&EPA
United States
Environmental Protection
Agency
Office of Water
4303T
Washington, DC 20460
EPA-823-R-05-005
April 2005
Quality Assurance Report for the National
Study of Chemical Residues
in Lake Fish Tissue:
Analytical Data for Years 1 through 4
Prepared/or:
U.S. Environmental Protection Agency
Office of Water
Office of Science and Technology
Prepared by:
CSC Environmental Programs Group
Prepared under:
MOBIS Contract No. GS-23F-9820, Task 68-C-00-137
April 2005
-------
-------
Quality Assurance Report
for the
National Study of Chemical Residues in Lake Fish Tissue:
Year 1 through Year 4 Analytical Data
April 2005
Prepared for
United States Environmental Protection Agency
Office of Water
Office of Science and Technology
Prepared by
CSC Environmental Programs Group
under
MOBIS Contract No. GS-23F-9820, Task 68-C-00-137
-------
-------
TABLE OF CONTENTS
Page
Chapter 1 Introduction 1
Section 1.1 Description of Study Objectives and Study Design 1
Section 1.2 Study Participants 3
Section 1.3 References 3
Chapter 2 Q A Program 5
Section 2.1 Collaborative Study Design 5
Section 2.2 Implementation of Approved Quality Assurance Project Plans for
Sampling and Analysis Activities 5
Section 2.3 Implementation of a Field Orientation/Training Program 6
Section 2.4 Development of Study-Specific Sample Documentation and
Sampling Kits 6
Section 2.5 Daily Monitoring of Sampling and Laboratory Activities 6
Section 2.6 Weekly to Monthly Reporting of Project Status as Appropriate 7
Section 2.7 Monthly Project Meetings among EPA Headquarters Staff and
Contractors Responsible for Coordination of Activities 7
Section 2.8 Implementation of Thoroughly Documented Methods That
Included All QC Elements Needed to Support Quality Objectives
Established for the Study 7
Section 2.9 Three Levels of Data Quality Assessment and Application of
Standardized Data Qualifiers 8
Section 2.10 Implementation of Standardized Data Format to Allow All
Results to be Reported Consistently and Accurately to Data Users ... 8
Chapter 3 Analytical Methods Employed 9
Section 3.1 Sample Prep Lab Procedures 9
Section 3.2 Dioxins/Furans 10
Section 3.3 Polychlorinated Biphenyls (PCBs) 11
Section 3.4 Total Mercury 11
Section 3.5 Arsenic Species 11
Section 3.6 Organochlorine Pesticides 12
Section 3.7 Organophosphorus Pesticides 12
Section 3.8 Semivolatile Organics 12
Chapter 4 Data Quality Assessment 15
Section 4.1 Pre-Qualification Review 15
Section 4.2 Individual Data Package Review 19
Section 4.3 Overall Data Quality Assessment 30
QA Report for the National Lake Fish Tissue Study
April 2005
-------
LIST OF EXHIBITS
Exhibit 1-1 Study Design Highlights 2
Exhibit 1-2 Overview of Study Participants 3
Exhibit 3-1 National Lake Fish Tissue Study Target Analytes and Corresponding
Analysis Methods 13
Exhibit 4-1A Summary (Target Analytes) of Data Review Qualifiers Used for Year 1
through Year 4 of the National Lake Fish Tissue Study 34
Exhibit 4-IB Summary (Non-Target Analytes) of Data Review Qualifiers Used for Year 1
through Year 4 of the National Lake Fish Tissue Study 48
April 2005 ii QA Report for the 'National Lake Fish Tissue Study
-------
Chapter 1
Introduction
This report documents the quality of data gathered during the National Study of Chemical
Residues in Lake Fish Tissue (hereafter referred to as either the "National Lake Fish Tissue
Study" or, more simply, "the Study." Please note that the abbreviated name for the study
changed from "National Fish Tissue Study" during 2004.). For reference purposes, this report
provides a brief overview of the study and outlines the primary participants and the analytical
parameters measured. Additional details concerning the design and implementation of this study
can be found in the references listed at the end of this chapter.
SECTION 1.1 DESCRIPTION OF STUDY OBJECTIVES AND STUDY DESIGN
The National Lake Fish Tissue Study is a screening-level study designed to estimate the national
distribution of selected persistent, bioaccumulative, and toxic (PBT) chemical concentrations in
fish tissue from lakes and reservoirs of the contiguous United States. The study involves the
collection of predator and bottom-dwelling fish from 500 randomly selected lakes and reservoirs
in the lower 48 states (excluding the Great Lakes) over a period of 4 years (-125 lakes per year).
The study began during the fall of 1999; however, full implementation did not commence until
2000. For this reason, tissue samples collected during the 1999 mobilization and 2000
implementation periods cumulatively represent Year 1 of the Study.
The study design resulted from a comprehensive planning effort that included a national
workshop involving more than 50 scientists from state, federal, and tribal agencies to obtain
technical input on sampling design, target analytes, sampling methods and data management.
The final study design is described in Reference 1 of this chapter and highlighted in Exhibit 1-1.
Implementation of the study is a collaborative effort being led by the Environmental Protection
Agency's (EPA) Office of Science and Technology (OST), within the Office of Water (OW),
with extensive support from participating states and tribes, EPA Regions, and the Office of
Research and Development's (ORD) Environmental Monitoring and Assessment Program
(EMAP). States, Tribes, and EPA Regional staff collected most of the fish for the study.
Contractor support is being provided by Tetra Tech (for study design, orientation/training,
workshops, field sampling activities, data compilation, and project reporting), CSC
Environmental Programs Group ([formerly DynCorp Environmental] for sampling kits, sample
coordination and tracking, data review and reporting, and database development), and the
following laboratories: Axys Analytical Services in Sydney, British Columbia, Canada; Battelle
Ocean Sciences in Duxbury, MA; Battelle Marine Sciences in Sequim, WA; and Pacific
Analytical Inc., in Carlsbad, CA (see Exhibit 1-2, Overview of Study Participants.)
The 500 lakes sampled for the study were statistically selected to meet study objectives. To
allow study objectives to be met in the event that all four years of sampling could not be
completed, the lakes were statistically sub-sampled and classified as Year 1, Year 2, Year 3, or
Year 4 lakes. These statistical designations did not always conform to resources available at the
state level in the assigned years. Therefore, some states requested and received permission to
collect samples from a lake in a year other than the year that the lake was designated to be
sampled. Consequently, it is important to note that the Year 1, Year 2, Year 3, and Year 4 data
QA Report for the National Lake Fish Tissue Study 1 April 2005
-------
sets described in this report do not correspond to the annual statistical subsets of lakes. To
distinguish the annual statistical subsets from the lakes actually sampled during the first through
the fourth years of the study, they are now being referred to as Panel 1, Panel 2, Panel 3, and
Panel 4 lakes. The 500 lakes sampled during the four years of the study consisted of 120 Panel 1
lakes, 109 Panel 2 lakes, 105 Panel 3 lakes, 109 Panel 4 lakes, and 57 lakes from a set of
statistically drawn "reserve" lakes.
Exhibit 1-1 Study Design Highlights
Objective
Estimate the national distribution of selected persistent, bioaccumulative, and toxic (PBT) chemical concentrations in fish
tissue from lakes and reservoirs of the contiguous U.S.
Sample Sources
500 randomly selected lakes and reservoirs (each defined as a permanent body of water having at least one hectare in
surface area with a minimum of 1,000 m2 of open (unvegetated) water, a minimum depth of one meter, and a permanent fish
population)
Sample Types
Edible tissue (i.e., skin-on fillet) composites of targeted predator species
Total body tissue (i.e., whole fish) composites of targeted bottom-dwelling species
Composite Definition
Five individual fish of the same species that:
Satisfy legal requirements of harvestable size or weight (or are of consumable size if no legal harvest requirements are in
effect)
Are of similar size so that the smallest individual within the composite is no less than 75% of the total length of the largest
individual
Are collected at the "same" time (i.e., as close to the same time as possible but no more than 1 week apart)
Are of adequate size to allow analysis of target study analytes
Target Species
Selected in accordance with EPA's Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories, Volume I:
Fish Sampling and Analysis, Third Edition (U.S. EPA 2000), with preferences as cited in References 1, 2, and 3 of this
chapter.
Field QC
Development and implementation of the Sample Collection Activities QAPP (Reference 2), Field Sampling Plan
(Reference 3), and SOPs (References 2 and 3)
Collection of replicate fish samples from 10% of the lakes to estimate sampling variability
Use of experienced fisheries biologists to ensure use of proper procedures
Distribution of standardized sampling kits to control contamination and ensure proper documentation
Daily tracking and coordination of sample shipments through a centralized source
Regional orientation/training workshops to ensure all field personnel understood objectives and design of study and to
ensure consistent application of required sample collection, handling, and shipping procedures
Field audits to ensure consistent sample collection, handling, and shipping procedures
April 2005 2 QA Report for the 'National Lake Fish Tissue Study
-------
Exhibit 1-1 Study Design Highlights
Laboratory QC
Development and implementation of the Analytical Control and Assessment Activities QAPP (Reference 4)
Use of centralized Sample Prep Laboratory to minimize variability during sample grinding, homogenizing, and compositing
Identification of quantifiable measurement quality objectives (MQOs)
Implementation of standardized sample tracking, lab analysis, data reporting, and data review procedures
Use of pure and traceable reference standards
Demonstration of instrument calibration and system performance
Periodic calibration verification
Verification that each laboratory could achieve the required detection and quantitation levels
Analysis of initial and ongoing QC samples to demonstrate each laboratory's ability to achieve precise and accurate
results with the method
Analysis of blanks to demonstrate the absence of contamination
SECTION 1.2 STUDY PARTICIPANTS
A detailed list of participants referenced in this report is provided in Reference 3 of this chapter.
Exhibit 1-2 below identifies the parties cited in this report.
Exhibit 1-2 Overview of Study Participants
Study Participants
EPA Office of Science and Technology:
Leanne Stahl, National Study Manager, Field Support Manager
Cynthia Simbanin, Analytical Project Manager
Sample Control Contractor (SCO: CSC Environmental Programs Group
(formerly DynCorp Environmental)
Study Design and Field Support Contractor: Tetra Tech, Inc.
Sampling Teams: State, Tribal, and Regional contacts listed in Reference 3
Sample Prep Laboratory: Axys Analytical Services
Analysis Laboratories: Axys Analytical Services (PCBs, Dioxins/Furans)
Battelle Ocean Sciences (Semivolatile Organics)
Battelle Marine Sciences (Mercury and Arsenic)
Pacific Analytical Inc. (Pesticides)
SECTION 1.3 REFERENCES
Additional information regarding the design and implementation of this study can be found in
the following references:
(1) U.S. Environmental Protection Agency. 1999. National Study of Chemical Residues in Lake
Fish Tissue: Study Design. Prepared by Tetra Tech, Inc. for USEPA, OW/OST, Washington,
DC.
QA Report for the National Lake Fish Tissue Study 3 April 2005
-------
(2) U.S. Environmental Protection Agency. 2000. Quality Assurance Project Plan for Sample
Collection Activities for a National Study of Chemical Residues in Lake Fish Tissue.
Prepared by Tetra Tech, Inc., for USEPA, OW/OST, Washington, DC.
(3) U.S. Environmental Protection Agency. 2000. Field Sampling Plan for the National Study
of Chemical Residues in Lake Fish Tissue. Prepared by Tetra Tech, Inc. for USEPA,
OW/OST, Washington, DC.
(4) U.S. Environmental Protection Agency. 2000. Quality Assurance Project Plan for
Analytical Control and Assessment Activities in the National Study of Chemical Residues in
Lake Fish Tissue. Prepared by CSC Environmental Programs Group (formerly DynCorp
Environmental) for USEPA, OW/OST, Washington, DC.
The first three documents are available from Leanne Stahl, National Study Manager, and the
final document is available from Cynthia Simbanin, Analytical Project Manager, at the following
addresses:
Leanne Stahl
USEPA/OST (4305T)
1200 Pennsylvania Avenue, NW
Washington, DC 20460
stahl.leanne@epa.gov
Cynthia Simbanin
USEPA/OST (4303T)
1200 Pennsylvania Avenue, NW
Washington, DC 20460
simbanin.cynthia@epa.gov
April 2005
4 QA Report for the 'National Lake Fish Tissue Study
-------
Chapter 2
QA Program
At the outset of the National Lake Fish Tissue Study, EPA managers recognized that data
gathered from the study would be used extensively by individuals responsible for making
environmental, economic, and policy decisions. Environmental measurements always contain
some level of uncertainty. Decision makers, therefore, must recognize the uncertainty associated
with the data on which their decisions are based. In recognition of this, the study managers
established a quality assurance (QA) program intended to ensure that data produced under the
National Lake Fish Tissue Study would meet defined standards of quality within a specified
level of confidence (see Chapter 1, References 2 and 4).
Implementation of the QA Program ensured that all Measurement Quality Objectives
(AAQOs) were met in Year 1, Year 2, Year 3, and Year 4 of the Study and that not a
single sample was lost or misidentif ied.
The study QA program prescribes minimum requirements to which all organizations that gather
data are required to adhere. All of these elements were followed during each year of the study,
and data quality was defined, controlled, and assessed through these QA program activities. The
remainder of this chapter highlights the QA program employed during the study.
SECTION 2.1 COLLABORATIVE STUDY DESIGN
Development of the study design was a collaborative effort among EPA's Office of Water (OW),
Office of Research and Development (ORD), and Office of Prevention, Pesticides, and Toxic
Substances (OPPTS), with significant involvement by biologists, chemists, and statisticians in
OW's Office of Science and Technology (OST), statisticians in ORD's Environmental
Monitoring Assessment Program (EMAP), and chemists in OPPTS. The draft design was
reviewed by experts throughout federal, state, and tribal organizations (including EPA, the
National Oceanic and Atmospheric Administration (NOAA), U.S. Geological Survey (USGS),
and state and tribal environmental, wildlife, and fisheries management agencies) via a two-day
workshop. Input obtained during this workshop was incorporated into the final study design.
SECTION 2.2 IMPLEMENTATION OF APPROVED QUALITY ASSURANCE PROJECT
PLANS FOR SAMPLING AND ANALYSIS ACTIVITIES
Two Quality Assurance Project Plans (QAPPs) were developed and approved by EPA to support
this study. The Quality Assurance Project Plan for Sample Collection Activities for a National
Study of Chemical Residues in Lake Fish Tissue (May 2000) establishes data quality goals for all
sample collection and handling activities and describes the QA/Quality Control (QC) techniques
employed by field teams and the field support contractor to support those goals. The Quality
Assurance Project Plan for Analytical Control and Assessment Activities in the National Study of
Chemical Residues in Lake Fish Tissue (September 2000) establishes MQOs for laboratory data
generated during the study and describes QA/QC techniques employed by laboratory and sample
control contractor (SCC) staff to ensure these goals are met.
QA Report for the National Lake Fish Tissue Study 5 April 2005
-------
SECTION 2.3 IMPLEMENTATION OF A FIELD ORIENTATION/TRAINING PROGRAM
Because the study design relied on a large number of state, tribal, and federal sampling teams,
EPA established a field orientation and training program to ensure that personnel responsible for
sampling activities within each organization understood the study objectives, were familiar with
paperwork developed specifically to document sample collection activities under the study, and
were prepared to collect, document, and ship samples in accordance with the standard operating
procedures (SOPs) and the sample collection QAPP.
