United States
Environmental Protection
»m Agency

Office of Water

www.epa.gov	March 2023

Quality Assurance Report for the National Pilot
Study of Pharmaceuticals and Personal Care
Products in Fish Tissue


-------
U.S. Environmental Protection Agency
Office of Water
Office of Science and Technology (4305T)
1200 Pennsylvania Avenue, NW
Washington, DC 20460

EPA 820-F-23-002


-------
Table of Contents

Page

Acknowledgements	iii

Disclaimer 	iii

Contact 	iii

Chapter 1 Introduction	1

Section 1.1 Background	1

Section 1.2 Study Design	1

Section 1.3 Study Participants	3

Chapter 2 Quality Assurance Program	5

Section 2.1 Quality Assurance Project Plans	5

Section 2.2 Field Sampling QA/QC	6

Section 2.3 Sample Analysis QA/QC	7

Section 2.4 Training	7

Section 2.5 QA Oversight of Laboratory Operations	8

Chapter 3 Analytical Methods	9

Section 3.1 Preparation of Fish Tissue Samples	9

Section 3.2 Determination of Lipid Content	10

Section 3.3 Analysis of Pharmaceuticals by HPLC-MS/MS	10

Section 3.4 Analysis of Personal Care Products by GC-MS/MS	12

Section 3.5 Quality Control	14

Chapter 4 Data Quality Assessment	15

Section 4.1 Data Review	15

Section 4.2 Analysis of Blanks	17

Section 4.3 Surrogate Spiking	18

Section 4.4 Matrix Spiking	20

Section 4.5 Other Quality Control Checks	21

Section 4.6 Overall Data Quality Assessment	21

Section 4.7 Completeness	22

References 	24

Appendix I Unique Qualifier Combinations Applied to Pilot Study Data by Analytical Method

List of Tables

Table 1.	PPCP Fish Pilot Study Design Element Summary	1

Table 2.	Target Analytes by Analytical Method	3

Table 3.	HPLC-MS/MS Operating Conditions	12

Table 4.	GC-MS/MS Operating Conditions	14

Table 5.	Individual Data Qualifiers Applied to Pilot Study Data	16

Table 6.	Sample/Analyte Combinations	22

PPCP Fish Pilot Study QA Report	i	March 2023


-------
List of Figures

Page

Figure 1. Project Organization Chart, circa 2012	4

Figure 2. Impacts of Blank Contamination on the Pharmaceutical Results	18

Figure 3. Impacts of Blank Contamination on the Personal Care Product Results	18

Figure 4. Impacts of Surrogate Recoveries on the Pharmaceutical Results	19

Figure 5. Impacts of Surrogate Recoveries on the Personal Care Product Results	19

Figure 6. Impacts of Matrix Spike Recoveries on the Pharmaceutical Results	20

Figure 7. Impacts of Matrix Spike Recoveries on the Personal Care Product Results	21

PPCP Fish Pilot Study QA Report	ii	March 2023


-------
Acknowledgements

This quality assurance report was prepared by the U.S. Environmental Protection Agency, Office of
Water, Office of Science and Technology. The EPA Project Manager for the pilot study was Leanne
Stahl, who provided overall project coordination and technical direction. Tetra Tech, Inc. provided
primary support for the study and subcontracted Baylor University's Center for Reservoir and Aquatic
Systems Research to conduct the fish tissue analysis for the pilot study under Contract Numbers
EP-C-04-030 and EP-C-09-019. Quality assurance support was provided by General Dynamics
Information Technology (GDIT) and several predecessor organizations, including Computer Sciences
Corporation (CSC) and CSRA (hereafter collectively referred to as GDIT) under Contract Numbers EP-
C-06-085, EP-W-06-046, EP-C-10-060, EP-C-12-008, and EP-C-17-024. GDIT was responsible for
production of this report under the direction of Leanne Stahl and John Healey.

Cover design by Harry McCarty, General Dynamics Information Technology, Falls Church, VA.

Disclaimer

The U.S. Environmental Protection Agency, Office of Water, Office of Science and Technology has
approved this report for publication. Mention of trade names, commercial products, or services does not
constitute official EPA approval, endorsement, or recommendation for use.

Contact

Please address questions and comments to:

John Healey

Standards and Health Protection Division
OW/Office of Science and Technology (4305T)
US Environmental Protection Agency
1200 Pennsylvania Ave, NW
Washington, DC 20460
healey.john@epa.gov

PPCP Fish Pilot Study QA Report

Hi

March 2023


-------
Chapter 1
Introduction

This report documents the quality of data generated for pharmaceuticals and personal care products
(PPCPs) under the National Pilot Study of Pharmaceuticals and Personal Care Products in Fish Tissue
(hereafter referred to as the "pilot study"). Below is a brief summary of the pilot study design and
implementation. The PPCP Fish Pilot Study final report, which is referenced at the end of this document,
provides additional information (USEPA, 2023).

Section 1.1 Background

PPCPs are a large and diverse group of chemicals consisting of all prescription drugs and over-the-
counter medications, as well as non-medicinal chemicals in consumer products, such as fragrances in
soaps and lotions and active ingredients in sunscreens and insect repellants. Studies completed by 2005
documented the occurrence of PPCPs primarily in surface waters, sediment, and effluents (discharges)
from wastewater treatment plants. Little data were available on PPCPs in fish at this time, and early
studies of PPCPs in fish focused on sample collection from a single site and analysis of tissue for a
specific chemical or chemical class. In 2006, EPA's Office of Science and Technology within the Office
of Water decided to expand investigation of the presence and accumulation of PPCPs in fish by
conducting this pilot study. EPA's pilot study was the first broad assessment of PPCPs in fish from
locations across the United States.

Section 1.2 Study Design

The pilot study is a screening-level study designed to investigate the occurrence of a broad suite of PPCPs
in freshwater fish. It incorporated the following design elements, which are summarized in Table 1:

•	Selection of sampling sites on five effluent-dominated streams in population centers across the
country and at a remote reference site in an area not influenced by wastewater discharges

•	Collection of six composite samples of resident fish species at each sampling site

•	Analysis of fish tissue composite samples for 36 PPCPs

Table 1. PPCP Fish Pilot Study Design Element Summary

Number of Sites

5 Effluent-dominated streams + 1 reference site

Site Selection Method

Targeted

Sampling Year

2006

Fish Composite Samples/Site

6

Fish Tissue Sample Types

Fillets and Livers for Pharmaceuticals
Fillets only for Personal Care Products

Number of Target Analytes

24 Pharmaceuticals
12 Personal Care Products

Total Tissue Samples Analyzed

72 for Pharmaceuticals
36 for Personal Care Products

1.2.1 Site Selection

EPA selected fish sampling locations on five effluent-dominated streams in densely populated areas of the
U.S. based on the assumption that PPCPs were more likely to occur in these areas. Other factors
considered for site selection included a larger percentage of elderly residents, higher median incomes (as

PPCP Fish Pilot Study QA Report

1

March 2023


-------
a surrogate for volume of PPCP sales), availability of fish, and representation of different wastewater
treatment technologies to evaluate their potential impact on PPCP removal.

The sampling sites for the pilot study are as follows:

•	North Shore Channel in Chicago, Illinois

•	Trinity River in Dallas, Texas

•	Little Econlockhatchee River in Orlando, Florida

•	Salt River in Phoenix, Arizona

•	Taylor Run in West Chester, Pennsylvania (a suburb of Philadelphia)

In addition, EPA received assistance from staff at New Mexico Environment Department to identify a
suitable reference site on the East Fork Gila River in the Gila River Wilderness Area of New Mexico.
The remote location of this reference site ensured minimal impacts of human influences.

1.2.2	Fish Sample Collection

Field sampling teams applied consistent methods to collect 18 or 24 adult fish of the same species and
similar size from each sampling location during the late summer and fall of 2006. Fish collection from
effluent-dominated streams focused on resident fish species that were continually exposed to discharges
from wastewater treatment plants. The field teams divided the fish from each site into six composites of
either 3 or 4 fish before shipping whole fish to the laboratory for tissue sample preparation and analysis.

1.2.3	Tissue Sample An alysis

Baylor University's Center for Reservoir and Aquatic Systems Research prepared and analyzed the fish
tissue samples for the pilot study. Laboratory technicians removed fillets and livers from individual fish
in each of the six composite samples collected at every site, and they homogenized the tissue to prepare
36 fillet and 36 liver composite samples for analysis (6 composites/site x 6 sites x two tissue types for a
combined total of 72 tissue composite samples for analysis). At the time of this study, Baylor
University's analytical methods for PPCPs in fish tissue could screen tissue samples for 24
pharmaceuticals using high performance liquid chromatography (HPLC) with a tandem mass
spectrometric (MS/MS) detector and for 12 personal care product chemicals using gas chromatography
(GC) with a tandem ion trap mass spectrometric (MS/MS) detector. Table 2 provides the name, CAS
Registry Number, and analytical method for each of the 36 PPCPs included in the pilot study. Fillet
composites were analyzed for both the pharmaceuticals and personal care product chemicals. Analysis of
the 36 liver composites was limited to pharmaceuticals only due to problems the laboratory encountered
with interferences from the high lipid content in the liver tissue.

