made in two ways.
1.For the Descriptive Questionnaire: 10% of the responses of
non-participating subject families will be verified by
readministration by phone.
2. For all other Questionnaires: At least 10% of the Stage 2
questionnaires will be readministered to respondents in Stage 3.
These results will be compared to evaluate discrepancies. If any
question has more than 5% non-explained and inconsistent responses,
the question will be eliminated from the data base for all
respondents.
B. Measurements (Physical or Chemical): Tables 10-12 describe methods and
associated known detection limits. Since the NHEXAS goals are the
identification of the upper 90th percentile of the distribution function
for each of the target pollutants, a relatively high error can be
associated with measurements at the lower limits of detection. If the
detection limits represent the upper 25% of the distribution then NHEXAS
can still meet its goals. Blanks and spiked samples will be needed for
analysis of each sample batch in every sample lab (including the Federal
Contract Labs). These will detect major lab errors. Other false
negatives and false positives most frequently occur at or near the
detection limit. These are not likely to be a problem. Their detection
would generally place the sample in the lower half of the distribution.
Critical false positives are those falling as upper end outliers. When
sufficient sample with adequate integrity is available, then outliers in
the upper 2% of the distribution will be re-evaluated to confirm the
value validity.
The laboratory SOPs outline the accuracy, precision and acceptance
limits for all lab generated data. Method Detection Limits are reported
in Table 12.
2.0 MANAGEMENT ASSESSMENT
2•1 Assessment Responsibility
A. The Principal Investigator will have overall direction of the project
(and coordination with the appropriate outside agencies {i.e., County
Departments, etc.). The Principal Investigator is ultimately
-------
Section No. 2.1
Revision #0
June 1995
Page; 60 of 116
responsible for all aspects of the project. The On-site Co-principal
Investigator will be responsible for ensuring that project control
documents are prepared by the project staff in a timely manner. The
Principal Investigator and/or On-Site Co-principal Investigator will be
responsible for reviewing and approving all project control documents as
well as all final project deliverables to ensure their compliance with
the project requirements. Responsibility may be delegated to project
personnel for implementation of the Quality Assurance program.
B. The On-Site Co-principal Investigator and/or Coordinator(s)/
Supervisor{s) will be responsible for preparing or delegating
responsibility for the preparation of project documents.
Responsibilities include ensuring that proper quality control
activities are implemented in the research program, in the office,
field, laboratory and data offices.
C. The Director of Quality Assurance is responsible for the maintenance of
this QA Plan. Responsibilities include verifying the proper
implementation of project control documents through the performance of
planned and scheduled quality assurance audits/inspections (see Appendix
C).
D. The Director of Quality Control is responsible for the overall
implementation of this QA Plan and for all QA/QC activities within the
project. Responsibilities include assisting the project technical
staff in the development of the required project control documents.
The Director of Quality Control is further responsible for developing
and implementing a plan for regularly monitoring the quality aspects of
project activity and providing periodic documentation of monitoring
activities to the On-Site Co-Principal Investigator and Project
Coordinator(s) and to the Principal Investigator as appropriate. The
Director will assist the project staff, as required, in the resolution
of quality related problems (Appendix C).
E. The appointed in-house Quality Assurance personnel are responsible for
verifying that quality activities are being implemented regularly by
the laboratory and field personnel as prescribed in the Standard
Operating Procedures. They will notify the appropriate Project
Coordinator and/or On-Site Co-Principal Investigator upon discovery of
any quality related problems or non-compliances.
2.2 Assessment Types
36 •
A. Informal QA Audits; Each Project Coordinator/Supervisor has specialized
QA responsibilities defined in the SOPs. Each will perform routine
in-house QA Audits on appropriate QA records and report in writing to
the On-Site Co-Principal Investigator. These audits will be performed
on a quarterly basis. Reports are due the last day of each quarter.
Errors will be sought in;
* technician understanding of procedures,
* equipment performance,
-------
Section No. 2.2
Revision #0
June 1995
Pages 61 of 116
* recording data,
* compiling data, and
* analysis of data.
Mechanisms of error searches includes
* direct visual inspection,
* comparison of forms, and
* statistical evaluation.
3 7
B. Formal QA Audits: These will be performed by the QA Unit of the
University of Arizona for all 3 members of the Consortium. The QA audit
will pertain to the specific QA requirements for each member. Field
Audits will occur once during the first 3 months; then at 6, 12 and 18
months. Laboratory and Data Audits will occur annually.
3 Assessment Usage
Quality Assurance assessment procedures will be used in all aspects of the
project including: document development, procedure evaluation, procedure
implementation, sample collection, sample shipment, sample analysis, data
reduction and data analysis.
Quality Assurance Audits will be performed to identify and correct faulty
conduct within the project.
4 Assessment Criteria
The University of Arizona, Office of Arid Lands Studies contains qualified
scientists certified as QA reviewers. This University of Arizona
department is independent. They will perform formal audits on all aspects
of the project.
Documents that will be enforced include this QSIP, all appendices and all
Standard Operating Procedures identified within this QSIP.
5 Assessment Document at ion
Internal QA audits will be submitted to the on-site Co-Principal
Investigator on the last Friday of each month. All notebooks containing QA
documents as described in the SOPs will be checked, signed and dated by
the appropriate Coordinator/Supervisor. Each Coordinator/Supervisor will
write a memo to the On-Site Co-Principal Investigator disclosing the
results of the audit in their area and proposed corrective action in
accordance with the SOPs if needed. Documentation of the informal QA
audits will reside with the On-site Co-Principal Investigator.
Copies of all publications and associated peer reviews of articles, theses
or dissertations must be kept on file with the principle or On-site
Co-Principal Investigator(s).
-------
Section No. 2.5
Revision #0
June 1995
Page: 62 of 116
Results of the External QA Audits (as specified in Appendix C, QA-02
section 7.2 B) will be provided to the on-site Co-Principal Investigator
and the Principal Investigator.
3.0 PROJECT IMPLEMENTATION PLAN
3.1 Project Design Criteria
3.1.1 Siting Criteria
Information and physical samples will be collected only from homes under
the jurisdiction of the State of Arizona. These homes will be selected
using a population-based probability design (see section 1.7.3).
Exclusionary criteria include: Military bases, institutions and residences
occupied by more than 9 unrelated persons and part time residents with a
permanent residence elsewhere. We are currently working through 1PA Region
9 to gain access to Native American Tribal lands and members
(Reservations).
Once a home is selected for monitoring, equipment will be sited inside and
outside the home to obtain specific measurements. Siting criteria for each
procedure are defined in the specific Field SOPs (see Figures 1-6). A
summary is presented below (Table 13).
3.1.2 Project Data Quality Objectives (DQOs)
Measurements never reflect the condition being measured perfectly. Data
Quality describes the uncertainty associated with any measurements taken
whether by questionnaire, description or an actual measurement. Data
Quality Objectives describe the degree of uncertainty that will be
acceptable for each portion of the project. A single analyte obtained from
a given medium may have several DQO measures associated with it. All must
fall within the defined limits for the sample to have validity. Some types
of uncertainty can be reported quantitatively; other conditions can only be
described qualitatively. The following discourse will describe DQOs as
they pertain to our study.
A. Representativeness of the Evaluated Population.
Our study design outlines the selection of the subjects to be evaluated
through a 4 phase process and characterizes the adequacy of the number
of households (primary respondents) that will be evaluated at each study
phase. Please refer to section 1 for details. The DQO for each stage
of our study is 75% of the targeted population. We expect to exceed
this for the Descriptive Questionnaire and all of our Stage 2 sampling.
B. Consistent Application of the Methodology.
Five components play a major role in the consistent application of the
study methodology. These are education, experience, training, clear
written procedures, implementation and supervision/evaluation.
-------
Section No. 3.1.2
Revision #0
June 1995
Pages 63 of 116
Table 13. Siting Criteria for samplers at households
-------
Section No. 3,1.2
Revision #0
June 1995
Pages 64 of 116
Table 13 (cont). Siting Criteria for samplers at households (M= metals analysis; P = Pesticide Analysis;
VOC).
Media Analyte Equipment Indoor Location Outdoor Location
& *Rationale & "Rationale
the wall on each side of the house.
•Foundations are treated with
chlorpyrifos; leakage may be
differential.
M+P
Ziploc bag
and stainless
steel trowel
none
Using each foundation
sample point as a reference, move
perpendicularly in to the yard 10'
and collect 30-60 g in each location.
* provides a composite that is
consistently similar among homes.
Dust
M+P
Vacuum Sample
Composite of up to 8 m
from the main room and
the bedroom of the
Primary Respondent;
Precise area sampled
is recorded* Maximizes
the view of subject
exposure.
None
M+P Window Sill
Wipe
Composite: 2 square foot None
of main room window plus
2 square foot of bedroom
window sill on each of 2
wipes. * good location
for Pb paint. Good view
of soiI component of house dust.
Dermal M 2x2 Wipe Both hands to wrist
(respondent 1) at Time 1
p 2x2 Wipe Both hands to wrist
(respondent 2) at Time 2
Water M,P,VQC Plastic/ Kitchen tap after 2 min None,
(tap) amber glass flush. Aerator removed
vials to minimize gas "Flushed
Sample is the majority
of cons lined water. Water
collection will be
consistent among houses.
Teams will visit houses
-------
Section No. 3.1.2
Revision #0
June 1995
Page: 65 of 116
Table 13 (cont). Siting Criteria for samplers at households (M= metals analysis; P = Pesticide Analysis;
VQC).
Media Analyte Equipment Indoor Location
& *Rationale & *Rationale
Outdoor Location
after variable water use.
An "unflushed" sample at
one home will not be
related to the same sample
at another depending on
time of day and water use.
Water MfP,V0C
(drinking)
Plastic/ Collect as is from None,
amber glass storage container at
vials the house * greater
potential for contaminant
from container.
Food M,P,V0C
Ziploc
bags
Duplicate of all food
consumed within 1-day
regardless of source.
None.
Beverage M, P,VOC
Plastic Duplicate of all None,
container beverage consumed within
1-day regardless of
source.
Blood and M,P,V0C
Urine
Vacutainer
specimen
cups
Primary Respondent and None,
other household residents
if resources are
avallable.
-------
Section No. 3.1.2
Revision #0
June 1995
Pages 66 of 116
1. All employees must understand their roles and tasks. Sometimes the
education needed is formal. At other times it is implemented through
on-the-job lectures, readings and demonstration. The educated
project employee will understand why methods must be applied
consistently and is more likely to conform.
2. Experienced employees may already possess the skills to perform the
job and realize the importance of applying the same techniques
throughout the project. They may be easier to train in the specifics
of our project protocols and needs.
3. Training is specific to each project regardless of the amount of
experience or education an employee may have. All employees will be
trained in terms of the specific procedures for the NHEXAS project.
Each employee will observe new protocols and perform them
successfully (without coaching) a number of consecutive times under
observation prior to certification in a task. Employee training will
be documented and checked periodically. The performance standards
and periodic evaluation will be determined for each task as a
function of its complexity.
4. Clear, concise, written procedural instructions must be prepared for
every task. All employees must follow these procedures and any
deviations must be documented. Major changes must be documented and
released as a revised procedure (SOP). A new raw data base must be
created for the data collected under the new procedure clearly
marking it as differing from the preceding data collected.
5. Once certified, the employee will perform the work as specified by
the supervisor. Employee proficiency will be documented and checked
periodically. Failure results in loss of certification and training
must be repeated and documented. All tasks performed by the employee
since the last evaluation will be evaluated. If problems arise the
data will be re-evaluated if possible or voided if they can not be
re-evaluated.
Supervisory/professional staff have attained credentials through
education, experience and periodic training seminars. Their work
should be documented. Qualifications of all personnel must reside
with the appropriate Co—P.I. and be updated annually.
Training SOPs
Field Staff; QA-T-1.0, QA-T-2.0, UA-T-3.0, UA-T-4.0
Laboratory Staff: UA-T-6.0, BCO-T-l.O
Data Staff: UA-T-5.0
Specific SOPs as they pertain to the various project components are
illustrated in Figures 1 through Figure 6. We have one SOP that
specifically addresses interviewing technique (UA-T-1.0). If a
-------
Section No. 3.1.2
Revision #0
June 1995
Page: 67 of 116
manual for uniform questionnaire administration is developed we will
append the manual to this SOP and require that administration of the
Descriptive and Baseline questionnaires are performed according to
the protocol for the Primary Respondent. (All other questionnaires
are self-administered in the NHBXAS Arizona project).
Completeness
NHEXAS is designed as a comprehensive exposure assessment study. The
following possible sources of error will be evaluated.
1. Survey Design. Once the grid is established and sampling units
selected, areas must be completely sampled from the PSU through the
individual (see Figure 8J for each sequential PSU. Partial sampling
of a PSU negates the validity of the design (See section 1 for
details). If resources are insufficient a leastwise deletion of PSUs
may occur (from the bottom up)1.6 If additional NHEXAS resources are
forthcoming, the study can be expanded by accessing the subsequent
PSUs on the list.
2. Field Collections. All questionnaires, samples and sample forms are
to be collected and completed according to the SOPs at all
participating houses. ALL information blanks on sample forms and
questionnaires should be filled. If data are not collected because
the medium is unavailable then that information is recorded on the
sample form and coded into the data base. If collection is attempted
and the item is unavailable or insufficient for analysis then the
information is recorded and the house is completed for available
aspects. Blood and urine will be requested, they will be permissible
nonresponse items. Respondents have the right to refuse all or
partial participation at any time. Minimal requirements for
household completion ares (1) 75 % of the solicited Descriptive,
Baseline and Follow—up questionnaires and (2) 75% of all
environmental samples completed for the sampling stage. The
completion status at each project stage and phase will be noted in
the overall data collection/custody/sample-integrity/analysis
tracking management system.
3. Laboratory Analysis. Data will be obtained for 90% of all samples
analyzed4. Field blanks, lab blanks, spiked and replicate samples
will be collectively obtained for 10% of the samples. Initially
(first 6 months) all 10% of these QA samples will be analyzed. The
QA sample collection will be reduced to a collective 5%.
4. Data Management. Every field in the data base will be filled. If a
procedure is not applicable, or data are missing or refused, then the
field will be coded accordingly (UA-D-5.0). All batches of data will
be appended to the master data bases. Missing data will be tracked
and identified.
-------
Section No. 3.1.2
Revision #0
June 1995
Page: 68 of 116
5. Analysis. The data will be analyzed according to appropriate SOPs
(IIT-D-1.0 through IIT-D-3.0) and a final report will be written and
submitted to the IPA Project Officer.
D. Accuracy
1. Study Design. Accuracy errors associated with the study design are
discussed in section 1.
2. Field implementation.
a. Housing unit location: all houses will be located by latitude and^
longitude using GPS equipment. This is accurate within a few
hundred meters. To preserve confidentiality the data will be
encrypted.
b. Questionnaires: A random selection of 10% of all Descriptive
questionnaires collected from each Arizona Counts will be
readministered by phone and accuracy evaluated.
c. Field collection: All data labels will be preprinted barcodes
minimizing errors of identification and transcription. when
possible field readings will be stored in a data logger and down-
loaded instead of transcribed. Some records will be transcribed to
make sure the data has been properly downloaded. When
transcription is required, all critical values will be read by one
technician and verified by another. QC checks are built into each
field SOP.
d. Equipment will be calibrated in the field according to the
specifics of each field procedure.
3. Laboratory Analysis:
a. Full QA will be practiced and documented throughout the project.
b. QC samples will be processed. Recovery of spikes will be greater
than 50-150%, recovery of surrogate standards (NIST) will be
greater than 80-120%. Variability of reference standards will be
less than 25% (RSD). NIST NHEXAS standards will be employed!2
c. The laboratory will be asked to re-evaluate the outliers in terms
of the QC samples collected and any laboratory spikes evaluated.
If a possible error is evident from the QC samples, the sample
will be reanalyzed if sufficient sample remains. If the sample
can not be reanalyzed, it will be coded as void.
4. Data Management:
a. Data loggers will be down-loaded into prepared data base batches.
Values will be compared with selected hand recorded field
observations to ensure that the batch was properly downloaded and
then appended to the master database.
b. Questionnaires and forms will be set up for scanning, thereby
limiting potential data entry errors.
c. Some data will have to be entered in a traditional fashion. These
data will be entered, verified (100% re-entry), cleaned, validated
and appended to the master data base. The re-entry step usually
reduces data error to under 1%.
-------
Section No. 3.1,2
Revision #0
June 1995
Page: 69 of 116
d. Outliers will be checked for data accuracy. If a second value is
obtained and both sample results are deemed valid, then a mean of
the sample values will reside in the database!3 If one sample is
invalidated, then the reasons for the invalidation error will be
documented and the new value will be entered in the database.
Internal comparability of our study methods
When finalized, the collection methods and the data management system
will remain the same throughout the study. Any unanticipated changes
will require a side-by-side methods comparison. If a collection method
change occurs, the data will be placed in different master databases.
These can be copied and combined for analysis, but the database
structure will reflect the field or laboratory procedure used.
We are employing both non-intensive and intensive methods within our
study stages.
1. Metals: Air, Soil and Dust
We will use a SPECTRACE 9000 XRF for the chemical analysis of
particulate matter captured on filters (PM10), soils and house dust.
We can obtain results for all of our target metals. This method will
be used on ALL air, soil and dust samples collected. The method is
non-destructive. In stage 3, 4, 5 of the study implementation, the
same samples will be sent to Battelle for AAS and ICAP-AES analysis?6
Our goal with the XRF is to identify the high end (upper 90%) of
exposure and maximize our sample numbers while minimizing costs. A
precise measure of the metal concentrations for the lower 50%-75% of
the distribution is not required to meet our NHEXAS goals.
2. Pesticides: Soil and Dust
We will use either an ELISA method or GC/ECD to identify chlorpyrifos
and diazinon in soils and house dust. This method will be used on
all Stage 2 soil and dust samples. In Stage 3,4,5 8 GC/MSD
techniques will be used and where there is sufficient sample for a
split, ELISA will be used as well*3
Our goal with the ELISA and GC/ECD methods is to identify the high
end (upper 90%) of exposure and maximize our sample numbers while
minimizing costs. A precise measure of the pesticide concentrations
for the lower 50%-75% of the distribution is not required to meet our
NHEXAS goals.
1 6
3. Total VOCs
We will use a Photo-Ionization Detector (PID) to assay total VOCs in
residences. The calibrated equipment can take many readings inside
homes very quickly. This method will be used inside and outside all
-------
Section No. 3.1.2
Revision #0
June 1995
Pages 70 of 116
Stage 2 homes. During Stage 3, passive VOC badges will be placed
inside and outside the homes and both individual and total VOCs will
be determined.1 The badges will be analyzed using GC/FID techniques.
We will also sample Stage 3 homes using actively-pumped multisorbent
tubes. The tubes will be analyzed using a thermal desorption GC-MS
technique.
Our goal with the PID is to identify the high end (upper 90%) of
exposure and maximize our sample numbers while minimizing costs.
Further, we can look separately at specific sites within a home. A
precise measure of the total VOC concentrations for the lower 50%-75%
of the distribution is not required to meet our NHEXAS goals.
EPA is interested in merging comparable data sets collected by the
different consortia. We will use all provided NIST standards to
facilitate this cross comparison. Further, we will cooperate and assist
EPA in evaluating differences between field collection and instrument
variations whenever possible in terms of subject and cost burden?0 42
F. Precision (Minimum Detection Limits) These are covered in Tables 10-12.
Precision is important for defining the distribution function.
Precision is a critical consideration when an analysis is performed on
readings with different levels of precision. Then, a simple analysis
can only be as precise as the least precise measure. In more complex
analyses weighted factors can be used to get around this constraint to
some degree.
G. Documentation.
Every aspect of the study will be documented to make it verifiable in
terms of design, field collection, laboratory analysis and data
evaluation.
3.1.3 Types of Samples Required
Up to four classes of samples will be obtained from each sample site
(household):
A. Questionnaires and Diaries: Subjects will be asked to respond to
written and verbal questions.
B. Environmental samples:3Dust, soil, air and household water samples will
be collected at homes and in some cases from the "personal air cloud".
C. Food and Beverage Samples: For a subset of the recruited subjects,
duplicate diet samples will be collected for a 24 hour period. These
will include duplicate portions of all food and beverages ingested by a
person on a day specified with the respondent and field team collecting
the samples. Each food type will be collected in its own sample bag and
composited upon return to the laboratory. Beverages will be collected
as a separate composite in a single container.
-------
Section No. 3.1.3
Revision #0
June 1995
Page: 71 of 116
D. Tissue and Body Fluid Samples: Each subject will be asked to provide
blood and urine samples during the sampling period.
1.4 Collection Media Criteria
A. Air: Outdoor air will be evaluated in 28% of the Stage 2 homes for
1 47
metals using a Hi-Vol Sampler. Indoor (100%) and outdoor (50%) air will
be collected during Stage 3. Indoor air will come from the main room.
The volume of air sampled will depend on the collection time, and the
flow rate of the pump. The proposed procedures will collect the volumes
of air described in Table 14. Outdoor air will be sampled on the north
side of the residence at the same rate and duration as indoor air.
Personal air is defined as the air the subject encounters during a
defined time period. Personal air depends on the location occupied by a
subject. Personal air"will be collected for a random 20% of the Stage 3
primary respondents.
B. Dust: Samples will be collected from carpets inside subject homes
during Stages 2 & 3. Surface dust will be collected in all Stage 3
homes.