SECTION 2.4 DEVELOPMENT OF STUDY-SPECIFIC SAMPLE DOCUMENTATION AND
SAMPLING KITS
The study design calls for collection offish samples by multiple teams from participating states,
tribes, and EPA Regions. To ensure samples were consistently documented by such a large and
diverse group, several documentation materials were custom-designed for the study. These
forms include a:
Field Record Form to document information about each lake sampled and individual
specimens collected from the lake,
Sample Identification Label to accompany and identify each fish specimen,
Chain-of-Custody Form to provide constant tracking information for all samples, and
Chain-of-Custody Label to seal each shipping container.
These forms were being provided annually in custom-made sampling kits prepared by the SCC.
The kits also contained contaminant-free materials needed to store each specimen (i.e., solvent-
rinsed aluminum foil and food grade polyethylene tubing), a reference instruction sheet with
contact phone numbers, and pre-completed forms needed to ship the specimens to the Sample
Prep Laboratory for homogenization and compositing. In addition, sample Traffic Reports were
created for use by the Sample Prep Laboratory to document each homogenized composite aliquot
that was sent to either an Analysis Lab or to the Sample Repository for long-term storage.
Implementation of these tools in all four sampling years of the study (2000-2003) was highly
successful.
SECTION 2.5 DAILY MONITORING OF SAMPLING AND LABORATORY ACTIVITIES
To ensure effective communication among all organizations involved in the study, the field
support and SCC contractors were tasked with establishing and implementing a series of
procedures for the following activities: preparing and distributing sampling kits; coordinating
and tracking sample shipments; identifying corrective actions in the event of lost shipments;
reviewing Field Record Forms to identify and notify EPA offish samples that deviated from the
sample criteria and recommending corrective actions; obtaining laboratory analyses; reviewing
laboratory data; and generating a STORET-compatible database of study results. These
activities were highly successful in controlling the quality of data during all four years of the
study. Not a single sample was lost during the study, and all potential deviations from the study
design were mitigated by these early identification techniques.
April 2005 6 QA Report for the 'National Lake Fish Tissue Study
-------
SECTION 2.6 WEEKLY TO MONTHLY REPORTING OF PROJECT STATUS AS
APPROPRIATE
Each of the contractors routinely reported the status of project activities to EPA so that the
National Study Manager could monitor study progress (e.g., samples collected, samples
analyzed, etc.), notify senior EPA management of potential problems and success stories, and
communicate project status to other organizations supporting the study. Such communications
provided a real time means through which the study manager could notify study participants of
important issues (e.g., the need to halt sample shipments until air traffic returned to normal
following the 2001 terrorist attacks; clarifications concerning the amount of dry ice needed when
shipping coolers; and the need for alternate documentation procedures to streamline the shipment
of samples from the field through Customs to the Sample Prep Laboratory).
SECTION 2.7 MONTHLY PROJECT MEETINGS AMONG EPA HEADQUARTERS STAFF
AND CONTRACTORS RESPONSIBLE FOR COORDINATION OF ACTIVITIES
During Year 1 of the study, the National Study Manager held monthly meetings with EAD staff
responsible for managing laboratory and data review activities and with each team of contractors
responsible for daily tracking of activities. With a full year of study experience behind the
teams, the frequency of these meetings was reduced to an as-needed basis in the remaining years
of the study. The purpose of each meeting was to review study progress, discuss upcoming
schedules, and identify and resolve issues. Depending on project activities, additional staff were
brought into these meetings to facilitate planning and resolve issues. Notably, OW
representatives responsible for developing and managing STORET were invited to meetings that
included discussions of database designs, STORET data upload, and data distribution plans.
Likewise, EAD statisticians responsible for interpreting study results were invited to participate
in discussions of the procedures used to review, qualify, and report laboratory results.
SECTION 2.8 IMPLEMENTATION OF THOROUGHLY DOCUMENTED METHODS THAT
INCLUDED ALL QC ELEMENTS NEEDED TO SUPPORT QUALITY
OBJECTIVES ESTABLISHED FOR THE STUDY
A suite of EPA 1600-series methods was employed to support the study. Each participating
laboratory, including the Sample Prep Laboratory, was required to demonstrate their ability to
practice these methods before preparing or analyzing samples collected in the study. Chapter 3
describes these methods in detail.
QA Report for the National Lake Fish Tissue Study 7 April 2005
-------
SECTION 2.9 THREE LEVELS OF DATA QUALITY ASSESSMENT AND APPLICATION OF
STANDARDIZED DATA QUALIFIERS
All analytical data generated during the study were subjected to three levels of review:
A pre-qualification review was performed on data submitted by each laboratory to
demonstrate that the labs were qualified to prepare and/or analyze tissue samples collected
during the study.
Each submission of tissue sample results was carefully scrutinized to verify that the samples
were analyzed as directed and that supporting QC results demonstrated the quality of results
generated. In evaluating these submissions, data reviewers employed a suite of standardized
data qualifiers and abbreviated qualifier codes to consistently and accurately document the
quality of all data generated so that both the primary data users (statisticians) at EPA
Headquarters and secondary data users within states, tribes, and other organizations could
make informed decisions regarding their use.
A third level of data review was performed at the conclusion of each of the four annual data
review processes to determine if overall data quality supported study objectives. These end-
of-year evaluations indicated that all MQOs were met for all four years of the study.
Chapter 4 describes the data quality assessment procedures employed in the study.
SECTION 2.10 IMPLEMENTATION OF STANDARDIZED DATA FORMAT TO ALLOW ALL
RESULTS TO BE REPORTED CONSISTENTLY AND ACCURATELY TO DATA
USERS
All data generated during the study are being compiled in a centralized, custom-developed
database designed for the following: eventual upload of results to the national STORET
database system, statistical manipulation of results, export of results to user-friendly formats
such as Excel spreadsheets, and consistency in data format and nomenclature across laboratories
and over time.
April 2005 8 QA Report for the 'National Lake Fish Tissue Study
-------
Chapter 3
Analytical Methods Employed
To control variability among tissue sample results, all samples collected during the study were
analyzed by a single set of methods, and all analyses performed with a given method were
performed by only one laboratory. Further control of variability was ensured by utilizing a
single laboratory to prepare, composite, homogenize, and aliquot samples in a strictly controlled,
contaminant-free environment. The methods employed by the Sample Prep Laboratory and by
each Analysis Laboratory are described below. A complete list of the chemicals measured by
these laboratories is provided at the end of this chapter in Exhibit 3-1.
SECTION 3.1 SAMPLE PREP LAB PROCEDURES
Each composite tissue sample prepared by the Sample Prep Lab consisted of five individual fish
(where available) of a single species. Bottom-dwelling species were prepared as whole fish
composites (i.e., the entire specimen, including the head, skin, internal organs, muscle, and
bones were thoroughly homogenized). Predator/gamefish were prepared as skin-on (scales
removed) fillet composites. All tissue sample preparation, filleting, and homogenization
activities were performed in accordance with EPA's Guidance for Assessing Chemical
Contaminant Data for Use in Fish Advisories, Volume 1: Fish Sampling and Analysis, Third
Edition, November 2000 (the "Fish Advisory Guidance") in a strictly controlled, contaminant-
free environment. Equipment rinsates were collected during each day of sample preparation
activities and analyzed to document the absence of contamination. Upon receipt of samples from
the field, the Sample Prep Lab:
Checked that each shipping container arrived undamaged and verified that samples were still
frozen and in good condition upon receipt. All samples collected and shipped throughout the
four years were received frozen. Fish composite samples that did not conform to established
study criteria (i.e., "nonroutine" composite samples) were frozen pending a determination by
EPA concerning procedures for analysis. EPA documented all processing and analysis
decisions for nonroutine fish composite samples (e.g., accepting composites with fewer or
greater than 5 specimens), and these decisions were entered into the study database.
Forwarded all associated paperwork to the field support contractor for full verification of
completeness and accuracy. All QA problems were successfully resolved and reported back
to the SCC for entry of field data into the database.
Verified that all specimens listed on the paperwork for each composite were included in the
shipment and were properly wrapped and labeled. Deviations were rare, and they were
immediately reported to EPA (via SCC) for resolution and noted in the study database.
Signed the chain-of-custody form and forwarded it to the Field Support Contractor with a
copy to the SCC.
Documented information about each specimen in a laboratory notebook.
QA Report for the National Lake Fish Tissue Study 9 April 2005
-------
All grinding and homogenization activities were performed in accordance with Section 7.2.2.9 of
the Fish Advisory Guidance (for fillet composites) and Appendix G of the guidance (for whole
fish composites) with the following exceptions:
The laboratory was required to use equipment known to be free from contamination of all
target analytes at the concentrations of interest.
For predator fillets, the Sample Prep Lab used the entire fillet from both sides of each fish
(skin-on, scales removed, with belly flap attached) instead of measuring and homogenizing
equal weights of each fillet.
For bottom feeders, each entire fish was homogenized and included in the composite instead
of compositing equal weights of individually homogenized fish.
Once homogenized and composited, the Sample Prep Lab aliquoted samples for distribution to
each of the Analysis Labs. If volume allowed, the Sample Prep Lab also prepared "extra
volume" aliquots for shipment to a sample repository for archiving. Aliquots intended for
organics analyses were placed into 125-mL trace-organics clean amber jars with fluoropolymer
(FEP)-lined lids. Metals aliquots were stored in 125-mL I-Chem Level III trace metals clean
(or equivalent) glass jars, also equipped with FEP-lined lids. Each aliquot was further stored
inside two food-grade plastic bags to avoid sample loss in the event of breakage. To avoid
breakage, the jars were filled to no more than 80% capacity. All aliquots were frozen (-20°C)
pending distribution to the Analysis Labs and the Sample Repository.
The Sample Prep Lab determined the lipid content of each sample. This lipid determination was
performed using the procedure described in EPA Methods 1613B and 1668A, and is the same
procedure used in EPA's National Dioxin Study.
The Sample Prep Lab assigned a unique five-digit EPA sample number to each composite tissue
sample and documented the sample number, and corresponding percent lipids result, on a Traffic
Report that accompanied each aliquot to the designated Analysis Lab. The Sample Prep Lab
also prepared a series of "blind composite duplicates" on 5% of the samples. These blind
composite duplicates were also assigned five-digit EPA sample numbers and sent to the Analysis
Labs in exactly the same manner as were true field tissue sample aliquots. The blind sample
aliquots were used by EPA to verify that Sample Prep Lab procedures were yielding
homogeneous aliquots and to characterize variability arising from the entire sample preparation,
re-distribution, and analysis processes.
SECTION 3.2 DIOXINS/FURANS
The presence and concentration of seventeen 2,3,7,8-substituted chlorinated dibenzo-^-dioxins
and dibenzofurans (CDDs and CDFs) in each sample was determined by a slightly modified
version of EPA Method 1613, Revision B (Tetra- through Octa- ChlorinatedDioxins and
Furans by Isotope Dilution HRGC/HRMS, EPA-821-B-94-005).
Modifications were made to the procedures specified in Method 1613B in order to allow for
determination of dioxins and furans at levels ten times lower than those specified in the method.
Specifically, the method was modified to increase the tissue sample size used for analysis and to
April 2005 10 QA Report for the 'National Lake Fish Tissue Study
-------
add a sixth calibration solution that contained all the method-specified analytes at levels lower
than the levels specified in the method to verify linearity at the lower concentrations targeted.
SECTION 3.3 POLYCHLORINATED BIPHENYLS (PCBs)
EPA Method 1668, Revision A (ChlorinatedBiphenyls Congeners in Water, Soil, Sediment, and
Tissue byHRGC/HRMS, EPA-821-R-00-002) was used to determine PCB congener
concentrations in tissue samples collected during the study. There are 209 possible congeners,
12 of which have toxicological significance (i.e., the "toxic" PCBs identified by the World
Health Organization). Method 1668A can unambiguously determine 126 of the 209 congeners
as separate chromatographic peaks. The remaining 83 congeners do not appear as separate
peaks, but elute from the gas chromatograph in groups of 2 to 6 congeners that cannot be
completely resolved by the instrumentation. Ten of the 12 "toxic" congeners are resolved, and
the remaining two congeners (PCB 156 and PCB 157) elute as a congener pair. Because PCB
156 and 157 have identical toxicity equivalency factors, however, it is possible to accurately
calculate PCB toxic equivalence based on the 12 toxic congeners.
For reporting purposes, each tissue sample is associated with 126 results that represent the 126
single PCB congeners, and another 33 results that represent co-eluting congener groups for the
remaining 83 congeners, for a total of 159 PCB congener "results." In addition, each sample is
associated with 10 values corresponding to the 10 possible levels of chlorination for the parent
biphenyl. Each of these 10 values represents the sum of the concentrations of all of the
congeners in a given level of chlorination (i.e., a total of the monochlorinated PCBs, a total of
the total dichlorinated PCBs, etc). Finally, each sample is associated with a total PCB value,
which represents the sum of the 126 congener results plus the 33 values for the co-eluting
congeners. All told, states and other study partners receive 170 unique PCB records for each
sample (126 + 33 + 10 + 1), and 11 of these records represent totals drawn from the first 159
records (126 + 33).
SECTION 3.4 TOTAL MERCURY
Total mercury (Hg) concentrations were determined by EPA Method 1631, Revision B (Mercury
in Water by Oxidation, Purge and Trap, and Cold Vapor Atomic Fluorescence Spectrometry)
and its Appendix (Digestion Procedures for the Determination of Total Mercury in Tissue,
Sludge, Sediment, and Soil).
SECTION 3.5 ARSENIC SPECIES
Total inorganic arsenic, arsenic (III), monomethylarsonic acid (MMA), and dimethylarsinic acid
(DMA) were directly determined by Method 1632, Revision A (Chemical Speciation of Arsenic
in Water and Tissue using Hydride Generation Quartz Furnace Atomic Absorption
Spectrometry). Arsenic (V) was determined by mathematically subtracting the measured
concentration of arsenic (III) from the measured concentration of total inorganic arsenic. Strictly
speaking, the techniques provided in Method 1632A allow for determination of the valence
states of arsenic (III) and arsenic (V) rather than the species of inorganic arsenic. The actual
species of inorganic arsenic are assumed to be those predicted by a geochemical equilibrium
model. Total arsenic (which includes organic forms such as arsenobetaine) was not measured.
Although it is commonly found in fish, arsenobetaine was not targeted in the study because of its
low toxicity to fish and humans.
QA Report for the National Lake Fish Tissue Study 11 April 2005
-------
SECTION 3.6 ORGANOCHLORINE PESTICIDES
Organochlorine pesticides and total Aroclors were determined by Method 1656, Revision A
(Organo-Halide Pesticides in Wastewater, Soil, Sludge, Sediment, and Tissue by GC/HSD),
except that tissue sample extracts were concentrated by a factor of five beyond method-specified
levels before instrumental analysis. This modification ensured that all target pesticides could be
quantified at levels equal to or lower than the screening values published in EPA's Fish
Advisory Guidance.
SECTION 3.7 ORGANOPHOSPHORUS PESTICIDES
EPA Method 1657, Revision A (Organophosphorus Pesticides in Water, Soil, and Tissue by
GC/FPD) was used to determine the presence and concentration of organophosphorus pesticides
listed in Exhibit 3-1.
SECTION 3.8 SEMIVOLATILE ORGANICS
The remainder of the target organic analytes were analyzed by a modified version of EPA
Method 1625, Revision C (Semivolatile Organic Compounds by Isotope Dilution GC/MS). The
modifications involved fractionating the samples by gel permeation chromatography (GPC) to
yield a fraction containing the phthalates and some of the lower molecular weight lipids and a
lipid-free fraction containing the polar target compounds. The phthalate/lipid fraction was
further cleaned using Alumina and then recombined with the lipid-free fraction so that all target
analytes could be determined in a single run. Please note that 3,3'-dichlorobenzidine was
originally included as a target analyte in this study. However, the nature of this compound (and
benzidines in general) often results in poor recoveries of this compound and its labeled analog,
3,3'-dichlorobenzidine-d6, from tissue samples and Ongoing Precision and Recovery (OPR)
samples. Historically, the recovery problems have led to exclusion of large numbers of
analytical results for 3,3'-dichlorobenzidine. Therefore, rather than exclude a large percentage of
these results and include other results for this compound with recoveries that are very disparate
from the other target analytes, EPA decided not to report results for 3,3'-dichlorobenzidine.