PPCP Fish Pilot Study QA Report

2

March 2023


-------
Table 2. Target Analytes by Analytical Method

Pharmaceuticals by HPLC-MS/MS

Analyte

CAS No.

Analyte

CAS No.

Acetaminophen

103-90-2

Ibuprofen

15687-27-1

Atenolol

29122-68-7

Lincomycin

154-21-2

Caffeine

58-08-2

Metoprolol

37350-58-6

Carbamazepine

298-46-4

Miconazole

22916-47-8

Cimetidine

51481-61-9

Norfluoxetine

54910-89-3

Codeine

76-57-3

Propranolol

525-66-6

Diltiazem

42399-41-7

Sertraline

79617-96-2

1,7-Dimethylxanthine

611-59-6

Sulfamethoxazole

723-46-6

Diphenhydramine

58-73-1

Thiabendazole

148-79-8

Erythromycin

114-07-8

Trimethoprim

738-70-5

Fluoxetine

54910-89-3

Tylosin

1401-69-0

Gemfibrozil

25812-30-0

Warfarin

81-81-2

Personal Care Products by GC-MS/MS

Analyte

CAS No.

Analyte

CAS No.

Benzophenone

119-61-9

Nonylphenol

104-40-5

Celestolide

13171-00-1

Octocrylene

6197-30-4

Galaxolide

1222-05-5

Octylphenol

1806-26-4

4-Methylbenzylidine camphor

36861-47-9

m-Toluamide

618-47-3

Musk ketone

81-14-1

Tonalide

1506-02-1

Musk xylene

81-15-2

Triclosan

3380-34-5

Section 1.3 Study Participants

EPA's Office of Science and Technology (OST) conducted the pilot study with support from two agency
contractors (Tetra Tech and GDIT) and the analytical laboratory at Baylor University. Figure 1 (on the
following page) identifies roles and responsibilities for the primary study participants from the initiation
of the study through 2012. OST also received voluntary logistical and field sampling assistance from
EPA's Great Lakes National Program Office and the Metropolitan Water Reclamation District of Greater
Chicago for the North Shore Channel site in Chicago and from the New Mexico Environment Department
for the reference site in the Gila Wilderness Area of southwest New Mexico.

PPCP Fish Pilot Study QA Report

3

March 2023


-------
Figure 1. Project Organization Chart, circa 2012

PPCP Fish Pilot Study QA Report

4

March 2023


-------
Chapter 2
Quality Assurance Program

Environmental measurements always contain some level of uncertainty, and decision makers must
recognize the uncertainty associated with the data on which their decisions are based. In planning the
pilot study, EPA managers recognized that there was significant potential that the study data might be
used by others within EPA and by other interested parties responsible for making environmental,
economic, and policy decisions. Therefore, the study managers established a quality assurance (QA)
program intended to ensure that data produced during the pilot study would meet defined and documented
standards of quality.

The pilot study QA program prescribed minimum requirements to which all organizations that gathered
data were required to adhere. Data quality was defined, controlled, assessed, and documented through
these QA program activities. The remainder of this chapter presents highlights of the QA program
employed during the study.

Section 2.1 Quality Assurance Project Plans

EPA decided to develop separate quality assurance project plans (QAPPs) to support field sampling and
laboratory analysis for the pilot study. This decision allowed the QAPP for sample collection activities to
be prepared and approved on an accelerated schedule, so field teams could complete sampling at all study
sites during the summer and fall of 2006. The analytical activities QAPP was completed and approved
about a month before field teams finished fish sample collection for the study. Tetra Tech provided
support for development of the sample collection activities QAPP, and both GDIT and Tetra Tech
supported development of the analytical activities QAPP.

2.1.1	Sample Collection Activities QAPP

The sample collection QAPP is formally referred to as the Quality Assurance Project Plan for Sample
Collection Activities for a Pilot Study to Investigate the Occurrence ofPharmaceuticals and Personal
Care Products (PPCPs) in Fish Tissue (USEPA, 2006a). This QAPP established data quality goals for
all sample collection and handling activities and described quality assurance/quality control (QA/QC)
techniques employed by field sampling teams to support those goals. EPA based the sampling procedures
for the pilot study on the approach successfully applied for sample collection in the agency's National
Lake Fish Tissue Study. Technical and QA staff from EPA and Tetra Tech approved the sample
collection QAPP on August 1, 2006. Everyone involved in the sample collection process received a copy
of this QAPP.

2.1.2	Analytical Activities QAPP

The analytical activities QAPP is officially known as the Quality Assurance Project Plan for Laboratory
Sample Preparation and Analysis Activities in the National Pilot Study of Pharmaceuticals and Personal
Care Products (PPCPs) in Fish Tissue (USEPA, 2006b). This QAPP established measurement quality
objectives (e.g., QC acceptance criteria) for laboratory data generated during the pilot study and described
the QA/QC procedures applied by laboratory staff and the other contractors supporting the study to ensure
these goals were met. Scientists and QA managers from EPA, GDIT, Tetra Tech, and Baylor University
approved the analytical activities QAPP on October 19, 2006. All contractors responsible for fish tissue
sample preparation and analysis, data quality review, or database development received copies of this
QAPP.

PPCP Fish Pilot Study QA Report

5

March 2023


-------
2.1.3 QAPP Revisions

Field sampling teams successfully implemented the sample collection activities QAPP as initially written,
so this QAPP did not undergo revision during the pilot study. In contrast, some approaches that were
planned and documented in the initial analytical activities QAPP were refined as work on tissue sample
analysis progressed during the study. Since this was a pilot study, all of the study participants anticipated
that changes to the original version of this QAPP would be necessary. Following is a list of the
substantive revisions to the analytical activities QAPP and a brief summary describing each change:

•	The laboratory revised the list of target analytes as follows:

After the original analytical activities QAPP was signed, the laboratory refined their HPLC-
MS/MS procedure and added diphenhydramine as a target analyte

The laboratory dropped clofibric acid from the target analyte list due to instrumental difficulties
encountered while switching ionization modes on the HPLC-MS/MS

The laboratory dropped three nonylphenol monoethoxylate isomers as target analytes due to
problems obtaining a suitable standard. The readily available standard was a mixture of all three
isomers, and the vendor could not certify the concentration of each isomer in the mixture or
supply a separate standard for each isomer

•	The laboratory revised the analytical techniques for the personal care products, particularly to
incorporate the use of a tandem mass spectrometric detector rather than using selected ion monitoring
on a single MS detector. Section 3.4 of this report provides more information about these changes

•	The laboratory extracted and analyzed only the fillet samples for personal care products after
encountering problems with lipid interferences in multiple attempts to analyze liver samples and
modify the extract cleanup procedures to address these interferences

•	Analytical results for the tissue samples were reported down to the laboratory's method detection
limits (MDLs) rather than the laboratory's project quantitation limits

For each case, EPA approved changes to the analytical approach described in the original QAPP prior to
their implementation. The laboratory delivered the final data package for pharmaceuticals in May 2007
and for personal care products in February 2008. At EPA's request, Tetra Tech updated the analytical
activities QAPP to document changes to the analytical procedures implemented during tissue sample
analysis and submitted the revised analytical activities QAPP in July 2010.

Section 2.2 Field Sampling QA/QC

EPA incorporated a number of QA/QC procedures to ensure consistency in fish sample collection and to
produce complete and accurate documentation of field sampling data. Collectively these procedures
contributed to reducing sampling variability and to improving sample representativeness and sampling
completeness. Key field sampling QA/QC procedures include the following:

•	Preparation and implementation of the sample collection activities QAPP

•	Development and application of standard operating procedures for sample collection and handling
activities

•	Preparation and use of standardized sampling kits containing field sampling supplies to control
contamination of fish samples and forms to allow consistent documentation of fish collection and
sample shipping

•	Use of the same experienced fisheries biologist to lead field sampling teams of fully trained
technicians and to ensure proper implementation of procedures

PPCP Fish Pilot Study QA Report

6

March 2023


-------
•	Daily tracking and coordination of fish sample shipments through a centralized source during fish
sampling operations

•	Implementation of standardized procedures for review and documentation of field data quality

Field teams documented sample collection and shipping information using four standardized forms and
labels: a field record form, a sample identification label, a chain-of-custody form, and a chain-of-custody
label for sealing each shipping container. Field data reviewers used a fifth standardized form, the
resolved action form, to record results and decisions for field data quality assessments. Assessment of
field data quality showed that sampling teams met the goal of collecting six fish composite samples that
adhered to composite criteria specified in the QAPP at each of the six study sites by November 2006. No
fish samples were lost or compromised during shipment to the laboratory at Baylor University, so field
teams achieved the sampling completeness goal of 100%.