C. Soil: In Stages 2 & 3, soil will be collected outside the homes of
subjects in two locations: (1) the foundation1 and (2) a composite of
front and back yard soil surface sub-samples to a depth of 1" (2.5 cm).
49
D. Water: Up to two water sources will be sampled from all Stage 3 (and a
random 25% of Stage 2) homes. These are collected from the kitchen
faucet and any treated or independent drinking water source. We
estimate that 50% of the homes have drinking water sources other than
tap water.
24 8
E. Food and Beverage: The Stage 3, 4 and 5 primary respondent will save
samples of all foods and beverage consumed during a 24 hour time period.
F. Subject's Body: Dermal wipes will be collected from selected subjects.
Blood and urine samples will be requested from all participating Stage 3
study respondents.
G. Target pollutant classes analyzed in the media are listed in Table 15.
1.5 Pre—assessment of the NHEXAS Arizona Field Study
The NHEXAS project is being submitted to the Office of Management and
Budget as one package for three consortia. Only nine houses can be
evaluated prior to the OMB approval. Four of these have been reserved by
RTI/EOHSI for pretesting the questionnaire^). RTI will use two for their
pre-assessment, Harvard/Johns Hopkins will use one, and NHEXAS Arizona will
use two. This means that a great deal of the preassessment will occur
within the field-office/lab.
-------
Section No. 3.1.5
Revision #0
June 1995
Page: 72 of 116
Table 14. Projected volume of air collected by active samplers. The sampling
schedule for the intermittent samplers is outlined in the
appropriate SOPs.
Sampler Compound Class
Flow Rate
Duration of
Sample
Collection
Volume of
Air
Sampled (m3)
Harvard PM10
Metals
4 L/min
.75 (72
hours)
12.96
3 9
URG PUP
Pesticides
4 L/min
.25 (72
hours)
4.32
Personal Air Metals
(Teflon Filters
SKC pump)
4 L/min
.75 (24
hours)
1.92
Personal Air
(URG PUF
SKC pump)
Pesticides
4 L/min
.25 (24
hours)
1.92
Active VOC
(Multisorbent
Carbo trap)
VOC
5 ml/min
.333 (24
i hours)
0.0024
Sentinel Air1
Metals
28.3 L/min
24 hours
40.75
-------
Section No. 3.1.5
Revision #0
June 1995
Pages 73 of 116
l
Table 15. Sampling Media for the Pollutant Classes:
Sample Medium Pollutant Class
Metals Pesticides VOCs
Air
Indoor
X
X
X
Outdoor
X
X
X
Personal
X
X
Dust
Carpet/Floor
X
X
Surface
X
X
Dermal
X
X
Soil
Foundation
X
X
Composite
X
X
Water* Kitchen tap, DW
X
X
X
Biological* Blood
X
X
Urine
X
X
Food/Beverage* 1-day dupl.
X
X
* Collection only - Analysis through IAGs or EPA Contracts
Although RTI will test the questionnaire, Spanish is widely spoken in
Arizona so we have special needs. The draft questionnaire was translated
into grammatically correct colloquial Arizona Border Spanish. A summary of
Arizona's preliminary testing of the questionnaire(s) is outlined in
Appendix B.
During the pre-assessment the study design will be evaluated. We will work
with the census files and complete the programming necessary to draw the
sample. We will identify an area segment and prepare to sample two houses.
Meanwhile, all previously used equipment is being examined and tested for
conformance with operating standards. If any fail to function properly,
they will be repaired to specifications (Harvard PM10, VOC pumps, SKC
personal air monitors, vacuum cleaners and particle collectors). New
equipment (GPS, PID and XRF) is being evaluated on samples collected and
stored from previous studies. These air, dust and soil samples are typical
of samples we expect to encounter throughout Arizona. SOPs written for
operation of the older equipment are being re-evaluated and upgraded. SOPs
written prior to the receipt of the new equipment are being evaluated and
upgraded. SOPs written prior to the receipt of the new equipment are being
evaluated and upgraded.
-------
Section No. 3,1.5
Revision #0
June 1995
Page; 74 of 116
Vast amounts of data will be collected. We have upgraded our LAN to
accommodate the incoming data. We have identified the "key" and important
variables that should be associated with each type of sample and are
creating a "tracking" data system (Figure 9). All data, data custody,
shipment, integrity, status and archiving issues will be tracked through
this system. Battelle will electronically update this file from Columbus.
The field, lab and data teams are working together to evaluate field forms
and make sure they are adequate at the time of entry into the field. Data
shells are being built to reflect these field forms.
Battelle has been testing all the materials planned for field use. Prior
to entry we will be sure that the vinyl and rubber products we use do not
contaminate our samples. We are testing our shipping and receiving
protocols and evaluating sample integrity. We are improving our sample
handling procedures by the use of bar codes and scanable field forms.
When we have accomplished these tasks, we will prepare to enter the field.
Letters will be mailed to the houses in one trial area segment. We will
select two houses for sampling.
Every phase of the field operation will be implemented and assessed. We
will evaluate all aspects from the letter through the data processing loop.
A team approach will be applied to the evaluation. All substantive
changes to protocols will be documented and resubmitted as necessary.
Field forms will be altered and we will discuss any problems detected in
the questionnaire with RTI and agree on any changes or clarifications while
not violating the spirit of any potential OMB approval.
After all problems have been discussed and an altered approach formulated,
we will perform the final dress rehearsal. We expect this to be about 2
weeks prior to OMB's potential approval. This will provide about 1 month
to resolve any additional problems that have been encountered in the second
home.
.1.6 Sampling Time and Frequency
A. General Strategy:
Housing units identified in the sampling design will be sent a letter of
introduction followed by a field interview within one week (see section
1.7.3). This letter is sent to potential respondents living on an
identified "block: and addressed to "Resident". The letter defines the
project in broad terms and provides respondents with time to consider
possible participation. This approach is consistent with obtaining the
necessary "informed consent" of the study respondents.
The sampling period will be defined in terms of person weeks. Some
measurements will be weekly; others will be collected for shorter
periods and weighted to represent a week. Certain techniques represent
a "point in time" (pit) sample related to an area (e.g. surface and
-------
Section No. 3.1.6
Revision #0
June 1995
Page; 75 of 116
Figure 9. Diagrammatic of the NHEXAS Arizona Project and Data Tracking
components.
Materials
Technician
PREPARED
MATERIALS
RECEIVED
Field
Coordinator
iniiil-iaiBSIIiiiUiUiiiiitii iiMS-:i:-Hi
Tracking
jjjjjjj Database
¦l\
FIELD ASSIGNMENT:
HMD
FIELD PERSONEL
Team
Leader
Field
Coordinator
Data
Coordinator
FIELD
COLLECTION
FIELD FORM AND SAMPLE LOGIN
TEMPORARY SAMPLE STORAGE
Data
Manager
Laboratory
Supervisor
DATA
READIED
FOR USE
-------
Section No. 3,1,6
Revision #0
June 1995
Pages 76 of 116
dermal samples. All data will be presentable in terms of person weeks
through direct measurements or a weighting factor applied to lesser
measurements. The sampling schedule is designed to permit sample
completion from 18-24 months.
B. Specifics:
Specific operational durations are outlined in the appropriate field
protocol. Projected duration of sampling within the field is
illustrated in Figures 10-13?° Sampling equipment to be used and
duration of use are presented in Tables 16-17.
Figure 10 identifies the various implementation stages relative to the
questionnaire and provides estimates of respondent burden. Figure 11
illustrates the total technician time needed in each home. The number
of technicians can be increased to keep the respondent "contact" hours
to a minimum. Figure 12 illustrates time lines for collection
10
implementation in Stages 1 and 2. Figure 13 illustrates collections
for Stage 3 (same for 4 and 5, but the number of households drop to 50).
3.1.7 Sample Collection
An overview of sample numbers and analysis methods are included in Table
16. Specific aspects of sample collection are outlined in the Field SOP(s)
(See Field SOP Volume #2).
Repeated/replicate or same sample splits will be obtained in a random 10%
of all homes sampled. All samples will be treated precisely the same and
results should be similar. The degree of similarity will be sample type
dependent. Some samples will vary by as little as 5% and other samples
will vary by as much as 50%. Tolerance limits will be defined in the
appropriate Laboratory SOPs. Over the first 6 months all 10% of the QC
samples will be analyzed. Once personnel are experienced on this project,
the analysis will be reduced by half and the remaining samples will be
archived unless problems arise.
3.1.8 Sample Handling
Samples must be handled to preserve identity and integrity. All samples
will be appropriately labeled in the field. Prior to leaving the field the
team leader will take a sample inventory. All labels will be checked and
samples will be transported and stored in accordance with specifications
outlined in each Field SOP. In general, sample handling is target
pollutant dependent. Filter samples collected for metals analysis are
robust, requiring little special treatment. Samples collected for
pesticides are semi-stable. Pesticides volatilize. Samples should be kept
cold and shipped for analysis within 5 days of collection. VOCs are
very volatile and must be kept cold and shipped within a week. Water and
food samples must be pretreated and shipped quickly. Serum and urine
samples must be frozen, and shipped on dry ice.
-------
Section No. ^ j
Revision #0
June 1995
Pages 77 of 116
Figure 10. study Stages and Questionnaire (Q) Collection burden. N equals the
Field Stage estimates of number of completed questionnaires and the time
(per individual respondent) required to complete them.
Field Visits and Subject Burden Assessment Estimates
for NHEXAS Arizona
Stage 2—Q
N = 450
t.JR,) = 3.3 hrs
Stage 1 —Q50
N = 1200
WR,} = 1.2 hrs
Stage 3—Q
N = 175
W/Q Personal Air
N = 150
= 5.3 hrs
W/ Persona! Air
N = 25
t^R,) = 29.8 hrs
Stage 4-5 -Q
N = 50
WR,! = 15.9 hrs
N; Total Households Evaluated t: Time of Interview
R,: Primary Respondent R„: Secondary, Tertiary Respondent
-------
Section No. 3.1,8
Revision #0
June 1995
Page: 78 of 116
Figure 11. Study Field Stages (P) and Technician Task Completion burden. N
equals the number of households a Field team will visit and the time
expenditure required to sample the household.
Field Visits and Field Time Expenditure Estimates
for NHEXAS Arizona
(Concurrent With Subject Interview)
N: Total Households Evaluated
R-: Primary Respondent
t: Time
Tn: Technicians, n = 1,2.
-------
Section No. 3 i g
Revision #0
June 1995
Page: 79 of 116
Figure 12. Time lines and information collected in Stages 1 & 2.
Sample Collection: Relative Timing
Questionnaires
Descriptive
Baseline 50
Day 1 Day 2 Day 3 Day 4 Day 5
10
Day 6
STAGE 2 (n=450) ta
i
I::::::::;;.:"::;;:;:;;:;!:!!!!::::::::::::::::;:::-:!,;;!
Questionnaires
50
Time/Activity-
Food Diary
Technician
Day 7
Follow-up
Questionnaire
Collection
1 1
Day 1 Day 2
Day 3
Day 4
Day5
Day 6
Day 7
t t
Sample Collection
All Homes
Soil Foundation
Soil Composite
PID
Dust Vacuum
28% of Homes
Phone
Call
Follow-up
Outdoor air
set-up
Drinking Water
Tap Water
Outdoor air
tear-down
-------
Section No. 3.1,8
Revision #0
June 199S
Page: 80 of 116
8 10
Figure 13. Time lines and information collected in Stages 3, 4 & S.
Sample Collection: Relative Timing
STAGE 3 (n=175), 4,5 (n=50)
10
Questionnaires
Time/Activity
Food Diary
Technician
Follow-up
4
Food follow-up
Questionnaire
Cnllytinn
Day 1 Day 2 Day 3 Day 4 Day 5
Day 6
Sample Collection
All Homes
PID
Drinking Water
Tap Water
VOC Active Setup
PM Metal Active
PM Pesticide Active
VOC Passive
Formaldehyde
Passive
Dermal Sample (M)
Phone
Call
Follom-up
tt
Subject
Food & Beverage
Collection
Personal Air
VOC Active
Tnifiatyi
Day 7
Air Sample
Collection
Badge
Collection
Food & Beverage
Morning Pickup
Urine Samples
Dermal Samp^ ^
51
Blood Draw
Dust Surface
Dust Vacuum
Soil Foundation
Soil Composite
-------
Section No. 3.1.8
Revision #0
June 1995
Page: 81 of 116
Table 16. Specific Equipment and operation parameters used in
Residential Sampling (Stages 2 through 5)8; I = Indoor o = Outdoor
M = Metals, P = Pesticides, PIT = Point in Time Collection.
STAGE 2(Households evaluated = 450)a 17 is
INITIAL VISIT Collection Comments
(Interviewer + 1 Technician) Interval
PID Measurements (1,0) (VOC) PIT
Vacuum Carpets (I, M, P) PIT
Foundation Soil (0,M,P) PIT
Composite Soil (0, M,P) PIT
Hi-Vol Sampler (25% homes) (0,M) 24 Hour
Water Sample(s) (28% homes) (I,M,P,VOC) PIT
Tap Water
Drinking water (if it comes from a separate source).
Concluding Visit
Questionnaire Collection 1 week record
Time/Activity
Food Diary-
Follow-up Questionnaire
-------
Section No.3.1.8
Revision #0
June 1995
Page: 82 of 116
Table 17. Specific Equipment and operation parameters used in
Residential Sampling (Stages 3, 4, 5); I = Indoor 0 = Outdoor
M = Metals, P = Pesticides, PIT * Point in Time Collection.
STAGS 3* (Performed with 1 interviewer
(Households evaluated = 175J1 17 is
INITIAL VISIT
Dual Head Harvard Sampler(I,0,M,P)
Active VOC Sampler (1,0)
Passive VOC badges (1,0)
Passive Formaldehyde badges (1,0)
PID Measurements (1,0)
Water Sample(s)
Dermal Sample (M)
Follow-up Phone Call
51
Mid-Week Follow-up
and 2 technicians)
Collection Comments
Interval
3-day integrated To assure Pesticide
PUF Stability
1-day integrated To assure VOC
(delay start) Stability
7-day integrated
7-day integrated
PIT
PIT
PIT
Food Collection
Beverage Collection
Dermal Samples (P)
Urine Sample
Dual Head Harvard Sampler
Active VOC Sampler
1-day Composite
1-day Composite
PIT
Morning Void
Separate Food
& Beverage
Post Food
Collection
Collection of
Active Samplers
Concluding Visit:
Passive VOC badges
Passive Formaldehyde badges
Blood Draw
Collection of
Passive Samplers
Supplemental Med-
Tech or Nurse
Surface Dust (1,0,P,M)
Vacuum Carpets (I,0,P,M)
Foundation Soil (1,0,P,M)
Composite Soil (I,0,P,M)
PIT
PIT
PIT
Follows collection
of Badges
*Note: Sampling in Stages 4 and 5 is the same as Stage 3 excluding the
VOC sampling (except for the PID).
-------
Section No. 3.1.8
Revision #0
June 1995
Pages 83 of 116
Table 18. Summary of samples collected for each medium in each stage.
UA/BATTELIE/IIT OVERVIEW Or SAMPLE DESIGN
I Sample Location or Active/Passive Stage 2t Stage 3 '"'s-.age"^3 Stage 5 8 !
[ Designation Medium Sample Contaminant (450) (175) (50) (50) |
I Air
Indoor
Active
Metals
175
50
50
i
Pesticides
175
50
50
1
VOCs
450
50
i
Passive
VOCs
175
50
50
I
Outdoor
Formaldehyde
175
1
Active
Metals
125
175
50
50
Pesticides
175
50
50
i
VOCs
450
50
1
Passive
VOCs
175
50
50
1
Formaldehyde
175
1
Personal
Active
Metals
25
Pesticides
25
Dust
Carpet/Floors
Metals
450
175
50
50
Pesticides
450
175
50
50
Surfaces
Metals
175
50
50
Dermal wipe
Pesticides
175
50
50
Metals
175
50
50
Pesticides
175
50
50
Soil
Foundation
Metals
450
175
50
50
Pesticides
450
175
50
50
Composite
Metals
450
175
50
50
Water
(Yard)
Pesticides
450
175
50
50
Tap
Metals
(125)
175
50
50
Pesticides
(125)
175
50
50
Drinking
VOCs
(125)
175
50
50
Metals
( 63)
63
25
25
Pesticides
( 63)
63
25
25
Food
VOCs
( 63)
63
25
25
Day
Metals
175
50
50
Biologicals
Pesticides
175
50
50
Blood
Metals
175
50
50
Urine
VOCs
175
50
50
Metals
175
50
50
Pesticides
175
50
50
Questionnaires
Weekly Activity
450-1200
175-375
50-150
50-150
Daily Diary
450-1200
175-375
50-150
50-150
Technician
450
175
50
50
Dietary Diary
450-1200
175-375
50-150
50-150
t Initial Moriitorina (Low Cost Methods - PID. XRT, ELISA)
Assumptions: (1) Average household contains 3 ntemijers (e.g., 175}|3 indicates 3 persons per home)
(2) Values indicate maximum number of houses in each category
(3) 50 Houses in Stages 4 and 5 will be a subset of the 175 houses from Stage 3
-------
Section No. 3.1.8
Revision #0
June 1995
Page: 84 of 116
In general, field and laboratory blanks will accompany a random 10% of
all homes sampled. Field blanks will be handled in precisely the same
manner as field samples. Evaluation of field and lab blanks should not
vary from the initial QA measures of the product by more than 10%.
Specific sampling, handling, and shipping requirements are outlined in each
of the appropriate Field SOPs (UA-F-4.0 through UA-F-20.0). Data flow and
control is also illustrated in Figures 14 - 22 for each media collected.
The custody is transferred as the data moves from sample to sample. s&s is
evident from these Figures, all samples will be shipped within 7 days.
3.1.9 Sample Custody
At the time of field collection, a sample custody form is completed for
each sample. The original custody form remains with each sample until it
is fully expended. Copies of the custody form are retained by each
individual who passes the sample to another person or entity. Aspects of
sample and data custody are covered in field and data SOPs. Overall issues
of sample custody are addressed in eight separate Custody SOPs (UA-C-1.0
through UA-C-8.0).
3.1.10 Sample Preparation
Samples will be prepared in accordance with written Standard Operating
Procedures developed for the laboratory evaluation of each sample type.
(See Laboratory SOPs).
3.1.11 Sample Analysis
Samples will be analyzed in accordance with written Laboratory Standard
Operating Procedures developed for each sampling type.
3.2 Data Quality Indicators
Data Quality Indicators flow from the Data Quality Objectives expressed in
section 3.1.2. If the DQIs are met, then the data has a consistently good
high quality. If DQIs are not met, then the data is not of sufficient
quality to meet the stated objectivesf3
3.2.1 Planning
All field, laboratory and analysis aspects of the study design will be
considered, tested and evaluated prior to field implementation.
A. Collect a functional, creative work team.
B. Define the problem in all of its details.
C. Plan how the project is to be approached.
D. Test the Plan. _
F. Decide how to handle errors.
E. Document, Document, Document 1
These steps form an
iterative do loop.
-------
Section No. 3.2.1
Revision #0
June 1995
Page: 85 of 116
Figure 14. Handling the Photo-Ionization Detector for Total VOC in air.1
PID SAMPLES
Team Leader
SAMPLE ID ASSIGNED
SAMPLES
TAKEN
X
Field
Coordinator
¦>
Data Coordinator
VALUES BATCHED
TO DATA ENTRY
ENTERED
VERIFIED
VALUES RETURNED
CLEANING
VALIDATION
V
Data Manager
VALUES APPENDED TO DATABASE
PHYSICAL DATA ARCHIVED
ELECTRONIC DATA TO IIT
-------
Section No. 3 2 1
Revision #0
June 1995
Page: 86 of 116
1
Figure 15. Handling of Soil Composites for yard and foundation.
SOIL COMPOSITE AND FOUNDATION SAMPLES
HELD
TEAM
TO HMD
Tearn Leader
SAMPLE PACKAGE CREATED
SAMPLE ID ASSIGNED
SAMPLES
TAKEN
X
X
Field
Coordinator
<7 Days
Laboratory
Supervisor
Data Coordinator
VALUES BATCHED
SENT TO DATA ENTRY
ENTERED
VERIFIED
VALUES RETURNED
CLEANING
VALIDATION
SAMPLE
PREPARATION
& ALIQUOTING
A
Stage
3, 4, 5
Stage 2, 3, 4, 5
X
V
Materials
Technician
V
VALUES
FROM
ANALYSIS
SHIPPED
ANALYSIS
XRF & SOIL CHARACTERIZATION
VALUES RETURNED
V
SAMPLE RESIDUAL
& FORMS ARCHIVED
SAMPLE
ARCHIVED
V
Data Manager
VALUES APPENDED TO DATABASE
PHYSICAL DATA ARCHIVED
ELECTRONIC DATA TO DT
-------
Section No. 3.
Revision #0
June 1995
Page: 87 of 116
2 . 1
Figure 16. Handling of dust samples from floors collected using a vacuum
cleaner!