During Year 4 of the study only, the laboratory employed a Florisil cleanup instead of an
Alumina cleanup in order to mitigate lipid interferences that were resulting in an excessive
number of reextractions and reanalyses in the Year 3 samples. The use of Florisil did reduce the
number of reanalyses required, suggesting that this approach should be considered in any further
studies.
April 2005 12 QA Report for the National Lake Fish Tissue Study
-------
Exhibit 3-1
National Lake Fish Tissue Study Target Analytes and
Analysis Method
Dioxins and Furans by Isotope Dilution High-
resolution Gas Chromatography (GC)/Mass
Spectrometry (MS) (Method 1613, Revision B)
Total Mercury by Oxidation, Purge and Trap,
and Cold Vapor Atomic Fluorescence
Spectrometry (Method 1631, Revision B with
Appendix A - Digestion procedures for Total
Mercury in Tissue, Sludge, Sediment, and
Soil)
Arsenic Speciation by Arsine Generation,
Chromatography, and Atomic Absorption
Spectrometry (Method 1632, Revision A )
Polychlorinated Biphenyls by Isotope Dilution
High-resolution GC/Mass Spectrometry
(Method 1668, Revision A)
Organochlorine Pesticides by GC/HSD
(Method 1656, Revision A)
Organophosphorus Pesticides by GC/FPD
(Method 1657, Revision A)
Corresponding Analysis Methods
Target Analyte
2,3,7,8-TCDD
2,3,7,8-TCDF
1,2,3,7,8-PeCDD
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,4,6,7,8-HpCDD
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
OCDD
OCDF
Mercury
Arsenic (III)
Arsenic (V)
Dimethylarsinic acid (DMA)
Monomethylarsonic acid (MMA)
Total inorganic arsenic
209 congeners, including the following 12 "dioxin-like" congeners:
3,3',4,4'-TeCB
3,4,4',5-TeCB
2,3,3',4,4'-PeCB
2,3,4,4',5-PeCB
2,3',4,4',5-PeCB
2',3,4,4',5-PeCB
2,4'-DDD (TDE) *
2,4'-DDE *
2,4'-DDT *
4,4'-DDD (TDE)
4,4'-DDE
4,4'-DDT
Aldrin
cis- and frans-Nonachlor
Dicofol
Dieldrin
Endosulfan sulfate
Endosulfan I
Endosulfan II
Endrin
Ethalfluralin (Sonalan)
Heptachlor
Heptachlor epoxide
Isodrin
Kepone (Chlordecone)
Methoxychlor
Mirex
Octachlorostyrene
3,3',4,4',5-PeCB
2,3,3',4,4',5-HxCB
2,3,3',4,4',5'-HxCB
2,3',4,4',5,5'-HxCB
3,3',4,4',5,5'-HxCB
2,3,3',4,4',5,5'-HpCB
Oxychlordane
Aroclor-1016
Aroclor-1221
Aroclor-1232
Aroclor-1242
Aroclor-1248
Aroclor-1254
Aroclor-1260
Pendamethalin (Prowl)
Pentachloronitrobenzene (PCNB)
ds-Permethrin
frans-Permethrin
Toxaphene
Trifluralin
a-BHC
a -Chlordane (c/s-Chlordane)
P-BHC
Y -BHC (Lindane)
Y -Chlordane (frans-Chlordane)
6 -BHC
Pentachloroanisole
* Analytes were added to the target analyte list after Year 1 of the study.
Chlorpyrifos
Diazinon
Disulfoton
Ethion
Paraoxon
Parathion (ethyl)
Terbufos
Terbufos sulfoxide
Terbufos sulfone
QA Report for the National Lake Fish Tissue Study 13
April 2005
-------
Exhibit 3-1
National Lake Fish Tissue Study Target Analytes and Corresponding Analysis Methods
Analysis Method
Target Analyte
Semivolatile Organic Compounds by Isotope
Dilution GC/MS (Method 1625, Revision C with
modifications for tissue) [See Section 3.8 for
note concerning the analyte
3,3'-dichlorobenzidine.]
1,2,4,5-Tetrachlorobenzene
1,2,4-Trichlorobenzene
1,2-Dichlorobenzene
1,3-Dichlorobenzene
1,4-Dichlorobenzene
2,4,5-Trichlorophenol
4,4'-Methylenebis(2-chloroaniline)
4-Bromophenyl phenyl ether
4-Nonylphenol
Acenaphthene
Acenaphthylene
Anthracene
Benzo[a]anthracene
Benzo[a]pyrene
Benzo[b]fluoranthene
Benzo[g,h,i]perylene
Benzojpuoranthene
Benzo[k]fluoranthene
Bis(2-ethylhexyl) phthalate
Butyl benzyl phthalate
Chrysene
Dibenzo[a,h]anthrancene
Di-/?-butyl phthalate
Diethylstilbestrol (DBS)
Fluoranthene
Fluorene
Hexachlorobenzene
Hexachlorobutadiene
lndeno[1,2,3-cd]pyrene
Naphthalene
Nitrobenzene
Pentachlorobenzene
Pentachlorophenol
Perylene
Phenanthrene
Phenol
Phenol, 2,4,6-tris(1,1 -dimethylethyl)-
Pyrene
Tetrabromobisphenol A
1,2,3-Trichlorobenzene*
1,2-Diphenylhydrazine*
2,3,6-Trichlorophenol*
2,4,6-Trichlorophenol*
2,4-Dichlorophenol*
2,4-Dimethylphenol*
2,4-Dinitrophenol*
2,4-Dinitrotoluene*
2,6-Dinitrotoluene*
2-Chloronaphthalene*
2-Chlorophenol*
2-Nitrophenol*
2-Picoline*
4-Chloro-3-methylphenol*
4-Chlorophenylphenyl ether*
4-Nitrophenol*
alpha-Terpineol*
Biphenyl*
Bis(2-chloroethoxy) methane*
Bis(2-chloroethyl) ether*
Bis(2-Chloroisopropyl) ether*
Carbazole*
Di-ff-octyl phthalate*
Di-/?-propylnitrosamine*
Dibenzofuran*
Dibenzothiophene*
Diethyl phthalate*
Dimethyl phthalate*
Diphenyl ether*
Diphenylamine*
Hexachlorocyclopentadiene*
Hexachloroethane*
Isophorone*
n-Decane*
n-Docosane*
n-Dodecane*
n-Eicosane*
n-Hexacosane*
n-Hexadecane*
n-Nitrosodimethylamine*
n-Nitrosodiphenylamine*
n-Octacosane*
n-Octadecane*
n-Tetracosane*
n-Tetradecane*
n-Triacontane*
/>Cymene*
2-Methyl-4,6-Dinitrophenol*
Styrene*
*This analyte was not considered a target analyte for study purposes, but
since it was listed as a target analyte in Method 1625C, it was determined as
a means of obtaining additional useful data for the study.
April 2005
14 QA Report for the National Lake Fish Tissue Study
-------
Chapter 4
Data Quality Assessment
Three levels of review were applied to all data generated in the National Lake Fish Tissue Study.
First, a pre-qualification review was performed prior to analysis of tissue samples to verify that
each laboratory was qualified to analyze the tissue samples in accordance with the prescribed
methods. Second, ongoing reviews were performed to verify that the results of each data
submission were, in fact, generated in accordance with all method and study requirements.
Finally, overall data quality was evaluated at the end of Year 1, Year 2, Year 3, and Year 4 to
verify that data as a whole were meeting established MQOs. The procedures employed for each
of these three data review levels are described in Sections 4.1 through 4.3. Section 4.1 describes
the procedures employed for reviewing laboratory prequalification data submitted prior to
analysis of tissue samples during the study. Section 4.2 describes our process for reviewing
individual data packages as they were submitted throughout the study, and describes our
assessment of overall data quality in terms of the items reviewed. Section 4.3 discusses initial
MQOs established for the study, and summarizes how well the overall study data set measured
up against those MQOs. Where applicable, quantitative measures (e.g., percent of data that met
Criterion X) of data quality are presented; these measures were calculated using results from
study target analytes. Specific data qualifiers applied to the target analytes are presented in
Exhibit 4-1 A, and specific data qualifiers applied to non-target analytes measured during the
study are presented in Exhibit 4-1B.
SECTION 4.1 PRE-QUALIFICATION REVIEW
Prior to preparing or analyzing tissue samples collected during the study, each laboratory was
required to submit data demonstrating their ability to achieve the sensitivity, precision, and
accuracy goals established for the study.
Labs did not analyze fish tissue samples until they submitted pre-qualification data
demonstrating they could achieve the sensitivity, precision, and accuracy goals defined
for the study.
4.1.1 Sensitivity
4.1.1.1 Sensitivity Goals
Analytical sensitivity reflects the minimum concentration of an analyte above which a data user
can be reasonably confident that the analyte was reliably detected and quantified. For this study,
the method detection limit (MDL) and the minimum level (ML) of quantitation were used to
define the sensitivity of each measurement process.
The MDL is defined as the measured concentration at which there is 99% confidence that a
given analyte is present in a given sample matrix. Prior to analyzing tissue samples collected in
Year 1 of the study, all laboratories were required to perform MDL studies in accordance with
the procedures specified by EPA at 40 CFR 136, Appendix B.
QA Report for the National Lake Fish Tissue Study 15 April 2005
-------
Quantitative sensitivity in this study was established by the ML. The ML is defined as the
lowest concentration at which the entire analytical system gives a recognizable signal and
acceptable calibration for an analyte. The ML is equivalent to the lowest calibration standard
analyzed by a specific analytical procedure, assuming that all the method-specified sample
weights, volumes, and processing steps have been employed. The EPA 1600-series methods
described in Chapter 3 specify MLs for tissue and/or aqueous samples. Generally speaking, MLs
are approximately three times greater than the MDL and are comparable to the American
Chemical Society's Limit of Quantitation.
In accordance with study objectives, each laboratory was required to demonstrate it could
achieve MLs that were equal to or lower than those listed in the analytical method they would be
using in the study. The only exceptions were as follows:
The laboratory tasked with analyzing tissue samples for dioxins/furans was required to
achieve MLs that were ten times lower than those specified in EPA Method 1613B. This
was accomplished by increasing the tissue sample size used for analysis (to increase
measurement sensitivity) and by adding a sixth calibration solution that contained all the
method-specified analytes at levels lower than the levels specified in the method (to verify
linearity at the lower target concentrations).
The laboratory tasked with analyzing the organochlorine pesticide tissue samples was
instructed to further concentrate its sample extracts in order to quantify all the target
pesticides at levels that were equal to or lower than the screening values published by EPA in
the Fish Advisory Guidance.
The laboratory tasked with determining total mercury was permitted to target a ML of 2 ng/g
instead of the 1 ng/g figure cited in the tissue appendix to Method 163 IB. The allowed ML
of 2 ng/g was considered to be acceptable because it is well below the EPA's recommended
screening value for mercury.
4.1.1.2 Sensitivity Assessments
The Sample Prep Lab was responsible for receiving, filleting (where appropriate),
homogenizing, aliquoting, and distributing samples to the analysis laboratories. Because these
processes could theoretically affect the results generated by all other laboratories, it was critical
to demonstrate that the Sample Prep Lab processes would not introduce contamination of any
target analytes at the levels of interest in this study. To do so, the Sample Prep Lab was required
to analyze equipment blanks (rinsates) before preparing any samples collected in the study. In
conjunction with this, the Sample Prep Lab was required to perform MDL studies to demonstrate
the lab's ability to measure blanks at the levels of interest in the study. These Sample Prep Lab
MDL studies were performed in reagent water. Reagent water was used as a reference matrix for
the equipment rinsates generated to demonstrate the absence of contamination.
The Analysis Labs were required to perform their MDL studies in a reference tissue matrix,
using the same analytical methods they would be using in the study (see Exhibit 1-2). When
possible, actual fish tissue was used for the MDL studies. However, it was often not possible to
locate fish that were free of the analyte of interest at the low detection limits targeted in this
study. In such cases, an alternative matrix, such as chicken breast or corn oil, was used.
April 2005 16 QA Report for the National Lake Fish Tissue Study
-------
Each MDL submission was reviewed to verify that:
The MDL procedures specified at 40 CFR 136, Appendix B were followed correctly, with
respect to the number of replicates, the spiking levels, and the statistics applied in
determining the MDL;
The laboratory used the analytical method(s) they would be employing in the study to
analyze actual tissue (or, for the Sample Prep Lab, equipment rinsate) samples when
performing their MDL studies;
The laboratory performed their MDL studies on a calibrated instrument;
The laboratory used an appropriate reference matrix when performing their MDL studies;
and
The MDLs determined by the laboratory supported the ML that would be targeted in the
study (i.e., the MDLs were at least three times lower than the target MLs).
After evaluating the MDL results and, where necessary, obtaining clarification or missing
information from the laboratories, the MDL data submitted by each laboratory for metals,
dioxins, PCB congeners, and pesticides were deemed to meet method and study requirements.
For the semivolatile organic analytes, all of the MDL study results were below the method-
derived MLs for tissue samples; however, a few instances occurred in which the measured MDL
was above this objective. These cases included the following target analytes:
1,2-dichlorobenzene, 1,3-dichlorobenzene, 1,4-dichlorobenzene, di-w-butyl phthalate, and
bis(2-ethylhexyl) phthalate. This was not considered to be an issue because the instrument was
calibrated at the ML and non-detects were reported at the ML rather than the MDL. To further
clarify, any result above the calculated MDL that met method-specified criteria would be
reported as a "hit"; however, any result detected below the calculated MDL would be reported
as a non-detect at the ML (see Section 4.2.2). Also, the measured MDLs for some of the
analytes, such as the phthalates, were considered to be elevated by analyte contributions from the
reference matrix used, and it was evident that other analytes, such as the dichlorobenzenes, could
be detected at much lower levels (the measured MDLs were elevated by variability among the
replicate measurements).
4.1.1.3 Reporting Thresholds for Sensitivity
As noted above, all of the 1600-series methods used for this study list MLs for aqueous and/or
tissue samples. The appendices to Methods 163 IB (mercury) and 1632A (arsenic) give tissue-
based MDLs and MLs. These method-specified thresholds were used as reporting thresholds in
the study for arsenic, but, as noted above, the laboratory-determined mercury ML of 2 ng/g was
approved for use in this study. For consistency, the laboratory-determined mercury MDL also
was used as the reporting threshold for detection limit sensitivity.
Method 1613B (dioxins/furans) provides MLs applicable to solids and tissues but, with the
exception of 2,3,7,8-TCDD, does not provide corresponding MDLs. Because the method was
modified as described in Section 3.2 to quantify dioxins/furans at levels 10 times lower than
specified in the method, the quantitation limit thresholds reported in the database reflect the
method-specified MLs divided by a factor of 10. Laboratory-determined MDLs were used as the
QA Report for the National Lake Fish Tissue Study 17 April 2005
-------
reporting threshold for detection limit sensitivity. As noted above, these laboratory-determined
MDLs were at least a factor of three lower than the target MLs.
Method 1668 A (PCB congeners) provides estimated MDLs and MLs that were derived based on
the standard deviation of single lab blank measurements. Because these levels have not been
finalized, laboratory generated MDLs and MLs were used as the detection and quantitation limit
thresholds, respectively. As noted above, the laboratory's calibration curve encompassed their
ML, and the laboratory-determined MDLs were at least a factor of three lower than the target
MLs.