Section 2.3 Sample Analysis QA/QC

EPA integrated several QA/QC activities and laboratory requirements into the pilot study to ensure data
comparability and generate analytical data of known quality during preparation and analysis of the fish
tissue samples and evaluation of analytical data quality. Following is a summary of critical QA/QC
components to achieve analytical data quality goals:

•	Development and implementation of the analytical activities QAPP

•	Use of one laboratory for sample preparation (filleting and liver extraction, tissue homogenization,
and preparation of tissue aliquots) and tissue sample analysis

•	Requirement for triplicate lipid analyses to test for tissue homogeneity during sample preparation

•	Identification of quantifiable measurement quality objectives

•	Use of pure and traceable reference standards

•	Demonstration of instrument calibration and system performance

•	Periodic calibration verification

•	Analysis of QC samples to assess performance of analytical methods

•	Specification of MDLs and method/chemical QC acceptance criteria that applied throughout the study

•	Use of a standardized data quality assessment process

The general measurement quality objective (MQO) for the pilot study was to satisfy method-specific
performance criteria. The analytical activities QAPP provides a summary of the method performance
criteria and specifies MQOs and QC acceptance criteria to assess the bias and precision associated with
the analytical methods used for this study. Chapter 4 of this report describes the process for data quality
assessment and presents the results of these assessments, which includes data from the following
laboratory QC samples or measures: blanks, recoveries for spiking surrogate chemicals into field-based
tissue samples, matrix spiking (matrix spike/matrix spike duplicate (MS/MSD)), laboratory control
samples, and calibration verifications. Chapter 4 also includes a discussion of data completeness for the
pilot study.

Section 2.4 Training

Planning for the pilot study included using experienced staff for sample collection, laboratory analyses,
and data validation and providing project-specific training, as necessary, to staff responsible for these
activities. The QAPPs covering the respective sampling and analysis activities describe training for these

PPCP Fish Pilot Study QA Report

7

March 2023


-------
areas of responsibility. Prior to initiating field sampling operations, sample collection staff received
training on pilot study sample collection and handling procedures. Laboratory personnel at Baylor
University consisted of senior research associates supervising technical staff with the education and skill
levels required to maintain consistent measurement system performance throughout the study. GDIT
assigned data reviewers to evaluate pilot study results that were trained in the application of data review
guidelines developed for EPA's National Lake Fish Tissue Study and adopted for this study. They were
also experienced in reviewing data generated with the instrumentation used in the pilot study. Each data
reviewer received a copy of the analytical activities QAPP, which specified the performance criteria and
MQOs applicable to this study.

Section 2.5 QA Oversight of Laboratory Operations

Tetra Tech was responsible for providing analytical services for the pilot study, and subcontracted with
Baylor University to obtain these services. As part of their contractual oversight, Tetra Tech assigned a
QA chemist to maintain oversight of laboratory operations for the duration of the contract and to work
collaboratively with laboratory staff to ensure optimal completion of tissue sample analysis for the pilot
study. The QA chemist provided technical and QA support to the laboratory through a variety of
activities. These activities included the following:

•	Conducting collaborative reviews of laboratory operations at critical points in the process of
preparing and analyzing the tissue samples, such as:

Visiting Baylor University in December 2006 to observe and provide guidance on the preparation
of fish tissue samples. Laboratory staff did not have prior experience preparing fillet samples
from the size of fish collected for the pilot study, and they improved the process for
homogenizing fillet tissue as a result of this visit. The laboratory also implemented actions to
improve freezer temperature monitoring and cataloging freezer contents.

Returning to the laboratory in February 2007 to observe the extraction of tissue samples and
review the HPLC-MS/MS and GC-MS instrumentation and analytical procedures. Discussions
between the Tetra Tech and Baylor University culminated in the laboratory adopting use of
GC-MS/MS techniques for personal care product analysis after EPA's concurrence.

Traveling to an onsite meeting with laboratory staff in September 2007 to discuss approaches for
effectively addressing the extensive lipid interferences the laboratory was encountering during
analysis of fillet samples for personal care products. The goal of the meeting was to identify a
solution that would allow the laboratory to complete these analyses and deliver analytical results
of acceptable quality. Using gel permeation chromatography (GPC) as a cleanup technique in the
analyses of fillet samples for personal care products provided adequate resolution of this issue.

•	Involving the laboratory in regular comprehensive reviews of the analytical procedures for both the
pharmaceuticals and personal care products. An outcome of these reviews was Tetra Tech working
with the laboratory to fine-tune several aspects of the laboratory's analytical effort and obtaining EPA
approval for the revised procedures.

•	Working extensively with the laboratory to resolve QA and data reporting issues, including:

Identifying and testing options to address lipid interferences in analyzing tissue samples for
personal care products. After this effort, Tetra Tech and Baylor University agreed that the use of
gel permeation chromatography (GPC) would be required as a cleanup technique for all
GC-MS/MS analyses in this study.

Participating in discussions with Baylor University and other study participants for resolving a
data reporting error for galaxolide results

PPCP Fish Pilot Study QA Report

8

March 2023


-------
Chapter 3
Analytical Methods

When EPA initiated the pilot study, several agency programs, including those within the Office of Water
and the Office of Research and Development, were beginning to develop methods for the analysis of
PPCPs in various matrices. At that time, however, there were no formal EPA analytical methods for the
PPCPs1, nor were methods available from any voluntary consensus standards bodies. Therefore, the
project team sought the assistance of Baylor University's Center for Reservoir and Aquatic Systems
Research because researchers in the center had developed separate analytical methods for analyzing
pharmaceuticals and personal care products in fish tissue. Scientists from Baylor joined the project team
when the university agreed to provide laboratory services for the pilot study, which included preparation
of fish tissue samples and analysis of the tissue samples for lipid content and for 24 pharmaceuticals and
12 personal care products. Section 3.1 summarizes the process for preparation of fish tissue samples, and
Section 3.2 describes the method for analyzing the lipid content of each tissue sample. Sections 3.3 and
3.4 outline procedures for the PPCP methods, respectively, that Baylor University used to analyze fish
tissue samples for the pilot study. Section 3.5 provides information about quality control procedures for
the PPCP methods.

Section 3.1 Preparation of Fish Tissue Samples

The laboratory prepared 36 fillet composite samples (from 6 fish composite samples per site at 6 sites)
following procedures described in EPA's Guidance for Assessing Chemical Contaminant Data for Use in
Fish Advisories, Volume I: Fish Sampling and Analysis, Third Edition, 2000 (USEPA, 2000). During the
homogenization process, Baylor staff ensured that there was minimal loss of tissue mass, and that each
sample was uniformly mixed, using lipid analysis to test for tissue homogeneity

The process for preparing fillet samples included the following steps:

•	Removing the entire fillet (including the skin and belly flap) from both sides of each fish in the
composite sample and using all the available tissue to prepare the fillet composite sample (i.e., the
batch method)

•	Grinding frozen cubes of fillet tissue to a fine powder using a high-speed blender and adding small
amounts of dry ice during grinding to facilitate consistent blending of the tissue

•	Applying quartering, mixing, and re-grinding techniques described in the guidance document to
produce a homogeneous composite mixture of fillet tissue and

•	Storing the homogenized fillet composite samples in a freezer at a temperature of -20°C until the
laboratory was ready to analyze them for PPCPs

To prepare the 36 liver composite samples, the laboratory applied tissue dissection and homogenization
techniques developed for prior studies conducted by Baylor University to characterize concentrations of
PPCPs in fish tissue (Brooks et al., 2005). These techniques involved the following steps:

•	Removing the liver from each fish in the composite (a total of three or four livers, depending on the
sampling location) and placing all of them in a clean glass container

1 The Engineering and Analysis Division (EAD) of the Office of Water released a draft EPA method for PPCPs in December
2007, well after the start of the pilot study. Method 1694: Pharmaceuticals and Personal Care Products in Water, Soil,

Sediment, and Biosolids by HPLC/MS/MS, addresses the analysis of a large suite of PPCPs in water and biosolids, but the method
does not include any procedures for the analysis of fish tissues, nor has EAD evaluated the method in tissue matrices.