VACUUM DUST SAMPLE
VACUUM FILTERS FROM BATTP.T.T.R
Team Leader
V
SAMPLE ID ASSIGNED
SAMPLE
TAKEN
X
Field
Coordinator
< 7 Days
X
Laboratory
Supervisor
SAMPLE
ALUQUOTS
HELD
FORM
VALUES
METALS
XRF
PESTICIDES
X
V
Materials
Technician
Data Coordinator
VALUES BATCHED
TO DATA ENTRY
ENTERED
VERIFIED
VALUES RETURNEE
CLEANING
VALIDATION
V
>
Data Manager
VALUES APPENDED TO DATABASE
PHYSICAL DATA ARCHIVED
ELECTRONIC DATA TO HT
-------
Figure 17,
Section No. 3.2.1
Revision #0
June 1995
Page: 88 of 116
Handling of particulate from the pumped air samplers. i 64
(Harvard PM10/URG and Personal Air PM10 metals and URG pesticides)
ACTIVE PUMP PARTICULATE MATTER
FILTERS FROM BATTELLE
Team Leader
V
SAMPLE ID ASSIGNED
APPARATUS
SETUP
APPARATUS
TAKEDOWN
PESTICIDE METALS
< 7 Pays
I
X
X
Field
Coordinator
FIELD
FORM
VALUES
¦>
Laboratory
Supervisor
XRF
VALUES
FROM
ANALYSIS
Data Coordinator
VALUES BATCHED
TO DATA ENTRY
ENTERED
VERIFIED
VALUES RETURNED
CLEANING
VALIDATION
X
Materials
Technician
SHIPPED
ANALYSIS
V
VALUES RETURNED
SAMPLE RESIDUAL
& FORMS ARCHIVED
V
>
Data Manager
VALUES APPENDED TO DATABASE
PHYSICAL DATA ARCHIVED
ELECTRONIC DATA TO HT
-------
Section No. 3.2.1
Revision #0
June 1995
Pages 89 of 116
1
Figure 18. Handling of all VOC samples (active and passive).
VOC SAMPLES
FIELD
TEAM
TOHfflD
vex: BADGES
MULTISORBENT CARBO-TRAPS
FORMALDEHYDE TUBES
FROM BATTELLE
Team Leader
< 7 Days
V
V
SAMPLES
TAKEN
SAMPLE ID ASSIGNED
X
Field
Coordinator
Data Coordinator
VALUES BATCHED
TO DATA ENTRY
ENTERED
VERIFIED
VALUES RETURNED
CLEANING
VALIDATION
X
Materials
Technician
SHIPPED
ANALYSIS
VALUES RETURNED
SAMPLE RESIDUAL
& FORMS ARCHIVED
V
Data Manager
VALUES APPENDED TO DATABASE
PHYSICAL DATA ARCHIVED
ELECTRONIC DATA TO HT
-------
Section No.
Revision #0
June 1995
Page: 90 of 116
3.2.1
Figure 19. Handling Wipe Samples (Dermal and Sill),
DERMAL AND WINDOWSILL WIPES
CURB AND WINDOWSILL BLOTS
FIELD
TEAM
TO HMD
Team Leader
WEPES FROM BATTELLE
\
/
SAMPLE ID ASSIGNED
SAMPLE
TAKEN
I
X
< 7 Days
Field
Coordinator
Data Coordinator
VALUES BATCHED
TO DATA ENTRY
ENTERED
VERIFIED
VALUES RETURNED
CLEANING
VALIDATION
X
V
Materials
Technician
SHIPPED
ANALYSIS
V
VALUES RETURNED
SAMPLE RESIDUAL
& FORMS ARCHIVED
V
>
Data Manager
VALUES APPENDED TO DATABASE
PHYSICAL DATA ARCHIVED
ELECTRONIC DATA TO IIT
-------
Section No.
Revision #0
June 1995
Pages 91 of 116
3.2.1
Figure 20. Handling Separate Food and Beverage Samples
23 24
FOOD AND BEVERAGE SAMPLES
sample containers]
Team Leader
SAMPLE ID ASSIGNED
SAMPLE
TAKEN
I
X
Field
Coordinator
FIELD
FORM
VALUES
->
X
< 7 Days
Laboratory
Supervisor
SAMPLES CHECKED
AGAINST FOOD DIARTKS
Data Coordinator
VALUES BATCHED
TO DATA ENTRY
ENTERED
VERIFIED
VALUES RETURNED
CLEANING
VALIDATION
X
V
Materials
Technician
SHIPPED
V
ANALYSIS
VALUES RETURNED
V
¦>
AVAILABLE SAMPLE
RESIDUALS RETURNED
& FORMS ARCHIVED
Data Manager
VALUES APPENDED TO DATABASE
PHYSICAL DATA ARCHIVED
ELECTRONIC DATA TO HT
-------
Section No. 3 _ 2 1
Revision #0
June 1995
Pages 92 of 116
Figure 21. Handling Tap and Separate Drinking Water Samples!"3
HELD
team
TOHHir
TAP AND DRINKING WATER
'Team Leader
' 7 Days
r
V
[SAMPLE BOTTT FS ]
V
[sample id assigned"]
SAMPLES
taken
x[
Field
Coordinator
X
Laboratory
Supervisor
SAMPLES
PRESERVED
I
X
[shipped
Materials
Technician
T analysis]
VALUES RETURNED
SAMPLE RESIDUAL
& FORMS ARCHIVED
FIELD
FORM
VALUES
Data Coordinator
VALUES BATCHED
TO DATA ENTRY
entered
VERIFIED
VALUES RETURNED
CLEANING
VALIDATION?
V
-------
Section No.
3.2.1
Revision #0
June 1995
Page: 93 of 116
Figure 22. Handling Urine and Blood Samples.
BLOOD AND URINE SAMPLES
FIELD
TEAM
ARRANGES
SAMPLING
SAMPLE CONTAINERS
Team Leader
V
SAMPLE ID ASSIGNED
SAMPLES
TAKEN
BY THIRD PARTY
1 Day
PAPERWORK
X
V
Field
Coordinator
SHIPPED
ANALYSIS
VALUES RETURNED
V
SAMLE RESIDUAL
& FORMS ARCHIVED
Data Coordinator
VALUES BATCHED
TO DATA ENTRY
ENTERED
VERIFIED
VALUES RETURNED
CLEANING
VALIDATION
V
Data Manager
VALUES APPENDED TO DATABASE
PHYSICAL DATA ARCHIVED
ELECTRONIC DATA TO HT
-------
Section No. 3.2.2
Revision #0
June 1995
Page: 94 of 116
3.2.2 Field Data
The following data quality indicators will be documented for the field
portion of the project.
A. Trainings Each field team member is completely trained in the
performance of all field procedures including collection, handling,
transport, storage and shipment. Each member is required to know and
refer to the SOPs governing each field procedure. Appropriate and
thorough training results in quality samples. Good training is
evidenced in superior performance evaluations and timely corrections in
performance. Technicians are evaluated monthly and corrective actions
are documented.
B. Field Staff Evaluation: The Project Field Coordinator is responsible
for incremental evaluation of all field staff in the performance of
sample and data collection. Performance is evaluated and documented on
the Field training SOPs. The project Field Coordinator is responsible
for maintaining these records for audit.
26 54
C. Field Cross Checks: All field technicians are cross trained. Field
Team Members are responsible for cross checking each other's work in the
field. 100% of all critical data are cross checked in the field and
documented on field collection forms. The team leader checks all sample
labels, packaging and transport storage prior to departure from the
household being evaluated.
D. Internal QA Audits by the Project Field Coordinator: Within 24 hours of
sampling, all completed field forms and data sheets are reviewed by the
Project Field Coordinator for completeness and appropriateness of
response?-* Sample labels on all QX, forms and samples are inspected for
completeness. Custody forms must accompany each sample and be
appropriately completed and signed. Field samples must be stored in
accordance with the SOPs and shipped or delivered by the Project Field
Coordinator or his/her designate. Internal monthly audits of all QA/QC
procedures, notebooks and documents will be made on the last Friday of
each month. A written report will be provided to the Co-PI in charge of
that project section.
E. Sample integrity is lost if the chain of custody is broken. (See Custody
SOPs). Poor custody transfer records will result in corrective action.
F. Repeated or replicate samples are collected routinely from homes.
Similar lab results are indicative of consistent collection techniques.
Repeated/replicate or split samples varying by more than 20% are
indicative of either poor field or poor lab practice and corrective
actions {like training review and increased internal and external
audits) will be implemented.
G. External Field Audits: A representative of the University of Arizona's
Quality Assurance Unit will perform periodic field audits. During the
first 3 months of field operation, audits will occur monthly.
-------
Section No. 3.2.2
Revision #0
June 1995
Page: 95 of 116
Thereafter, quarterly audits will be performed. Results of these audits
will be communicated directly to the Principal Investigator. Appendix C
outlines the audit procedures that will be followed. 37
3.2.3 Laboratory Data
The following additional data quality indicators will be documented for the
laboratory.
A. Internal quality control procedures to monitor measurement precision and
accuracy will be implemented for sample collection in the field and for
sample analysis in the laboratory. These internal quality controls will
consist of the analyses of field blanks, lab blanks, lab spikes, and
reference matrix samples, in addition to replicate analyses of selected
samples. The data reduction for these sample types has been discussed
previously. The data quality objectives that will be met through these
analyses are listed at the beginning of this document. QA samples
(blanks, spikes, reference samples, replicates) will encompass all
sample matrices and will constitute approximately 10% of the total
sample number. Each lab supervisor will perform an internal monthly QA
audit (last Friday of each month) and provide a written report to the
Co-PI supervising that section of the project. 37
B. External Quality Assurance Measures: National Institute of Science and
Technology (NIST) will provide materials for performance evaluation and
standards. These will be run with samples as independent QC checks in
accordance with Laboratory SOPs.
C. The QA officer will review the results of these control samples.
Aspects of each individual analytical procedure also serve as checks of
internal QC. These are discussed below.37
D. Pesticides: Each sample will be spiked with a recovery surrogate. The
recovery of the spiked surrogate standard will be reported for each
sample as a running check on the reliability of the extraction and
analysis procedures. Recovery of analytes will not be corrected by
recovery of the surrogate standard. The same surrogate standard will be
spiked into each sample matrix, so that method performance can be
assessed broadly, on the basis of its average recovery, for each matrix.
Recovery of this surrogate standard will also help to identify samples
where data may be suspect due to poor recovery of the surrogate.
The GC/MS instrument performance is also monitored in the laboratory
through the use of an internal standard for quantification. Significant
differences, during the course of a day, in the area of the internal
standard quantification ion (>50% DM) signal instrument difficulties
that require operator attention.
1. Metals: The analysis of one or more internal calibration verification
standards will be used as a periodic check on instrument drift. These
data will be used to signal the need for a new calibration curve, or
reanalysis of samples.
-------
Section No. 3.2.3
Revision #0
June 1995
Page: 96 of 116
F. VOCs: The GC/HS instrument performance will be assessed by comparison
of current data with historical VOC data. Peak areas of compounds in
reference standards will be used as a check of this performance.
3.2.4 Databases
The following additional data quality indicators will be documented for
electronic data.
A. The structure of the databases is defined in the planning phase and all
necessary key indicators are recorded on all forms, samples,
questionnaires and records.
B. Each database undergoes multiple procedures to ensure that it reflects
what is reported in the questionnaire or on the data form. A detailed
explanation of these is offered in SOP# UA-D—25.0 and UA—D-26.0.
c. Prior to any data analysis, univariate and bivariate analysis are
undertaken and the outliers are checked against field and laboratory
records for any discrepancies.
3.3 Project Responsibilities
The ultimate responsibility for the success of the project falls on the
Principal Investigator. However, he cannot supervise such a large project
alone. All project employees are responsible for immediately reporting
problems to their direct supervisor. If the problems cannot be readily
solved then they are reported further up the chain of command (Figure 24).
The solutions must be in conformance with all project documents. The
project QA officer can always be consulted for an outside perspective to
resolve problems. Each Co-Principal Investigator (Co-PI) is responsible to
the Principal Investigator for a portion of the project.
3.3.1 Overall administrative responsibility for the project will be assumed by
the Principal Investigator.
3.3.2 The Principal and Co-Principal Investigators will supervise and document
the implementation of the study and population selection.
3.3.3 Day-to-day field operations will be implemented and documented in
Arizona and are the responsibility of the Arizona Co-Principal
Investigator and the Project Field Coordinator.
A. Changes in field study design will be made and documented by the
Principal and Co-Principal Investigators.
B. Changes in the Questionnaire will be made by the PI in conjunction
with the Pi's of the other consortia and EPA and conforming with
office of management and budget regulations.
-------
Section No. 3.3.3
Revision #0
June 1995
Page: 97 of 116
C. Major changes in types of equipment used to sample a medium will be
made and documented by the Principal and Co-Principal Investigators.
D. Minor substantive changes in field protocol implementation will be
made by the Arizona Co-PI and the Project Field Coordinator with
documentation sent to the other Co-PIs and the Principal
Investigator.
E. Collection, preliminary sample preparation and stabilization,
shipping and maintenance of samples and questionnaires surveys and
field collections will be the responsibility of the Project Data
Coordinator.
F. The Field Coordinator is responsible for the post field transfer,
storage and shipment of all collected samples. These are to be
performed according to the appropriate protocols.
G. The Field Coordinator is responsible for maintaining and calibrating
equipment during the field assessment.
H. The Field Coordinator is responsible for training, supervising
monitoring, evaluating, enforcing all QC/QA requirements for his area
and assigning tasks to all the field staff.
J. The Field Coordinator is responsible for the field and questionnaire
data during the field procedures and transferring all original data
records to the data section?8
3.3.4 Day-to-day laboratory operations will be implemented and documented in
in each of the responsible labs and are the responsibility of the Local
Co-Principal Investigator and the Laboratory Supervisors. Several labs
participate in this study.
A. Major changes in Laboratory Methods either in terms of equipment used
or approach will be recommended by the local laboratory to the
Principal and Co-Principal Investigators, and decided and documented
by them.
B. Minor changes in laboratory protocol implementation will be made by
the Arizona Co-PI and the Project Field Coordinator with
documentation sent to the other Co-PIs and the Principal
Investigator.
1. Arizona and Battelle
a. The Lab Supervisor is responsible for the post field
evaluation/preparation of all collected samples. These are to be
performed according to the appropriate protocols and include
sample storage, shipment, and archiving.
-------
Section No. 3,3,4
Revision #0
June 1995
Page: 98 of 116
b. The Lab Supervisor is responsible for maintaining and calibrating
support equipment for lab analysis and field assessment.
c. The Lab Supervisor is responsible for training, supervising
monitoring, evaluating, adhering to all QA/QC requirements for his
area and assigning tasks to the other laboratory personnel.
d. The Lab Supervisor is responsible for keeping laboratory data
records and providing copies of all data records to the data
section.
e. Equipment and materials evaluation and certification will be
performed by Battelle.
2. Laboratories performing Food, Water, Blood and Urine Analysis:
a. The labs will be responsible for performing all analysis according
to the specifications set up by 1PA in the various contracts or in
accordance with this document at the request of EPA.
b. Arizona will provide 10% field QA samples for water and food. For
water 5% of the QA samples will be replicates and 5% will be
blanks. For food 5% will be "blanks" and 5% will be spikes.
Sufficient sample is provided to CDC for splits of biological
samples for QA.
c. The Federal Labs will be responsible for adhering to Standard QA
procedures.
d. The labs will run a total of 10% QA samples for field blanks, Lab
/blanks and spiked samples.
e. Unexpended blood will be archived and co-owned by CDC and the
Consortium for future research as we jointly decide. Unexpended
urine, food and water samples are owned by the recipient
laboratory. We request return of two 125 ml sample aliquots of
solid food if available after analysis.
f. NIST will provide standards and blanks analyzed in accordance
with laboratory protocols as controls.
g. NIST will provide samples for analysis by laboratories across the
consortia to assess comparability of techniques.
h. Federal and Contract laboratories will send data to EPA Cincinnati
Office for verification. Simultaneous transfer will be made to
the University of Arizona NHEXAS Project Data Manager.
i. Verified laboratory data will be transferred to the NHEXAS Arizona
Project Data Manager from EMSL Cincinnati in a timely fashion.
-------
Section No. 3.3.5
Revision #0
June 1995
Page: 99 of 116
3.3.5 Day-to-day data processing operations will be implemented and documented
in each of the project areas defined above. The responsible parties
will transfer their data to the University of Arizona Data office in
accordance with EPA Order 2180.2.
A. Major changes in data processing methods either in terms of equipment
used or approach will be determined by the the Principal and
Co-Principal Investigators and documented by them.
B. Minor substantive changes in data protocol implementation will be
made by the Local Co-PI and the Data Officer and the Arizona Data
Manager with documentation sent to the other Co-PIs and the Principal
Investigator.
c. The University of Arizona Data Manager will provide SPA with NHEXAS
records at the written instruction of the PI or Arizona Co-PI in the
format that EPA requests. All data transfers and receipts will be
recorded.
D. Once the data is provided to 1PA in the specified format, subsequent
manipulations are the responsibility of EPA with the cooperation of
the Consortium.
3.3.6 Data Analysis for Exposure Assessment. This responsibility will be
fulfilled by Arizona and IIT.
Notes Figure 23 illustrates the organizational structure of NHEXAS ARIZONA
and interacting agencies.
4.0 DATA ACQUISITION AND MANAGEMENT
The specific techniques employed for data acquisition, reduction,
processing and validation will be dependent on the situations presented.
Data reporting methods also may vary. Specific approaches to data
acquisition are covered in the Field and Data SOPs. "Key" variables have
been identified and combined to form a unique verifiable sample
identification number.
In the context of this discussion, data means all collected and project
generated information. This means questionnaires, field forms, samples,
lab analysis, tracings or extracts from lab analysis, reports and documents
maintained for QA purposes, electronic files, unexpended samples are all
data. For discussion purposes these will be characterized as:
QA Documents = Paper documents (i.e. the QSIP) required for or generated
by a QA audit. This also includes UA, Battelle and IIT notebooks needed
to perform any QA audit. These records are readily accessible at all
times. They are kept in ordered, sequentially numbered notebooks and/or
files in the area where they are needed.
Lab forms = Paper forms generated in the lab and containing records of
results. These are sequentially ordered in notebooks and/or files in
the laboratory where they are generated.
-------
Section No. 4 _ g
Revision #0
June 1995
Pages 100 of 116
Figure 23. The organizational structure of NHEXAS Arizona, Shading
represents those areas subject to QA audit and oversight.
"r / 7 > f / t1 j > ) f j t t * j f f f
1 * f r t /
s \ s \ \ s \
\ \ N
S S /
/ s /
PRINCIPAL INVESTIGATOR
/ /
. S '
/ /
, s •
' '
s s s
\ N V
f s /
S \ \
S S S
¦s. S. \
S f /
S % N
/* /• /
•s V V
^ > N-vS\NSN%%
\ > %S\H,SSS\V\S
' * * ' f f * f * f f f f ? S f * / f / t f J
>>>,"*>> << < <<< <<
-------
Section No. 4.0
Revision #0
June 1995
Page: 101 of 116
Physical forms = Other papers (generated primarily by the field or data)
that contain project information. These are filed
sequentially by household identification number or topic in
file cabinets.
Samples = raw material collected from the field and evaluated in a
laboratory. These are labeled and placed sequentially in
storage boxes or other suitable labeled containers.
Lab Data Files= These are primarily downloaded electronic analysis
results files, regardless of source. These will be retained
in their original batches in a storage directory. After
validation they will be sequentially appended into a master
database structure. The master data base will reside with
others in the electronic file tree.
Archive samples = retained residue after a sample has been processed.
Extracts, tracings, filters; these are materials or records
that are a by-product of the analysis. These will be
numbered, dated and placed in appropriate containers and
suitably stored under the appropriate conditions as
specified in the applicable SOP.
To track and archive these data, we have constructed a master tracking
system which includes the location of all the data at any point in time.
The system has cross programs that match newly entered records against
their previous entry components. Location of each unprocessed sample is
listed in the system for easy retrieval at a time when processing is
possible. This tracking system will facilitate evaluation of sample
integrity and sample shipment.
The following will be included in Project Control Documents (SOPs) either
directly or indirectly by reference to other approved documents:
A. Methods used for data acquisition.
B. Data reduction scheme(s) for collected data, including all equations
used to perform required calculations.
C. Methods used to process raw data.
D.Principal criteria used to validate data integrity during
acquisition, reduction, processing, and reporting of data.
4.1 Data Acquisition
A. Data acquisition, instrumentation, analytical standards and methods,
and data collection and reduction methods will be controlled. They
will be based on recognized standards and techniques, and will be
evaluated prior to use and periodically during use, to verify
accuracy, stability, and repeatability. This evaluation will be
documented by each laboratory performing the analysis.
-------
Section No. 4 > j
Revision #0
June 1995
Page: 102 of 116
B. The activities related to field questionnaires, sample collection,
analysis and reporting shall be controlled in accordance with the
appended SOPs.
4.2 Control and Calibration of Measurements and Testing Equipment
4.2.1 Balances, instruments, and other measuring and test equipment used for
activities affecting quality will be controlled and, at specified
periods, calibrated and adjusted to maintain accuracy within necessary
limits (See Lab SOPs).
4.2.2 The specific measuring and equipment calibration procedures used, will
comply with manufacturer's recommendations. If the equipment calibration
procedures are complex, or if the equipment is leased, outside
calibration services may be employed (See lab and Field SOPs).
4.2.3 The appended SOPs specify the equipment and instrument requirements and
the calibration procedures to be employed, their frequency and their
sources.
4.2.4 Records of measuring and test equipment calibrations will be maintained
by the responsible Project Coordinator or his designee and periodically
reviewed by the Project Manager, in-house Quality Control Director and
the Director of Quality Control.
