Mercury Science Inc.
4802 Glendarion Drive, Durham NC 27713
(866) 861-5836
mercuryscience.com
Domoic Acid ELISA Test
Kit
ISSC Method Application
Submitted by Mercury Science Inc.
11/15/2010
2
SLV Report & Results
Table of Contents
Executive Summary ........................................................................................................................ 4
Method Application and Single Lab Validation Checklist ............................................................. 5
Single Laboratory Validation (SLV) Protocol ................................................................................ 7
SLV Documents for Marine Biotoxin and Non-MPN Based Microbiological Methods ............. 13
ELISA Test Kit User Guide .......................................................................................................... 21
Appendix I. DAK-36 Kit Stability ................................................................................................ 36
Appendix II. Mercury Science Lab Validation Results ................................................................ 37
Appendix III. Accuracy and Trueness Data and Calculations ...................................................... 40
Appendix IV. Ruggedness Data and Calculations ........................................................................ 50
Appendix V. Precision & Recovery Data and Calculations ......................................................... 60
Appendix VI. Specificity Data and Calculations .......................................................................... 69
Appendix VII. Comparing Methods Data and Calculations ......................................................... 80
Appendix VIII. Limit of Quantitation/Limit of Detection Data and Calculations ....................... 91
Appendix IX. Definitions............................................................................................................ 108
Acknowledgements ..................................................................................................................... 109
4
Executive Summary
DAK-36 Domoic Acid Test Kit
Method Application and Single Lab Validation Report
This document contains the results of validation testing for the Domoic Acid Test Kit (product #
DAK-36) manufactured by Mercury Science Inc. Testing was performed on seven species of
shellfish at the Center for Coastal Fisheries and Habitat Research laboratory in Beaufort, NC.
The graph in Figure 1 shows the cumulative results of over 1500 data points obtained during this
study. Precision of replicate assays was approximately 15% which is not unusual for
immunoassays. Although not as precise as HPLC analysis, the DAK-36 assay demonstrates an
ability to provide a rapid, sensitive and low cost alternative to HPLC analysis of shellfish. All
unspiked samples analyzed showed low levels of domoic acid (less than 1 ppm). The DAK-36
assay demonstrated low levels of matrix interference across all species tested. No interfering
compounds were found at naturally occurring levels.
The limit of detection for the method was less than 3 ppm for all species.
The body of this report tabulates the performance of the test for seven species. A Table of
contents is provided to assist in navigating the document.
An appendix containing detailed analysis of all the data collected is provided at the end of this
report. The appendix also includes results from validation testing performed independently at
the laboratory of Mercury Science Inc.
Figure 1. More than 1500 analyses of
spiked samples ranging from 2 ppm
DA to 40 ppm DA from seven different
species were tested during this study.
The resulting correlation with the
spiked value indicates good accuracy.
Quantitative values were calculated
from DAK-36 data using the
calibration curve provided by Mercury
Science Inc.
Executive Summary
Method Application and Single Lab Validation Checklist
(http://www.issc.org/client_resources/lmr%20documents/i.%20issc%20lab%20method%20application%20checklist.pdf)
ISSC Method Application and Single Lab Validation Checklist For Acceptance of a Method for Use in the NSSP
The purpose of single laboratory validation in the National Shellfish Sanitation Program (NSSP) is to ensure that the
analytical method under consideration for adoption by the NSSP is fit for its intended use in the Program. A Checklist
has been developed which explores and articulates the need for the method in the NSSP; provides an itemized list of
method documentation requirements; and, sets forth the performance characteristics to be tested as part of the
overall process of single laboratory validation. For ease in application, the performance characteristics listed under
validation criteria on the Checklist have been defined and accompany the Checklist as part of the process of single
laboratory validation. Further a generic protocol has been developed that provides the basic framework for
integrating the requirements for the single laboratory validation of all analytical methods intended for adoption by the
NSSP. Methods submitted to the Interstate Shellfish Sanitation Conference (ISSC) Laboratory Methods Review
(LMR) Committee for acceptance will require, at a minimum, six (6) months for review from the date of submission.
DOMOIC ACID RAPID ENZYME-LINKED
IMMUNOSORBENT ASSAY – 96 Well
Format
Name of the New Method
Name of the Method Developer
Mercury Science Inc. and the National
Oceanic and Atmospheric Administration
Attn: Tom Stewart
4802 Glendarion Dr.
Durham, NC 27713
Phone: (866) 861-5836
Developer Contact Information
Checklist
A. Need for the New Method
1. Clearly define the need for which the
method has been developed.
2.
What is the intended purpose of the method?
3.
Is there an acknowledged need for
this method in the NSSP?
4.
What type of method? i.e. chemical,
molecular, culture, etc.