PPCP Fish Pilot Study QA Report

9

March 2023


-------
•	Homogenizing the liver tissue using a Tissuemiser® set to rotate at 30,000 revolutions per minute
(rpm) and

•	Storing the liver homogenate samples in the freezer at a temperature -20°C until the laboratory was
ready to analyze them for pharmaceutical compounds

Section 3.2 Determination of Lipid Content

The laboratory analyzed three replicate aliquots (individually consisting of approximately 2 g) from each
of the 36 fillet composite samples for lipids using the method described in Mottaleb et al. (2009). This
method was modified slightly for liver samples. Due to limited sample mass, triplicate lipid
measurements were made for only one of the six liver composite samples from each sampling site. Lipid
analysis involved the following steps:

1)	weighing out 2 g of tissue and combining it with 15 mL of a 1:1 mixture of dichloromethane:hexane
in a borosilicate vial

2)	homogenizing each mixture for 3 minutes using a Tissuemiser®

3)	Placing the vials in an incubator for 24 hours at 35 °C and periodically agitating by gentle end-over-
end rotation

4)	adding 2 g of solid anhydrous sodium sulfate to each 1 g of sample following extraction

5)	filtering the mixture through Grade 415 filter paper

6)	washing the solid residue with an additional 15 mL of 1:1 dichloromethane:hexane

7)	collecting the combined filtrate for each sample in a pre-weighed test tube

8)	evaporating the solvent with dry nitrogen for 8 hours at 45 °C using a Zymark® Turbovap LC
Concentration Workstation and

9)	drying the lipid residue after evaporation to a constant weight in a vacuum oven at 40 °C

Lipid content was determined gravimetrically by weighing three replicates of each sample. Percent lipid
determinations were then calculated as shown in the following equation:

weight of lipid residue (g)

% lipid=	—	—	—	x 100

weight of tissue (g)

Section 3.3 Analysis of Pharmaceuticals by HPLC-MS/MS

Baylor analyzed for 24 pharmaceuticals in fish tissue by high performance liquid chromatography-tandem
mass spectrometry (HPLC-MS/MS) using the method described in Ramirez et al. (2007). Table 2
(Chapter 1) provides a list of the target analytes. This method utilizes matrix-matched calibration
standards (aliquots of control matrix from outside of the study area that are expected to be reasonably free
of target compounds) spiked at a minimum of five concentrations, and extracted and analyzed along with
study samples. By extracting standards, matrix effects and bias were minimized in the final analytical
results.

PPCP Fish Pilot Study QA Report

10

March 2023


-------
3.3.1	Sample Extraction for Ph armaceuticals

Prior to extraction, each sample was spiked with a set of surrogate compounds to assess extraction
efficiency. Extraction of tissue homogenates for analysis involved the following steps:

1)	weighing out 1.0g±0.01gof fillet tissue and 0.5g±0.01gof liver tissue for each standard or
sample composite and placing each tissue aliquot into a 20-mL borosilicate glass screw-cap vial

2)	spiking the samples with the appropriate surrogates (acetaminophen-d4, diphenhydramine-d3,
carbamazepine-dio, fluoxetine-d6, and ibuprofen-13C3)

3)	combining sample homogenates with 8 mL of 1:1 mixture of 0.1 M acetic acid buffer (pH 4) and
methanol extraction solvent, and tightly replacing the cap

4)	sonicating the mixture in an ultrasonic bath for 15 minutes at 25 °C

5)	shaking the mixture vigorously by hand for 20 seconds to further ensure mixing and extraction

6)	quantitatively transferring each extract to a separate 50-mL polypropylene copolymer round-
bottomed centrifuge tube with several rinses of the extraction solvent

7)	centrifuging the extracts at 16,000 rpm for 40 minutes at 4 °C to achieve a full separation of residual
solid and liquid phases

8)	transferring the supernatant into a clean 18-mL disposable borosilicate glass culture tube with rinses
of methanol using disposable Pasteur pipettes

9)	evaporating the sample extracts to dryness under a stream of dry nitrogen at 45 °C

10)	reconstituting the extracts in 1 mL (for fillets) or 0.5 mL (for livers) of mobile phase (0.1% formic
acid in reagent water)

11)	adding internal standards (7-aminoflunitrazapam-d7, fluoxetine-d6, and meclofenamic acid)

12)	sonicating for 1 min at 25 °C

13)	filtering extracts using a 0.2-f.im PTFE-membrane syringe filter into an amber HPLC injection vial
and

14)	sealing the vial with a fluoropolymer-lined cap

To maintain similar method sensitivity for both tissue types, the final extract volumes were 1.0 mL for
fillet samples and 0.5 mL for liver samples.

The laboratory prepared samples in batches of 20 or fewer samples, which were accompanied by the
required batch quality control samples, including a method blank, low- and high-level control samples,
and a pair of MS/MSD samples from each site.

3.3.2	Preparation of Calibration Standards for Pharmaceuticals

Fish tissue samples generally contain large amounts of lipids and other materials that can interfere with
the tissue analyses. Cleanup techniques, such as gel permeation chromatography (GPC), can remove
much of the lipid material and other interferences. However, significant "matrix effects" can remain,
which may limit the overall accuracy of the measurement process. To address potential matrix effects,
the laboratory prepared their instrumental calibration standards in a clean tissue matrix, extracted those
standards in the same manner as samples were extracted, and analyzed the extracted standards. Some
EPA methods apply this approach for other analytes and matrices, such as the analysis of herbicides in
drinking water. The laboratory used tissue from smallmouth bass samples collected at the New Mexico
reference site to prepare separate extracted calibration standards for liver and fillet tissues. Each standard

PPCP Fish Pilot Study QA Report

11

March 2023


-------
for liver analysis required a 0.5-g aliquot of clean liver tissue and each standard for fillet analysis required
a 1.0-g aliquot of clean fillet tissue.

3.3.3 HPLC-MS/MS Analysis of Pharmaceuticals

HPLC-MS/MS has advantages over gas chromatographic (GC) methods for the analysis of
pharmaceuticals because GC methods involve introducing the analytes into the instrumentation in a
gaseous form and many of the pharmaceuticals are not easily volatilized. For example, some
pharmaceuticals have boiling points that are above the operating temperatures of a GC system, while
others will break down when heated.

Tandem mass spectrometry (MS/MS) involves the use of two quadrapole mass spectrometers in series,
with a collision cell between them, such that selected ions produced in the first MS unit are directed into
the collision cell and further fragmented before being sent to the second MS for detection. The only ions
passed through the collision cell are those selected by the instrument as representing the analytes of
interest, thus minimizing the effects of many potential interferences. These fragments, or "daughter"
ions, are characteristic of the "parent" or "mother" compound, and they are used to positively identify the
analyte in the presence of other analytes and potential interferences. The MS/MS detector can be
operated in several ionization modes, including one that produces positive ions and another that produces
negative ions from the analytes of interest.

Table 3 provides a brief summary of Baylor University's instrumental operating conditions. These
conditions may not be applicable to instruments from other manufacturers or to different lists of target
analytes.

Table 3. HPLC-MS/MS Operating Conditions

HPLC

Varian ProStar Model 210

MS/MS

Varian Model 1200L triple quadrapole mass analyzer equipped with an electrospray interface

Guard column

Agilent Extend-C18 column, 12.5 mm x 2.1 mm, 5 |jm particle size

Analytical column

Agilent Extend-C18 column, 15 cm x 2.1 mm, 5 |jm particle size

Injection volume

10 |jL

Elution gradient

0.1% (v/v) formic acid in water and 100% methanol at 350 pL/min and 30 °C

Collision gas

Argon, at 2.0 mTorr

Run time

50 minutes

Baylor staff selected the optimal electrospray ionization (ESI) mode (either positive or negative) and
MS/MS parameters by infusing standards of each individual target analyte and selecting the conditions
that yielded the most intense precursor ion for each analyte. Immediately prior to analysis, they fortified
each sample extract with internal standards, adding 7-aminoflunitrazapam-d7 and fluoxetine-d6 as the
internal standards for the ESI+ analyses and meclofenamic acid for the ESI- analyses. For routine sample
analyses, laboratory staff identified the target analytes on the basis of chromatographic retention time
(compared to an authentic standard) and the presence of both the parent and daughter ions for each
analyte. They quantified each analyte by an internal standard calibration approach using the extracted
calibration standards.

Section 3.4 Analysis of Personal Care Products by GC-MS/MS

Baylor determined 12 personal care products in fillet tissue samples by gas chromatography-tandem mass
spectrometry (GC-MS/MS) using procedures described by Mottaleb et al. (2009). Baylor originally
planned to use a single MS unit and monitor selected ions (i.e., GC-MS with SIM), but modified those
plans during the method development efforts before the study samples were analyzed. Table 2 (Chapter
1) provides the list of target analytes. Unlike the pharmaceutical method, the personal care product (PCP)

PPCP Fish Pilot Study QA Report

12

March 2023


-------
method did not involve preparation and extraction of matrix-matched calibration standards. Instead,
standards were prepared in solvent, as is common practice for most GC procedures.