4.3 Identification of Data
4.3.1 Raw, reduced, and processed data, whether obtained via literature
reconnaissance from the public domain, university sources, or during
field and laboratory activities will be identified and controlled to
assure that only correct and acceptable items or services are used by
the project. Identification will, at a minimum, be maintained on
documents traceable to the items, data or services.
4.3.2 Controlled storage shall be provided and maintained for all samples to
ensure that only acceptable samples will be used.
4.3.3 Reports
Drafts of reports including calculations and supporting documentation
for reviews will be paginated and identified by the author/preparer as
to project number, draft or revision number, date of preparation, and
total number of pages.
4.4 Control of Erroneous Data
4.4.1 Activities, services or data failing to conform with established
requirements will be controlled to prevent their inadvertent use or
installation. These controls will provide for the identification,
documentation, evaluation and disposition of nonconformances.
-------
Section No. 4.4.1
Revision #0
June 1995
Pages 103 of 116
Nonconforming items will be segregated to prevent their inadvertent use.
(See UA—6-2.0, UA-D-2.0).
4.4.2 A nonconformance relating to written material will be defined as an
identified or suspected deficiency in an approved or verified document
(e.g., technical report, analysis, calculation, computer program) where
the quality of the end item itself or subsequent activities using the
document would be affected by the deficiency.
4.4.3 All project staff members are responsible for promptly reporting
suspected nonconformances in writing to the responsible Project
Coordinator, Project Co-Principal Investigator, and the Director of
Quality Control. Nonconformances that cannot be readily resolved by the
responsible Project Coordinator, Project Co-Principal Investigator, and
Director of Quality Control will be reported to the Principal
59
Investigator for resolution.
4.5 Data Evaluation
Once master data bases are generated, 10% of the data will be randomly
selected and compared against the questionnaires, field or laboratory forms
as a Quality Assurance check. If more than 5 % of the data is erroneous,
corrective action will be taken as per the data protocols.
Further, data will be examined for errors using uni- and bivariate
statistical procedures. Outliers will be checked for accuracy.
4.6 Procedures
A control document describing each field, data and laboratory procedure
will be written and known as a Standard Operating Procedure (SOP) [ SOP #
UA-G-1.0]. The SOP will describe the Purpose, application,
responsibilities, steps of each procedure, the records to be kept and
quality control and assurance requirements. Current SOPs are attached and
listed in the Table of Contents. SOPs fall into eight major categories,
General, Training, Field, Laboratory & Calibration, Custody, Data, Analysis
and Miscellaneous.
5.0 RECORDS USAGE AND MANAGEMENT
5.1 Data Records
Records furnishing documentary evidence of data integrity and validity will
be specified as appropriate in each SOP under the "Records" (SOP UA-G-1.0
Section 8.0) (Project Control Documents). Records will be legible,
identifiable, retrievable, and protected against damage, deterioration or
loss. Requirements and responsibilities for record identification,
storage, and retrieval will comply with recognized procedures.
Project Records fall into two classes: written and electronic.
5.2 Records Management System
-------
Section No. 5.2
Revision #0
June 1995
Page: 104 of 116
A records index {written sign-in and out notebook) will be prepared by the
responsible Project Coordinator or his designee using the information in
the applicable SOPs. The Project Coordinator or his designee will maintain
current working files for documents originating within his/her area of
responsibility during the research project. All records will be filed and
labeled in accordance with the appropriate SOP descriptions.
Quality assurance records generated by the Quality Assurance Unit and its
staff (e.g., audit/inspection reports, corrective action requests,
nonconformance reports, personnel quality assurance and indoctrination and
training records) will be maintained by the Quality Assurance Unit and/or
its designee(s).
The On-site Co-Principal Investigator (aka local Project Manager) will have
copies of all Project Control Documents on file as well as copies of all
quality assurance related documentation as enumerated in this document.
Data input received from other individuals or organizations will be
retained by the organization using such data.
Access to all data record files written and electronic will be controlled.
Removal of records from the files will be controlled by both access and
completion of a sign-out log. Upon return of the document, the responsible
custodian will co-sign and date the entry, indicating the document was
returned.
Records of external transmittal of all project records will be maintained
by the On-site Co-Principal Investigator or designee.
Generic records such as test procedures, work instructions, software
program documentation and verification records will be maintained by the
originating organization. Revisions to existing documents such as
procedures, work instructions, and computer programs will be filed with the
original, together with any and all previous revisions. Copies of
superseded versions will be retained for at least three (3) years.
5.3 Record Validation
Written records: All written records will be maintained by the project for
a minimum of three years following project completion. Detailed records
are kept in the packets and filed by study stage and the household
identification number. Collection of all data is documented, initialed by
the technicians and supervisors, dated and has chain of custody forms
attached documenting the transmittal of the material. All forms are
initially completed in black ink. Any changes are made in red or purple on
the form, initialed and dated by the project member making the change.
Electronic records: Changes in data bases are documented extensively. The
change must be noted on the original physical form and in the "Data Base
Change Log Book". Changes must be approved by the Project Data Coordinator
-------
Section No. 5.3
Revision #0
June 1995
Page: 105 of 116
and documented. Errors found in the Master Data Bases must be written up
in a memo and passed on the the Project Data Manager. Only he/she is
authorized to make changes in any Master Data Bases. The Project Data
Manager logs the changes on the "Master Data Base Log Form" and files it in
the appropriate notebook. Accurate (but suspicious looking) outlier values
may be noted in the Data Dictionary to save time during future analyses.
These Validation methods are covered in Detail in SOPs # UA-D-4.0,
UA-D-16.0, UA-D-25.0.
5•4 Records Identification. Indexing and Retention
A records index will be prepared by the responsible Project Coordinator or
his designee using the information in the applicable Project Control
Document. The Project Coordinator or his designee will maintain current
working files for documents originating within his area of responsibility
during the research project. All records will be filed and labeled in
accordance with the SOPs.
In general, data packets and laboratory results will be filed by the study
stage and household identification number (SOP # UA-F-1.0). Other
documents will be filed by type and date; they will reside in each area of
the project as defined in the SOPs.
Quality assurance records generated by the Quality Assurance Unit and its
staff (e.g., audit/inspection reports, corrective action requests,
nonconformance reports, personnel quality assurance and indoctrination and
training records) will be maintained by the Quality Assurance Unit and/or
its designee(s).
The On-site Co-Principal Investigator will have copies of all Project
Control Documents on file as well as copies of all quality assurance
related documentation as enumerated above.
All records and unexpended samples (retaining integrity) will be kept at
least three (3) years after the project ends.
5.5 Records Distribution and Storage
Extensive custody SOPs have been developed to describe how samples,
physical forms and custody issues are handled in the project (SOP #
UA-C-1.0 through UA-C-8.0). All current and archive locations will be
recorded in the Tracking System (Figure 9).
Documents and physical forms will be stored and cataloged for at least
three (3) years from the end of the project in the Respiratory Sciences
Offices and Laboratories at the University of Arizona. Project forms will
be placed in a secure, locked, dry storage environment on metal shelving.
At the completion of the project, all master data bases will be transferred
to magnetic tape and reside in the control of the Principal Investigator.60
-------
Section No. 5.5
Revision #0
June 1995
Page; 106 of 116
Data input received from other individuals or organizations will be
retained by the organization using such data.
Access to all data record files will be controlled.
Records of external transmittal of all project records will be maintained
by the Principal Investigator or On-site Co-Investigator or designee.
These procedures are outlined in the Data SOPs (Volume IV).
Generic records such as test procedures, work instructions, and software
program documentation and verification records will be maintained by the
originating organization. Revisions to existing documents such as
procedures, work instructions, and computer programs will be filed with the
original, together with any and all previous revisions. Copies of
superseded versions will be retained for at least three (3) years.
Many of the physical samples lose integrity quickly. These samples are
kept refrigerated or frozen prior to shipment and analysis in other
laboratories (as specified in the SOPs). Suitable refrigerator and storage
space is available in the field lab for short term storage. The recipient
laboratories (outside the consortium) and the EPA NHEXAS Administrative
hierarchy, in conjunction with the Principal Investigators, must agree to
long term storage procedures. (We believe a committee has been formed to
handle this aspect of storage. After the committee makes recommendations a
formal SOP describing Sample Archiving Procedures [UA-G-4.0] will be
developed.) Unexpended sample residues will be retained with field and
custody forms in the Respiratory Sciences Offices and Laboratories at the
University of Arizona for at least 3 years following completion of the
project.
6.0 Routine Controls and Procedures
6.1 Maintenance of Equipment
Each piece of equipment used in the study will undergo four classes of
maintenance. The maintenance level will depend on the type of equipment
employed.
A. Routine maintenance will be performed according to the operating
instructions of the equipment document.
B. Preventive maintenance will be performed annually if equipment
maintenance is not described in the equipment documentation.
C. Lab checks and equipment calibration will be performed as each piece of
equipment is returned from the field and preceding reuse in the
field.
D. Field checks and equipment calibration will be performed by the field
technician when the equipment is installed in the field.
-------
Section No. g # i
Revision #0
June 1995
Pages 107 of 116
E. Specifics:
1. Balances, instruments, and other measuring and test equipment used
for activities affecting quality will be controlled and, at specified
periods, calibrated and adjusted to maintain accuracy within
necessary limits.
2. The specific measuring and equipment calibration procedures used will
comply with manufacturer's recommendations. If the equipment
calibration procedures are complex, or if the equipment is leased,
outside calibration services may be employed.
3. The SOPs specify the equipment and instrument requirements and the
calibration procedures to be employed, their frequency and their
sources.
4. Records of measuring and test equipment calibrations will be
maintained by the responsible Project Coordinator or his designee and
periodically reviewed by the On—Site Co—Principal Investigator, and
the Director of Quality Control.
F. Identification and isolation of non-functioning field and lab equipment
is described in SOP # UA-G-2.0, UA-L-1.0 through UA-L-4.0.
Identification and isolation of non-functioning data equipment is
described in SOP # UA-D-1.0, UA-D-2.0.
6-2 Quality of Consumables
Quality control of consumable products will occur at two levels.
A. A visual inspection of all consumables will occur when received from the
supplier and when used.
B. A contaminant analysis of specific scientific consumables will be
performed for 1 sample per batch or a sample aliquot per lot number for
chemical reagents and standards.
Specific consumable materials to be tested include:
Filters: PM10, FUF, Personal Air PM10, Personal Air PUF,
Air Sentinel, Vacuum
VOC Collectors: Multisorbent Carbo Traps, OVM badges, F-l tubes
Other: Gauze for sill and dermal wipes, Ziploc Freezer Bags
Water Containers: Cubitainers, Vials, Bottles
Lab Analysis Containers: Test tubes, vials
Reagents: chemicals needed for field and laboratory protocols
-------
Section No. g , 3
Revision #0
June 1995
Pages 108 of 116
6.3 Labeling
Laboratory labels will be attached to all primary reagents. They will
identify the material by composition, stability, storage requirements,
safety handling requirements, safety hazards, and date of receipt.
All secondary reagents (and those of subsequent generations) will have
labels that with materials identification, concentration, date of
preparation, identification of preparer and when appropriate, safety
information and expiration date.
Sample Labels will be indelible and preprinted. Labels will be attached
to all questionnaires and samples collected.
Unlabeled materials will not be used; Unlabeled chemicals will be disposed
of through the hazardous waste unit of the cooperating agency.
6.4 Acceptance of Equipment and Materials
All equipment and materials purchased will be evaluated to see if they meet
critical specifications as outlined in the project specifications.
Changes in the specifications will not be permissible without review and
consent by the Principal, Co-Principal and Cooperating Investigators. If
changes in materials specifications are agreed to, they will be specified
in writing and the specifications signed by the Principal and Co-principal
investigators.
6.5 Storage of Eguipment and Materials
The storage of equipment and materials, will comply with manufacturer's
recommendations.
SOPs indicate storage requirements for some media. These SOP
specifications must be followed as some samples loose their integrity.
7.0 TECHNICAL ASSESSMENT AND RESPONSE
7.1 Assessment Procedures
7.1.1 Pre-Field Assessment
A. Equipment: Equipment will be calibrated in the laboratory prior to
transport into the field. Equipment must be able to maintain
operating standards during the course of operation. Equipment must
be tested at the time of set-up and removal to assure that it
performed to standards as defined in the specific SOPs. All critical
values associated with equipment operation must be 100% independently
validated by another field team member.
-------
Section No. 7.1.1
Revision #0
June 1995
Pages 109 of 116
B. SOPs and Formss All SOPs must be field tested prior to project
implementation. Field forms must be pretested. SOPs and field forms
found inadequate will be revised and finalized prior to entry into
the field.
7.1.2 Field
A. Duplicates will be collected for 10% of all samples with the
exceptions of food, beverage, blood and urine. For the exceptions,
the samples will be split in the laboratory (when possible) to
provide QA split samples. All samples will be analyzed using the
same methods.
B. Field blanks will be used in conjunction with 10% of all samples
collected. They will be prepared and analyzed using the same
methods as study samples.
C. Questionnaire results obtained during field visits will undergo 100%
QC check by technicians in the field, 100% QA check for completeness
by the Project Field Coordinator (or designee) acting as an in house
QA auditor for field forms.
D. 10% of the subjects contacted during Field Stage 1 will be
re-contacted by phone and the Descriptive Questionnaire
readministered as a QA check.61
E. Coordinator(s)/Supervisor(s) will routinely evaluate technicians as
outlined in the Training Procedures (SOP # UA-T-1.0 through
UA-T-6.0).
F. The QA Unit will perform periodic field audits as per their protocol
presented in Appendix C?^
7.1.3 Laboratory
A. Coordinator(s)/Supervisor(s) routinely evaluate technicians as
outlined in the Training Procedures (SOP # UA-T-1.0 through
UA-T-6.0).
B. NIST Standards will be used in conjunction with all laboratory
evaluations. Measurement deviations beyond acceptable limits
described in the SOPs will be documented, investigated and corrected.
Any deviations in field or laboratory procedures will be corrected
, 12
and documented.
C. Repeated/replicate, split and blank samples (field and lab) will be
treated as all study samples and evaluated in accordance with all
pertinent SOPs. Values may vary from samples by some designated
amount (see appropriate lab SOPs for value). Excessive deviation
between samples will result in re-evaluation of the field and lab
procedures. All implemented corrective actions will be dated and
documented.
-------
Section No. 7.2
Revision #0
June 1995
Page: 110 of 116
7.2 Assessment Evaluation
7.2.1 The QA unit recognizes the need for project data to meet specific
criteria for scientific validity and technical defensibility, and to
be of defined precision and accuracy. The QA systems to be
implemented provide a planned and systematic management approach of
procedures and controls to ensure that personnel, equipment,
activities, and documentation comply with U.S. SPA requirements.
7.2.2 To assist in meeting these objectives, the UA maintains a Quality
Assurance Program to ensure that activities affecting the quality and
integrity of data are appropriately planned and coordinated. The
University of Arizona QA Unit will monitor all three members of the
consortium. The QA system audits will be used to assure that QA/QC
plans are prepared, approved and fully implemented, that QA/QC
procedures are fully understood by field and lab personnel, and that
data are reported in a manner reflecting the data quality objectives
of the program. To assist task leaders in tracking critical QA
issues, a Quality Assurance Checklist will be developed for each
aspect of sampling and analysis. These will be distributed to field
and lab personnel with instructions.
7.2.3 QA performance audits (Appendix C) will be used to assure compliance
with the project plan. These audits will have the form of
verification surveillances and will be performed by the assigned QA
Officer. These surveillances will be performed to ensure that a
specified requirement is being met. These audits will include both
real time observations during the work or analytical process to
ensure that specific applicable procedures are being implemented, and
traceability checks through data to ensure that project data can be
tracked back through the analytical process, through sample handling
and transportation, back to the date, location, staff, and technique
used to collect the sample. At least one QA surveillance audit will
be performed for each of the following key activities: sampling,
sample tracking from field to lab, analytical measurement, and data
reporting.
7.3 Assessment Response and Follow-up
A. Field: Conditions adverse to quality, such as variances, unusual
occurrences or abnormal conditions, and deviations from the SOPs and
contractual requirements will be identified promptly and corrected as
soon as practicable. In the case of a significant condition adverse to
quality, the cause of the condition will be determined and corrective
action taken to preclude recurrence. The identification, cause, and
corrective action for significant conditions adverse to quality will be
documented and reported to the appropriate levels of management.
Follow-up action will be taken by the Director of Quality Control to
verify implementation of corrective action.
-------
Section No. 7.3
Revision #0
June 1995
Page: 111 of 116
B. Laboratory: The need for corrective action will be identified by the
technical staff during the course of their work or review of data. We
expect that problems will be identified and corrected prior to QA
audits, although this additional measure of oversight will provide
further opportunity for identification of problems prior to data
reporting. Each individual staff performing laboratory or data
processing activities will be sufficiently well-trained in their
operations that they will be responsible for notifying the appropriate
supervisory personnel of any circumstance that would affect the quality
or integrity of the data. Deviations from approved procedures that
require corrective actions typically result from unforeseen
circumstances. All deviations will be documented in field or lab
notebooks, and will be dealt with as expeditiously as possible by the
responsible task leader.
C. Data: The need for corrective action will be identified by the technical
staff during the course of their work or review of data. We expect that
problems will be identified and corrected prior to QA audits, although
this additional measure of oversight will provide further opportunity
for identification of problems prior to data reporting. Each individual
performing data processing activities will be sufficiently well-trained
in their operations that they will be responsible for notifying the
appropriate supervisory personnel of any circumstance that would affect
the quality or integrity of the data. Deviations from approved
procedures that require corrective actions typically result from
unforeseen circumstances. All corrective changes in the data bases must
be approved by the Project Data Coordinator (Working Data Bases) or
Project Data Manager (Master Data Base). All corrections will be noted
on the appropriate "Data Change" form as outlined in the Data SOPs.
-------
Section No. DLIST
Revision #0
June 1995
Page: 112 of 116
DISTRIBUTION LIST AND STATUS
PRINCIPAL AND CO-INVESTIGATORS
M.D. Lebowitz (P.I.)
D. Moschandreas (Co-Principal Investigator)
S.M. Gordon (Co-Principal Investigator)
M.K. O'Rourke (Co-Principal Investigator)
University of Arizona
S. Hopf (UA-QAU)
M. Karpiscak (UA-QAU)
S. Rogan (Field)
L. Piorentino (Data)
P. Van de Water (Lab)
U.S. EPA
T. Buckley (SPA Collaborator)
D. Kieffman (Project Officer)
-------
Section No. REF
Revision #0
June 1995
Page: 113 of 116
References
Anderson, L.G., P. Wolfe, R. Barrett, and J.A. Lanning, "The Effects of Using Oxygenated Fuels on the
Atmospheric Concentrations of Carbon Monoxide and Aldehydes in Denver," Preprint of paper presented at
the 205th ACS National Meeting, Denver, CO, March 28- April 2, 1993, Division of Environmental
Chemistry, 33 (1), 243-246 (1993).
Bernick, M.B., J. Corcoran, P.R. Campagna, P. Stanislaw, P.F. Berry, and S.R. Little,
"Field-Screening Filters Used in Monitoring Air Duality for Metals with a
Field-Portable X-Ray Fluorescence Spectrometer" in Measurement of Toxic and
Related Air Pollutants, Air & Uaste Management Association, Pittsburgh,
Pennsylvania, 1992.
Buckley, T.J., A.B. Lindstrom, V.R. Highsmith, W.E. Bechtold, and L.S. Sheldon, "The Time-Course and
Sensitivity of Muconic Acid as a Biomarker for Hunan Environmental Exposure to Benzene," in
Proceedings of the 1992 U.S. EPA/A&WMA International Symposium on the Measurement of Toxic and Related
Air Pollutants, VIP-25, Air & Uaste Management Association, Pittsburgh, PA, pp. 981-986, 1992.
"Carbaryl" in Hazardous Substances Databank, National Library of Medicine,
February 13, 1987.
"Carbaryl" in Pesticide Fact Handbook, U.S. Environmental Protection Agency,
Noyes Data Corporation, Park Ridge, NJ, pp. 130-139, 1988.
Chromy, JR. Sequential Sample Selection Methods, American Statistical Association
Proceedings of the Section on Survey Research Methods, 401-406 (1979).
Clickner, R, et al. Statistical Design Issues in Human Exposure Assessment Surveys:
Sampling Design and Options. Contract #6S-W1-019, USEPA, 1993.
Clobes, A.L., G.P. Ananth, A.L. Hood, J.A. Schroeder, and K.A. Lee, "Human Activities as Sources of
Volatile Organic Compounds in Residential Environments," Annals NY Acad. Sci.641, 79-86 (1992).
Cooke, T.F., "Indoor Air Pollutants: A Literature Review," Rev. Environ. Health, 9, 137-160 (1991).
Cote, I.L. and S.P. Bayard, "Cancer Risk Assessment of 1,2-Butadiene," Environ. Health Perspect. 86,
149-153 (1990).
Dann, T. and D. Wang, "Volatile Organic Compound Measurements in Canadian Urban and Rural Areas:
1989-1990," Paper 92-75.16 presented at 85th Annual Meeting & Exhibition of Air & Waste Management
Association, Kansas City, MO, June 21-26, 1992.
Davis, J. R., R. C. Brownson, R. Garcia, B. J. Bentz, and A. Turner, "Family Pesticide Use and Childhood
Brain Cancer", Arch. Environ. Contam. Toxicol., 24, 87-92, 1993.