Methods 1656A (organochlorine pesticides) and 1657A (organophosphorus pesticides) provide
recommended MLs for tissue samples and list recommended calibration standards intended to
support the ML. These method-specified MLs were used as reporting thresholds for
Method 1657A. For Method 1656A, the method MLs were divided by a factor of 5 to account
for the five-fold extract concentration step described in Section 3.6 of this report and in the
sensitivity goals discussion above. The laboratory MDLs, determined in fish tissue, were used
as the threshold for reporting detection limit sensitivity. As noted above, these MDLs supported
the MLs targeted in the study.
Method 1625C (semivolatile organics) does not specify MLs in tissue, nor does the method
provide MDLs in either tissue or aqueous samples. Therefore, target tissue-based MDLs were
mathematically derived from the method-specified aqueous MLs by converting aqueous units to
solid units, accounting for the sample mass used in the tissue measurements, and dividing the
resulting tissue MLs by 3. These method-derived MDLs and MLs were used as reporting limits
for the semivolatile analytes.
4.1.2 Initial Demonstration of Precision and Accuracy
Prior to analyzing tissue samples collected during all four years of the study, each laboratory was
required to demonstrate its ability to achieve precise and accurate results with the required
analytical method. To do so, laboratories were required to prepare and analyze Initial Precision
and Recovery (IPR) samples as described in each method.
IPR samples consisted of four aliquots of a reference matrix spiked with a known level of the
target analytes. The reference matrix was chosen to serve as an indicator of method performance
that could be expected for the tissue samples collected in the study. Accuracy was measured by
determining the average recovery in the replicate IPR samples; precision was measured by
calculating the relative standard deviation (i.e., RSD = standard deviation/mean) of the measured
levels in the IPR samples.
Each laboratory's IPR submission was reviewed to verify the following:
An appropriate reference matrix and spiking levels were used to prepare the four replicate
samples.
The designated 1600-series method was used to analyze the samples.
The samples were generated on a properly calibrated instrument.
Calculations of analyte recovery and precision were performed correctly.
April 2005 18 QA Report for the National Lake Fish Tissue Study
-------
After evaluating these factors and, where necessary, obtaining clarification or missing
information from the laboratories, the IPR data submitted by each laboratory was deemed to
meet method and study requirements and the laboratories were considered to be pre-qualified to
analyze samples collected in Year 1 through Year 4 of the study.
SECTION 4.2 INDIVIDUAL DATA PACKAGE REVIEW
Upon completion of tissue sample preparation activities, each sample was assigned to a "sample
delivery group" (SDG) for analysis purposes. Each SDG consisted of an arbitrarily assigned
group of 20 samples assembled to simplify sample distribution and minimize the number of
Quality Control (QC) samples that had to be prepared by each analytical laboratory. (Most
methods specify that QC samples be prepared at a frequency of one per 20 tissue samples or one
per analytical batch, whichever is smaller.) Laboratories were also asked to report their data by
SDG to standardize the size of each data package and expedite data review processes.
All Year 1, Year 2, Year 3, and Year 4 field results and supporting QC data were carefully
reviewed to determine if the method had been properly followed and if all systems were
in control during sample preparation and analysis. QC deviations were assessed to
determine if they had an impact on the data, and these deviations and assessments were
coded directly into the study database with flags and descriptive comments.
When submitting results for each SDG, the laboratories were required to submit all results
associated with analysis of the samples in the SDG. This included results of the fish tissue
samples analyzed, as well as results associated with any supporting QC measurements (e.g.,
instrument calibration, blank, and spike data). Results were to include both summary level data
(the final measurement) and raw data (spectra, chromatograms, bench worksheets, etc.).
Each data package was thoroughly reviewed to ensure the following:
All samples listed on the corresponding field and Sample Prep Lab documentation were
analyzed and that results were provided.
When possible, each analyte was uniquely identified with a Chemical Abstract Registry
Service Number (CAS Number) to eliminate any ambiguity in naming conventions.
All required QC samples were analyzed and these QC samples met specified acceptance
criteria.
Results were provided for each sample analyzed, including any dilutions and reanalyses, and
for all associated QC samples.
Required data reporting forms and/or electronic formats were provided for each of the field
tissue sample and/or associated QC analyses.
Raw data associated with each field tissue sample and QC sample was provided with each
data package, and the instrument output (emission intensity, peak height, area, or other signal
intensity) was traceable from the raw data to the final result reported.
Any problems encountered and corrective actions taken were clearly documented.
QA Report for the National Lake Fish Tissue Study 19 April 2005
-------
If anomalies were found, the laboratory was contacted and asked to provide the missing data,
clarifications, and/or explanations so that a comprehensive data review could be performed to
verify the quality of their results. Results of these data reviews were documented directly in the
database through the application of standardized data qualifier flags and descriptive comments
concerning the reliability of the flagged results. Exhibit 4-1A (Target Analytes) at the end of
this report summarizes flags and comments applied to the data for target analytes as a result of
the review process described in subsections 4.2.1 - 4.2.7. Exhibit 4-1B (Additional or Non-
Target Analytes) summarizes the flags applied to the 49 "non-target," semivolatile organic
analytes measured as part of this study.
It is important to note that, because several of the methods used in the National Lake Fish Tissue
Study contain a large number of analytes that are being tested simultaneously, there is always a
small statistical probability that QC results for some of these compounds will occasionally fail to
meet method specifications. Likewise, the large number of samples collected and the complex
matrices being analyzed suggest some probability of occasional QC failures. In other words, the
presence of QC failures and data qualifiers in the Year 1, Year 2, Year 3, and Year 4 data sets
does not automatically suggest poor data quality in this study. To the contrary, EPA believes
that the overall quality of data generated in this study was high, as evidenced by the limited
number of failures described in Sections 4.2.1 through 4.2.7, and by the fact that the overall data
set met all MQOs established for the study, as described in Section 4.3.
4.2.1 Assessment of Instrument Calibration
All of the methods used in the study required the laboratory to calibrate their instruments using a
series of standards that covered a range of target analyte concentrations. This initial calibration
provides a quantitative determination of instrument response and generates qualitative criteria
for analyte identification. Methods 163 IB (mercury), 1632A (arsenic), 1656A (organochlorine
pesticides), and 1657A (organophosphorus pesticides), require a three-point calibration (i.e., the
use of three calibration standards that contain the target analytes at low, medium, and high
concentrations). Methods 1625C (semivolatile organics) and 1668A (PCBs) specify the use of
five standards to calibrate the instrument. Method 1613B (dioxins/furans) requires the use of
five standards, but as noted above in Section 4.1.1, a sixth calibration standard was used in this
study to achieve lower quantitation limits.
The relationship between the response of an analytical instrument to the concentrations or
amounts of an analyte introduced into the instrument is referred to as the "calibration curve."
The 1600-series methods used in the study contain specific criteria for determining the linearity
of calibration curves. When the applicable criterion is met, the calibration curve is considered to
be sufficiently linear to permit the laboratory to use an average response factor or calibration
factor, and it is assumed that the calibration curve is a straight line that passes through the
zero/zero calibration point. If the calibration curve is not linear, an alternative approach must be
used to quantify sample results. This means that a regression line or other mathematical function
must be employed to relate instrument response to the concentration.
Each data submission was reviewed to verify that the appropriate number of calibration
standards were used and that the resulting calibration curve either met the linearity specifications
in the method or that the calibration curve was used to quantify samples. All Year 1, Year 2,
April 2005 20 QA Report for the National Lake Fish Tissue Study
-------
Year 3, and Year 4 data were generated on instruments that met the linear calibration
requirements specified in the referenced method.
Initial calibration data submitted with each data package also were examined to verify that the
calibration curve encompassed the MLs targeted in this study. The use of the ML as a point on
the calibration curve is the principal means by which to assure that measurements made at the
quantitation level are reliable. All Year 1, Year 2, Year 3, and Year 4 data were generated on
instruments that were properly calibrated at or below the MLs in this study. In addition, all but
two of the results reported in the study were analyzed within the instrument calibration range.
These semivolatile organic results were coded with "REXC" to indicate that the result exceeded
the calibration range and were further qualified as an "Estimated Value."
Because analytical instruments are subject to drift over time, analytical methods typically require
periodic analysis of standards to verify the instrument remains calibrated throughout the duration
of analysis. The 1600-series methods used in the National Lake Fish Tissue Study specify that
the calibration verification be performed by analyzing a mid-point standard. The concentration
of each analyte in this standard is determined using the initial calibration data and compared to
the specifications in the method. If the results are within the method specifications, the
laboratory is allowed to proceed with analysis without recalibrating and to continue using the
initial calibration data to quantify sample results. If calibration verification results fall outside
the required limits, the laboratory is required to recalibrate their instrument before proceeding
with sample analysis. The frequency of this calibration verification varies according to the
instrumentation (more frequent verification is required for instruments that are highly prone to
drift) and is specified in the method. Verifying calibration at the method-specified frequency
allows for early identification of problems and minimizes the need to reanalyze samples that
might otherwise have been analyzed on an uncalibrated instrument.
Each Year 1, Year 2, Year 3, and Year 4 data package received was reviewed to verify that:
Calibration verification was performed at the required frequency using appropriate
calibration standards, and
Results of the calibration verification met method-specified acceptance criteria, or
If results did not meet method specified criteria, the laboratory re-calibrated the instrument
before proceeding with sample analysis.
If the calibration verification requirements were not met, the reviewer evaluated the data package
to verify that the laboratory followed the corrective action dictated by the method and that results
were not affected. Although 284,973 field tissue sample results were generated during the first,
second, third, and fourth years of the study, only 107 (6 samples in Year 1; 63 samples in Year 2;
0 samples in Year 3; and 38 samples in Year 4) of these results were flagged with calibration
verification (CalVer) qualifiers. In Year 1 of the study, the six results flagged were for
organochlorine pesticides, an outcome that is not surprising given the large number of pesticides
targeted by the method and the fact that each pesticide is determined on two separate columns.
When the analyte was found using both columns, the lowest of the measured results was reported
since that value could be supported by both measurements. Reported results that failed to meet
the recovery specifications for Method 1656A were qualified with either "LVER" to indicate a
low recovery of the calibration standard or "HVER" to indicate a higher recovery of the
standard. Failure to recover the analyte from the CalVer sample at all resulted in the application
of a "NVER" flag. In Year 2 of the study, 63 of the arsenic speciation results (Method 1632A)
QA Report for the National Lake Fish Tissue Study 21 April 2005
-------
were qualified with "LVER" indicating a low recovery of the calibration standard. In Year 4 of
the study, all of the 38 "LVER" and "NVER" codes were also associated with other data
qualifiers.
99.96% of the 284,973 field tissue sample results generated in the four years of the
study met all instrument calibration requirements.
4.2.2 Reporting Thresholds
Each laboratory was instructed to report all positive results that met all method-specified criteria
(i.e., "hits") down to the MDL. The labs were further instructed to apply a "J" flag to any results
reported above the MDL but below the ML.1 The purpose of the "J" flag was to indicate that,
although the analyte was detected, it was detected at a value below the quantitation limit. In
other words, the presence of a " J" flag suggests the reported value is qualitative (the analyte is
definitely present) but not quantitative (the reported value is an estimate of the true
concentration).
Note: The PCB lab was instructed to include these "J" flagged results when reporting congener
totals for the PCBs. Because nearly all samples had at least a few "J" flagged PCB
congeners, nearly all the samples (964 samples out of a total of 1,003 samples analyzed
for PCBs) also have a "J" flag on total PCB congener values.
The MDL is designed to provide a 99% level of confidence that when an analyte is reported as
being present, it really is present. The converse is not true, however. If an analyte is reported as
not being present at the MDL level, a 50% possibility exists that the result is a false negative. To
ensure that results reported as non-detects are reliable indicators of the true concentration at
which the analyte could not be detected, the reporting threshold for non-detects was set at the
ML. (For further clarification, see final paragraph of Section 4.1.1.2)
Positive results were reported to the MDL and flagged with a "J" if the results were
below the ML. J-f lagged results were further annotated as "Estimated Values" to caution
users that the results are qualitative rather than quantitative.
All field tissue sample results submitted during Year 1, Year 2, Year 3, and Year 4 of the
National Lake Fish Tissue Study were carefully reviewed to verify that the laboratory adhered to
these reporting conventions.
1MDLs reported in the study database reflect the study MDLs approved for the study, as discussed in
Section 4.1.1. Because MDLs are determined in a single laboratory, measured MDL results are subject to slight
changes over time and between laboratories. For analytes, such as hexachlorobenzene and terbufos, where the
recommended screening value was below the "study MDL," but above a lab-determined MDL, any results reported
between the study MDL and the lab-determined MDLs were reported in the database with a "LJS" (e.g., "J flagged
result was between the lab and study MDLs").
April 2005 22 QA Report for the National Lake Fish Tissue Study
-------
4.2.3 Analysis of Blanks
Blanks are used to verify the absence of
contamination that may occur at any point
in the measurement process. In the
National Lake Fish Tissue Study, many
analytes were targeted at extremely low
concentrations comparable to or lower than
those typically found in the ambient
environment. Therefore, frequent analysis
and assessment of blanks was critical to
determine if measured sample
concentrations were biased by the presence
of contamination during sample collection,
handling, or analysis. In this study, the
following blanks were used:
Impact of Blank Qualifiers - Years 1,2,3, and 4
Result changed to non-detect (0.99%)
Result not affected or alread/ applied to individual congener (1.
Result considered maximum value (0.18%)
Result with no blank qualifiers (96.84%)
Equipment blanks were used by the Sample Prep Lab to verify that the procedures and
equipment used to prepare, fillet, homogenize, and or aliquot the samples were not
introducing contamination. These blanks consisted of reagent water (i.e., water known to be
free of the target analytes) that was run through all equipment used to process the samples in
the facility where sample processing occurred. Each year, blanks were prepared and
analyzed before any samples were prepared to verify the equipment and procedures to be
used in the study were clean. The laboratory was required to reduce contamination to a level
below all target MLs, and requested to reduce contamination below the target MDLs to the
maximum extent possible. (It was understood that complete elimination of some ubiquitous
contaminants, such as mercury and certain PCB congeners, at the low levels targeted in this
study would be extremely difficult, if not impossible.) After lab cleanliness was verified, the
lab continued to prepare and analyze an equipment blank on each day of sample
preparation/homogenization activity.
Calibration blanks were used during metals analysis to verify that contamination was not
being introduced by the analytical system. These calibration blanks were to be analyzed
immediately after calibration verification.
Method blanks were used by each Analysis Lab to verify that contamination was not being
introduced by the analytical process (i.e., the combination of the sample digestion or
extraction procedures and the analytical system).
All equipment blank results submitted by the Sample Prep Lab were evaluated to verify that the
equipment blanks were free of contamination below the MLs targeted in this study, and that any
contamination reported below the ML but above the MDL was considered when evaluating
corresponding field tissue sample results. To evaluate the potential contribution of equipment
contamination to corresponding field tissue samples, the mass of contaminant detected in the
equipment blank was determined by multiplying the concentration of the contaminant reported in
the equipment blank by the volume of rinse water used to generate the blank. The worst-case
concentration of the contaminant in each corresponding tissue composite was calculated by
QA Report for the National Lake Fish Tissue Study 23
April 2005
-------
dividing the mass of the contaminant in the equipment blank by the total tissue mass of each
composite. This calculation assumes that any equipment-related contamination in the composite
sample would be equally distributed through the sample by the compositing process. The effects
of equipment blank contamination were then assessed as follows:
If the analyte was detected in the equipment blank but was not detected in the associated
field tissue samples, the sample data were considered to be acceptable.