3.4.1 Sample Extraction for Personal Care Products

Prior to extraction, each sample was spiked with a set of surrogate compounds to assess extraction
efficiency. The extraction procedures include derivatization of sample extracts to enhance measurement
response. This involved use of N-methyl-N-(trimethylsilyl)trifluoroacetamide (MSTFA) to derivatize
some target analytes to more volatile forms (e.g., musk ketone and triclosan). Extraction of fillet
homogenates for analysis involved the following steps:

1)	weighing out 1.0g±0.01gof fillet tissue for each sample composite and 1.0 g of control matrix for
each QC sample individually to the nearest 0.01 g into a 20-mL borosilicate glass screw-cap vial

2)	spiking the samples with the appropriate surrogates (benzophenone-dio, and />n-nonylphcnol-l3C,)

3)	adding 10 mL of acetone to each spiked homogenate aliquot

4)	sonicating samples for 15 min at 25 °C

5)	shaking the samples vigorously by hand for 20 seconds to further ensure mixing and extraction

6)	transferring extracted samples into 50-mL polypropylene copolymer round-bottomed centrifuge
tubes using lmL acetone as a rinse

7)	centrifuging extracted samples at 16,000 rpm for 40 min at 4 °C

8)	transferring the supernatant into 18-mL disposable glass test tubes

9)	evaporating the solvent to dryness under a stream of nitrogen at 30 °C

10)	reconstituting the samples in 200 (iL of 65:35 (v/v) hexane:acetone in preparation for silica gel
clean-up

11)	loading sample extracts onto a preconditioned (8 mL of 65:35 hexane:acetone by volume) silica gel
column (1 g), and eluting with 30 mL of hexane:acetone

12)	exchanging the solvent by evaporating the resultant extract to near-dryness and reconstituting in
700 |_iL of methylene chloride

13)	injecting one half (350 |_iL) of the extract into the GPC, which was equipped with a cross-linked
styrene divinylbenzene copolymer guard column (30 mm x 4.6 mm) and a similar analytical
(150 mm x 19 mm) column connected in series

14)	eluting the GPC column with methylene chloride and collecting the fraction eluting between
11.4 and approximately 19.4 minutes

15)	concentrating the methylene chloride extract to near dryness and reconstituting it to approximately
200 (.iL in hexane:acetone in a GC injection vial with a PTFE-lined septum

16)	adding 100 (.iL of MSTFA derivatizing agent, capping the GC vial, and heating the mixture in an
oven at 60 °C for 45 minutes

17)	concentrating the derivatized extract to near dryness at room temperature under a stream of nitrogen
then reconstituting it in 180 (.iL of n-hexane and

18)	spiking the extracts prior to analysis with 20 (.iL of the internal standards (phenanthrene-dio and
mirex)

The laboratory prepared samples in batches of 20 or fewer samples, which were accompanied by the
required batch quality control samples, including a method blank, low- and high-level control samples,
and a pair of spiked MS/MSD samples from each site.

PPCP Fish Pilot Study QA Report

13

March 2023


-------
3.4.2 (iC-MS/MS An alysis of Person al Care Products

Table 4 provides a brief summary of Baylor University's instrumental operating conditions. These
conditions may not be applicable to instruments from other manufacturers or to different lists of target
analytes.

Table 4. GC-MS/MS Operating Conditions

GC

Varian Model CP3800

MS

Varian Model 1200 triple-quadrapole mass spectrometer

GC column

Restek XTI-5, 30 m* 0.25 mm * 0.25 |jm film thickness

Carrier gas

Helium, at 1.0 mL/min

Injection volume

1 |jL

1st Temperature gradient

100-180 °C at 15 °C/min

2nd Temperature
gradient

180 - 290 °C at 6 °C/min

Final temperature

290 °C for 6 min

Ionization potential

250 eV

For routine sample analyses, laboratory staff identified the target analytes on the basis of chromatographic
retention time (compared to an authentic standard) and the presence of a primary quantitation ion and at
least one secondary ion. Baylor researchers investigated the use of extracted calibration standards during
the development of their GC-MS/MS procedure for the personal care products, but found that it did not
offer the same advantages that it did for the HPLC-MS/MS analyses of the pharmaceuticals. Therefore,
they quantified the target analytes by an internal standard calibration approach using traditional
calibration standards prepared in pure solvent.

Section 3.5 Quality Control

Although the procedures used by Baylor University were not formal EPA or voluntary consensus
standard body methods, they did include the same types of quality control parameters found in these types
of methods. Laboratory staff implemented the following procedures:

•	Multi-point calibration of all instruments (minimum of five points) with a linearity requirement

•	Calibration verification (each analysis shift)

•	Method blanks prepared with each batch of samples

•	Addition of surrogate compounds to every sample as a measure of extraction efficiency

•	Spiked sample analyses in duplicate (e.g., matrix spike and matrix spike duplicate samples)

•	Laboratory control samples in duplicate (e.g., blank spikes)

Results from all of these QC operations were reported in the final data packages submitted to Tetra Tech
and evaluated by GDIT.

PPCP Fish Pilot Study QA Report

14

March 2023


-------
Chapter 4
Data Quality Assessment

Section 4.1 Data Review

All of the data from the pilot study were subjected to three distinct levels of review by different study
participants. First, all of the laboratory data were reviewed by Baylor's Laboratory Quality Manager,
Kevin Chambliss, before they were reported to Tetra Tech. Tetra Tech staff reviewed the data that the
laboratory submitted to track progress and completeness (e.g., all samples sent to the laboratory were
analyzed) and to assess the effectiveness of the analytical procedures. Finally, the data reviewers
examined the results for each field-based tissue sample and the available quality control data to assess and
document the quality of the data relative to the objectives of the pilot study.

Each data package was thoroughly reviewed to ensure the following:

•	All samples were analyzed and results were provided for each sample analyzed, including results for
any dilutions and reanalyses, and for all associated QC samples

•	All required QC samples were analyzed and these QC samples met specified acceptance criteria

•	Data reporting forms and/or electronically formatted data were provided for each of the field-based
tissue samples and/or associated QC analyses

•	Raw data associated with each field-based tissue sample and QC sample were provided with each
data package, and the instrument output (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

When anomalies were identified, GDIT contacted the laboratory through Tetra Tech and asked them 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.

GDIT developed a database to capture results for each sample and entered results of the data reviews
directly in the database through the application of standardized data qualifier flags and descriptive
comments concerning the reliability of the flagged results. Table 5 contains the individual data qualifiers
that were applied to results from the pilot study and provides an explanation of the implications of each
qualifier for the use of the data.

Note: The presence of data qualifiers is not intended to suggest that data are not useable; rather, the
qualifiers are intended to caution the user about an aspect of the data that does not meet the
acceptance criteria established in the project QAPP.

PPCP Fish Pilot Study QA Report

15

March 2023


-------
Table 5. Individual Data Qualifiers Applied to Pilot Study Data

SCC Code

Definition

Implication





This flag was applied to any analyte detected in the associated method blank. The
data reviewers applied the 5x rule to determine the implication of the levels in the
blank and samples.





If the amount detected in the sample was at least 5xthe blank level, the RNAF flag
was added, indicating that the sample result was not affected.

B

Blank

contamination

If the amount detected in the sample was greater than 1x and less than 5x of the
blank level, then it was likely that the sample result was inflated by background
contamination in the lab that showed up in the blank. In that case, the RMAX flag
was assigned, indicating that the sample result may have been a maximum value
due to the blank contribution. Data users should consider this in making decisions
about RMAX values.

If the amount detected in the sample was less than or equal to the blank amount, the
sample result was reset to a nondetect at the method detection limit, adjusted for
sample size and percent solids, if applicable.

EXC

Excluded

This flag was applied to the results for analytes where the laboratory determined that
analytical difficulties indicated that results could not be reported with any reliability.
For the personal care products analyses by GC-MS/MS, the analytes benzophenone
and octocrylene were excluded. For the pharmaceuticals analyses by HPLC-
MS/MS, the results for miconazole in liver tissue samples were excluded. (The
miconazole results in the fillet tissue samples were not affected.)

HLCS

High LCS
recovery

The lab control sample (LCS) was a clean reference matrix. If recovery in the LCS
was high, there may be a high bias for that analyte. If the analyte was not detected
in a field-based tissue sample, there was no concern and the RNAF flag was applied
as well.

HMSR

High matrix spike
recovery

High matrix spike (MS) recovery indicated a positive interference or a high bias.
Isolated instances of high recovery are not uncommon, and patterns across multiple
MS samples are more of a concern. When high matrix spike recovery was observed
for an analyte, the results for that analyte in all of the samples in the batch with the
matrix spike sample were qualified. However, if the analyte was not detected in a
given sample in the batch, then there was no concern and RNAF was added to the
HMSR flag.

HSSR

High surrogate
spike recovery

Assignment of this flag indicated surrogate recovery above the acceptance limits.
Large exceedances suggested a positive interference or "matrix effect." If the analyte
was not detected in a field-based tissue sample, there was no concern and the
RNAF flag was added to the HSSR flag. Even for detected analytes, exceedances
of a few percent were not cause for concern, given that the methods were being
developed during the study, and therefore, reasonable method performance criteria
and acceptance limits were not well established.

HVER

High VER
recovery

Results for the calibration verification were above the acceptance limit. If the analyte
was not detected in a field-based tissue sample, there was no concern and the
RNAF flag was added to the HVER flag. Results for detected analytes may have a
possible high bias.