Ducos, P., R. Gaudin, J. Bel, C. Mai re, J.M. Francin, A. Robert, and P. Wild, "trans,trans-Muconic Acid, a
Reliable Biological Indicator for the Detection of Individual Benzene Exposure Down to the ppm Level,"
Int. Arch. Occup. Environ. Health 64, 309-313 (1992).
EXTOXNET: Extension Toxicology Network, A Pesticide Information Project of Coorperative
Extension Offices of Cornell University, The University of California, Michigan
State University, and Oregon State University, January 1989.
-------
Section No. kef
Revision #0
June 1995
Page: 114 of 116
Fleiss J. Statistical Methods for Rates and Proportions. (2nd Ed.) New York: Wiley, 1981.
Goldberg SJ, Lebowitz HO, Graver EJ, and Hicks S. "An association of human congenital cardiac malformations
and drinking water contaminants." Jj. Amer. Coll. Cardiol. 16(15:155-164, 1990
Fortmann, R. C., L. S. Sheldon, D. Smith, K. Perritt, and D. E. Camann, "House Dust/Infant Pesticides
Exposure Study (HIPES)", final report to U.S. Environmental Protection Agency, Contract No. 68-02-4544,
prepared by Research Triangle Institute and Southwest Research Institute, October 1, 1991.
Fujii, Y., and S. Asaka, "Metabolism of Diazinon and Diazoxon in Fish Liver Preparations", Bull. Environm.
Contam. Toxicol., 29, 455-460, 1982.
Glotfelty, D. E., M. S. Hajewski, and J. N. Seiber, "Distribution of Several Organophosphorus Insecticides
and Their Oxygen Analogues in a Foggy Atmosphere", Environ. Sci. TechnoI., 24, 353-357, 1990.
Gordon, S. M., D. V. Kenny, and T. J. Kelly, "Continuous Real-Time Breath Analysis for the Measurement of
Half-Lives of Expired Volatile Organic Compounds." J. Expos. Anal, and Environ. Epidemiol. Suppl. 1,
41-54 (1992).
Gordon, S. M., L. A. Wallace, E. D. Pellizzari, and H. J. O'Neill, "Human Breath Measurements in a
Clean-Air Chanter to Determine Half-Lives for Volatile Organic Compounds." Atmos. Environ., 22(10),
2165-2170 (1988).
Gordon, S. M., "Identification of Exposure Markers in Smokers' Breath." J. Chromatogr. 511,291-302 (1990).
Grover, R., A. J. Cessna, N. I. Muir, D. Riedel, C. A. Franklin, and K. Yoshida, "Factors Affecting the
Exposure of Ground-Rig Applicators to 2,4-D Oimethylamine Salt", Arch. Environ. Contam. Toxicol., 15,
677-686, 1986.
Hallenbeck, W.H., "Cancer Risk Assessment for the Inhalation of 1,3-Butadiene Using Physiologically Based
Pharmacokinetic Modeling," Bull. Environ. Contam. Toxicol. 49, 66-70 (1992).
Hattemer-Frey, H.A., C.C. Travis, and M.L. Land, "Benzene; Environmental Partitioning and Human Exposure,"
Environ. Res. 53, 221-232 (1990).
Hodgson, A.T. and J.D. Wooley, "Assessment of Indoor Concentrations, Indoor Sources and Source Emissions of
Selected Volatile Organic Compounds," Final Report to the California Air Resources Board, Copntract Ho.
A933-063, prepared by Lawrence Berkeley Laboratory, University of California, Berkeley, CA, March 1991.
Iverson, F., D. L. Grant, and J. Lacroix, "Diazinon Metabolism in the Dog", Bulletin of Environmental
Contamination & Toxicology, Vol. 13, No. 5, 1975.
Kirchherr, H., H.D. Koester, and H.U. Schiwara, "Determination of Formaldehyde in Indoor Dust, in Activated
Charcoal After Exposure, and in Indoor Air," Klin. Labor 38, 412-417 1992).
Lewis, R.G., J.D. Hulik, R.U. Coutant, G.W. Wooten, and C.R. McMillin, "Thermally Desorbable Passive
Sampling Device for Volatile Organic Chemicals in Ambient Air," Anal. Chem. 57, 214-219 (1985).
Lewis, R. G., A. E. Bond, D. E. Johnson, and J. P. Shu, "Measurement of Atmospheric
Concentrations of Common Household Pesticides: A Pilot Study", Environmental Monitoring and Assessment, 10,
59-73, 1988.
-------
Section No. REF
Revision #0
June 1995
Page: 115 of 116
Lewis, R. G., A. E. Bond, R. C. Fortmann, L. S. Sheldon, and D. E. Camann, "Determination of Routes of
Exposure of Infants and Toddlers to Household Pesticides: A Pilot Study to Test Methods", AWMA 84th
Annual Meeting, Vancouver, British Columbia, June 16-21, 1991.
Lin, D.C.K., R. G. Melton, F. C. Kopfler, and S. V. Lucas, "Glass Capillary Gas Chromatographic/Mass
Spectrometrlc Analysis of Organic Concentrates from Drinking and Advanced Waste Treatment Waters",
Advances in the Identification and Analysis of Organic Pollutants in Water, Vol. 2, Chapter 46, 861-906,
1981.
Lindstrom, A.B., V.R. Highsraith, T.J. Buckley, W.J. Pate, L.C. Michael, and R.M. Johnson, "Household
Exposures to Benzene from Showering with Gasoline-Contaminated Ground Water,"in Proceedings of the 1992
U.S. EPA/A&WMA International Symposium on the Measurement of Toxic and Related Air Pollutants, VIP-25,
Air & Waste Management Association, Pittsburgh, PA, pp. 39-44, 1992.
Lofgren, L. and G. Petersson, "Butenes and Butadiene in Urban Air," Sci. Total Environ. 116, 195-201
(1992).
Lofroth, G., R.M. Burton, L. Forehand, S.K. Hamnond, R.L. Seila, R.B. Zweidinger, and J.
Lewtas, "Characterization of Environmental Tobacco Smoke," Environ. Sci. Technol. 23,610-614 (1989).
Moschandreas, D. J. and S. M. Gordon. "Volatile Organic Compounds in the Indoor Environment:Review of
Characterization Methods and Indoor Air Quality Studies," in Organic Chemistry of the Atmosphere, L. D.
Hansen and D. J. Eatough (Eds.), CRC Press, Boca Raton, FL, 1991, pp. 121-153.
Murphy, R. S., F. W. Kutz, and S. C. Strassman, "Selected Pesticide Residues or Metabolites in Blood and
Urine Specimens from a General Population Survey", Environmental Health Perspectives, Vol 48, pp 81-86,
1983.
Needhsm, L. L., R. E. Cline, S. L. Head, and J. A. Liddle, "Determining Pentachlorophenol in Body Fluids by
Gas Chromatography After Acetylation", J. Analyt Toxicol., 5:283-286, 1981.
Nishioka, M., M. Brinkman, and H. Burkholder, "Evaluation and Selection of Analytical Methods for
Lawn-Applied Pesticides", final report to U.S. Environmental Protection Agency, Contract No. 68-D0-Q007,
prepared by Battelle Memorial Institute, September 1992.
PeLlizzari, E. D., L. A. Wallace, and S. M. Gordon, "Elimination Kinetics of Volatile Organics using Breath
Measurements." J. Expos. Anal, and Environ. Epidemiol. 2, 341- 355 (1992).
Sabourin, P.J., L.T. Burka, A.R. Dahl, W.E. Bechtold, and R.F. Henderson, "Species Differences in
Metabolism of Inhaled Butadiene," Toxicologist 11, 50 (1991).
Sabourin, P.J., L.T. Burka, W.E. Bechtold, A.R. Dahl, M.D. Hoover, I.Y. Chang, and R.F. Henderson, "Species
Differences in Urinary Butadiene Metabolites; Identification of 1,2-
dihydroxy-4-(N-acetylcysteinyI)butane, a Novel Metabolite of Butadiene," Carcinogenesis 13, 1633-1638
(1992).
Salem, H., and E. J. Olajos, "Review of Pesticides: Chemistry, Uses and Toxicology", Toxicology and
Industrial Health, Vol. 4, No. 3, pp 291-321, 1988.
Schattenberg, III, H. J., and J.-P. Hsu, "Pesticide Residue Survey of Produce from 1989 to 1991", Jour of
AOAC International, Vol. 75, Mo. 5, 1992.
-------
Section No. REF
Revision #0
June 1995
Page: 116 of 116
Sexton K, S.G. Selevan, D.K. Wagener, J.A. Lybarger, "Estimating Human Exposures to Environmental
Pollutants: Availability and Utility of Existing Databases", Archives of Environ. Health 47(6), 398-407
(1992).
Shehata, A.T., "A Multi-Route Exposure Assessment of Chemically Contaminated Drinking Water," Toxicol.
Ind. Health 1, 277-298 (1985).
Sheldon, L.S. and P. Jenkins, "Indoor Pollutant Concentrations and Exposures for Air Toxics -- A Pilot
Study," Proceedings of the 5th International Conference on Indoor Air Quality and Climate, Toronto,
Canada, July 29-August 3, 1990, Vol. 2, pp.759-764.
Sittig, M., "Handbook of Toxic and Hazardous Chemicals and Carcinogens," 2nd Ed., Noyes Publications, Park
Ridge, NJ, 1985.
U.S. EPA, Compendium Chaper IP-9, "Determination of reactive acidic and basic gases and particulate matter
in indoor air," Atmospheric Research and Exposure Assessment Laboratory, U.S. EPA, Research Triangle
Park, NC, September 1989.
Wallace, L.A., "The TEAM Study; Summary and Analysis: Volume I," U.S. Environmental Protection Agency,
EPA-6O0/6-87/QO2a, Washington, D.C., 1987.
Wallace, L.A., "Comparison of Risks from Outdoor arid Indoor Exposure to Toxic Chemicals," Environ. Health
Perspect. 95, 7-13 (1991).
Wallace, L.A., "Recent Field Studies of Personal and Indoor Exposures to Environmental Pollutants," Ann.
N.Y. Acad. Sci. 641, 7-16 (1992).
Wallace, L. A., "Comparison of Risks from Outdoor and Indoor Exposure to Toxic Chemicals", Environmental
Health Perspectives, Vol. 95, pp 7-13, 1991.
Weisskopf, C.P., J. N. Seiber, N. Maizlish, and M. Schenker, "Personnel Exposure to Diazinon in a Supervised
Pest Eradication Program", Arch. Environ. Contain. Toxicol., 17, 201-212, 1988.
WHO. Environmental Health Criteria. Geneva.
WHO/EURO. Ai r Qua I i tv Guidelines for Europe. R. Turck and D. (Cello (Eds.) (WHO Regional Publ., Euro,
Series No. 23), 1987.
WHO/EURO. Indoor Air Quality: Organics. Copenhagen: WHO/EURO, 1989.
WHO (World Health Organization), "Environmental Health Criteria 89, Formaldehyde", Geneva, 1989. ATS
(American Thoracic Society). Epidemiology Standardization Project. Am Rev Respir Dis 118 (Suppl.), 1978.
WHO (World Health Organization) Estimating Human Exposure to Air Pollutants. Geneva: WHO, 1982.
WHO Guidelines on Studies in Environmental Epidemiology (EHC 27) Geneva: WHO, 1983.
WHO/EURO Indoor Air Pollutants: Exposure and Health Effects (Report 78, 1983); Indoor Air Research (Report
103, 1986).. Copenhagen: WHO/EURO, 1983.
-------
APPENDIX A: STANDARD OPERATING PROCEDURE LIST
-------
Appendix A; Standard Operating Procedure Volumes and Contents
VOLUME I. GENERAL PROCEDURES
Preparation of Standard Operating Procedures (SOPs) UA-G-1.0
Isolation of Malfunctioning or Damaged Equipment UA-G-2.0
Assurance of Respondent Confidentiality UA-G-3.0
Archive Procedure for Study Samples UA-G-4.0
Receipt of Equipment and Implementation Supplies UA-G-5.0
Procedure for the Transmittal of Sampling Materials from
Battelle BCO-G-l.O
Procedure for the Receipt of Samples or Reference Compounds
for Laboratory Analysis at Battelle BCO-G-2.0
Verification and Transfer of Data to UA BCO-G-3.0
TRAINING PROCEDURES
Training Guide and Exposition of Interviewer
Responsibilities and Behavior UA-T-1.0
Administering Field Qx— General UA-T-2.0
Field Personnel Training— General UA-T-3.0
Field Personnel Training Plan UA-T-4.0
Student Data Assistant Training Plan UA-T-5.0
Laboratory Assistant Training Plan—General (UA) UA-T-6.0
General Laboratory Training Plan— Battelle BCO-T-l.O
DATA CUSTODY PROCEDURES
Custody of Field Samples UA-C-1.0
Form QA/QC Checks UA-C-2.0
Packet Assembly, Custody and Transfer to Data Section UA-C-3.0
Batching of Field Data Forms UA-C-4.0
Flow and Custody of Field Data Forms UA-C-5.0
Keypunch Tracking, Custody & Data Transfer UA-C-6.0
Batching of Lab Data UA-C-7.0
Flow & Custody of UA Laboratory Data UA-C-8.0
VOLUME II. FIELD PROCEDURES
HHID & IRN Assignment UA-F-1.0
Preparation of Field Bucket & Equipment Assignment UA-F-2.0
Field Use of the Particulate Sampler UA-F-3.0
Operation, Calibration & Maintenance of the Sentex
Scentogun Portable Photoionization Detector UA-F-4.0
Field Collection of Yard Composite Soil Samples UA-F-5.0
Field Collection of Residential Foundation Soil Samples UA-F-6.0
Field Collection of Indoor Floor Samples UA-F-7.0
Collection of Surface Wipe Samples for Pesticides or Metals UA-F-8.0
Collection of Dermal Wipe Samples for Pesticides or Metals UA-F-9.0
Field Use of the Modified Air Sentinel UA-F-10.0
1
-------
Use of an Active VOC Sampler for the Collection of Airborne
VOCs at Fixed Indoor & Outdoor Sites UA-F-11.0
Use of a Passive Sampling Devise for the Collection of Airborne
VOCs at Fixed Indoor & Outdoor Sites UA-F-12.0
Collection of Fixed Site Indoor & Outdoor Formaldehyde
Passive Samples UA-F-13.0
Collection of Personal Air Samples for Analysis of Pesticides
or Metals UA-F-14.0
Field Collection of Food £ Beverage Samples UA-F-15.0
Collection, Storage and shipment of Drinking and Tap
Water Samples for Trace Metals (EPA Method 200.8) UA-F-16.0
Collection, Storage & Shipment of Drinking & Tap Water Samples
for Pesticides (EPA 225.2) & Carbaryl (EPA 531.1) UA-F-17.0
Collection, Storage and Shipment of Drinking and Tap
water Samples for VOCs (EPA Method 524.2) UA-F-18.0
Collection, Storage and Shipment of Blood Samples for
Selected Metals and VOCs UA-F-19.0
Collection, Storage and Shipment of Urine Samples for
Selected Metals and Pesticides UA-F-20.0
Thin Film Collection of Dust and Soil for Metals Analysis UA-F-21.0
Operation & Initialization of the Magellan GPS
Satellite Navigator UA-F-22.0
Reimbursement for Field Collection of Food Samples UA-F-23.0
VOLUME III. LABORATORY PROCEDURES
Volume IIIA. Laboratory Procedures (Arizona)
Volume IIIB. Laboratory Procedures (Battelle)
Calibration and Operation of NHEXAS Balances
Weight Room Operation & Maintenance
Still Operation and Maintenance
Maintainence S Temperature Verification
of Refridgerated Units for Sample Storage
Standard Protocol for Cleaning Laboratory
and Field Sampling Apparatus
Calibration of Harvard PM Samplers
Harvard PM Impactor Calibration and Leak Testing
Preparation of PM & URG Impactors & Impaction Plates
Filter Weighing
Operation, Calibration & Routine Use of the spectrace
9000 Field Portable X-Ray Fluorescence Analyzer
Soil Characterization
Vacuum Dust Sampling
Food Sample Comparison with Diet & Shipment
UA-L-1.0
UA-L-2.0
UA-L-3.0
UA-L-4.0
UA-L-5.0
UA-L-6.0
UA-L-7.0
UA-L-8.0
UA-L-9.0
UA-L-10.0
UA-L-11.0
UA-L-12.0
UA-L-13.0
Analysis of Soil or House Dust Samples using Chlorpyrifos
ELISA Samples
Preparation of filters & PUF for Field Collection of Metals
and Pesticides in Air
Extraction of Metals from Soil, Dust, Air Filter, and
Surface & Dermal Samples for AA (Graphite Furnace or Flame)
BCO-L-1.0
BCO-L-2.0
2
-------
or ICP-AES Analysis BCO-L-3.0
*No Longer in Use by the NHEXAS Project BCO-L-4.0
Operation, Calibration and Maintenance of the Perkin-Elmer
1100B Atomic Absorption Spectrometer BCO-L-5.0
Operation, Calibration and Maintenance of the Perkin-Elmer
Zeeman/5000 System Atomic Absorption spectrometer BCO-L-6.0
Operation, Calibration and Maintenance of the Jobin-Yvon Model
70 Inductively Coupled Plasma Atomic Absorption
Spectrometer BCO-L-7.0
Operation, Calibration and Maintenance of the Thermo Jarrell
ICAP 61-975 Plasma AtomComp Emission Spectrometer BCO-L-8.0
Operation, Calibration and Maintenance of Fixed and Adjustable
Volume Pipette Guns BCO-L-9.0
Procedures for Cleaning Glassware to be used for inorganic
Metals Analysis BCO-L-IO.Q
Extraction of Air samples for GC/MS Analysis of Pesticides BCO-L-ll.O
Extraction of Pesticides from Dermal Wipe Samples BCO-L-12.0
Extraction of Pesticides from Surface Wipe Samples BCO-L-13.0
Extraction of Soil/House Dust for GC/MS Analysis of
Pest ic ide s BCO-L-14.0
Analysis of Pesticide Samples by GC/MS BCO-L-15.0
Analysis of Passive Formaldehyde Samplers BCO-L-16.0
Analysis of Volatile Organic Compounds Collected with
a Passive Sampler BCO-L-17.0
Preparation of Multisorbent Tubes for Actively-Pumped
VOC Samplers BCO-L-18.0
*No Longer in Use by the NHEXAS Project BCO-L-19.0
*No Longer in Use by the NHEXAS Project BCO-L-20.0
Preparation of Calibration & Surrogate Recovery Solutions
for GC/MS Analysis of Pesticides BCO-L-21.0
Analysis of Multisorbent Samplers for VOCs BCO-L-22.0
Calibration, Maintainence and Operation of
Electronic Balances BCO-L-23.0
Analysis of Pesticide Samples by GC/ECD BCO-L-24.0
VOLUME IV. DATA PROCEDURES
Operation and Maintenance of the LAN & Related Microcomputer
Environment for University of Arizona NHEXAS UA-D-1.0
Performance of Computer Software: Verification and Validation UA-D-2.0
Defining Working Databases and Data Entry Forms (Hand Entry) UA-D-3.0
The Generation and Operation of Data Dictionaries UA-D-4.0
Global Coding used by NHEXAS Arizona (Hand Entry) UA-D-5.0
Coding: Descriptive Questionnaire (Scanned) UA-D-6.0
Coding: Baseline Questionnaire (Scanned) UA-D-7.0
*No Longer Applicable to NHEXAS AZ UA-D-8.0
Coding: Time Diary & Activity Questionnaire (Scanned) UA-D-9.0
Coding: Food Diary Follow Up (Scanned) UA-D-10.0
Coding: Follow Up Questionnaire (Scanned) UA-D-11.0
*No Longer Applicable to NHEXAS AZ UA-D-12.0
Coding: Arizona Lab Data (Scanned) UA-D-13.0
Coding & Coding Verification (Hand Entry) UA-D-14.0
3
-------
Data Entry and Data Verification (Hand Entry) UA-D-15.0
First Stage of Cleaning Electronic Data (Hand Entry) UA-D-16.0
Cleaning: Descriptive Questionnaire (Scanned) UA-D-17.0
Cleaning: Baseline Questionnaire (Scanned) UA-D-18.0
*No Longer Applicable to NHEXAS AZ UA-D-19.0
Cleaning: Time Diary & Activity Questionnaire (Scanned) UA-D-20.0
Cleaning: Food Diary Follow Up (Scanned) UA-D-21.0
Cleaning: Follow Up Questionnaire (Scanned) UA-D-22.0
*No Longer Applicable to NHEXAS AZ UA-D-23.0
Cleaning: Arizona Lab Data (Scanned) UA-D-24.0
Correcting Electronic Data (Hand Entry & Scanned) UA-D-25.0
Electronic Data QA Check (Hand Entry & Scanned) UA-D-26.0
Instructions for the Addition of Individual Cleaned Data
Batches to Master Databases for the University of
Arizona NHEXAS UA-D-27.0
Electronic Tracking System: Samples, Custody Transfer,
Data Archive, Appendage UA-D-28.0
Transfer and Receipt of NHEXAS Data UA-D-29.0
Scanable Form and Data Base Definition UA-D-30.0
Global Coding for Scanned Forms UA-D-31.0
Operation and Maintenance of Data Scanners UA-D-32.0
Light Pen Operation 6 Verification of Scanned Bar Codes UA-D-33.0
Scanning & Verifying Forms and Questionnaires UA-D-34.0
Coding: Technician Walk-Through Questionnaire (Scanned) UA-D-35.0
Cleaning: Technician Walk-Through Questionnaire (Scanned) UA-D-36.0
Coding: Field Forms (Scanned) .. . UA-D-37.0
Cleaning: Field Forms (Scanned) '^vU^Q UA-D-38.0
Coding: 24 Hour Food Diary Sc
-------
APPENDIX B: EVALUATION OF THE NHEXAS QUESTIONNAIRE (9/94)
BY ARIZONA
-------
METHODS TESTING FOR NHEXAS
DRAFT VERSION OF THE NHEXAS QUESTIONNAIRE (9/94): Many of the questions
found in the Draft version of the questionnaire were previously
validated in other studies and by the OS Census. To date we have
evaluated two complete households and will perform the final dress
rehearsal with the revised protocols and questionnaire. Questionnaires
administered during training exercises, staff trials and in other
populations are as follows:
Quest ionnaire
Number
Administered/completed
Descriptive
Baseline
Follow-up
Food Diary &
Time Activity
Technician Qx
Food Follow-up
9
42
35
5
27
2
Approx.