If the analyte was detected in the associated field tissue samples at levels far greater (i.e., at
least ten times more) than the levels detected in the equipment blank, the effect of the blank
was considered to be negligible and the field tissue sample data were considered to be
acceptable. Such data were qualified with "B" and "RNAF" to indicate that equipment blank
contamination was present but the sample result was not affected by it.
If the analyte was detected in the field tissue samples at levels close to (i.e., within 5 times)
the level detected in the equipment blank, there are no means by which to ascertain if the
tissue result reported was due to contamination. In such cases, the result reported by the lab
was changed to a non-detect at the ML and coded with "B" and "RNON" qualifiers to
indicate the change.
If the analyte detected in the field tissue sample was more than 5 times but less than 10 times
the concentration detected in the equipment blank, the field tissue sample result was coded
with "B" and "RMAX" qualifiers to indicate a possible high bias from contamination.
Blanks submitted by each Analysis Lab were reviewed and flagged according to the same
approach. In applying these rules, data reviewers were careful to consider the impact of dilution
on the field tissue sample results (i.e., the concentration of a diluted sample was compared to the
blank result multiplied by the dilution factor that was applied to the sample). For example, if
12 ppb of contaminant was found in the blank, and the associated sample extract was diluted by
a factor of 6 relative to the extract from the blank, then the sample result would have to be
greater than 12 x 6 x 10, or 720 ppb to be considered acceptable. (The result times the dilution
factor times the 10 times rule described above.) If the sample result was reported to be between
360 ppb and 720 ppb, it would be flagged with "B" and "RMAX" qualifiers to indicate a
possible high bias resulting from contamination as dictated by the between 5 times and 10 times
rule described above.
In all, 9,094 of the 284,973 field tissue sample results generated during Year 1 through Year 4 of
the study were qualified due to blank contamination, and more than 6,000 of these qualifiers
were applied to PCB congeners (other than the 12 dioxin-like congeners), which are ubiquitous
in the environment and very difficult to eliminate at the low levels targeted in this study. In
addition, most of the qualifiers were used to indicate that the results were determined to be not
affected by the blank contamination (i.e., flagged with "B, RNAF") or were duplicatively applied
as a result of PCB congener summation (i.e., if an individual congener received a blank qualifier
flag, the qualifier also was applied to the corresponding congener total). It is important to note
that no blank subtraction was performed to eliminate the effects of blank contamination detected
April 2005 24 QA Report for the National Lake Fish Tissue Study
-------
in the equipment, calibration, or method blank.2 Instead, data associated with contaminated
blanks were qualified so that data users could make decisions regarding data usability.
4.2.4 Spiked Sample Recoveries
Percentage of Data Qualified Due to Matrix Interferences -
Years 1. 2. 3. and 4
All laboratories were required to spike field
tissue samples to estimate the recovery of
target analytes from the field tissue samples
analyzed in this study. The GC/MS
methods used to analyze dioxins/furans,
PCB congeners, and semivolatile organics
required that isotopically labeled analogs of
the target analytes be spiked into each and
every sample, including QC samples. This
technique, known as isotope dilution,
provides an extremely accurate means of
quantifying a large number of analytes in
the presence of matrix interferences, and
each method specifies acceptable recovery
windows for the labeled compounds.
Because the isotope dilution technique
incorporates recovery-correction into calculations of target analyte concentration, any results that
fail to meet the method-specified recovery windows are considered to be estimated values. The gas
chromatography (GC), atomic absorption (AA), and atomic fluorescence (AF) methods used to
analyze pesticides, Aroclors, arsenic species, and mercury require that a matrix spike (MS) and a
matrix spike duplicate (MSD) pair be prepared and analyzed with each batch of 20 field tissue
samples.
QC samples are not subjected to the matrix spiking requirement. The methods provide precision
and accuracy criteria that should be met for each analyte. Precision criteria are expressed as the
relative percent difference (RPD) between the MS and MSD results, and accuracy criteria are
expressed as acceptable recovery of each analyte. RPD is determined as:
Results not affected by matrix interferences (99.73%)
Results affected by matrix interferences (0.27%)
RPD =
C1+C2
x 100
where, Cl is the concentration of the first spiked aliquot of the sample and C2 is the concentration
of the second spiked aliquot of the sample.
Unlike isotope dilution GC/MS techniques, the calculations involved in measuring analytes by
GC, AA, and AF techniques do not include a recovery-correction component. Therefore, the
direction of the MS recovery failure is used to estimate the directional bias of any associated
sample results.
Mercury results were corrected for bubbler blank contamination as per sample calculation procedures
described in Method 163 IB. As with results from other analyses, however, none of the mercury results were
corrected for contamination observed in the equipment, calibration, or method blanks.
QA Report for the National Lake Fish Tissue Study 25
April 2005
-------
All labeled compound and MS/MSD results reported in Year 1, Year 2, Year 3, and Year 4 of the
National Lake Fish Tissue Study were carefully reviewed to evaluate the quality of data. In
reviewing these results, reviewers verified that appropriate spiking compounds and spiking
levels were used, that samples were spiked at the correct frequency, and that spiked sample
results met method-specified criteria. In evaluating spiked sample results, data reviewers applied
the following rules:
If the isotopically labeled analog of a target compound was recovered in a sample above
method-specified criteria, the associated native compound was coded with "HLBL" to
indicate the presence of a high labeled compound recovery. The result also was qualified
with an "Estimated Value" descriptor to indicate that, although there was no question as to
the identify of the analyte, there is some doubt as to the reliability of the measured
concentration. In Year 1 through Year 4, only two results received this qualifier.
If the isotopically labeled analog of a target compound was recovered below the method-
specified criteria, the associated native compound in that sample was coded with "LLBL" to
indicate the presence of a low labeled compound recovery. The result also was qualified
with an "Estimated Value" descriptor. In Year 1 through Year 4, eighteen (18) results
received this qualifier.
In rare cases, a labeled analog is not recovered from the sample at all. Due to the extreme
nature of this QC failure, any associated native (i.e., target) compound result is considered
unreliable and should, therefore, be excluded from the database. In Year 1 through Year 4,
this situation occurred only for semivolatile organic and PCB congener analytes in 39
samples (a total of 49 results). The 49 affected Year 1, Year 2, Year 3, and Year 4 results
were excluded from the database, and the associated sample records were coded as "NLBL"
and further qualified with "Exclude" to explain the absence of these results.
If an analyte was recovered from an MS or MSD sample above method-specified criteria, all
associated samples with positive results were coded with "HMSR" to indicate the high MS
recovery. The associated samples also were qualified to indicate a "Potential High Bias"
unless other flags (such as a "J") applied to the same result suggested that the sample should
be qualified as an "Estimated Value." In Year 1, Year 2, Year 3, and Year 4, only 25 results
were qualified with the "HMSR" flag (the analytes affected were arsenic (III); 5-BHC;
cis-nonachlor; 4,4'-DDT; and dicofol). Ten of these results were qualified as having
"Potential High Bias;" the remaining 15 results were reported below the quantitation limit or
associated with "RPDX" and/or "RNF2" flags and were qualified as "Estimated Values."
Non-detects in samples associated with the high MS/MSD recovery were not flagged
because it was clear that the potential high bias indicated by the MS or MSD had no adverse
impact on the sample result.
If an analyte was recovered from an MS or MSD sample below method-specified criteria, all
associated samples were coded with "LMSR" to indicate the low MS recovery. The
associated samples also were qualified to indicate a "Potential Low Bias." All of the
"LMSR" codes applied during Year 1, Year 2, Year 3, and Year 4 of the study were applied
to those analytes determined by Methods 1632A and 1656A, with the majority applied to
MMA (during Year 1 of the study). The number of MMA qualifiers in Year 1 of the study
suggested that method improvements were needed to yield optimal results when determining
April 2005 26 QA Report for the National Lake Fish Tissue Study
-------
this form of arsenic and certain pesticides in tissue matrices. Improvements were
implemented during the analysis of Year 2, Year 3, and Year 4 samples, and zero (0) MS
failures were reported for MMA during Year 2, Year 3, and Year 4 of the study.
Disulfoton was not recovered at all in the first MS/MSD pair analyzed in Year 1 of the study
and in another pair analyzed in Year 4; also, for the Year 1 samples, there were low
recoveries of the same compound in the ongoing precision and recovery sample (OPR, see
section 4.2.5) analyzed with the batch. Disulfoton was not reported in any of the samples nor
could it be seen in a careful review of the sample chromatograms. Because it is impossible
to determine if the non-detects reported in the samples are valid, results for all 28 samples
were excluded from the database. These samples were flagged with "LOPR, NMSR"
(Year 1 samples) to indicate that there was no MS recovery and low OPR recovery and "NMSR"
(Year 4 samples) to indicate that there was no MS recovery, respectively; all samples were
further qualified with "Exclude" to explain the absence of results. Data users should note,
however, that this compound was not detected in any of the samples analyzed in Year 1,
Year 2, Year 3, or Year 4, even when laboratory and method performance was in control.
4.2.5 Ongoing Precision and Recovery
Summary of OPR Qualifiers - Years 1,2,3, and 4
Results not affected by OPR issues (99.51 %)
Qualified results with OPR flags (0.32%)
Excluded results with OPR flags (0.17%)
All laboratories that participated in the
National Lake Fish Tissue Study were
required to prepare and analyze an ongoing
precision and recovery, or OPR, standard
with each sample set and to meet OPR
acceptance criteria specified in the 1600-
series method used to analyze the samples.
The OPR standards are identical to those
used in the IPR demonstrations discussed in
Section 4.1.2.
A single OPR standard is analyzed with
each sample batch to verify that laboratory
performance is in control during the
analysis of field tissue samples. Whereas
the calibration verification (discussed in
Section 4.2.1) allows verification that the instrument remains in control during analysis of each
batch, the OPR allows verification that the entire analytical process, including the instrumentation,
is in control. Likewise, the OPR differs from the MS/MSD or labeled analog spikes described in
section 4.2.4 in that the OPR is performed in a reference matrix to verify that all laboratory systems
are in control, whereas the MS/MSD and labeled analogs are spiked into actual samples to verify
that the method is working as expected in the actual matrices analyzed.
Each Year 1, Year 2, Year 3, and Year 4 data package generated in the National Lake Fish
Tissue Study was reviewed to verify that OPR samples were prepared and analyzed at the
required frequency and that recoveries met acceptable performance criteria. All of the methods
used in this study list OPR acceptance criteria that are applicable to fish tissue, except for
Method 1625C. Therefore, OPR data generated for Method 1625C were assessed against the
OPR criteria specified in the method for aqueous samples. In evaluating OPR results, data
reviewers applied the following rules:
QA Report for the National Lake Fish Tissue Study 27
April 2005
-------
If the OPR results were only marginally outside the method criteria (i.e., results were within
10% of the method-specified windows), the data were considered to be acceptable without
qualification. This is because all OPR windows established for the 1600-series methods used
in this study reflect a 95% confidence interval.
If the OPR recovery was below method-specified criteria, all tissue data associated with that
OPR were qualified with either "LOPR" to indicate that a low OPR recovery of the target
analyte, "LLRO" to indicate a low labeled analog recovery from the OPR, or "LNRO" to
indicate a low native compound recovery from the OPR. Results with low target analyte
recoveries (i.e., "LOPR" results) were further qualified to indicate a "Potential Low Bias."
In Year 1, Year 2, Year 3, and Year 4, "LOPR" results were observed for 1.61% of the
pesticides measurements, several of which were associated with other data qualifiers (e.g.,
data also were reported below the ML). One semivolatile organic result was flagged with
"LLRO" due to low labeled compound recovery in the OPR. This sample was also flagged
with other data qualifiers ("B", "LLBL", and "RNAF"), as well. Results with low native
compound recoveries in the OPR (i.e., "LNRO" results) were further qualified to indicate a
"Potential Low Bias." In Year 1, Year 2, Year 3, and Year 4, "LNRO" results were observed
for 0.12% of the semivolatile organics measurements.
In rare cases, the native compound was not recovered from the OPR sample or there was no
OPR spike recovery. For these cases, the associated results were excluded from the
database. The associated sample records were coded as either "NNRO" or "NOPR,"
respectively, and further qualified with "Exclude" to explain the absence of any results. This
situation occurred for 1.14% of the semivolatile organic measurements made by Method
1625C (the vast majority of which were associated with the analyte tetrabromobisphenol A)
and 0.03% of the pesticide measurements made by Method 1656A and 1657A.
4.2.6 Holding Time Assessments
Each data submission was reviewed to verify that all samples were received and maintained in
frozen condition until analysis. When samples were thawed, the time between thawing and
sample digestion, extraction, or analysis was assessed to verify that the holding times specified
in each method were met. Samples that failed to meet the required holding times were coded
with "HTEX" to indicate the holding time exceedance. If no other QC failures occurred, the data
were coded to indicate a "Potential Low Bias." One semivolatile organics result (in Year 1) was
analyzed outside the holding time and was reported below the ML. This result was coded with
both "HTEX" and "J," and was further qualified as "Estimated Value." Thirteen other results
also were coded with an "HTEX" flag, but these results were excluded for other reasons
previously cited in this report. Finally, one semivolatile organics result (in Year 4) was analyzed
outside the holding time and was associated with a low native recovery in the OPR. This result
was coded with both "HTEX" and "LNRO," and was further qualified as "Potential Low Bias."
99.99% of the 284,973 sample results reported during Year 1 through Year 4 were
determined within analytical holding times.
April 2005 28 QA Report for the 'National Lake Fish Tissue Study
-------
4.2.7 Method-Specific Considerations
Pesticides
Methods 1656A and 1657A rely on the use of GC techniques coupled with selective detectors to
determine organochlorine and organophosphorus pesticides, respectively. The advantage of
these techniques is that they allow for measurement of pesticides at lower concentrations than is
possible with GC/MS. The disadvantage is that compound identification with the selective
detectors is not as reliable as compound identification via GC/MS. To overcome this
disadvantage, the methods require that any analytes detected be verified by analysis on a second
column and that the results from the second column closely agree with the primary column
results (the results measured on the two columns must agree within a factor of two).
When such verification occurs, the laboratory reports the lower of the two results. In reviewing
the data package for each pesticide analysis, reviewers evaluated the data to verify that all
positive results were detected on two columns and that the results agreed within a factor of two.
When the analyte was detected on both columns, but the results did not closely agree, the lower
result was reported in the database and coded with "RNF2" to indicate that the results did not
agree within a factor of two and further qualified as an "Estimated Value." When the analyte
was detected on one column but could not be confirmed on a second column, the result was
coded "NCNF" to indicate the lack of confirmation and further qualified as an "Estimated
Value." This approach was taken to advise data users about the possibility that an analyte might
be present so that informed decisions can be made about the value of gathering additional data.
For all pesticides, except kepone, unconfirmed results were reported as non-detects unless
confirmation on the second column was not possible due to co-elution. Because kepone cannot
be independently resolved on the second column, unconfirmed kepone results were reported as
detects, but qualified as described above. Several possible kepone hits are reported and flagged
as "NCNF" in the database. One Year 1 result for trans-nonachlor and one 2,4'-DDE result from
Year 2 could not be confirmed because the compound co-eluted with another analyte on the
second column. The co-eluted results also differed from the first column result by more than a
factor of two. These results were flagged with "NCNF, RNF2" and further coded as an
"Estimated Value."
In addition, for Year 2, Year 3, and Year 4 of the study, several dicofol results were coded with
"NCNF." The Year 2 results were also associated with high and low MS recoveries ("HMSR"
and "LMSR", respectively), and one of the sample results was further coded with a "J" flag.