LLCS

Low LCS
recovery

If the recovery in the LCS was low, there may have been a low bias for that analyte.
Nondetects in field-based tissue samples may be false negatives and detects may
have a low bias.

LMSR

Low matrix spike
recovery

Low recovery in the matrix spike indicated a potential low bias for the analyte,
possibly due to poor extraction efficiency in the sample matrix. Isolated instances of
low recovery are not uncommon, and patterns across multiple MS samples are more
of a concern. When low matrix spike recovery was observed for an analyte, the
results for that analyte in all of the samples in the batch with the matrix spike sample
were qualified.

PPCP Fish Pilot Study QA Report

16

March 2023


-------
Table 5. Individual Data Qualifiers Applied to Pilot Study Data

SCC Code

Definition

Implication

LSSR

Low surrogate
spike recovery

This flag was assigned for surrogate recovery below acceptance limits. Large
failures suggest a "matrix effect." If extraction efficiency was not sufficient, then all
analytes may have a low bias. Recoveries outside of the acceptance limits by a few
percent were not a serious concern, given that the methods were being developed
during the study, and therefore, reasonable method performance criteria and
acceptance limits were not well established.

LVER

Low VER
recovery

Results for the calibration verification were below the acceptance limit. If the analyte
was detected in a field-based tissue sample, there may be a low bias for that
analyte. Nondetects in field-based tissue samples may be false negatives.

REXC

Result exceeded
calibration range

This flag was applied when the reported result exceeded the instrument calibration
range, but the laboratory noted that it did not saturate the detector. Results above
the demonstrated calibration range may have greater uncertainty associated with the
numerical value.

RMAX

Result is a
maximum value

This flag was applied when other qualifiers indicated a potential positive bias, such
as method blank contamination. Data users should consider these values as upper
limits of the actual concentration.

RNAF

Result not
affected

Assignment of this flag indicated that the overall assessment of the qualifier applied
was that the sample result was not affected. Common examples included analytes
not found in the sample, but found in the associated blank or with high recovery in
the associated matrix spike.

RPDX

Relative percent
difference (RPD)
between MS and
MSD exceeded
criteria

This flag was applied when the precision of the recoveries in the MS and MSD
analyses, measured as the RPD, exceeded the acceptance limits. This may have
been due to a failure in one of the two analyses, and is often the result of a positive
interference in one of the two analyses. The poor precision may have been due to
sample nonhomogeneity, or it may have been an analytical issue. While data users
should be aware of this issue, it is often not a serious data quality concern.

When the 14 individual codes were applied to results from the study, there were a total of 41 unique
combinations of codes. Of those 41 combinations, 22 were applied to the HPLC-MS/MS results for the
pharmaceuticals, 23 were applied to the GC-MS/MS results for the personal care products, and 4 of the
codes overlapped between the two types of analyses. Appendix 1 presents a summary of these qualifier
flag combinations, which includes the frequency of occurrence of each combination by analysis type (e.g.,
a list for pharmaceuticals and a list for personal care products).

Section 4.2 Analysis of Blanks

Blanks are used to verify the absence of contamination that may occur at any point in the measurement
process. In the pilot study, target analytes were expected to occur at low concentrations. Therefore,
frequent analysis and assessment of blanks was critical to determine if measured sample concentrations
were biased by the presence of contamination during analysis.

The data reviewers evaluated each sample result in comparison to the result for that analyte in the method
blank prepared in the same extraction batch. For those analytes reported as present in the method blank,
the data reviewers applied the 5x rule described in Table 5 to determine the potential impact of the blank
contamination on the study results.

The impacts of blank contamination are discussed separately for the two analytical procedures (HPLC-
MS/MS and GC-MS/MS) in Sections 4.2.1 and 4.2.2.

4.2.1 Blanks for the HPLC-MS/MS Analyses

Overall, there were few data quality issues with the blanks from the HPLC-MS/MS analyses of the
pharmaceuticals, as illustrated in Figure 2 on the following page.

PPCP Fish Pilot Study QA Report

17

March 2023


-------
Figure 2 shows that more than 99% of the
pharmaceutical results were not affected by blank
contamination, either because the analyte was not
detected in the blank (83.74%) or because the
concentration in the sample was more than 5x times
the level observed in the blank (16.20%). For
0.06% of the results (1 result for 1 analyte), the data
reviewers judged that the sample result is likely a
maximum value (RMAX) because there is some
chance that the sample result was inflated by the
background contamination from the lab that is
evident in the blank. None of the pharmaceutical
results were changed to nondetects (RNON)
because of blank contamination.

4.2.2 Blanks for the GC-MS/MS Analyses

Figure 3 illustrates the impacts of blank
contamination on the results for the personal care
products. There were even fewer issues with the
blanks from the GC-MS/MS analyses than for the
HPLC-MS/MS analyses. More than 97% of the
personal care product results had no blank qualifier
at all, and all of remaining results (2.78%) were
more than 5x times the level observed in the blank,
and therefore not affected by blank contamination.
These percentages do not include two analytes,
benzophenone and octocrylene, that were detected
often in method blanks associated with a number of
samples. Due to other analytical considerations, all
results for these two analytes were excluded from
the project data set, including these assessments.

Section 4.3 Surrogate Spiking

As noted in Chapter 3, the laboratory spiked each sample with a set of surrogate compounds that were
used to assess extraction efficiency. Surrogates often are isotopically labeled variants of the target
analytes or compounds with similar structures that are not expected to be found in environmental samples.
The analytical activities QAPP (USEPA, 2006b) specified acceptance limits of 60-150% for the recovery
of the surrogates. The impacts of surrogate recoveries are discussed separately for the two analytical
procedures (HPLC-MS/MS and GC-MS/MS) in Sections 4.3.1 and 4.3.2.

4.3.1 Surrogate Recoveries for the HPLC-MS/MS Analyses

During the pilot study, the laboratory used five surrogate compounds in the analysis of the
pharmaceuticals: acetominophen-d4, carbamazephine-dio, diphenylhydramine-ds, fluoxetine-d6, and
ibuprofen-13C3. These five surrogates were recovered within acceptance limits for all but a small
percentage of the samples. Figure 4 illustrates the impacts of surrogate recoveries on the
pharmaceuticals.

0.06%

16.20%

i No blank
qualifier

i Results not
affected

Results
considered a
maximum
value

Figure 2. Impacts of Blank Contamination on the
Pharmaceutical Results

2.78%

i No blank
qualifier

i Results not
affected

Figure 3. Impacts of Blank Contamination on the
Personal Care Product Results

PPCP Fish Pilot Study OA Report

18

March 2023


-------
There were no occurrences of low surrogate
recoveries for the pharmaceutical analyses. Figure
4 shows that over 99% of the samples were not
affected by surrogate recoveries. This includes
92.65% of the samples that met the surrogate
recovery acceptance criterion and 6.89% of the
samples that were associated with higher-than
expected surrogate recovery, but there was no
impact on data quality because the corresponding
analytes were not detected in those samples. Only
0.46% of the samples had detected results that are
associated with high surrogate recoveries. In these
cases, the recovery exceeded the upper acceptance
criterion of 150%. This may indicate either a
positive interference in the sample that inflates the
apparent result for the surrogate, or an issue with
the extracted internal standards where a lower-than-
expected internal standard response could cause similar inflation of the results.

4.3.2 Surrogate Recoveries for the GC-MS/MS Analyses

The laboratory used two surrogates in the analysis of personal care products: benzophenone-dio,
and />nonylphcnol-nC,. More surrogate recovery problems were observed with the personal care
products analyses than for the pharmaceutical analyses. Figure 5 illustrates the impacts of surrogate
recoveries on the personal care products.

Figure 5 shows that more than 63% of the personal
care product results were not affected by surrogate
recovery problems. Of these, 43.52% of the sample
results met the acceptance criterion for surrogate
recoveries and 19.91% were associated with high
surrogate recoveries which did not affect sample
results because the associated analytes were not
detected in the tissue samples.

Surrogate recoveries exceeded the upper
acceptance criterion of 150% for 1.85% of the
samples in which the associated analyte was
detected. For these detected results, the high
surrogate recovery may indicate positive
interferences that could inflate both the surrogate
result and the target analyte result.

As noted in Section 3.4, the laboratory did not use extracted internal standards in the GC-MS/MS analyses,
and the laboratory calibrated the target analytes using solutions prepared in pure solvent. As a result,
quantifying the surrogates in each GC-MS/MS analysis may have been subject to more instrument
variability than in the LC-MS/MS analysis of the pharmaceuticals, yielding more recoveries outside of the
acceptance limits.

Approximately one third (34.7%) of the sample results were affected by low surrogate recoveries. The
low recoveries are indicative of either difficulties in extracting the sample, or unexpected losses of
analytes during any cleanup steps. Both suggest the potential for low bias in the target analyte results,
and an increased chance of not detecting a target analyte that is present at low concentration.