Completion Time
22 min
60 min
35 min
30 min/day
8 min/day
20 min
English QX Evaluation
We have taken a two pronged approach with our questionnaire trials. (1)
We have performed complete field evaluations including questionnaire
administration and field testing (excluding biological sampling) in two
households [a household where Finnish is the native language and English
is spoken as a second language; and the household of a Pacific Islander
where English is the only language spoken]. (2) We have administered
questionnaire segments to participants in our NIH funded study to help
develop our training documents, evaluate subject understanding of each
question and identify subtleties in word-meaning associated by regional
differences. 24 Anglos ranging in age from 5 to 67 years old; 5
Hispanics ranging in age from 8 to 38 years and 2 Asian/Pacific
Islanders (ages 8 & 42} responded to the questionnaire.
Spanish QX Version
A border Spanish rough draft of the NHEXAS questionnaire has been
prepared. The translation was prepared by a project employee born in
Nogales, Sonora who speaks Spanish as his first language. This draft
translation reflects the current questions in the questionnaire. The
draft was administered to a bilingual 86 year old, female, legal
resident alien (green card holder) born in Empalmae, Son. Mexico and
currently residing in Tucson, Arizona. The final version of the NHEXAS
questionnaire will be edited and corrected by our 3 Hispanic staff
members who are all native speakers.
QUESTIONNAIRE SUMMARY STATEMENT (Version 9/94):
For the most part these questionnaires went well. They are less
burdensome than many of the questionnaires we have used.
-------
2. All of the Hispanlcs we encountered were bilingual and had no particular
problems with the questionnaire as a whole.
3. The biggest problem was self-identification of ethnic background. They
do not know whether they are white, Spanish, Mexican American, or
Chicano. Most are "Mexican" a Spanish & Mexican Indian mix.
4. Do Mexican Indians count as Native Americans? Tribes from Mexico do not
register their members.
5. Most other problems were minor and could be handled by the training
document.
Problems encountered are described below. If a question is labeled "OK"
it means none of the responding subjects had a problem with the
question.
Descriptive Questionnaires (number administered =9)
(Time of administration range = 15-30 min; mean = 22 min)
1. OK
2. OK
3. OK
4. OK
5. OK
COMMENT A: Responses to these QX are made for all people living in a
household, but reported by the 1 person interviewed. The response categories
need to be given to the interviewee on a card. Subjects have difficulty
remembering the response options when the information is asked with no
visual prompt.
BIA. OK
BIB. OK
BIC. OK
B2. OK
B3. OK
B4. SEE COMMENT A ABOVE
B5. OK
B6. SEE COMMENT A ABOVE
B7A. GENERALLY ok
OCCASIONAL RESPONSE = NOT THAT I AM AWARE OF (PARENT RESPONDING FOR
CHILD)
B7B. OK
8A. OK
PROBLEM WITH TEACHERS WHO HAVE EXTENDED SUMMER BREAKS
Handle in the Q x Q.
2
-------
SB. OK
BSC. OK
BSD. OK (MOST COMMON RESPONSE - DON'T KNOW)
BSE. OK (MOST COMMON RESPONSE - DON'T KNOW)
CI. OK (SEE COMMENT A ABOVE)
C2. OK
C3. OK
Dl. OK
D2. OK
*************************************************************************
Baseline Questionnaire (number administered = 42)
(Time of administration range = 45-90 min; mean = 60 min)
Perhaps it would be better to ask about ethnicity first and then deal with
the question of race. This would eliminate some problems like "Afro-
American Haitians." Arizona has a problem with "Spanish" or "Mexican"
since the Spanish came to Mexico.
Al. Problems with 1/2 white & 1/2 spanish/hispanic
Problems with "enrolled tribe " many tribes in Mexico with
no "enrollment"
A2. The public confuses race and ethnicity.
A3. OK
A4. OK
A5. OK
A6. APPROXIMATION OF HEIGHT NOT FACTUAL HEIGHT
Handle in the Q x Q
A7. RELUCTANCE OF WOMEN TO ANSWER THIS QUESTION (11 OUT OF 42)
EXPLAINING WHY THE INFO WAS NEEDED ELIMINATED THE PROBLEM
ASA. OK
ASB. OK
A9A. OK
A9B. OK
A9C. OK
A9D. OK
BlA-f. OK
B2. SEE COMMENT A ABOVE.
THIS QX NEEDS A TIME HORIZON POINT OF REFERENCE IN THE QX
(Last month, last 6 months, last year, or in your lifetime?)
B2B. This question has comma problems
Do you want to know whether people east snacks outdoors? or
-------
Do you want to know where people eat snacks outdoors (at
home or elsewhere)?
B2F. Eat Produce from a local market or stand (does this include local
supermarkets?)
B2H. Eat fish caught in the ocean? Does this include canned tuna?
B3. OK
B4. HARD QUESTION FOR PEOPLE—Interviewer sensed that subjects were
giving the "right" answer not necessarily what they did. This
question may need a lead in statement relative to the importance
of "true reporting" or some wording that is very non -judgemental.
B5. OK
B6. People laughed at this question and seemed a little embarrassed.
Interviewer sensed that people answered this truthfully unlike QX B4.
People asked:
b6a. Does this include pens or pencils?
Why is b6e included under the adult QX. Isn't this a child behavior?
B7a. OK
B7b. problems
# hours at work was a problem for school teachers in the summer.
# hours at school problem in summer.
B7c. OK
B7d. OK
B7e. Problem: most important job activity/duty or most frequent job duty?
"What do you do most often?"
B7f. Many people have jobs (5 out of 42 respondents) evenly split
among multiple environments.
B7g. comments:
What constitutes "regular" Should define specific timeline.
The headers (i.e., Dust, Fumes, Pesticides, Other Chemicals) need
a yes/no class response and then the specific types of exposure need a
yes/no response. Include a "don't know" category and an "other"
category in the specific types responses.
B8
(see
B9a.
OK
B9B
OK
BIO.
OK
Bll.
OK
CI.
OK
C2.
OK
C3.
OK
C4.
OK
4
-------
Dl. OK
D2. OK
D3. OK
D4. OK
D5A. OK
D5B. OK
D5C. OK
D6A. OK
D6B. Different water sources were used for cooking and drinking.
Frequently tap water is used for cooking and bottled water is used
for drinking. (This affected 14 out of 42 respondents).
D6C. OK
DSD. OK
D7A. OK
D7B. OK
D7C. OK
D7D. OK
D8. OK
D9A. OK
D9B. OK
DSC. OK
D10A. OK
D10B. OK
D10C. Problems ask only # changes per year (eliminate other choices)
ask Qx. regarding a bleeder line on the cooler. Yes/No
D10D. OK
D10E. OK
D10F. OK
Notes There should be an equivalent of D9C for the Evap cooling
question (D10). This will be added.
Dll. OK
D12. OK
D13. OK
D14A. OK
D14B. OK
D14C. OK
D15A. OK
D15B. OK
D15C. OK
D16A. OK
D16B. OK
D16C. OK
D16D. OK
5
-------
D17A. OK
D17B. OK
D17C. OK
D17D. OK
D18. Problems: Table Configuration is problematical,
D19. Use of the word lawn implies grass. Arizona doesn't
have many grass "lawns". Yard is a better
term. This can be handled in the Q by Q instructions.
D20. OK
D21. What defines a "pet". Horses are petsi
Handle in Q x Q.
D21A-C. Problems with overlap. Rework this question to
identify pets who are indoors and pets who transport
particulate from the outdoor environment into the house.
D21D. OK
************************************************************************
FOLLOW-UP QX (number administered = 35)
(Self-Completed « 20-40 min; mean = 30 min)
This questionnaire is set up by RTI as an administered QX.
Arizona will have this as a self completed QX and as such
it will have to be reformatted to be user friendly. We
will address the question content problems. Be aware
that we must change the format.
A1 A-O. OK
A2 A-1. OK
A3. OK
A4. OK
A5 A-B. OK
Bl. The medicine subdivisions are causing problems. People
take drugs but they don't know what class they fall into.
It would be better to have subjects list the drugs, vitamins,
pills etc. and then we are able to group them.
(17 subjects were taking some medication and 9 subjects
had problems with the question; example: Benadryl is an
antihistamine & benadryl has diuretic side effects)
B2A-B. OK
6
-------
B3.
B4A-B
B5.
OK
OK
OK
B6A-B. OK
Cl-33. Good Table.
Dl. Income: people do not want to answer.
About 20% refused to answer.
************************************************************************
Time Activity Questionnaire: (number administered = 27)
(Self-Completed = < 10 min per day)
In the group of NIH Subjects who completed the diary, all agreed
that this was far easier to complete than diaries from our
previous studies.
Daily time expenditure has a formatting line-up problem.
There were no other problems with the questions asked.
************************************************************************
24-Hour Food Diary (number administered = 5)
(Self-Completed = < 30 min per day)
The example provided conflicts with the instructions given.
The level of detail varies greatly among individuals.
************************************************************************
24 hour food diary follow-up. (number administered = 5)
(Self-Completed = 5 min per day)
no problems
************************************************************************
Technician Questionnaire (number completed = 2)
(Technician Completed * 20 min)
1. OK
2. OK
3. OK
4. OK
5. OK
6. OK
7A-I. OK
Az only: Tracking Info. Very important for Temporal Pollow-up
7
-------
APPENDIX C: QAU AUDIT DOCUMENTS
-------
QUALITY ASSURANCE UNIT
University of Arizona
FonnTP-1
0394
Page 1 of 2
Title: Procedure for the Performance of Informal Quality Assurance Audits
Document Ho.: qa-05
Approval/Date:
HFull SOP ~ Working SOP
No. of Pages: 5
Dir., Office of Arid Lands Studies: \J XV. -1^
March 1994
Dir., Quality AssuranaK^^ &
Martin H. Karpiscak 6 A A" '
Revision No.: ^
doft
Revision No.:
Revision Date:
Revisions Made:
Revision No.:
Revision Date:
Revisions Made:
Revision No.:
Revision Date:
Revisions Made:
Dir., Office of Arid Lands Studies:
Dir., Quality Assurance:
Preparer:
Dir., Office of Arid Lands Studies:
Dir., Quality Assurance:
Preparer:
Dir., Office of Arid Lands Studies:
Dif., Quality Assurance:
Preparer:
This page is a record of all revisions of this document Each time this document is changed, only the new or revised pages
will be issued as appropriate.
DISTRIBUTION: XX Controlled D ^controlled
If distribution is controlled, see page 2 (reverse side) of this document.
-------
FormTP-1
QUALITY ASSURANCE UNIT 0394
University of Arizona page 2 of 2
DOCUMENT CONTROL FOR:
Title: Procedurecfor the Performance of Informal Quality Assurance Audits
ID No: QA-05
Re
vision No
m
DISTRIBUTED TO:
Glenn Sipes, Pharm/Tox
Mary Kay O'Rourke, Resp. Sciences
Judy Ulreich, Surg. Biol.
0
03-94
03-94
03-94
03-94
1
2
3
4
Michael Mayersohiu Pharm. Sci.
William Remers, Pharm. Scl.
03-94
03-94
* Entry date of distribution in revision number box.
-------
SOP #QA-05
Rev. 0
Mar 1994
Page 1 of 5
PROCEDURE FOR THE PERFORMANCE OF
INFORMAL QUALITY ASSURANCE AUDITS
1.0 Purpose
This procedure describes audit activities such as audit preparation, audit scheduling, conduct
of the audit and documentation of audit results for informal/minor audits.
2.0 Applicability
This procedure is to be used for all informal quality audits performed by the Quality
Assurance Unit (QAU) for research projects on-campus at The University of Arizona and
for off-campus facilities/projects as requested.
3.0 Definitions
3.1 Audit: a formal, independent and unbiased official examination.
3.2 Verification: a methodical examination and review.
3.3 Performance Audit: an examination of the compliance of a program to its written
requirements.
3.4 Systems Audit: an examination of the technical proficiency of a program.
3.5 Internal Audit: an audit by individuals for the same parent organization as those whose
program is being audited.
3.6 External Audit: an audit by individuals NOT of the same parent organization as those
whose program is being audited.
3.7 Major Audit: a formal audit with written evaluations of all aspects of a program which
must be responded to by the audited organization.
3.8 Minor Audit: an informal examination/inspection of any aspect of a program.
4.0 References
4.1 Jose A. Rivera, Basic Quality Course, NUS Corporation (1985).
4 2 Frederick M. Garfield, Quality Assurance Principles f°r A™Wcal Oratories,
Association of Official Analytical Chemists, Arlington, VA (1984).
-------
SOP #QA-05
Rev. 0
Mar 1994
Page 2 of 5
5.0 Discussion
Planned and periodic audits must be performed to verify all aspects of quality assurance or
good laboratory practice requirements which are appUcable to a project to determine fte
effectiveness of the program implementation. Audits shall be conducted by qualified
individuals who are familiar with, but not directly responsible for, the work or activities
being audited. Auditors shall be knowledgeable in the methods of auditing project work to
established requirements.
6.0 Responsibilities
6.1 The Quality Assurance Director or his designee shall be responsible for:
6.1.1 Planning and executing a minor audit in accordance with the specifications
outlined in this procedure;
6.1.2 Preparing a final report listing the findings of the audit.
6.2 The Quality Control Director shall be responsible for:
6.2.1 Assisting the Quality Assurance Director in planning and executing a minor
audit in accordance with this procedure;
6.2.2 Performing minor audits.
7.0 Materials and Equipment
None
8.0 PROCEDURE (The following steps apply to minor or informal audits only. Reference
SOP #QA-02 for major or formal audits.)
5.1 Scheduling
8.1.1 Minor audits will be scheduled by the Quality Assurance Unit frequently enough
to allow for review of all critical project elements.
8.1.2 Minor audits will be scheduled at the request of a Project or Study Director.
8.2 Written Audit Checklist
8.2.1 The Quality Assurance Director or his designee shall be responsible for the
preparation of a written audit checklist.
8.2.2 The audit checklist shall include the following information:
-------
SOP #QA-05
Rev. 0
Mar 1994
Page 3 of 5
a) Unique Audit ID Number
b) Organization to be audited
c) Name of contact person
d) Project element to be audited
d) Date, time and location
e) Applicable documents) or section(s) of document(s)
8.3 Audit Notification
8.3.1 Unscheduled audits shall be performed at a date and time agreed to between the
project individuals to be audited and the audit team leader.
8.3.2 Notification for scheduled audits will be made by
a) Telephone with follow-up written confirmation
b) Written memorandum
8.3.3 The notification document will include
a) Date, time and location
b) Project element(s) to be audited
c) Auditor's name
8.4 Performance of Audit
8.4.1 The checklist as described above (8.2) will be used to guide the audit.
However, the checklist will not restrict the audit when evidence raises further
questions not specifically included in the checklist.
8.4.2 The auditor shall record each finding (observation, concern or deficiency).
8.4.3 When a finding is identified, sufficient investigation shall be conducted to
identify the basic cause of the finding.
8 4 4 Deficiencies found in procedures or plans not designated to be audited shall be
noted but not reported as findings. However, the appropriate management will
be notified in writing of the deficiency so it can be corrected.
8 4 5 Any findings identified that require immediate corrective action shall be reported
immediately to the audited organization's management and shall be recorded and
described in the Audit Final Report.
8.5 Evaluation of Findings
8.5.1 Findings shall be categorized as follows:
-------
SOP #QA-05
Rev, 0
Mar 1994
Page 4 of 5
al Observation' a situation that is weak and should be strengthened.
b) Concern: a situation that could potentially become a deficiency and must be
corrected in a timely manner. ..
c) Deficiency; a situation that is totally unacceptable and must be corrected
immediately.
8.5.2 The auditor shall list his findings in the Final Audit Report.
8 5 3 Findings shall be stated in clear, concise, revealing, and irrefutable statements
of feet that identify the problem. Persons in the audited organization who can
to the validity of the findings shall be identified by name.
8.5.4 Observations shall be reported, but no specific corrective action commitments
shall be required.
8.6 Post-audit Meeting
8.6.1 At the conclusion of an audit, a short post-audit meeting will be scheduled m)
held with the appropriate members of the audited organization who can verify
the validity of the findings.
8.6.2 The meeting will be conducted by the auditor.
8.6.4 The objectives of this meeting will be to:
a) Discuss the audit findings. r „
b) Determine and resolve any errors or misunderstandings regarding the findings.
c) Achieve agreement on the validity of the findings and reach agreement on those
findings that constitute non-compliance. . .
d) Recommend improvements or corrective actions to the audited organization.
e) Establish a tentative plan and schedule for corrective action development and
implementation.
f> Schedule a follow-up audit if appropriate.
8.7 Audit Reporting
8 7 1 The results of each audit and the preliminary corrective actions taken will be
documented and reported to the Project or Study Director.
8 7 2 If deficiencies or non-compliances have been found, the Prqject or Study
DireS must then respond in writing in a timely manner to the Director of
Quality Assurance and describe corrective actions to be taken.
8.8 Audit Follow-up
-------
SOP #QA-05
Rev. 0
Mar 1994
Page 5 of 5
The Quality Assurance Unit will maintain an open file on all audits until such time that
corrective action has been completed by the audited organization and verified by the
Director of Quality Assurance or his designee.
9.0 Records
All documents generated as a result of the audit function will be maintamed by the Director
of Quality Assurance or his designee. Such documents include at a minimum:
9.1 Audit Notification
9.2 Audit Checklist
9.3 Audit Final Report .
9.4 Response by audited organization to findings of concern or deficient
-------
DIVERSITY OF ARIZONA
QUALITY ASSURANCE UNIT
DOCUMENT CONTROL
3SUER
Quality Assurance Unit
University of Arizona
CLIENT ®ee con*-r°l list
below
Number: QA-02
Title and/or Description
of Documents
ProcedurefortheJIerformance^^^JJualit^^Assurance^Audits
(SOP)
*Mar 94: J^rocedure_JEoi^^h£—Performanc^^of^^oirmal^^ualit^^ssurance Audits
This page Is a record of all revisions of this document. Each time this document is changed, only the new
or revised pages will oe issued.
Eev
Date and Signature
Preparer Reviewed/Approved-Date
* f M XT a e^alr 1C
Issue
Date
Description of Revisions{s)
Affecte
Pages
0
~ C B U~
S.B. Hop1
Martin M. Karpiscak
7-22-86
original issue
1-9
Mar 94
New Title and revisions appro-
priate to new title
1-8
Distribution xgx controlled
August 1986:
~ Uncontrolled
March 94: __
Glenn Sipes , Pharm/Tox
John Barr r,Pharm/Tox
Dean Carter , Pharm/Tox
Jim Quackenboss,Resp. Sci.
Milos Chvaoil, Surg. Biol.
Judy Ulreich ,Surg. Biol.
Martha Anderson, Risk Management
Glenn Sipes, Pharm/Tox
Mary Kay O'Rourke, Resp. Sci.
Judy Ulreich, Surg.Biol.
Stuart Williams, Surgery
M. Mayersohn, Pharm. Sci.
Wm. Renters, Pharm. Sci.
I
-------
University of Arizona
Quality Assurance Unit
Page of
REVIEW RECORD FOR STANDARD OPERATING PROCEDURES
ID NO. QA-02 1 REV. 1 1 ISSUE DATE: Mar 94
PREPARED BY:
Susan B. Hopf
TITLE: Procedure for the Performance of Formal Quality
Assurance Audits
Working SOP i—
Fu 1 1 SOP it
ISSUED BY: Quality Assurance Unit
University of Arizona
Original r-n r=-.
Issue 1—1 Revision 1—1
(check as applicable)
SPECIFICALLY
Purpose
GENERALLY
Appticabi1ity
Def in i t ions
References
Discussion
Respons i b iIi 11es
Equipment
Sequence of Operation
Records
Technical Approach
CommunI cat ion
Effect iveness
QA/QA Requirements
Other:
SAT
/
y
/
/
/
S
/
/
'/
UNSAT N/A
below or on reverse
COMMENTS:
, -,~S*
Zjr
/Sl ,
APPROVED
APPROVED WITH
REVISIONS ABOVE
NOT APPROVED
'(date)
(signature)
(date)
(signature) (date)
PREPARER; Please respond to comments above and/or on reverse side. 0292
-------
SOP #QA-02
Rev. 1
Mar 1994
Page 1 of 8
procedure for the performance of
FORMAL QUALITY ASSURANCE AUDITS
1.0 Purpose
This procedure describes audit activities such as audit preparation, audit scheduling, conduct
of the audit and documentation of audit results.