These two results were given the overall qualifier of "Estimated Value." Finally, several 2,4'-
DDE, alpha-chlordane, c/'s-nonachlor, ethalfluralin, gamma-chlordane, trans-nonachlor, and
trifluralin results in the Year 2, Year 3, and Year 4 data sets were coded with "NCNF" (and
possibly other flags), which produced an overall qualification of "Estimated Value."
Note: EPA reminds all data users that the identification of any pesticide qualified with "NCNF"
or "RNF2" was not confirmed. Appropriate caution should be exercised when using such
results.
Also, in Year 1 of the study, seven pesticide results were coded "MTRX" to indicate that the
sample chromatograms suggested the presence of possible matrix interferences. In three of these
cases, the pesticide was detected on both columns, but the results did not agree within a factor of
two, so the data were further coded with an "RNF2" flag. One of these three results also was
QA Report for the National Lake Fish Tissue Study 29 April 2005
-------
reported below the ML on one of the columns, so this result was further coded with a "J" flag.
All six of these MTRX flagged results were qualified to indicate that they should be considered
"Estimated Values." In Year 4 of the study, one pesticide result for dicofol was qualified with
the "MTRX" code and should be considered an "Estimated Value." In Year 3 of the study, 25
PCB congener results were flagged with the "MTRX" code. These results were further coded
with "LLBL" or "NLBL" flags (See Section 4.2.4 - Spiked Sample Recoveries), and were
qualified as an "Estimated Value" or "Exclude" from the database, respectively. There were no
"MTRX" codes used for the results obtained from Year 2 of the study.
Semivolatile Organics
During Year 4 of the study, the laboratory initially reported "hits" for 4-nonylphenol in several
of the samples. Since only one sample had a detected result for 4-nonylphenol during the first
three years of the study, the semivolatile organics laboratory investigated the situation and
identified several possible explanations, including 1) the possibility of contamination, 2) the fact
that the interpretation was performed by a different lab staff member than in previous years, and
3) the less likely possibility that the hits were actually present in the samples.
Based on re-evaluation of Year 4 data and comparison to results from previous years, it was
concluded that, for all but 7 samples, the compound detected did not meet appropriate
identification criteria and should not be reported as 4-nonylphenol. It is not possible to tell with
certainty whether or not 4-nonylphenol was actually present in any of the samples. The
laboratory reported that the remaining "hits" for 4-nonylphenol were probably due to
interference and/or contamination. The laboratory performed an independent study in an attempt
to isolate the source of the contamination, evaluating the following:
the procedures used to clean and prepare the Soxhlets for extraction
the soap used to clean the extraction glassware
the Florisil/alumina clean-up columns used during the extraction process
The results of all of their tests proved to be inconclusive. The sample results in which
4-nonylphenol was detected were reported as "hits" based on best professional judgement.
However, these samples with concentrations of 4-nonylphenol detected above the MDL were
qualified with a "SLIC" code to signify the result was probably due to interference and/or
contamination.
SECTION 4.3 OVERALL DATA QUALITY ASSESSMENT
Upon completion of all data review and database development activities for Year 1, Year 2,
Year 3, and Year 4, the full data set was evaluated to determine if the results overall were falling
within the MQOs established in the study QAPP 3. Assessment of the data against these MQOs
is described in Sections 4.3.1 through 4.3.5 below.
Quality Assurance Project Plan for Analytical Control and Assessment Activities, September 2000,
Prepared for the U.S. EPA OW by CSC Environmental Programs Group (formerly DynCorp Environmental).
April 2005 30 QA Report for the National Lake Fish Tissue Study
-------
4.3.1 Precision
Precision is the degree of agreement among replicate measurements of the same property, under
prescribed similar conditions. It can be expressed either as a range, a standard deviation, or a
percentage of the mean of the measurements.
Ideally, precision is measured by subdividing samples in the field, preserving and numbering
each split separately, and sending the aliquots to the analysis laboratory as "blind" duplicates. In
this study, however, samples have to be homogenized, composited, and subdivided in a strictly
controlled, clean laboratory environment. Therefore, the Sample Preparation Laboratory was
required to prepare "duplicate composite pairs" on 5% of the samples analyzed. These duplicate
composite pairs were sent to each analysis laboratory as "blind duplicates" (e.g., labeled with
separate EPA sample numbers) and used to assess variability arising from the sample
homogenization, compositing, aliquoting, shipping, and laboratory analysis processes.
Because agreement between results was expected to be better at higher concentrations, two
MQOs were established. For sample results that were close to (i.e., less than 5 times) the ML,
the MQO was that 90% of results from the original sample and the blind composite duplicates
should agree within ±100%. For sample results that were well above (i.e., more than 5 times
higher than) the ML, the MQO was that at least 90% of the results from the original and blind
composite duplicate samples agree within ±50%.
A total of 54 blind composite duplicates were prepared and analyzed for all of the target analytes
during Year 1, Year 2, Year 3, and Year 4 of the study. These analyses yielded approximately
15,350 pairs of results; of these, more than 7,050 pairs of results were detected in both the field
tissue sample and its blind composite duplicate. (Agreement between non-detects could not be
quantitatively analyzed and therefore is not included in this discussion.)
The study MQO for precision was exceeded during Year 1, Year 2, Year 3, and Year 4 of
the study. Over 99% (99.8%) of the duplicate composite pairs agreed within ±50% when
the measured results were more than 5 times the ML, and 99.7% of the duplicate
composite pairs agreed within ±100% when measured results were less than 5 times the
ML.
Approximately 4,240 of these paired results were detected at concentrations that were at least 5
times greater than the ML, and all but eight (99.8%) of these results met the 50% MQO
established for agreement between analytes detected at this level. The remaining pairs
(approximately 2,810 paired results) were detected within 5 times the ML and subjected to the
±100% objective described above. All but five of these paired results (99.8% of the results),
three in Year 1, one in Year 2, and three in Year 4 of the study, met the MQO. All of these
paired results met the MQO for Year 3 of the study.
QA Report for the National Lake Fish Tissue Study 31 April 2005
-------
4.3.2 Bias
Bias is the systematic distortion of a measurement process that causes errors in one direction.4 In
this study, bias from the analytical process was measured by preparing and analyzing field tissue
samples spiked with 1) the analytes of interest (i.e., MS samples), 2) isotopically labeled analogs
of the target analytes, or 3) surrogate compounds that are expected to behave in a manner similar
to the target analytes. Assessment of these spiked sample results was described in Section 4.2.4.
The MQO for overall analytical accuracy in this study was for 80% of these spiked field tissue
sample results to fall within the acceptance criteria specified for each method.
This goal was easily exceeded in all four years of the study, suggesting that overall, the methods
selected for use in the study were appropriate for the analytes and matrices targeted.
The study MQO for bias was exceeded during Year 1, Year 2, Year 3, and Year 4 of the
study; 99.7% of the spiked sample results fell within method-specified acceptance
criteria.
Although the MQO was established for the entire data set, it is useful from a QA perspective to
evaluate bias across individual methods to verify that each method is working as planned. On
this basis, it is clear that the individual methods are working as intended. No MS failures were
observed for mercury; minimal spike (MS and surrogate samples) failures (0.19%) were
observed for organophosphorus pesticides, and 98.91% of the spiked (MS and surrogate
samples) organochlorine pesticide results met method-specified criteria.
In Year 1 of the study, MS samples performed for arsenic species also showed that, although the
method works well in tissue for total inorganic arsenic and most of the other species measured,
additional improvements were needed to yield optimal results for MMA. Such improvements
were implemented during the Year 2 analysis, and no (0) MS failures were reported for MMA
during Year 2, Year 3, or Year 4 of the study. Finally, all of the methods that involve spiking
isotopically labeled compounds into each sample yielded a success rate of 99.6%, with only
0.02%, 0.02%, and 0.15% of the labeled analog recoveries failing to meet method specified
recoveries for PCBs, dioxins/furans, and semivolatile organics, respectively.
4.3.3 Accuracy
Accuracy is a measure of the closeness of an individual measurement or the average of a number
of measurements to the true value. Accuracy includes a combination of random error (precision)
and systematic error (bias) components that result from sampling and analytical operations.
Accuracy is determined by analyzing a reference material of known analyte concentrations or by
reanalyzing a sample spiked with a known amount of analyte. Study objectives dictated that
certified reference materials (CRMs) be sent to each laboratory to assess bias when available,
feasible, and warranted. No certified reference materials were sent during Year 1, Year 2,
Year 3, or Year 4 of the study.
4EPA Guidance for Quality Assurance Project Plans, EPA QA/G-5, EPA Office of Research and
Development, Washington, DC, 20460. EPA/600/R-98/018.
April 2005 32 QA Report for the National Lake Fish Tissue Study
-------
4.3.4 Sensitivity
As described in Section 4.1.1, all laboratories demonstrated their ability to achieve the study
MQOs established for sensitivity, and 100% of the results generated during Year 1, Year 2,
Year 3, and Year 4 of the study were generated on instruments that were calibrated to encompass
MLs. In addition, all results reported below the ML have been qualified in the database to alert
data users that, although the presence of the analyte was confirmed, the concentration reported in
the database is an estimate because it falls below the quantitation limit.
4.3.5 Completeness
Completeness is defined in terms of the percentage of data that are collected and deemed to be
acceptable for use in the study. Three measures of completeness can be defined, as follows:
Sampling Completeness: The number of valid samples collected relative to the number of
samples planned for collection;
Analytical Completeness: The number of valid sample measurements relative to the number
of valid samples collected; and
Overall Completeness: The number of valid sample measurements relative to the number
of samples planned for collection.
The analytical completeness goal in this study was that EPA obtain valid measurements from
90% of the valid samples collected. This goal was exceeded during Year 1, Year 2, Year 3, and
Year 4 of the study. Only two of the 1003 valid samples collected during the four years of the
study could not be completely analyzed because they did not provide sufficient tissue. This
resulted in EPA obtaining valid measurements from 99.8% of the valid samples collected in
Year 1 through Year 4 of the study.
EPA obtained valid measurements from 99.8% of the samples collected in Year 1, Year 2,
Year 3, and Year 4 of the study, thereby exceeding the MQO of 90% for analytical
completeness during the study.
QA Report for the National Lake Fish Tissue Study 33 April 2005
-------
Exhibit 4-1A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
Full Length
Description or
Recommendation
(Comment)
Explanation of Code (Reason)
Notes on When This Occurs
B
Blank
Contamination
The target analyte was detected in one or more of the
blanks associated with this sample.
Applied to fewer than 0.05% of Year 1, Year 2, Year 3,
and Year 4 results and only occurs for PCB congener
totals in which some of the individual totals were
associated with B flags.
B,J
Blank
Contamination,
Estimated Value
One or more of the PCB congeners contributing to the total
was associated with a contaminated blank and one or more
of the congeners was reported above the MDL and below
the ML. In most cases, the impact of the flag on the total is
negligible due to the relatively minor impact of the
individual congener(s) the flags represent.
Applied to 0.78% of Year 1, Year 2, Year 3, and Year 4
results, but only for PCB congener totals.
This combination only occurs for PCB congener totals in
which B-flagged and J-flagged results from individual
congeners are mathematically summed with unflagged
results.
B,J,
RMAX
Blank
Contamination,
Estimated Value,
Result is a
Maximum Value
The target analyte was detected in one or more of the
blanks associated with this sample; the result was reported
above the MDL but below the ML and is, therefore, an
estimated value; and blank contamination was observed
and the target analyte was reported in the sample at a
concentration between 5 and 10 times higher than the blank
value. The result was considered to be of acceptable
quality, but data users are cautioned that it may be a
maximum value due to possible influence of contamination.
Applied to 0.43% of the dioxin results in Year 1 through
Year 4 of the study. This qualification occurs only
because the sample result was exactly five times or ten
times the blank contamination and the dioxin MDLs were
more than ten times lower than the MLs at which the
instruments were calibrated.
B,J,
RNAF
Blank
Contamination,
Estimated Value,
Result Not Affected
The target analyte was detected in one or more of the
blanks associated with this sample; the result was reported
above the MDL but below the ML and is, therefore, an
estimated value; and blank contamination was present but
was not considered to adversely impact the sample result.
The presence of the analyte in the blank is not considered
to adversely affect the data in cases where the sample
results are more than 10 times the associated blank results
or where the analyte is not detected in associated samples.
Applied to one 1,2,3,4,6,7,8-HpCDD result and three
2,3,7,8-TCDD results in the Year 1 data set.
April 2005
34
QA Report for the National Lake Fish Tissue Study
-------
Exhibit 4-1A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
Full Length
Description or
Recommendation
(Comment)
Explanation of Code (Reason)
Notes on When This Occurs
B, LLBL,
RNAF
Blank
Contamination,
Estimated Value,
Result Not Affected
Blank contamination was present but was not considered to
adversely impact the sample result. The presence of the
analyte in the blank is not considered to adversely affect the
data in cases where the sample results are more than 10
times the associated blank results or where the analyte is
not detected in associated samples. However, the labeled
analog of the target analyte was recovered from the field
tissue sample below method-specified criteria, resulting in
an estimated value flag.
Applied to two OCDD results in the Year 2 data set.
B, REXC,
RNAF
Blank
Contamination,
Estimated Value,
Result Not Affected
The target analyte was detected in one or more of the
blanks associated with this sample; the field tissue sample
result exceeded the instrument calibration range; and the
blank contamination that was present was not considered to
adversely impact the sample result. The presence of the
analyte in the blank is not considered to adversely affect the
data in cases where the sample results are more than 10
times the associated blank results or where the analyte is
not detected in associated samples.
Applied to two bis(2-ethylhexyl)phthalate results in the
Year 2 data set.
B,RMAX
Blank
Contamination,
Result is a
Maximum Value
Blank contamination was observed and the target analyte
was reported in the sample at a concentration between 5
and 10 times higher than the blank value. The result was
considered to be of acceptable quality, but data users are
cautioned that it may be a maximum value due to possible
influence of contamination.
Applied to 0.15% of Year 1, Year 2, Year 3, and Year 4
results, including PCB congener, dioxin/furan, pesticide,
and target semivolatile organic results.
QA Report for the National Lake Fish Tissue Study
35
April 2005
-------
Exhibit 4-1A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
Full Length
Description or
Recommendation
(Comment)
Explanation of Code (Reason)
Notes on When This Occurs
B,RMAX,
RNF2
Blank
Contamination,
Result is a
Maximum Value,
Estimated Value
Blank contamination was observed and the target analyte
was reported in the sample at a concentration between 5
and 10 times higher than the blank value; the result was
considered to be of acceptable quality, but data users are
cautioned that it may be a maximum value due to possible
influence of contamination; and although the analyte was
found in the field tissue sample using two different
columns, the results reported on the two columns differed
by a factor of more than two.
Occurred with one beta-BHC result in the Year 4 data set.
B, RNAF
Blank
Contamination,
Result Not Affected
Blank contamination was present but was not considered to
adversely impact the sample result. The presence of the
analyte in the blank is not considered to adversely affect the
data in cases where the sample results are more than 10
times the associated blank results or where the analyte is
not detected in associated samples.
Applied to 1.19% of the Year 1, Year 2, Year 3,
Year 4 results.
and
B,RNON
Blank
Contamination,
Result Reported as a
Non-detect
When the sample result is less than five times the blank
result, there are no means by which to ascertain whether or
not the presence of the analyte may be attributed to
contamination. Therefore, SCC recommends that the data
be reported in the database as a non-detect at the ML,
adjusted for dilution.
Applied to 0.98% of Year 1, Year 2, Year 3, and Year 4
results, with most of the occurrences for PCB congeners
other than the 12 dioxin-like congeners.
HLBL
Estimated Value
The labeled analog of the target compound was recovered
above method-specified criteria, suggesting a possible
matrix interference (High Labeled Compound Recovery).
Applied to two results in the Year 1 PCB data set.
HMSR
Potential High Bias
High analyte recovery was observed in one or more MS
samples associated with this result.