Y/N
Y
Y
Y
Y
Submitter Comments
Faster, more affordable DA analysis
Monitoring shellfish and water samples for DA
Faster analysis decreases public health risks
Enzyme-Linked Immunosorbent Assay
(ELISA)
B. Method Documentation
1.
Method documentation includes the
following information:
Method Title
Method Scope
References
Principle
Any Proprietary Aspects
Equipment Required
Reagents Required
Sample Collection, Preservation and
Storage Requirements
Safety Requirements
Y
Y
Domoic Acid Rapid EnzymeLinked ImmunoSorbent Assay
(ELISA) – 96 Well Format
Y
Y
Y
Y
Y
For the analysis of food, phytoplankton, and
water
Peer Reviewed Publication, Independent
Correlation Study
Competitive ELISA
Unique Antibody and Enzyme Conjugate
Equipment is listed for this method
Reagents are listed for this method
Requirements are described for this method
Y
Normal Good Lab Practices
Y
6
C.
1.
2.
3.
Clear and Easy to Follow Step-by-Step
Procedure
Quality Control Steps Specific for this
Method
Validation Criteria
Accuracy / Trueness
Measurement Uncertainty
Precision Characteristics (repeatability and
reproducibility)
Y
Described below or see User Guide
Y
Described below
Y
Y
Tested 7 species of shellfish
Tested 7 species of shellfish
Y
Tested 7 species of shellfish
4. Recovery
Y
5. Specificity
Y
6. Working and Linear Ranges
N/A
7.
8.
9.
10.
11.
Y
Y
Y
Y
Limit of Detection
Limit of Quantitation / Sensitivity
Ruggedness
Matrix Effects
Comparability (if intended as a substitute
for an established method accepted by the
NSSP)
Tested 7 species of shellfish but used
alternative method for calculating statistics
Tested 7 species of shellfish
Ranges were tested and established prior to
the validation study
Tested 7 species of shellfish
Tested 7 species of shellfish
Tested 7 species of shellfish
Y
Tested 4 species of shellfish from two different
locations using the established HPLC method
of detection
Y
$200 per 36 duplicate samples
Y
Some ELISA experience or training required
Y
See list
Y
See list
Y
90 minutes
Y
See attached
D. Other Information
1.
2.
3.
4.
5.
6.
Cost of the Method
Special Technical Skills Required to
Perform the Method
Special Equipment Required and
Associated Cost
Abbreviations and Acronyms Defined
Details of Turn Around Times (time
involved to complete the method)
Provide Brief Overview of the Quality
Systems Used in the Lab
Submitters Signature
Date:
November 15, 2010
Submission of Validation Data and
Draft Method to Committee
Date:
Reviewing Members
Date:
Accepted
Date:
Recommendations for Further Work
Date:
SLV Report & Results
7
Single Laboratory Validation (SLV) Protocol
For Submission to the Interstate Shellfish Sanitation Conference (ISSC)For Method Approval
(http://www.issc.org/client_resources/lmr%20documents/iii.%20slv%20protocol%20for%20method%20approval.pdf)
Information: Applicants shall attach all procedures, with materials, methods, calibrations and
interpretations of data with the request for review and potential approval by the ISSC. The ISSC
also recommends that submitters include peer-reviewed articles of the procedure (or similar
procedures from which the submitting procedure has been derived) published in technical
journals with their submittals. Methods submitted to the ISSC LMR committee for acceptance
will require, at a minimum, 6 months for review from the date of submission.
Note: The applicant should provide all information and data identified above as
well as the following material, if applicable:
Justification for New Method
Name of the New Method.
Domoic acid rapid enzyme-linked immunosorbent assay -96 well format
(Marketed by Mercury Science Inc. as Product # DAK-36 Domoic Acid Test Kit.)
Specify the Type of Method (e.g., Chemical, Molecular, or Culture)
Enzyme linked immunosorbent assay (ELISA) using an anti-domoic acid monoclonal
antibody
Name of Method Developer
The DA assay kit was developed jointly by NOAA’s National Centers for Coastal Ocean
Science, National Ocean Service, and the Northwest Fisheries Science Center, together
with an industry partner Mercury Science, Inc., Durham, North Carolina
Developer Contact Information [e.g., Address and Phone Number(s)]
Mercury Science Inc.
Attn: Tom Stewart
4802 Glendarion Dr.
Durham, NC 27713
Phone: (866) 861-5836
Date of Submission
November 15, 2010
SLV Report & Results
8
Purpose and Intended Use of the Method
The method is an accurate, rapid, cost-effective tool for use by environmental managers
and public health officials to monitor Domoic Acid concentrations in environment
samples.
Need for the New Method in the NSSP, Noting Any Relationships to Existing Methods
The regulatory method for DA detection sanctioned by the Interstate Shellfish Sanitation
Conference is a high performance liquid chromatography (HPLC) assay. Though
accurate, these analyses are generally run by centralized state facilities with results
typically not available for 3 to 14 days after the samples are collected. In more remote
communities, many of which depend heavily on subsistence clam harvests, these long
delays and the costs of sample analysis are causes for public health concern. The average
cost of approximately $100 per sample limits the number of samples that can be analyzed
(Harold Rourk, Washington State Department of Health, personal communication).