2.0 Appucamuty
This procedure is to be used for all formal quality audits performed by the Quality
Assurance Unit (QAU) for research projects on-campus at The University of Arizona and
for off-campus facilities/projects as requested.
3.0 Definitions
3.1 Audit.' a formal, independent and unbiased official examination.
3.2 Verification: a methodical examination and review.
3.3 Performance Audit: an examination of the compliance of a program to its written
requirements.
3.4 Systems Audit", an examination of the technical proficiency of a program.
3.5 Internal Audit: an audit by individuals for the same parent organization as those whose
program is being audited.
3.6 External Audit: an audit by individuals NOT of the same parent organization as those
whose program is being audited.
3.7 Major Audit: a formal audit with written evaluations of all aspects of a program which
must be responded to by the audited organization.
3.8 Minor Audit: an informal examination/inspection of any aspect of a program.
4.0 References
4.1 Jose A. Rivera, Basic Quality Course, NUS Corporation (1985).
4.2 Frederick M. Garfield, Quality Assurance Principles for Analytical Laboratories,
Association of Official Analytical Chemists, Arlington, VA (1984).
-------
SOP #QA-02
Rev. 1
Mar 1994
Page 2 of 8
5.0 Discussion
Planned and periodic audits must be performed to verify all aspects of q^ ^ur^^
Eood laboratory practice requirements which are applicable to a project to^determine the
effectiveness of the program implementation. Audits shall be conducted y Q
individuals who are familiar with, but not directly response for, die wor*
being audited. Auditors shall be knowledgeable in the methods of auditing project work to
established requirements.
6.0 Responsibilities
6.1 The Quality Assurance Director or his designee shall be responsible for:
6.1.1 Planning and executing a major audit in accordance with the specifications
outlined in this procedure;
6.1.2 Preparing a written audit plan;
6.1.3 Preparing a final report listing the findings of the audit.
6.2 The Quality Control Director shall be responsible for:
6.2.1 Assisting the Quality Assurance Director in planning and executing a major
audit in accordance with this procedure;
6.2.2 Performing minor audits.
7.0 Materials and Equipment
7.1 Quality Assurance Audit Guide (Figure 1)
7.2 Quality Notice Form (Figure 2)
8.0 Procedure (The following steps apply to major or formal audits only. Reference SOP
#QA-G5 for minor or informal audits.)
8 1 The Quality Assurance and Quality Control Directors of the Quality AssuranceUnit
sh^ WL a schedule of planned and periodic audits. A major audit will be
scheduled^ at least annually with minor audits scheduled frequently enough to allow for
review of all critical project elements.
8.2 Written Audit Plan
8.2.1 The Quality Assurance Director or his designee shall be responsible for the
preparation of a written audit plan.
-------
SOP #QA-02
Rev. 1
Mar 1994
Page 3 of 8
8.2.2 The audit plan shall include the following information:
a) Unique Audit ID Number
b) Organization and project element to be audited
c) Subject of audit
d) Scope of the audit
e) Activities to be audited
!) Names of the audit team members
g) Schedule of dates, time and location
h) Applicable documents
8.3 Audit Guides or Checklists (see Figure 1 of this procedure)
8.3.1 The Quality Assurance Director or his designee as Audit Team Leader shall be
responsible for directing and approving the preparation of audit checklists.
8.3.2 The following guidelines shall be used in preparing audit checklists:
a) Initial Baseline Audits: Checklists will be based on applicable project control
documents, e.g., QA Plan or GLP Plan, Protocols, or Standard Operating
Procedures.
b) Follow-up Audits: Checklists shall be based on a review and evaluation findings
from previous audits, responses to those findings and available objective
evidence of implementation of corrective action.
c) Periodic Audits (minor or informal audits): Checklists shall be based on items
similar to those noted above with specific emphasis on areas considered critical
to the project at the time of the major audit or found to be weak but not
reported as a finding during a previous audit.
8.4 Audit Team Organization
8.4.1 The audit team leader shall prepare his team, if one is required, prior to
initiation of the audit.
8.4.2 He shall assign specific areas for each member to audit in accordance with the
checklists which have been prepared.
8 4.3 Each member of the team shall be provided with copies of the audit plan,
procedures and checklists as well as any other pertinent documents as necessary
to ensure an orderly audit.
8.5 Audit Notification
-------
SOP #QA-02
Rev. 1
Mar 1994
Page 4 of 8
8.5.1 All organizations involved in a scheduled audit shall be notified in writing at
least 15 days prior to the date of the audit.
8.5.2 Notification shall include the following:
a) Date of Audit
b) Time and place of pre-audit meeting
c) Individuals who should attend pre-audit meeting
d) Scope and duration of the audit
e) Name of the audit team leader
8.6 Pre-audit Meeting
8.6.1 The audit team leader shall conduct a pre-audit meeting at the audit site with the
audit team members and cognizant management of the organization to be
audited.
8.6.2 The purpose of the pre-audit meeting shall be to:
a) Introduce the auditors
b) Meet counterparts
c) Confirm the scope of the audit
d) Present the audit plan
e) Discuss the audit checklists
f) Discuss the audit sequence
g) When appropriate, review previous audit results
h) Establish channels of communication
i) Arrange post-audit meeting
8.7 Performance of Audit
8.7.1 The checklists as described above (8.3) will be used to guide the audit and to
ensure ad«T""» depth and continuity. However, the checklists will not restrict
the audit when evidence raises further questions not specifically included in the
checklists.
8.7.2 Audit team members shall record each finding (observation, concern or
deficiency) on a Quality Notice Form (see Figure 2).
8.7.3 When a finding is identified, sufficient investigation shall be conducted to
identify the basic cause of the finding.
8 7 4 Deficiencies found in procedures or plans not designated to be audited shall be
noted but not reported as findings. However, the appropriate management will
be notified of the deficiency so it can be corrected.
-------
SOP #QA-02
Rev. 1
Mar 1994
Page 5 of 8
8.7.5 Any findings identified that require immediate corrective action shall be reported
immediately to the audited organization's management and shall be recorded on
a Quality Notice Form.
8.8 Evaluation of Findings
8.8.1 Findings shall be categorized as follows:
a) Observation: a situation that is weak and should be strengthened.
b) Concern: a situation that could potentially become a deficiency and must be
corrected in a timely manner.
c) Deficiency: a situation that is totally unacceptable and must be corrected
immediately.
8.8.2 Each member of the audit team shall draft his own findings on Quality Notice
Forms. These be reviewed by other team members. The audit team
IfarW will then categorize the findings under general headings and rank them
according to relative importance.
8.8.3 Findings shall be stated in clear, concise, revealing, and irrefutable statements
of fact that identify the problem. Persons in the audited organization who can
attest to the validity of the findings shall be identified by name.
8.8.4 Observations shall be reported, but no specific corrective action commitments
shall be required.
8.9 Post-audit Meeting
8.9.1 At the conclusion of an audit, a post-audit meeting will be scheduled and held
with the appropriate members of the audited organization.
8.9.2 The meeting will be conducted by the audit team leader.
8.9.3 The meeting attendance should include those members of the organization who
can verify the validity of the findings as well as those members of management
who can correct the deficiencies as defined in the audit.
8.9.4 The objectives of this meeting will be to:
a) Discuss the audit findings.
b) Determine and resolve any errors or misunderstandings regarding the findings.
c) Achieve agreement on the validity of the findings and reach agreement on those
findings that constitute non-compliance.
d) Recommend improvements or corrective actions to the audited organization.
-------
SOP #QA-02
Rev. 1
Mar 1994
Page 6 of 8
e) Fst^hiish a tentative plan and schedule for corrective action development and
implementation.
f) Schedule a follow-up audit if appropriate.
8.10 Audit Reporting
8.10.1 The results of each audit and the preliminary corrective actions taken will be
documented and reported to the Project or Study Director.
8.10.2 The Project or Study Director must then respond in writing in a timely
manner to the Director of Quality Assurance and describe the corrective
actions that will be taken.
8.11 Audit Follow-up
The Director of Quality Assurance will maintain an open file on all audits until such
time that corrective action has been completed by the audited organization and verified
by the Director of Quality Assurance or his designee.
9.0 Records
All documents generated as a result of the audit function will be maintained by the Director
of Quality Assurance or his designee. Such documents include at a minimum.
9.1 Audit Notification
9.2 Audit Plan
9.3 Audit Checklists
9.4 Audit Final Report
9.5 Quality Notice Forms
9.6 Response to quality notices by audited organization
-------
SOP #QA-02
Rev. 1
Mar 1994
Page 7 of 8
-------
SOP #QA-02
Rev. 1
Mar 1994
Page 8 of 8
Figure 2. Quality Notice Form.
1
5
0
I
1
o
I
a
a
i
1
a
*
a
i
IV
J
IS
i
m
I
h
|
2
' i
i
'{
1
i
a
fl
I
a
3
i
a
Is.
s
i!
f
i
r-
s
1
1
i
L
as
O
fib
o
*
u
<
m
I
l III
I
!
x
at
o
&*
6m
O
z
o
as
Cm
-------
University of Arizona
QoatiTX Assurance Unit
Page 1 of 1
REVIEW RECORD FOR QUALITY ASSURANCE PLAN
ID NO: UA-G-
QSIP
REV;
0
ISSUE DATE:
June 1995
PREPARED BY:
Mary Kay O'Rourke
title: Preliminary Quality Systems and Implementation Plan for Feasibility
Studies in Support of the NHEXAS Project
REVIEWED BY:
Susan B. Hopf
issued by: Health and Respiratory Sciences
The University of Arizona
Original Issue
~
Revision
VERIFICATION SCOPE:
(check as applicable)
Title Page
Table of Contents
Project Description
Project Organization and Responsibility
GA Objectives for Measurement of Data
(DQOs and DQIs)
Sampling Procedures
Sampling Custody
Calibration Procedures and Frequency
Analytical Procedures
Data Reduction, Validation, and Reporting
Internal Quality Control Checks and Frequency
Performance and System Audits and Frequency
Preventive Maintenance Procedures and
Schedules
Specific Routine Procedures for Assessment of
Data Precision, Accuracy, and Completeness
Corrective Action Procedures
Quality Assurance Reports to Management
Other
SAT
UNSAT
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
N/A
If UNSAT, explain below or on
reverse side
COMMENTS:
m APPROVED
~ APPROVED WITH
REVISIONS ABOVE
(Signature)*
(Signature)
~
NOT APPROVED
(Signature)
Preparer: Please respond to comments above and/or on reverse side.
-29
/9?S~
(Date)
(Date)
(Date)
0695
-------
ENDNOTES
1. (QSIP page 6) "We plan to analyze each media for all target analytes." At the time
of the survey not all media were deemed amenable to analysis for each analyte (e.g.
VOCs in any media except air); the statement was an over simplification. Further, on
July 26 of 1996, the cost effectiveness of sample analysis in selected media was
examined. The project investigators, with EPA Project Officer or Project
Collaborator concurrence, eliminated analysis of selected media for specified
analytes as described below.
Analyte Class-*
Metals
Pestic-
ides
VOCs
Last
Date
Collected
Action taken/
Comments
Sample Collection
strategy change i
PID(indoor/outdoor)
X
07-21-96
160 Homes BDL
Eliminate sample
Outdoor OVM
Badges
X
07-21-96
90% Reduction
Outdoor PF1 Tubes
X
07-21-96
90% Reduction
Vacuumed Floor
Dust
X
X
n/a
07-21-96
Tradeoffs were
made between
area sampled and
dust yield:
8 m2 until 03/96;
4 m2 until 08/96;
6 m2 until 07/97;
3 m2 Stage 6
12/97
All media except
food and water
X
n/a
07-21-96
Discontinue the
analysis of
carbaryl
Foundation Soil
X
X
n/a
07-21-96
Cease collecting
samples
Yard Soil
X
n/a
07-21-96
Eliminate sample
collection for
pesticides in
Stage II—Keep
metals byXRF
Surface thin films
X
n/a
n/a
07-21-96
Eliminate
experimental
Procedure for
analysis by XRF
Sill wipe
X
n/a
07-21-96
Eliminate
Sample
Collection
NHEXAS AZ QSIP Endnotes. Page 1 of 15
-------
Outdoor Fixed-site
X
(kept in
Cochise
Co.)
X
(reinst-
ated in
1997)
07-21-96
Eliminate
Sample
Collection
Food (pre & post)
preparation analysis
07-21-96
There is
insufficient
funding to
undertake the
this experimental
approach
These changes reflect the collection schedule. Some samples were collected and not
analyzed at the end of the project when funds were expended; (project officer
concurrence).
2. (QSIP page 6) TheQSIP was written prior to implementation of the project. We
hoped additional funds would be found to evaluate PAHs. PAHs were never
undertaken in our NHEXAS project. They were undertaken in the Arizona Border
Survey (data and QSIP to be available at a later time).
3. (QSIP page 6) Iron was never one of the primary or secondary analytes. This is a
typographical error in the QSIP. Iron is evaluated along with 17 other metals by XRF
in air, soil and house dust samples.
4. (QSIP page 6) XRF was employed to evaluate 17 supplemental elements in air, soil
and house dust samples. These elements were: potassium (K, CAS: 7440-09-7),
calcium (Ca, CAS: 7440-70-2), cobalt (Co, CAS: 7440-48-4), iron (Fe, CAS: 7439-
89-6), molybdenum (Mo, CAS: 7439-98-7), thallium (Tl, CAS: 7440-28-0), silver
(Ag, CAS: 7440-22-4), strontium (Sr, CAS: 7440-24-6), uranium (U, CAS: 7440-
61-1), thorium (Th, CAS: 7440-29-1), tin (Sn, CAS: 7440-31-5), tungsten (W, CAS:
7440-33-7), titanium (Ti, CAS: 7440-32-6), rubidium (Ri, CAS: 7440-17-7), iridium
(Ir, CAS: 7439-88-5), and mercury (Hg, CAS: 7439-97-6).
5. (QS IP page 9) Three of the 4 pesticides were available using a single extraction
technique. Carbaryl required separate treatment and added extensive cost to the
project. Preliminary analysis of initial samples were dominated by BDL values with
only a few detects in the very low range. Analysis of carbaryl in media other than
food and water was eliminated on 7-21-96 with EPA Project Officer or Project
Collaborator concurrence.
6. (QSIP page 10) Assessment of Total VOCs ceased when collection of data using the
photoionization detector (PID gun) ceased on 7-21-96 with EPA Project Officer or
Project Collaborator concurrence.
7. (QSIP page 13) See endnote 4 for additional analytes.
8. (QSIP page 15) By 7-21 -96 it was apparent there would be insufficient funds to
complete two stages of the proposed study design. Stages 4 and 5 would consist of
50 homes that would be evaluated in seasons other than that of primary collection
(during stage 3). These stages were eliminated with EPA Project Officer or Project
Collaborator concurrence.
NHEXAS AZ QSIP Endnotes. Page 2 of 15
-------
9. (QSIP page 15) The original proposal described use of an instrument designed by
one of the Pis to accurately identify the precise location of a study subject. This
objective was written into the QSIP; the instrument was never used (funding would
not support the effort); no data were collected to support this objective.
10. (QSIP page 16). Cost effectiveness and subject burden very quickly became issues
within the project and caused us to scale back several proposed approaches. By
January of 1995, we decided to simultaneously collect Stage 2 and Stage 3 samples
in any houses selected for Stage 3 sampling. We could not afford the added cost of
three additional visits to each Stage 3 home. We were also concerned of the
additional burden on the subjects. With EPA Project Officer or Project Collaborator
concurrence, we have a group of Stage 2 only homes and a group of Stage 2& 3
collapsed homes where samples for each stage were taken during the same visit.
11. (QSIP page 18) Changes in SOPs used for administrative purposes. (BCO-C-l.O-
removed and replaced by BCO-G-l.O; BCO-G-2.0; and BCO-G-3.0 date: 9-15-95).
UA-D-42.0 was added to the training protocols date 9-15-95).
12. (QSIP page 19) Changes in Food SOPs—UA-F-23 and 24 were added to field
procedures.
13. (QSIP page 20) BCO-L-1.0 was eliminated as a method for analysis. Battelle
informed Arizona that it was as cost effective to use GC-ECD as an analytical
approach as working with the ELISA test kits. The ELISA approach was eliminated
11/08/95. Protocol BCO-L-24 was issued. FDA provided Laboratory protocols for
food analysis FDA 101, 102 and 103 for metals and FDA 202, 203, 204 for pesticides
(8/22/95). Listed water methods were corrected 525.1 was replaced with 524.2 and
554 was replaced with 531.1 for pesticide analysis (8-22-95). These changes
occurred prior to collection of samples.
14. (QSIP page 21) Many of the Data SOPs were written based on past projects and
expectations of new data processing procedures to provide adequate QA and
documentation. Once the project began several procedures were eliminated and
others were instituted. System definition included the addition of UA-D-30, 41 and
42. UA-D-32 and 33 were added as data entry procedures. UA-D-8.0 and 12.0 were
never used and replaced with UA-D-31, 35, 37 and 39. UA-D-19 and 23 were
eliminated and replaced with UA-D-34, 36, 38, and 40 for cleaning and verification.
All were implemented by 9-15-95, prior to data entry of these forms. Additional
SOPs were developed to cover other functions UA-D-43 through 47. These SOPs
were instituted prior to data entry of these forms.
15. (QSIP page 22) Developing specifics of the analytical plan was difficult before the
data were available. Three general SOPs were written. Once data were available,
Illinois Institute of Technology developed the additional 12 SOPs (1IT-A-7.0 through
15.0)
16. (QSIP page 24) At the time the Cooperative Agreement was awarded, EPA requested
the "entire state of Arizona" be evaluated. To achieve this goal, 50 census tracts were
selected expanding the households to be contacted to 1250 (October 1995). PSUs
were later cut to 49, with EPA Project Officer or Project Collaborator concurrence,
for fiscal reasons. The last tract (PSU) identified in the sampling design was cut. The
final number of homes approached for enrollment was 1225.
NHEXAS AZ QSIP Endnotes. Page 3 of 15
-------
17. (QSIP page 25) We planned to approached over 900 homes for Baseline
questionnaire completion. We ended up with baseline questionnaires from 534
households (140 from Stage 1 + 141 Stage 2+179 Stage 3 + 74 Stage 6).
18. (QSIP page 25) We approached the 955 enrolled homes to participate in further
participation 534 agreed.
19. (QSIP page 25) With the net gain of 1 PSU the number of homes receiving intensive
sampling changed from 175 to 179 homes.
20. (QSIP page 25- add the following as section 1.3.2.e) After the initiation of the
project, EPA funded Research Triangle institute to conduct a comparability study of
analytical techniques. Arizona provided samples of VOCs in air and water; Metals
and Pesticides from air, floor dust, sill wipes, water and food, these samples were
collected at 30 homes sampled from May to August of 1996.
21. (QSIP page 28-29) At the time the project was undertaken, certain analyses were
envisioned. Virtually all funding went into collection, analysis of samples and
preparation/finalization of data. Within the project funding was insufficient to realize
all analytical goals. Work continues.
22. (QSIP page 31) With the change in approached homes (1200 to 1225) response rates
change to 1103 and 919 (at the 90th and 75th percentile, respectively).
23. (QSIP page 33) Water is a media category used throughout the QSIP. The category
includes "tap" water collected from an untreated tap in all homes and "separate
source drinking water" which may come from multiple sources outside the home (i.e.,
bottled water, cistern, etc); "separate source drinking water" also includes treated tap
water (i.e., water processed through an in-line filter, "britta" type filter, etc). Water is
not collected as a "beverage" or in the "liquid food" sample (see Endnote 24).
24. (QSIP page 33) The terms "food" and "beverage" are used throughout the QSIP.
Better conceptual terms are "solid food" and "liquid food" respectively. Respondents
placed "solid food" in ziplock bags and plastic food containers. "Liquid foods" like
ice cream or soup were placed in the "beverage" container along with fluids like
coffee, tea, milk, soda, and juice to name but a few. Water was evaluated separately;
water was not added to the beverage container unless as an ingredient of a beverage
or a small bottle purchased outside the home. Exposure from water consumption can
be modeled using information in the diet diary and analytical results for "tap" and
"separate source drinking water."
25. (QSIP page 35) Stage 2 visits required two visits to each home with subject
involvement. We were able to interest more subjects by reducing the burden. We cut
completion of the diet diary, time activity diary, and follow-up questionnaire from the
Stage 2 protocol and referred to these as Stage 6. The single visit approach involved
more subjects and reduced field burden and cost. The reduction of stage 2 to Stage 6
was initiated in late April of 1997. Stage 6 samples were collected through the end of
NHEXAS.
26. (QSIP 39) The project had about 35 people working on it at any one time. Students
employees performed most of the field an laboratory functions (under supervision) at
the University of Arizona. Field staff turn-over was great. More than 100 people
worked on the project in some capacity. Training in this project was continuous and
scrupulously enforced.
NHEXAS AZ QSIP Endnotes. Page 4 of 15
-------
27. (QSIP 43) We found it better to work with Counties and Census tracts as primary
sampling units (PSUs) instead of region by state. Secondary sampling units (SSUs)
were combined block groups.
28. (QSIP 44) For some parts of Arizona, seasonal evaluation of Stage 3 sampling
would provide no usable information and add to financial burden. In these regions
(e.g., Kingman, Arizona) very few (e.g., 3) homes were intensively evaluated. With
only one or two homes evaluated in different seasons, investigators could not
ascertain whether differences were caused by individual variability among homes or
seasonal variability. (Seasonal sampling would defeat the purpose of multistage
"cluster sampling" designed to reduce costs while minimizing varience. The project
as a whole samples across all seasons. All seasons are not represented for each
climatic regime whether caused by differences in elevation or differences in season.