6.89%

0.46%

No surrogate
qualifier

High surrogate
recovery,
results not
affected
High surrogate
recovery
suggesting
high bias

Figure 4. Impacts of Surrogate Recoveries on the
Pharmaceutical Results

¦ No surrogate
qualifier

1.85%

¦ High surrogate
recovery, results
not affected

High surrogate
recovery
suggesting high
bias

Low surrogate
recovery

suggesting low bias

Figure 5. Impacts of Surrogate Recoveries on the
Personal Care Product Results

PPCP Fish Pilot Study OA Report

19

March 2023


-------
Section 4.4 Matrix Spiking

The laboratory used matrix spike (MS) and matrix spike duplicate (MSD) samples to assess the bias and
precision of the analytical methods in a tissue matrix. The laboratory included a MS/MSD pair with each
batch of samples prepared for analysis, spiking the samples with all of the analytes of interest at
concentrations in the upper third of the calibration range. The pilot study analytical activities QAPP
(USEPA, 2006b) specified acceptance limits of 60-150% for MS/MSD recovery. The acceptance limit
for the precision of the MS/MSD analyses was a relative percent difference (RPD) of 40%. The impacts
of matrix spike recoveries and precision are discussed separately for the two analytical procedures
(HPLC-MS/MS and GC-MS/MS) in Sections 4.4.1 and 4.4.2.

4.4.1 Matrix Spike Recoveries and Precision for the HPLC-MS/MS Analyses

Figure 6 shows that more than 95% of the
pharmaceutical analytes were associated with
MS/MSD samples that met the acceptance criteria
for recoveries of the spiked compounds (88.77%)
or were nondetects (6.71%) that were not affected
by the high recoveries. Another 3.1% of the results
were detected analytes associated with high matrix
spike recoveries that suggest a potential high bias
for the target analytes. Low matrix spike recoveries
were associated with 1.39% of the results. As with
the surrogate recoveries, the low matrix spike
recoveries are indicative of either difficulties in
extracting the sample or unexpected losses of
analytes during any cleanup steps. Both suggest a
potential for low bias in the target analyte results,
and an increased chance of not detecting a target
analyte that is present at a low concentration.

For the pharmaceuticals, none of the results are associated with MS/MSD results where the precision did
not meet the acceptance criterion for the relative percent difference (RPD).

A large portion of the 3.13% of the detected results with high matrix spike recoveries is related to the
amount of analyte spiked into the MS/MSD samples relative to the "background" concentration in the fish
liver samples. Although the laboratory spiked the target analytes at concentrations that would be in the
upper third of the calibration range, some of the unspiked liver samples had background concentrations of
a few analytes that were greater than the spike amounts.

The situation was particularly difficult for sertraline, where a number of samples contained 5 to 15 times
the amount of this analyte that was spiked into the MS and MSD aliquots. In these cases, the variability
in background concentration can mask the amount spiked, resulting in the calculation of very large matrix
spike recoveries. For sertraline, an unspiked sample contained over 500 ppb of this analyte, but the spike
level was only 40 ppb. Fluoxetine, norfluoxetine, and gemfibrozil also exhibited high matrix spike
recoveries in some liver MS/MSD samples, although not as high as those for sertraline. Given the
analytical schedule for this project, it was not possible to determine an exact cause for these very high
matrix spike recoveries in the liver samples. Therefore, data users should take the high matrix spike data
qualifiers into account in interpreting the liver sample results for the pharmaceuticals.

6.71 %_

3.13% _ 1.39%

¦ No matrix spike
qualifier

' High matrix spike
recovery, result
not affected

1 High matrix spike
recovery
suggesting high
bias

Low matrix spike
recovery
suggesting low
bias

Figure 6. Impacts of Matrix Spike Recoveries on
the Pharmaceutical Results

PPCP Fish Pilot Study OA Report

20

March 2023


-------
4.4.2 Matrix Spike Recoveries and Precision for the GC-MS/MS Analyses

Figure 7 shows that the matrix spike results for the
personal care products exhibited different patterns
than the pharmaceuticals. About 85% of the
personal care product analytes were associated
with MS/MSD samples that met the acceptance
criteria for recoveries of the spiked compounds
(83.33%), or were nondetects (1.39%) that were
not affected by the high recoveries.

Only 1.39% of the detected results were associated
with high matrix spike recoveries that suggest a
potential high bias for the target analytes. None of
the matrix spike recoveries for the personal care
products were as high as those noted above for
certain pharmaceuticals.

Low matrix spike recoveries were associated with Figure 7. Impacts of Matrix Spike Recoveries on
9.72% of the results, indicative of either difficulties	Personal Care Product Results

in extracting the sample or unexpected losses of analytes during any cleanup steps. Both suggest the
potential for low bias in the target analyte results and an increased chance of not detecting a target analyte
that is present at a low concentration.

An additional 4.17% of the results are associated with MS/MSD results where the precision did not meet
the acceptance criterion of 40% for the RPD. In all those instances, the actual MS/MSD recoveries met
the acceptance criteria, but the RPD between the recoveries in that MS/MSD pair did not meet the
precision criterion.

Section 4.5 Other Quality Control Checks

As part of the data review effort, the data reviewers examined all of the other QC results generated during
the analyses, which included results for laboratory control samples (LCS) and calibration verifications
(VER). the data reviewers compared these QC results to the acceptance criteria in the analytical QAPP
and, when appropriate, flagged the associated sample results using qualifiers shown in Table 5. As with
the QC results described above, there were generally fewer data quality issues with these other QC checks
for the pharmaceuticals analyzed by LC-MS/MS than for the personal care products analyzed by GC-
MS/MS. For example, there were no instances of low LCS results for pharmaceuticals, while LCS and
VER issues were associated with 5.56% of the qualified data for the personal care products.
Pharmaceutical results associated with calibration verifications that did not meet the acceptance criteria
accounted for about 1.57% of the results (1.22% high and 0.35% low), and the personal care products
had 2.31% of the results similarly affected (all low).

Given the relatively low occurrences of the qualifiers for the LCS and VER, particularly for the
pharmaceutical analyses, this section of the report does not present pie charts similar to those in Figures 2
through 7. However, Appendix I reports the percentages of affected results for each analysis type.

Section 4.6 Overall Data Quality Assessment

The approach that EPA used to review the data is designed to maximize the amount of useful data, while
maintaining a high degree of transparency in the process. The end result is that EPA applies a data
qualifier flag to a measurement result when one aspect of the analytical process does meet the QA
acceptance criteria to advise data users. However, many of the data qualifier combinations in Appendix I

1.39%
1.39%

' No matrix spike
qualifier

i High matrix
spike recovery,
result not
affected
High matrix
spike recovery
suggesting
high bias
Low matrix
spike recovery
suggesting low
bias

PPCP Fish Pilot Study OA Report

21

March 2023


-------
also include the "RNAF" qualifier, indicating that the measurement result is not affected by the QA
concern. Common, readily understandable examples include instances where an analyte is found in a
method blank at a very low level, and the sample result for that analyte is much higher (see Section 4.2),
or when the recovery of a spiked analyte is higher than expected, but analyte is not detected in the
unspiked sample (see Section 4.4).

Table 6 illustrates the breakdown of analytes and samples across the two types of analyses
(pharmaceuticals, and personal care products). EPA, Baylor, and Tetra Tech ultimately abandoned
attempts to analyze the fish liver samples for the personal care products, therefore, that cell in Table 6 is
blank.

Table 6. Sample/Analyte Combinations

Analyte Class

Tissue Type

36 Fillets

36 Livers

24 Pharmaceuticals

864

864

12 Personal Care Products

432

-

Total

1296

864

Grand Total

2160

Overall, approximately 93% of the pharmaceutical results either had no qualifiers at all (68.68%) or had
all qualifiers that were applied and also include the "RNAF" code which signifies that the results were not
affected (24.08%). Approximately 2% of the pharmaceutical results were excluded. The excluded results
represent the decision to not pursue analysis of miconazole in the liver samples because of analytical
problems associated with the high lipid content of the liver samples.

In contrast, approximately 34% of the personal care product results either had no qualifiers at all
(18.76%) or had all qualifiers that were applied and also include the "RNAF" code which signifies that
the results were not affected (15.28%). Also, based on pervasive problems with the levels of two personal
care products observed in the laboratory method blanks, all the results for benzophenone and octocrylene
were excluded. These two analytes account for 16.67% of the results for the personal care products.
Considering all the analytes and samples in both analysis types, over 81% of the results had no data
quality issues that affect their use in meeting EPA's study objectives. Appendix I provides additional
details regarding the frequency at which each flag or flag combination was applied to the study results.

Section 4.7 Completeness

Completeness is a measure of the amount of data that are collected and deemed to be acceptable for use
the intended purpose. The sample collection QAPP (USEPA, 2006a) and the analytical activities QAPP
(USEPA, 2006b) for the pilot study identified three measures of completeness:

Sampling Completeness: The number of 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 completeness goal in this study was to obtain valid measurements from 95% of the samples planned
for collection.