Applied to two Arsenic (III) results and two 6 -BHC
results in the Year 1 data set; and applied to six cis-
nonachlor results in the Year 2 data set.
HMSR, J
Estimated Value
A high recovery of this result was observed in an associated
MS sample suggesting the J-flagged result may have a high
bias and the sample result reported was above the MDL but
below the ML.
Applied to three Arsenic (III) results and two 6 -BHC
results in the Year 1 data set.
April 2005
36
QA Report for the National Lake Fish Tissue Study
-------
Exhibit 4-1A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
Full Length
Description or
Recommendation
(Comment)
Explanation of Code (Reason)
Notes on When This Occurs
HMSR, J,
NCNF
Estimated Value
High analyte recovery was observed in one or more MS
samples associated with this result; the sample result
reported was above the MDL (detection limit) but below
the ML (quantitation limit); and the presence of the
reported analyte was not confirmed on a second column.
Applied to one dicofol result in the Year 2 data set.
HMSR, J,
RNF2
Estimated Value
High analyte recovery was observed in one or more MS
samples associated with this result; the sample result
reported was above the MDL (detection limit) but below
the ML (quantitation limit); and although the analyte was
found in the field tissue sample using two different
columns, the results reported on the two columns differed
by a factor of more than two.
Applied to one c/'s-nonachlor result in the Year 2 data set.
HMSR,
RNF2
Estimated Value
High analyte recovery was observed in one or more MS
samples associated with this result and although the analyte
was found in the field tissue sample using two different
columns, the results reported on the two columns differed
by a factor of more than two.
Applied to three c/'s-nonachlor results in the Year 2 data
set.
HMSR,
RNF2,
RPDX
Estimated Value
High analyte recovery was observed in one or more MS
samples associated with this result and although the analyte
was found in the field tissue sample using two different
columns, the results reported on the two columns differed
by a factor of more than two. Also, the RPD between the
MS and MSD exceeded criteria.
Applied to one 4,4'-DDT result in the Year 2 data set.
HMSR,
RPDX
Estimated Value
High analyte recovery was observed in one or more MS
samples associated with this result and the RPD between
the MS and MSD exceeded criteria.
Applied to four 4,4'-DDT results in the Year 2 data set.
HSSR
Potential High Bias
A high surrogate spike recovery was observed, suggesting a
possible high bias in the result reported.
Occurred for one 4,4'-DDE result in the Year 1 data set.
QA Report for the National Lake Fish Tissue Study
37
April 2005
-------
Exhibit 4-1A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
Full Length
Description or
Recommendation
(Comment)
Explanation of Code (Reason)
Notes on When This Occurs
HSSR, J
Estimated Value
A high recovery of the surrogate spike compound was
associated with the reported result and the sample result
was reported above the MDL but below the ML on at least
one of the two columns on which the pesticide was
detected.
Applied to one cc-chlordane result in the Year 1 data set.
HTEX, J
Estimated Value
The sample result reported was above the MDL (detection
limit) but below the ML (quantitation limit) and the holding
time was exceeded when analyzing the sample.
Associated with one bis(2-ethylhexyl) phthalate result in
the Year Idata set.
HTEX,
LNRO
Potential Low Bias
The holding time was exceeded when analyzing the sample
and the native compound of the target analyte was
recovered below method criteria in the OPR associated
with the sample.
Occurred for one naphthalene result in the Year 4 data
set.
HTEX,
NLBL
Exclude
The holding time was exceeded when analyzing the sample
and the labeled analog was not recovered from the sample,
suggesting severe matrix interferences.
Occurred for one pentachlorophenol result in the Year 4
data set.
HTEX,
NNRO
Exclude
The holding time was exceeded when analyzing the
sample; and the target analyte (native compound) was not
recovered in the OPR sample associated with this result.
Applied to seven tetrabrormobisphenol A results in the
Year 4 data set.
HVER, J,
RNF2
Estimated Value
The sample result was determined on both pesticide
columns, but the lower of the two results reported was
below the quantitation limit, and the results reported were
not within a factor of two of each other (possibly due to the
low level reported). In addition, the lower of the two
values was associated with a high calibration verification
standard recovery.
Occurred for one 4,4'-DDT result in the Year 1 data set.
HVER,
RNF2
Estimated Value
The sample result was determined on both pesticide
columns, but the results reported were not within a factor of
two of each other and the lower of the two values was
associated with a high calibration verification standard
recovery.
Occurred for two methoxychlor results in the Year 1 data
set.
April 2005
38
QA Report for the National Lake Fish Tissue Study
-------
Exhibit 4-1A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
Full Length
Description or
Recommendation
(Comment)
Explanation of Code (Reason)
Notes on When This Occurs
J
Estimated Value
The sample result reported was above the MDL (detection
limit) but below the ML (quantitation limit).
Occurred throughout the Year 1, Year 2, Year 3, and
Year 4 data sets.
J, LLBL
Estimated Value
The target analyte was reported above the MDL but below
the ML; and the labeled analog of the target was recovered
below method-specified criteria suggesting the possible
presence of matrix interferences.
Applied to one HxCDD result and one PCB congener
(PCB-3) result in the Year 1 and Year 3 data sets,
respectively.
J, LLBL,
MTRX
Estimated Value
The target analyte was reported above the MDL but below
the ML; the labeled analog of the target was recovered
below method-specified criteria, suggesting the possible
presence of matrix interferences; and the chromatogram
suggested possible matrix interferences with the sample.
Applied to three PCB congener (two PCB-3 and
one PCB-8) results in the Year 3 data set.
J, LMSR
Estimated Value
The target analyte was reported above the MDL but below
the ML in the sample and low analyte recovery was
observed with one or more MS samples associated with this
result.
Applied to one oxychlordane result in the Year 2 data set
and four trifluralin results in the Year 4 data set.
J, LMSR,
LOPR
Estimated Value,
Potential Low Bias
The target analyte was reported above the MDL but below
the ML in the sample; low analyte recovery was observed
with one or more MS samples associated with this result;
and the target analyte was recovered below method-
specified criteria in the OPR associated with the sample.
Occurred for one ethalflurarin result in the Year 4 data
set.
J, LMSR,
RNF2
Estimated Value
The target analyte was reported above the MDL but below
the ML in the sample; low analyte recovery was observed
with one or more MS samples associated with this result;
and the sample result was determined on both pesticide
columns, but the results reported were not within a factor of
two of each other.
Applied to one oxychlordane and one 2,4'-DDT result in
the Year 2 data set. Also, applied to six trifluralin results
in the Year 4 data set.
QA Report for the National Lake Fish Tissue Study
39
April 2005
-------
Exhibit 4-1A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
Full Length
Description or
Recommendation
(Comment)
Explanation of Code (Reason)
Notes on When This Occurs
J, LMSR,
RNF2,
RPDX
Estimated Value
The target analyte was reported above the MDL but below
the ML in the sample; low analyte recovery was observed
with one or more MS samples associated with this result;
the sample result was determined on both pesticide
columns, but the results reported were not within a factor of
two of each other; and the RPD between the MS and MSD
exceeded criteria.
Occurred for one methoxychlor result in the Year 3 data
set and one trifluralin result in the Year 4 data set.
J, LMSR,
RPDX
Estimated Value
The target analyte was reported above the MDL but below
the ML in the sample; low analyte recovery was observed
with one or more MS samples associated with this result;
and the RPD between the MS and MSD exceeded criteria.
Applied to two 4,4'-DDT results in the Year 3 data set.
J, LOPR
Estimated Value,
Potential Low Bias
The target analyte was reported above the MDL but below
the ML and the target analyte was recovered below
method-specified criteria in the OPR associated with the
sample.
Applied to two 4,4'-DDT results in the Year 1 data set,
one methoxychlor result in the Year 2 data set, and four
ethalfluralin results in the Year 1, Year 3, and Year 4 data
sets.
J, LOPR,
NCNF
Estimated Value
The target analyte was reported above the MDL but below
the ML; the target analyte was recovered below method-
specified criteria in the OPR associated with the sample;
and the sample result was not confirmed on a second
column.
Applied to five kepone results in the Year 1 data set and
one ethalfluralin results in the Year 3 data set.
J, LOPR,
NCNF,
RNF2
Estimated Value
The target analyte was reported above the MDL but below
the ML; the target analyte was recovered below method-
specified criteria in the OPR associated with the sample;
the sample result was not confirmed on a second column;
and although this analyte was found in the field tissue
sample using two different columns, the results reported on
the two columns differed by a factor of more than two.
Applied to one etharfluralin result in the Year 3 data set.
April 2005
40
QA Report for the National Lake Fish Tissue Study
-------
Exhibit 4-1A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
Full Length
Description or
Recommendation
(Comment)
Explanation of Code (Reason)
Notes on When This Occurs
J, LOPR,
NCNF,
RNF2,
RPDX
Estimated Value
The target analyte was reported above the MDL but below
the ML; the target analyte was recovered below method-
specified criteria in the OPR associated with the sample;
the sample result was not confirmed on a second column;
although this analyte was found in the field tissue sample
using two different columns, the results reported on the two
columns differed by a factor of more than two; and the
RPD between the MS and MSD exceeded criteria.
Applied to four ethalfluralin results in the Year 3 data set.
J, LOPR,
NCNF,
RPDX
Estimated Value
The target analyte was reported above the MDL but below
the ML; the target analyte was recovered below method-
specified criteria in the OPR associated with the sample;
the sample result was not confirmed on a second column;
and the RPD between the MS and MSD exceeded criteria.
Occurred for one ethalflurarin result in the Year 3 data
set.
J, LOPR,
RNF2
Estimated Value,
Potential Low Bias
The target analyte was reported above the MDL but below
the ML; the target analyte was recovered below method-
specified criteria in the OPR associated with the sample;
and although this analyte was found in the field tissue
sample using two different columns, the results reported on
the two columns differed by a factor of more than two.
Applied to two 4,4'-DDT results in the Year 1 data set
and twenty-two ethalfluralin results in the Year 3 and
Year 4 data sets.
J, LVER
Estimated Value,
Potential Low Bias
The target analyte was reported above the MDL but below
the ML and low analyte recovery was observed in a
calibration verification associated with this sample,
suggesting the possibility of a low bias in the result.
Applied to three MMA results in the Year 2 data set.
J,MTRX,
RNF2
Estimated Value
The target analyte was reported above the MDL but below
the ML; the chromatogram suggested possible matrix
interferences with this sample; and although this analyte
was found using two different columns, the results reported
on the two columns differed by a factor of more than two.
Applied to one c/'s-permethrin result in the Year 1 data
set.
J, NCNF
Estimated Value
The target analyte was reported above the MDL but below
the ML and the result was not confirmed on a second
column.
Applied to a few pesticide results (0.23%) in Year 1,
Year 2, Year 3, and Year 4.
QA Report for the National Lake Fish Tissue Study
41
April 2005
-------
Exhibit 4-1A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
Full Length
Description or
Recommendation
(Comment)
Explanation of Code (Reason)
Notes on When This Occurs
J, NCNF,
RNF2
Estimated Value
The target analyte was reported above the MDL but below
the ML; the result was not confirmed on a second column;
and although this analyte was found using two different
columns, the results reported on the two columns differed
by a factor of more than two.
Applied to one cis-nonachlor, six 2,4'-DDE results, and
six trifluralin results in the Year 3 data set.
J, NCNF,
RNF2,
RPDX
Estimated Value
The target analyte was reported above the MDL but below
the ML; the result was not confirmed on a second column;
although the analyte was found using two different
columns, the results reported on the two columns differed
by a factor of more than two; and the RPD between the MS
and MSD exceeded criteria.
Applied to four trifluralin results in the Year 3 data set.
J, RNF2
Estimated Value
The target analyte was reported above the MDL but below
the ML and although this analyte was found using two
different columns, the results reported on the two columns
differed by a factor of more than two.
Applied to a few pesticides results (0.
Year 2, Year 3, and Year 4.
/o) in Year 1,
J, RNF2,
RPDX
Estimated Value
The target analyte was reported above the MDL but below
the ML in the sample; the sample results reported on the
two columns differed by a factor of more than two; and the
RPD between the MS and matrix spike duplicate exceeded
criteria.
Applied to four 2,4'-DDE results, one heptachlor result,
one pendimethalin result, eight ethalfluralin, and seven
trifluralin results in the Year 2, Year 3, and Year 4 data
sets.
J, RPDX
Estimated Value
The target analyte was reported above the MDL but below
the ML in the sample and the RPD between the associated
MS and MSD exceeded criteria.
Applied to two 2,4'-DDE results, one 4,4'-DDT result,
two endosulfan sulfate results, three pendimethalin
results, and two trifluralin results in the Year 2, Year 3,
and Year 4 data sets.
LLBL
Estimated Value
The labeled analog of the target analyte was recovered
below method-specified criteria.
Applied to two PCB congener results (PCB-4 and
PCB-19) in the Year 3 data set.
LLBL,
MTRX
Estimated Value
The labeled analog of the target analyte was recovered
below method-specified criteria and the chromatogram
suggested possible matrix interferences with the sample.
Applied to six PCB congener results in the Year 3 data
set.
April 2005
42
QA Report for the National Lake Fish Tissue Study
-------
Exhibit 4-1 A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
LMSR
LMSR,
LOPR
LMSR,
LOPR,
RNF2
LMSR,
NCNF
LMSR,
RNF2
LNRO
LNRO,
LVER
LOPR
LOPR,
NCNF
Full Length
Description or
Recommendation
(Comment)
Potential Low Bias
Potential Low Bias
Estimated Value
Estimated Value
Estimated Value
Potential Low Bias
Potential Low Bias
Potential Low Bias
Estimated Value
Explanation of Code (Reason)
Low analyte recovery was observed with one or more MS
samples associated with this result.
Low analyte recovery was observed with one or more MS
samples and with the OPR sample associated with this
result.
Low analyte recovery was observed with one or more MS
samples and with the OPR sample associated with this
result; also, the sample results reported on the two columns
differed by a factor of more than two.
Low analyte recovery was observed with one or more MS
samples and the result was not confirmed on a second
column.
Low analyte recovery was observed with one or more MS
samples and the sample results reported on the two
columns differed by a factor of more than two.
The native compound of the target analyte was recovered
below method-specified criteria in the OPR sample.
The native compound of the target analyte was recovered
below method-specified criteria in the OPR sample and low
analyte recovery was observed in a calibration verification
associated with this sample, suggesting the possibility of a
low bias in the result.
Low analyte recovery was observed with the OPR sample
associated with this result.
Low analyte recovery was observed with the OPR sample
associated with this result and the presence of the reported
analyte was not confirmed on a second column.
Notes on When This Occurs
Applied primarily to MMA results and a small number of
total inorganic arsenic results in the Year 1 data set; also
applied to a few pesticide results (0.43%) in the Year 2,
Year 3, and Year 4 data sets.
Applied to a few pesticide results (0.28%) in the Year 1,
Year 2, Year 3, and Year 4 data sets.
Occurred for two ethalfluralin results in the Year 4 data
set.
Applied to one dicofol result in the Year 2 data set and
two kepone results in the Year 3 data set.
Occurred for five trifluralin results in the Year 4 data set.
Applied to five indeno(l,2,3-cd)pyrene and twenty -one
naphthalene results in the Year 4 data set.
Applied to nineteen phenol results in the Year 4 data set.
Applied to several organochlorine and organophosphorus
pesticide results (1.1%) in the Year 1, Year 2, Year 3, and
Year 4 data sets.
Applied to four kepone results (kepone cannot be
confirmed on the method-specified second column) in the
Year 1 data set.