29. (QSIP 44) PSUs = Census Tracts which are, in fact, "combined census block
groups."
30. (QSIP 48) Both primary and secondary respondents were selected. In some cases,
the primary respondent was not available when we arrived at the home for sampling.
In these cases, the preselected secondary respondent became the focus of the field
collection (for all intents, the secondary respondent became the primary respondent).
To keep this simple in the field collection and data bases, "Individual Respondent
Numbers" were changed to reflect the substitution. As a result, IRN 01 is always the
person in the household treated as the "primary" respondent.
31. (QSIP 49) In some cases, the selected family was unable to arrange schedules for
precisely 7 days. In a few cases, household sampling duration is shorter or longer
(range 5-8 days). This is one reality of field based projects.
32. (QSIP 57-58) Arsenic was evaluated using Hydride Generation (extraction) and
Atomic Absorption Spectroscopy or HG-AAS. Total arsenic was measured; arsenic
was not speciated.
33. (QSIP 58) At one time we planned to evaluate sources of contaminant in prepared
food. We expected to buy food in the grocery, then sample the food prior to storage
in the home; next, have the resident prepare the food and finally, sample the food
prior to service. We referred to that as Pre/Post-Cooked Food. That sub project was
eliminated (with concordance of the EPA Project Officer or Project Collaborator)
prior to entering the field.
34. (QSIP 58) FDA metals evaluation methods in food changed to ICP MS.
35. (QSIP 58) FDA Pesticide analysis in food aliquots was performed using GC/FPD.
36. (QSIP 60) Routine QA audits were made. Reports were not always filed. Verbal
reports were made at monthly staff meetings and issues were resolved.
37. (QSIP 61) Four, formal QA audits were performed. 1 by EPA and 3 internal QA
audits throughout the field project.
38. (QSIP 62) Tribal Councils determine access to tribal lands. One of the census tracts
selected included the Tohono O'odham Reservation. For this tribe, permission to
work on tribal lands is obtained by (1) obtaining sponsorship by one of the tribal
committee's, (2) the sponsoring committee takes the proposal to all other committees
and obtains their approval, (3) recommended projects then seek approval from the
tribal council. Investigators approached both the Environmental Committee and the
Health Committee. Directors of these committees viewed NHEXAS as "too
NHEXAS AZ QSIP Endnotes. Page 5 of 15
-------
intrusive." Neither committee would sponsor the project effectively prohibiting
implementation on tribal lands. This tract was replaced with the next tract on the list.
Native Americans living outside reservations and in selected census tracts were
included in the survey.
39. (QSIP 63) We planned to use Harvard Black Box Impactors and pumps exclusively.
As pumps failed they were replaced with the more dependable and less expensive
SKC pumps and operated at the same mass flow rate.
40. (QSIP 67) Many of the samples evaluated in the laboratory reflect values "below the
limit of detection'1 or BDL. These are valid measures. Other samples were collected
and unanalyzed because of budgetary limitations.
41. (QSIP 68) NHEXAS respected subject rights, and conformed to protocols for the
protection of human subjects in research projects. The project was reviewed and
approved by the Human Subjects Committee at the University of Arizona. Reporting
precise location of homes reveals subject identity. Although the measures were
made, we are reporting locations by zip code when possible. Zip codes and PSUs arc
not reported when fewer than 1000 households are in a zipcode and/or PSU.
42. (QSIP page 68) NIST developed samples for a round-robin evaluation of the
analytical labs. Specific standards to be used in the field and lab were anticipated but
ultimately not available.
43. (QSIP page 69) When samples or sample aliquots were reanalyzed, all values reside
in the data base. The initial value is reported in the results data base and flagged.
Replicate values are reported in the replicate data base and are flagged. Replicate and
original values are linked by sample id in both databases. Samples split for QA
purposes are reported in a separate QA data base. The analyst can then independently
decide how to handle the value.
44. (QSIP page 69) Method of collection varied only for air with the substitution of SKC
pumps for Harvard mass flow controlled pumps. Careful side by side evaluations
were performed and no significant differences were observed. The SOP (UA-F-3.X)
was altered (12-96) to reflect the use and specifications of both pump types.
45. (QSIP page 69) No other equipment changes were implemented. However, problems
were encountered in obtaining sufficient amounts of window sill dust and floor dust
in selected samples. Samples with low dust yield or expanded sampling area are
flagged in the data bases.
46. (QSIP page 69) Important typo AAS should be HG-AAS and ICAP-AES should be
ICP-AES.
47. (QSIP page 71) Concerns surrounding pump failure in the Arizona climate were so
great, that we undertook outdoor air sampling in all cases (except as modified in
Endnote #1), with the expectation of having values at least 50% of the time.
48. (QSIP page 71) Refusal rates were very high with personal air. We attempted
sampling 20% of the Stage 3 population and obtained samples for 13% of the
population.
49. (QSIP page 71) In a very few homes more than two water sources were routinely
used (n = 4). In those homes we collected samples from 3 and 4 water sources.
50. (QSIP page 76) Stage 1 estimates for burden were predicated on administering both
the Descriptive (10-15 minutes) and Baseline questionnaires (1 hour). We found few
subjects willing to spend 1.2 hours accommodating unscheduled interviewers. We
NHEXAS AZ QSIP Endnotes. Page 6 of 15
-------
altered the plan to administer the Baseline questionnaire only during Stage 2 ,3, 6
scheduled sample visits and by mail-in return.
51. (QSIP page 80) We originally envisioned three household visits within one week for
each Stage 3 household. We found that phone calls and special instructions could
eliminate the need for the midweek visit, thus reducing travel expenses considerably.
These changes were discussed and approved by our EPA Project Officer or Project
Collaborator. Subjects were coached over the phone to cap air impactors samplers
and place them in the refrigerator, freeze urine samples within special containers and
keep collected food in sealed containers within the refrigerator for pick-up during the
tear down visit at the end of each week.
52. (QSIP page 84) Depending on field progress it was not always practical to ship all
samples within 7 days. We carefully attended to hold times and tried to ship within
hold times. Instances of hold time violation are documented on Chain of Custody
forms and flagged within the databases.
53. (QSIP page 84) An overall Data Quality Indicator is placed in each analytical data
base. This is a simple number indicating fitness of the data point as a function of
field, lab and data protocols. 1 = valid, 2 = questionable, minor deviation from SOP
(e.g. soil from only 3 sides of a house due to construction parameters), 3 = invalid
(e.g. issue with calibration, sample contaminated),
54. (QSIP page 94) Initially all field personnel were completely cross trained. All Field
Team Leaders were completely cross trained throughout the project. By Spring of
1995, we began specialized training. Individuals were trained at everything, but
specialized at the collection of specific media. Training was sufficient for
verification of field readings.
55. (QSIP page 54) By the end of the first month in the field, complex QA checks on
forms could not be completed by the Project Field Coordinator. Instead, the Field
Coordinator debriefed each field team after every household visit and the Field Team
Leader performed QA checks in a timely fashion. Additional checks were performed
by a trained assistant to the Project Field Coordinator prior to transfer of forms to the
data entry section of the project.
56. (QSIP page 96) Section 3.3.3 A, All changes were discussed with the EPA Project
Officer or Project Collaborator and made with his concurrence.
57. (QSIP page 97) Section 3.3.3 E. This function was performed by the Project Field
Coordinator not by the Project Data Coordinator. We found that the field coordinator
had a clearer view of what was collected. This change was made in March of 1995.
58. (QSIP page 97) Section 3.3.3 J. The Project Field Coordinator retrieved data from
the Project Data Coordinator and maintained the hard copy records (physical Forms).
This change was made in March of 1995.
59. (QSIP page 103) Specific non conformance issues were recorded in the daily field
log, communicated verbally and communicated through e-mail. All problems were
documented on field or data sheets.
60. (QSIP page 105) Final Project Data Files have been transferred to the PI on CD-
ROM.
61. (QSIP page 109) Descriptive Questionnaires were readministered in the field for all
Stage 2, 3, 6 homes and serve as the 10% QA check on recruitment.
NHEXAS AZ QSIP Endnotes. Page 7 of 15
-------
62. (QSIP page 112) EPA pioneered a new concept of having Project Collaborators work
with the investigators on the project. They in turn reported to the Project officer. The
Arizona had a shift in Project Collaborators from Timothy Buckley to Mr. Gary
Robertson. Near the end of the project the positions were merged and Mr, Robertson
was both the Collaborator and the Project Officer. The EPA contact called weekly
for updates and visited annually.
63. (QSIP-Appendix A Revised)
VOLUME I. GENERAL PROCEDURES
Preparation of Standard Operating Procedures (SOPs) UA-G-1.1
Isolation of Malfunctioning or Damaged Equipment UA-G-2.0
Assurance of Respondent Confidentiality UA-G-3.1
Archive Procedure for Study Samples UA-G-4.0
Receipt of Equipment and Implementation Supplies UA-G-5.0
Procedure for the Transmittal of Sampling Materials from Battelle BCO-G-l.O
Procedure for the Receipt of Samples or
Reference Compounds for Laboratory Analysis at Battelle BCO-G-2.1
Verification and Transfer of Data to UA BCO-G-3.1
TRAINING PROCEDURES
Training Guide and Exposition of
Interviewer Responsibilities and Behavior UA-T-1.0
Administering Field Qx— General UA-T-2.0
Field Personnel Training— General UA-T-3.0
Field Personnel Training Plan UA-T-4.0
Student Data Assistant Training Plan UA-T-5.0
Laboratory Assistant Training Plan—General (UA) UA-T-6.0
General Laboratory Training Plan— Battelle BCO-T-1.0
NHEXAS AZ QSIP Endnotes. Page 8 of 15
-------
DATA CUSTODY PROCEDURES
Custody of Field Samples
Form QA/QC Checks
Packet Assembly, Custody and Transfer to Data Section
Batching of Field Data Forms
Flow and Custody of Field Data Forms
Keypunch Tracking, Custody & Data Transfer
Batching of Lab Data
Flow & Custody of UA Laboratory Data
UA-C-1.0
UA-C-2.0
UA-C-3.0
UA-C-4.0
UA-C-5.0
UA-C-6.G
UA-C-7.0
UA-C-8.0
VOLUME II. FIELD PROCEDURES
HHID & IRN Creation, Use and Assignment
Preparation of Field Bucket & Equipment Assignment
Collection of Airborne Particulate
Samples for Metals and Pesticides Analysis
Operation, Calibration & Maintenance of the
Sentex Scentogun Portable Photoionization Detector
Field Collection of Yard Composite Soil Samples
Field Collection of Residential Composite Foundation Soil Samples
Field Collection and Post Field Collection
Sample Handling of Indoor Floor Dust Samples
Collection of Surface Wipe Samples for Pesticides or Metals Analysis
Collection of Dermal Wipe Samples for Pesticides or Metals Analysis
Field Use of the Modified Air Sentinel
Use of an Active Sampling Device for the
Collection of Airborne VOCs at Fixed Indoor & Outdoor Sites
UA-F-1.0
UA-F-2.1
UA-F-3.1
UA-F-4.1
UA-F-5.1
UA-F-6.1
UA-F-7.1
UA-F-8.1
UA-F-9.1
UA-F-10.1
UA-F-11.1
NHEXAS AZ QSIP Endnotes. Page 9 of 15
-------
Use of a Passive Sampling Device for the
Collection of Airborne VOCs at Fixed Indoor & Outdoor Sites UA-F-12.0
Use of a Passive Sampling Device for the
Collection of Formaldehyde at Fixed Indoor and Outdoor Sites UA-F-13.0
Collection of Personal Air Samples for Analysis of Pesticides or Metals UA-F-14.1
Collection of Food & Beverage Samples UA-F-15.1
Collection, Storage and Shipment of Drinking and
Tap Water Samples for Trace Metals (EPA Method 200.8) UA-F-16.1
Collection, Storage & Shipment of Drinking &Tap Water
Samples for Pesticides (EPA 225.2) & Carbaryl Analysis (EPA 531.1) UA-F-17.1
Collection, Storage and Shipment of Drinking and
TapWater Samples for VOC Analysis (EPA Method 524.2) UA-F-18.1
Collection, Storage and Shipment of
Blood Samples for Selected Metals, Pesticides and VOC Analysis UA-F-19.1
Collection, Storage and Shipment of
Urine Samples for Selected Metals and Pesticides Analysis UA-F-20.1
Thin Film Collection of Dust and Soil for Metals Analysis UA-F-21.1
Operation & Initialization of the Magellan GPS Satellite Navigator UA-F-22.1
Reimbursement for Field Collection of Diet Samples UA-F-23.1
VOLUME III. LABORATORY PROCEDURES
Volume IIIA. Laboratory Procedures (Arizona)
Calibration and Operation of Balances UA-L-1.2
Weight Room Operation & Maintenance UA-L-2.1
Still Operation and Maintenance UA-L-3.1
Maintenance & Temperature Verification of
Refrigerated Units for Sample Storage UA-L-4.1
Standard Protocol for Cleaning Laboratory and Field Sampling Apparatus UA-L-5.1
NHEXAS AZ QSIP Endnotes. Page 10 of 15
-------
Calibration of Harvard PM Samplers
Harvard PM Impactor Calibration and Leak Testing
Preparation of PM & URG Impactors & Impaction Plates
Filter Weighing
Operation, Calibration & Routine Use of the
Speetrace9000 Field Portable X-Ray Fluorescence Analyzer
Soil Characterization
Vacuum Dust Characterization
Food Sample Comparison with Diary & Shipment
Volume IIIB. Laboratory Procedures (Battelle)
UA-L-6.1
UA-L-7.1
UA-L-8.1
UA-L-9.1
UA-L-10.1
UA-L-11.1
UA-L-12.1
UA-L-13.1
Analysis of Soil or House Dust
Samples using Chlorpyrifos ELISA Samples
Preparation of filters & PUF for
Field Collection of Metals and Pesticides in Air
Extraction of Metals from Soil, Dust, Air Filter, and Surface & Dermal
Samples for AA (Graphite Furnace or Flame) or ICP-AES Analysis
BCO-L-4.0
Operation, Calibration and Maintenance
of the Perkin-Elmer 1100B Atomic Absorption Spectrometer
Operation, Calibration and Maintenance of the
Perkin-Elmer Zeeman/5000 System Atomic Absorption Spectrometer
Operation, Calibration and Maintenance of the Jobin-Yvon
Model 70 Inductively Coupled Plasma Atomic Absorption Spectrometer
Operation, Calibration and Maintenance of the Thermo
Jarre 11ICAP 61-975 Plasma AtomComp Emission Spectrometer
Operation, Calibration and Maintenance
of Fixed and Adjustable Volume Pipette Guns
BCO-L-l.O
BCO-L-2.0
BCO-L-3.1
BCO-L-5.1
BCO-L-6.0
BCO-L-7,1
BCO-L-8.0
BCO-L-9.0
NHEXAS AZ QSIP Endnotes. Page 11 of 15
-------
Procedures for Cleaning Glassware
to be used for Inorganic Metals Analysis BCO-L-10.0
Extraction of Air samples for GC/MS Analysis of Pesticides BCO-L-11.1
Extraction of Pesticides from Dermal Wipe Samples BCO-L-12.1
Extraction of Pesticides from Surface Wipe Samples BCO-L-13.0
Extraction of Soil/House Dust for GC/MS Analysis of Pesticides BCO-L-14.0
Analysis of Pesticide Samples by GC/MS BCO-L-15.1
Analysis of Passive Formaldehyde Samplers BCO-L-16.0
Analysis of Volatile Organic Compounds Collected with a Passive Sampler BCO-L-17.1
Preparation of Multisorbent Tubes for Actively-Pumped VOC Samplers BCO-L-18.1
BCO-L-J 9.0
BCO-L-20,0
Preparation of Calibration & Surrogate
Recovery Solutions for GC/MS Analysis of Pesticides BCO-L-21.1
Analysis of 3-Phase Multisorbent Samplers for VOCs BCO-L-22.1
Calibration, Maintenance and Operation of Electronic Balances BCO-L-23.0
Analysis of Pesticide Samples by GC/ECD BCO-L-24.0
Freezer Maintenance & Temperature Verification for Sample Integrity BCO-L-25,0
VOLUME IV. DATA PROCEDURES
Operation and Maintenance of the LAN &
Related Microcomputer Environment UA-D-1.1
Performance of Computer Software: Verification and Validation UA-D-2.0
Defining Working Databases and Data Entry Forms (Hand Entry) UA-D-3.0
The Generation of Data Dictionaries UA-D-4.0
NHEXAS AZ QSIP Endnotes. Page 12 of 15
-------
Global Coding used by NHEXAS Arizona (Hand Entry)
UA-D-5,0
Coding: Descriptive Questionnaire
UA-D-6.0
Coding: Baseline Questionnaire (Household)
UA-D-7.0
Baseline Questionnaire -Individual
UA-D-8.0
Coding: Time Diary & Activity Questionnaire
UA-D-9,0
Coding: Food Diary Follow Up
UA-D-10.0
Coding: Follow Up Questionnaire
UA-D-11.0
Weekly Activity Follow up-Individual
UA-D-12.0
Coding: Arizona Lab Data
UA-D-13.0
Coding & Coding Verification (Hand Entry)
UA-D-14.0
Data Entry and Data Verification (Hand Entry)
UA-D-15.0
First Stage of Cleaning Electronic Data (Hand Entry)]
UA-D-16.0
Cleaning: Descriptive Questionnaire
UA-D-17.0
Cleaning: Baseline Questionnaire
UA-D-18.0
Cleaning: Baseline Questionnaire -Individual
UA-D-19.0
Cleaning: Time Diary & Activity Questionnaire
UA-D-20.0
Cleaning: Food Diary Follow Up
UA-D-21.0
Cleaning: Follow Up Questionnaire
UA-D-22.0
Cleaning Weekly Activity Follow up—Individual
UA-D-23.0
Cleaning: Arizona Lab Data
UA-D-24.0
Correcting Electronic Data (Hand Entry & Scanned)
UA-D-25.0
Electronic Data QA Check (Hand Entry & Scanned)
UA-D-26.0
NHEXAS AZ QSIP Endnotes. Page 13 of 15
-------
Instructions for the Addition of Individual
Cleaned Non Scanned Data Batches to Master Databases UA-D-27.0
Tracking System UA-D-28.0
Transfer and Receipt Data UA-D-29.0
Scanable Form and Data Base Definition UA-D-30.1
Global Coding for Scanned Forms UA-D-31.1
Operation and Maintenance of Data Scanners UA-D-32.1
Light Pen Operation & Verification of Scanned Bar Codes UA-D-33.0
Scanning & Verifying Forms and Questionnaires UA-D-34.0
Coding: Technician Walk-Through Questionnaire UA-D-35.0
Cleaning: Technician Walk-Through Questionnaire UA-D-36.0
Coding: Field Forms UA-D-37.0
Cleaning: Field Forms UA-D-38.0
Coding: 24 Hour Food Diary
(Hand Entry) Replaced By UA-D-43.0 UA-D-39.0
Cleaning: 24 Hour Food Diary
(Hand Entry) replaced by UA-D-45.0 UA-D-40.0
Database Tree & Data Sources UA-D-41.0
Instructions on the Completion of Scanable Forms UA-D-42.1
Operation Manual of the Mass Data Massage Program UA-D-44.G
Coding: Questionnaire Feedback Form UA-D-46.0
Cleaning: Questionnaire Feedback Form UA-D-47.0
NHEXAS AZ QSIP Endnotes. Page 14 of 15
-------
VOLUME V. ANALYSIS: STATISTICAL AND EXPOSURE ASSESSMENT
Operation and Maintenance of Computer Equipment for IIT NHEXAS IIT-A-1.0
Performance of Computer Software: Verification & Validation IIT-A-2.0
Performance of Analyses on NHEXAS Data IIT-A-3.0
Standard Operating Procedure for treatment of censored data IIT-A-4.0
Standard Operating Procedure for IIT-A-5.0
Estimating Inhalation Exposures to Chlorpyrifos and Diazinon
Standard Operating Procedure for Calculating Ingestions Exposure from
Day 4 composite measurements, the direct method of exposure estimation IIT-A-6.0
Standard Operating Procedure for
Calculating Ingestion Exposure Estimating Ingestion
Exposure, The Indirect method of Exposure Estimation IIT-A-7.0
Standard Operating Procedure for Conversion of Servings to Kilograms IIT-A-8.0
Standard Operating Procedure for Sampling Weight Calculation IIT-A-9.0
Standard Operating Procedure for Sampling Weight Adjustment IIT-A-10.0
Standard Operating Procedure for Dermal Exposure Calculation IIT-A-11.0
Standard Operating Procedure for Time Activity Calculation IIT-A-12.0
Standard Operating Procedure for Probabilistic
Approach of Exposure Calculation of Dermal Exposure IIT-A-13.0
Standard Operating Procedure for Probabilistic Approach
for Estimating Inhalation Exposures to Chlorpyrifos and Diazinon IIT-A-14.0
Standard Operating Procedure for Probabilistic
Approach for Calculating Ingestion Exposure from Day 4
Composite Measurements, The Direct Method of Exposure Estimation IIT-A-15.0
NHEXAS AZ QSIP Endnotes. Page 15 of 15
------- |