PPCP Fish Pilot Study QA Report

22

March 2023


-------
4.7.1 Sampling Completeness

The sample collection QAPP (USEPA, 2006a) states that "the completeness goal is achieved when five
effluent-dominated sites and one reference quality site found to contain target fishes are sampled, and the
fish tissue samples are shipped with no errors in documentation or sample handling procedures. "

Despite the challenges encountered by the sampling team, particularly at the remote reference site,
samples were collected at all 6 sites as planned, (yielding 36 fillet composites and 36 liver composites).
Therefore,, sampling completeness was 100%.

4.7.2	Analytical Completeness

For multi-analyte methodologies, analytical completeness is best calculated on the basis of the number of
possible sample/analyte combinations. Otherwise, a problem with a single analyte could be seen as
invalidating an entire field sample.

For the pilot study, there were 36 fillet composites and 36 liver composites, for a total of 72 samples.
There were 24 pharmaceutical analytes and 12 personal care product analytes. Thus, the number of
planned sample/analyte measurements was 2592 (72 x 36).

However, as noted elsewhere in this report, there were significant difficulties with the analysis of the liver
samples for the personal care products. Therefore, EPA made an informed decision not to pursue the
liver analyses for the personal care products. For the purposes of this report, the total number of planned
observations is 2160, as shown in Table 6 above.

Baylor reported the results for miconazole in the liver samples as "NA" for not analyzed. However, that
decision was based on significant analytical issues for this one pharmaceutical that affected data quality.
During the data quality assessment, these 36 miconazole results from the liver samples were excluded
from the data set. Similarly, as noted in Section 4.6, all 36 results for two personal care products,
benzophenone and octocrylene, were excluded from the data set (72 results in total). Therefore analytical
completeness was calculated based on 2052 valid sample/analyte combinations out of a possible 2160,
which is 95%.

4.7.3	Overall Completeness

Overall completeness can be calculated as the product of the sampling completeness and the analytical
completeness. Overall completeness for this study is 95% (e.g., 100% sampling completeness times 95%
analytical completeness). Thus, EPA met its completeness goal.

PPCP Fish Pilot Study QA Report

23

March 2023


-------
References

Brooks, B. W., C. K. Chambliss, J. K. Stanley, A. J. Ramirez, K. E. Banks, R. D. Johnson, and R. J.
Lewis. 2005. Determination of select antidepressants in fish from an effluent dominated stream.

Environmental Toxicology and Chemistry 24:464-469.

Mottaleb, M. A., S. Usenko, J. G. O'Donnell, A. J. Ramirez, B. W. Brooks, and C. K. Chambliss. 2009.
Gas chromatography-mass spectrometry screening methods for select UV-filters, synthetic musks,
alkylphenols, an antimicrobial agent, and an insect repellent in fish. Journal of Chromatography A
1216:815-823.

Ramirez, A. J., 2007. Determination of Pharmaceuticals and Personal Care Products in Fish Using High
Performance Liquid Chromatography-Tandem Mass Spectrometry and Gas Chromatography-Mass
Spectrometry, Department of Chemistry and Biochemistry. Baylor University, p. 158.

Ramirez A. J., R. A. Brain, S. Usenko, M. A. Mottaleb, J. G. O'Donnell, L. L. Stahl, J. B. Wathen, B. D.
Snyder, J. L. Pitt, P. Perez-Hurtado, L. L. Dobbins, B. W. Brooks, and C. K. Chambliss, 2009,
Occurrence of pharmaceuticals and personal care products in fish: Results of a national pilot study in the
United States, Environmental Toxicology and Chemistry, 28(12):2587-2597, 2009.

USEPA, 2000, Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories,

Volume 1: Fish Sampling and Analysis, Third Edition. EPA-823-B-00-007. U.S. Environmental
Protection Agency, Office of Water, Washington, D.C.

USEPA, 2006a, Quality Assurance Project Plan for Sample Collection Activities for a Pilot Study to
Investigate the Occurrence of Pharmaceuticals and Personal Care Products (PPCPs) in Fish Tissue,
August 1, 2006. U.S. Environmental Protection Agency, Office of Water, Washington, DC

USEPA, 2006b, Quality Assurance Project Plan for Laboratory Sample Preparation and Analysis
Activities in the National Pilot Study of Pharmaceuticals and Personal Care Products (PPCPs) in Fish
Tissue, October 19, 2006. U.S. Environmental Protection Agency, Office of Water, Washington, D. C

USEPA, 2023, Pilot Study of Pharmaceuticals and Personal Care Products in Fish Tissue. EPA 820-R-
23-003. U.S. Environmental Protection Agency, Office ofWater, Washington, DC, March 2023.

PPCP Fish Pilot Study QA Report

24

March 2023


-------
Appendix I

Unique Qualifier Combinations Applied to Pilot Study Data

by Analytical Method

PPCP Fish Pilot Study QA Report

A-l

March 2023


-------
Table A-1. Unique Qualifier Combinations Applied to Pharmaceutical Data

SCC Code

% of Results

% Not Affecting Data Quality

B, RMAX

0.06



B, RNAF

10.76

10.76

B, RNAF

HMSR

2.43



B, RNAF

HMSR, RNAF

1.39

1.39

B, RNAF

HMSR, RNAF; HSSR, RNAF

0.23

0.23

B, RNAF

HMSR; HSSR

0.35



B, RNAF

HMSR; HSSR; HVER

0.23



B, RNAF

HMSR; HVER

0.12



B, RNAF

HSSR

0.23



B, RNAF

HSSR, RNAF

0.12

0.12

B, RNAF

HSSR, RNAF; LVER

0.06



B, RNAF

LVER

0.29



EXC

2.08



HLCS, RNAF; HSSR, RNAF

0.35

0.35

HLCS, RNAF; HVER, RNAF

0.12

0.12

HMSR, RNAF

4.51

4.51

HMSR, RNAF; HSSR, RNAF

0.41

0.41

HMSR, RNAF; HSSR; RNAF; HVER, RNAF

0.17

0.17

HSSR, RNAF

5.21

5.21

HSSR, RNAF; HVER, RNAF

0.35

0.35

HVER, RNAF

0.46

0.46

LMSR

1.39









No qualifier codes

68.68



No qualifier codes, plus those not affecting data quality

92.76



The codes are listed in this table in alphabetical order. A detailed explanation of each code can be found
in Table 5. The frequencies are based on a total of 1728 sample/analyte combinations for this analysis
type. All frequencies are rounded to two decimal places.

If all of the data qualifiers in a particular combination of codes are followed by the "RNAF" code, then
the data quality of the results for that sample/analyte combination is not affected, and the frequency
appears in the third column as well.

For HPLC-MS/MS analyses, 68.68% of the results were not qualified at all (e.g., the "SCC Code" field in
the database is blank). An additional 24.08% of the results had the "RNAF" code for each qualifier, and
therefore, the net effect of the qualifiers did not affect data quality.

PPCP Fish Pilot Study QA Report

A-2

March 2023


-------
Table A-2. Unique Qualifier Combinations Applied to Personal Care Product Data

SCC Code

% of Results

% Not Affecting Data Quality

B, RNAF

LSSR; REXC; RPDX

0.23



B, RNAF

LSSR; RPDX

1.85



B, RNAF

RPDX

0.69



EXC

16.67



HMSR

0.46



HMSR, RNAF

1.39

1.39

HMSR; HSSR

0.93



HSSR

0.93



HSSR, RNAF

13.89

13.89

HSSR, RNAF

LLCS

2.31



HSSR, RNAF

LMSR

1.39



HSSR, RNAF

LMSR; LVER

0.93



HSSR, RNAF

LVER

1.39



LLCS

1.39



LLCS

LMSR

0.46



LLCS

LMSR; LSSR

2.31



LLCS

LSSR

1.39



LMSR

1.16



LMSR; LSSR

3.47



LSSR

26.62



LSSR; REXC; RPDX

0.46



LSSR; RPDX

0.69



RPDX

0.23









No qualifier codes

18.76



No qualifier codes, plus those not affecting data quality

34.04



The codes are listed in this table in alphabetical order. A detailed explanation of each code can be found
in Table 5. The frequencies are based on a total of 432 sample/analyte combinations for this analysis
type. All frequencies are rounded to two decimal places.

If all of the data qualifiers in a particular combination of codes are followed by the "RNAF" code, then
the data quality of the results for that sample/analyte combination is not affected, and the frequency
appears in the third column as well.

For GC-MS/MS analyses, 18.76% of the results were not qualified at all (e.g., the "SCC Code" field in
the database is blank). An additional 15.28% of the results had the "RNAF" code for each qualifier, and
therefore, the net effect of the qualifiers did not affect data quality.

PPCP Fish Pilot Study QA Report

A-3

March 2023


-------