QA Report for the National Lake Fish Tissue Study
43
April 2005
-------
Exhibit 4-1A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
Full Length
Description or
Recommendation
(Comment)
Explanation of Code (Reason)
Notes on When This Occurs
LOPR,
NCNF,
RNF2
Estimated Value
Low analyte recovery was observed with the OPR sample
associated with this result; the presence of the reported
analyte was not confirmed on a second column; and the
sample results reported on the two columns differed by a
factor of more than two.
Applied to six ethalfluralin results in the Year 3 data set.
LOPR,
NCNF,
RNF2,
RPDX
Estimated Value
Low analyte recovery was observed with the OPR sample
associated with this result; the presence of the reported
analyte was not confirmed on a second column; the sample
results reported on the two columns differed by a factor of
more than two; and the RPD between the MS and matrix
spike duplicate exceeded criteria.
Applied to five ethalflurarin results in the Year 3 data set.
LOPR,
NCNF,
RPDX
Estimated Value
Low analyte recovery was observed with the OPR sample
associated with this result; the presence of the reported
analyte was not confirmed on a second column; and the
RPD between the MS and matrix spike duplicate exceeded
criteria.
Applied to one ethalflurarin result in the Year 3 data set.
LOPR,
NMSR
Exclude
Low analyte recovery was observed with the OPR sample
associated with this result and the analyte was not
recovered at all in the MS sample associated with the
result. The analyte was not detected in the sample, but the
recovery problems observed in the OPR and the MS make
it impossible to confirm the reliability of these non-detects.
Applied to eighteen disulfoton results in the Year 1 data
set.
LOPR,
RNF2
Estimated Value
The analyte was detected on both columns, but the results
were not within a factor of two of each other and the lower
of the results reported was associated with a low OPR
standard recovery.
Applied to one ethalflurarin result in the Year 1 data set;
eight methoxychlor and one endosulfan sulfate result in
the Year 2 data set; fifteen ethalflurarin results in the
Year 3 data set; and one ethalflurarin results in the Year 4
data set.
LOPR,
RNF2,
RPDX
Estimated Value
The analyte was detected on both columns, but the results
were not within a factor of two of each other; the lower of
the results reported was associated with a low OPR
standard recovery; and the RPD between the associated MS
and MSD exceeded criteria.
Applied to five 4,4'-DDT results in the Year 2 data set.
April 2005
44
QA Report for the National Lake Fish Tissue Study
-------
Exhibit 4-1A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
Full Length
Description or
Recommendation
(Comment)
Explanation of Code (Reason)
Notes on When This Occurs
LOPR,
RPDX
Estimated Value
The lower of the results reported was associated with a low
OPR standard recovery and the RPD between the
associated MS and MSD exceeded criteria.
Applied to six 4,4'-DDT results in the Year 2 data set.
LVER
Potential Low Bias
Low analyte recovery was observed in a calibration
verification associated with this sample, suggesting the
possibility of a low bias in the result.
Applied to two 4,4'-DDT results in the Year 1 data set
and several (5.2%) arsenic species results in the Year 2
data set.
LVER,
RNF2
Estimated Value
The analyte was detected on both columns, but the results
on the two columns differed by more than a factor of two,
and the lower of the two results reported was associated
with a low recovery of calibration verification standard.
Applied to one 4,4'-DDT result in the Year 1 data set.
MTRX
Estimated Value
The chromatogram suggested possible matrix interferences
with the sample.
Applied to one methoxychlor and two /raws-permethrin
results in the Year 1 data set; also applied to one dicofol
result in the Year 4 data set.
MTRX,
NLBL
Exclude
The chromatogram suggested possible matrix interferences
with the sample; and the labeled analog was not recovered
from the sample, suggesting severe matrix interferences.
Because it is impossible to determine if the analyte is
present or not present, the reported target analyte result was
excluded from the database.
Applied to sixteen PCB congener results in the Year 3
data set.
MTRX,
RNF2
Estimated Value
The chromatogram suggested possible matrix interferences
and although this analyte was found using two different
columns, the results reported on the two columns differed
by a factor of more than two.
Applied to one methoxychlor and one c/s-permethrin
result in the Year 1 data set.
NCNF
Estimated Value
The result was not confirmed (either no elution or co-
elution on a second column).
Applied to several analytes in the Year 1, Year 2, Year 3,
and Year 4 data sets. Occurred for 0.17% of all the
pesticide results in the study. (Please note that kepone
cannot be confirmed on the method-specified second
column.)
QA Report for the National Lake Fish Tissue Study
45
April 2005
-------
Exhibit 4-1 A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
NCNF,
RNF2
NCNF,
RNF2,
RPDX
NCNF,
RPDX
NLBL
NLRO,
NNRO
NMSR
NNRO
NNRO,
NVER
NOPR
Full Length
Description or
Recommendation
(Comment)
Estimated Value
Estimated Value
Estimated Value
Exclude
Exclude
Exclude
Exclude
Exclude
Exclude
Explanation of Code (Reason)
The result was found on one column but co-eluted with
another compound on the second column and the co-eluted
results differed from the other column results by a factor
greater than two.
The result was found on one column but co-eluted with
another compound on the second column; the co-eluted
results differed from the other column results by a factor
greater than two; and the RPD between the associated MS
and MSD exceeded criteria.
The result was found on one column but co-eluted with
another compound on the second column and the RPD
between the associated MS and MSD exceeded criteria.
The labeled analog was not recovered from the sample,
suggesting severe matrix interferences. Because it is
impossible to determine if the analyte is present or not
present, the reported target analyte result was excluded
from the database.
Neither the target analyte (native compound) nor the
labeled analog were recovered in the OPR sample
associated with this result.
No analyte recovery was observed with one or more MS
samples.
The native compound was not recovered in the OPR
sample suggesting that the method may not be working for
this analyte.
The native compound was not recovered in the OPR
sample suggesting that the method may not be working for
this analyte and there was no recovery of the analyte in the
calibration verification sample associated with this result.
There was no recovery of the analyte in the OPR sample
associated with this result.
Notes on When This Occurs
Applied to one /raws-nonachlor result in the Year 1 data
set; three 2,4'-DDE results in the Year 2 and Year 3 data
sets; and one trifluralin result and one alpha-chlordane
result in the Year 3 data set.
Applied to five trifluralin results in the Year 3 data set.
Applied to one trifluralin result in the Year 3 data set.
Applied to a few (0.08%) semivolatile organic results in
the Year 1, Year 2, Year 3, and Year 4 data sets.
Applied to three phenol results in the Year 1 data set.
Occurred with ten disulfoton results in the Year 4 data
set.
Applied to 1.06% of the semivolatile organic results in
Year 1, Year 2, Year 3, and Year 4 data sets.
Applied to nineteen tetrabromobisphenol A results in the
Year 4 data set.
Applied to eighteen endosulfan sulfate results in the Year
2 data set.
April 2005
46
QA Report for the National Lake Fish Tissue Study
-------
Exhibit 4-1 A (Target Analytes ) -Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
KNF2
RNF2,
RPDX
RPDX
Full Length
Description or
Recommendation
(Comment)
Estimated Value
Estimated Value
Estimated Value
Explanation of Code (Reason)
Although the analyte was found using two different
columns, the results reported on the two columns differed
by a factor of more than two.
Although the analyte was found using two different
columns, the results reported on the two columns differed
by a factor of more than two and the RPD between the MS
and MSD exceeded criteria.
The RPD between the MS and MSD exceeded criteria.
Notes on When This Occurs
Applied to a few (0.68%) organochlorine pesticide results
in the Year 1, Year 2, Year 3, and Year 4 data sets.
Applied to one 2,4'-DDE result in the Year 2 data set;
three 2,4'-DDE results, one 4,4'-DDT result, two
heptachlor, and two trifluralin results in the Year 3 data
set; and one ethalfluralin result in the Year 4 data set.
Applied to seven 2,4'-DDE results in the Year 2 data set
and one 2,4'-DDE result, one 4,4'-DDT result, and two
trifluralin results in the Year 3 data set.
QA Report for the National Lake Fish Tissue Study
47
April 2005
-------
Exhibit 4-1B (Additional or Non-Target Analytes) - Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
Full Length
Description or
Recommendation
(Comment)
Explanation of Code (Reason)
Notes on When This Occurs
B, HTEX,
RNON
Blank
Contamination,
Result Reported as a
Non-detect,
Potential Low Bias
The analyte was detected in one or more of the blanks
associated with this sample; the holding time was exceeded
when analyzing the sample; and when the sample result is
less than five times the blank result, there are no means by
which to ascertain whether or not the presence of the
analyte may be attributed to contamination. Therefore,
SCC recommends that the data be reported in the database
as a non-detect at the ML, adjusted for dilution.
Occurred for one w-octadecane and one w-tetradecane
result in the Year 4 data set.
B,RMAX
Blank
Contamination,
Result is a
Maximum Value
Blank contamination was observed and the analyte was
reported in the sample at a concentration between 5 and 10
times higher than the blank value. The result was
considered to be of acceptable quality, but data users are
cautioned that it may be a maximum value due to possible
influence of contamination.
Applied to six non-target analytes (one w-decane, two
n-hexadecane, and three w-octadecane results) in the
Year 3 and Year 4 data sets.
B, RNAF
Blank
Contamination,
Result Not Affected
Blank contamination was present but was not considered to
adversely impact the sample result. The presence of the
analyte in the blank is not considered to adversely affect the
data in cases where the sample results are more than 10
times the associated blank results or where the analyte is
not detected in associated samples.
Applied to 0.47% of the non-target, semivolatile analytes
in the Year 2 and Year 4 data sets.
B, RNON
Blank
Contamination,
Result Reported as a
Non-detect
When the sample result is less than five times the blank
result, there are no means by which to ascertain whether or
not the presence of the analyte may be attributed to
contamination. Therefore, SCC recommends that the data
be reported in the database as a non-detect at the ML,
adjusted for dilution.
Applied to 0.82% of the non-target, semivolatile analytes
in the Year 2, Year 3, and Year 4 data sets.
HNRO
Potential High Bias
High native compound recovery in the OPR associated with
the sample.
Applied to one w-octadecane result in the Year 4 data set.
April 2005
QA Report for the National Lake Fish Tissue Study
-------
Exhibit 4-1 B (Additional or Non-Target Analytes ) - Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
HNRO,
HTEX,
LLRO
HNRO, J
HTEX
HTEX, J
HTEX,
NLBL
HTEX,
NLRO,
NNRO
J
Full Length
Description or
Recommendation
(Comment)
Exclude
Estimated Value
Potential Low Bias
Estimated Value
Exclude
Exclude
Estimated Value
Explanation of Code (Reason)
A combination of QC failures that suggest the reliability of
the result is questionable. In this case, the native
compound was recovered above method-specified criteria
and the labeled analog was recovered below method-
specified criteria in the OPR associated with the sample
(HNRO and LLRO). In addition, the holding time was
exceeded (HTEX) and, although the compound was
reported in a sample dilution, it was not found in the neat
(undiluted) analysis.
High native compound recovery in the OPR associated with
the sample; the sample result reported was above the MDL
(detection limit) but below the ML (quantitation limit).
The holding time was exceeded when analyzing the
sample.
The sample result reported was above the MDL (detection
limit) but below the ML (quantitation limit) and the holding
time was exceeded when analyzing the sample.
The holding time was exceeded when analyzing the sample
and the labeled analog was not recovered from the sample,
suggesting severe matrix interferences.
The holding time was exceeded when analyzing the sample
and neither the analyte (native compound) nor the labeled
analog were recovered in the OPR sample associated with
this result.
The sample result reported was above the MDL (detection
limit) but below the ML (quantitation limit).
Notes on When This Occurs
Applied to one 2,6-dinitrotoluene result in the Year 1
data set.
Applied to seven n-octadecane results in the Year 4 data
set.
Associated with one di-w-octyl phthalate, two n-eicosane,
seven w-hexadecane, and six w-octadecane results in the
Year 1 and Year 3 data sets. Also, associated with
several non-target analytes for seven samples in the
Year 4 data set.
Associated with five w-decane , one w-eicosane, fourteen
w-hexadacane, seven w-octadecane, and six w-tetradecane
results in the Year 1, Year 3, and Year 4 data sets.
Applied to four 2-methyl-4,6-dintrophenol results in the
Year 4 data set.
Applied to seven 4-nitrophenol results in the Year 4 data
set.
Occurred throughout the Year 1, Year 2, Year 3, and
Year 4 data sets.
QA Report for the National Lake Fish Tissue Study
49
April 2005
-------
Exhibit 4-1 B (Additional or Non-Target Analytes ) - Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
J, LLBL
J, LLRO
LLBL
LLRO
LNRO
NLBL
NLBL,
NLRO,
NNRO
NLRO
Full Length
Description or
Recommendation
(Comment)
Estimated Value
Estimated Value
Estimated Value
Result Not Affected
Potential Low Bias
Exclude
Exclude
Result Not Affected
Explanation of Code (Reason)
The analyte was reported above the MDL but below the
ML and the labeled analog of the target was recovered
below method-specified criteria, suggesting the possible
presence of matrix interferences.
The analyte was reported above the MDL but below the
ML in the sample and the labeled analog of the analyte was
recovered below method criteria in the OPR associated
with the sample.
The labeled analog of the analyte was recovered below
method-specified criteria.
The labeled analog of the analyte was recovered below
method-specified criteria in the OPR sample.
The native compound of the analyte was recovered below
method-specified criteria in the OPR sample.
The labeled analog was not recovered from the sample,
suggesting severe matrix interferences. Because it is
impossible to determine if the analyte is present or not
present, the reported analyte result was excluded from the
database.
The labeled analog was not recovered from either the native
sample nor from the Ongoing Precision and Recovery
(OPR) sample associated with this result, suggesting severe
matrix interferences; also, the analyte was not recovered
from the OPR. The failures suggest poor method
performance for this analyte.
The labeled analog was not recovered in the OPR sample
associated with this result.
Notes on When This Occurs
Applied to one w-tetracosane result in the Year 1 data set.
Applied to one 2,4-dimethylphenol result in Year 1.
Applied to one w-hexadecane result in the Year 1 data set.
Associated with nineteen carbazole, nineteen
w-nitrosodiphenylamine, and nine w-triacontane results in
the Year 4 data set.
Applied to twelve 2,3,6-trichlorophenol, nineteen
hexachlorocyclopentdiene, three w-docosane, three n-
hexacosane, and fourteen w-octacosane results in the
Year 4 data set.
Applied to a few (0.24%) non-target organic results in the
Year 2, Year 3, and Year 4 data sets.
Applied to six 2,4-dinitrophenol results in the Year 1 data
set; and sixteen 4-nitrophenol results in the Year 3 data
set.
Applied to two 4-nitrophenol results in the Year 3 data
set.
April 2005
50
QA Report for the National Lake Fish Tissue Study
-------
Exhibit 4-1 B (Additional or Non-Target Analytes ) - Summary of Data Review Qualifiers Used for Year 1 through Year 4 of the
National Lake Fish Tissue Study
Qualifiers
Applied
(SCC
Code)
NLRO,
NNRO
NNRO
REXC
Full Length
Description or
Recommendation
(Comment)
Exclude
Exclude
Estimated Value
Explanation of Code (Reason)
Neither the analyte (native compound) nor the labeled
analog were recovered in the OPR sample associated with
this result.
The native compound was not recovered in the OPR
sample, suggesting that the method may not be working for
this analyte.
The result exceeded the instrument calibration range.
Notes on When This Occurs
Applied to a few non-target organic results (0.18%) in the
Year 1, Year 3, and Year 4 data sets.
Applied to 0.45% of the non-target organic results in
Year 1, Year 2, Year 3, and Year 4 data sets.
Applied to one di-n-octyl phthalate result in the Year 1
data set.
QA Report for the National Lake Fish Tissue Study
51
April 2005
------- |