Resource managers in coastal communities have expressed their desire for a costeffective method for rapid and accurate determination of DA concentrations in shellfish
and phytoplankton samples. The high throughput capacity of the assay also allows for
much faster response times when domoic acid events occur. The relatively low cost of the
assay means that significantly more sampling is also possible on the same or smaller
budget.
Method Limitations and Potential Indications of Cases Where the Method May Not Be
Applicable to Specific Matrix Types
This ELISA is sensitive to organic solvents such as methanol. Sample extracts that
contain methanol can be diluted with Sample Dilution Buffer (provided in the kit) to
reduce methanol concentrations to less than 1%.
Other Comments
The implementation of this ELISA system required the development and validation of
two essential reagents, a high avidity monoclonal antibody to DA and a stable DA-HRP
conjugate recognized by the same monoclonal antibody.
Method Documentation
Method Title
Domoic Acid Rapid Enzyme-Linked ImmunoSorbent Assay (ELISA) – 96 Well Format
Method Scope
The method is a sequential competitive enzyme linked immunosorbent assay (ELISA)
utilizing a high avidity monoclonal antibody (mAb) to DA to ensure assay specificity and
SLV Report & Results
9
consistency across production lots. The assay is specific for Domoic Acid and can be
used for the analysis of tissue extracts, phytoplankton samples, and water samples.
References
Litaker, R.W, Stewart, T.N., Eberhart, B.-T. L., Wekell, J.C., Trainer, V.L., Kudela,
R.M., Miller, P.E., Roberts, A., Hertz, C., Johnson, T.A., Frankfurter, G., Smith, J.G.,
Schnetzer, A., Schumacker, J., Bastian, J.L, Odell, A., Gentien, P., Le Gal, D., Hardison,
R.D., Tester, P.A. 2008. Rapid enzyme-linked immunosorbent assay for detection of the
algal toxin domoic acid. Journal of Shellfish Research 27(5): 1-10.
Available online at: http://mercuryscience.com/LitakerStewartDec2008.pdf
User Guide Available Online at:
http://www.mercuryscience.com/DAUserGuide2009B.pdf
Principle
A fixed number of anti-DA mAb binding sites are incubated with dissolved DA in the
sample followed by the addition of a DA – horseradish peroxidase (HRP) conjugate. As
these binding events occur, the anti-DA mAb molecules are simultaneously captured by
anti-mouse antibodies affixed to the surface of the microtiter plate wells. Following a
wash step, subsequent HRP derived color development, readable on standard microplate
readers, was inversely proportional to the concentration of DA in the sample matrix. The
assay reagents were titrated so that the amount of mAb and the DA–HRP conjugate
added produced a maximal absorbance signal of approximately 2.5 absorbance units
when no DA was present.
Analytes/Measurands
Domoic Acid
Proprietary Aspects
The assay uses a unique monoclonal antibody and enzyme conjugate developed by
Mercury Science Inc.
Equipment
Microtiter plate orbital shaker
Automated microtiter plate washer
Multichannel pipette
Pipetman (P20, P200, P1000) or equivalent
Microtiterplate reader (capable of reading at 450nm)
Reagents
SLV Report & Results
10
1.
2.
3.
4.
5.
6.
7.
anti-DA antibody
DA-HRP conjugate
Assay Buffer
Control Solution
Wash solution
TMB substrate
Stop solution
Media
Tissue samples are extracted using a solvent of Methanol:Water (50:50, v:v)
Extracts are diluted into an aqueous sample buffer prior to analysis by the ELISA.
Water samples are filtered and diluted in sample buffer prior to analysis by the ELISA.
Phytoplankton samples are ruptured by appropriate methods in aqueous sample buffer
prior to analysis by the ELISA.
Matrix or Matrices of Interest
Butter clam (Saxidomus giganteus), blue mussel (Mytilus edulis), California mussel
(Mytilus californianus), cockles (Clinocardium nuttalli ), geoduck (Panopea abrupta),
littleneck clam (Protothaca staminea), manila clam (Venerupis japonica), oyster
(Crassostrea gigas and Crassostrea virginica), and razor clam (Siliqua patula) tissues, as
well as phytoplankton and water samples.
Sample Collection, Preservation, Preparation, Storage, Cleanup, etc.
Shellfish preparation: In the case of shellfish, pooled samples of 10-12 individuals are
cleaned, and ground to a smooth and uniform homogenate in a commercial blender.
Approximately 1 g of homogenized tissue is added to a tared 15 mL conical tube and the
weight recorded to the nearest 0.01g. Next, 9 mL of 50% methanol are added and the
samples mixed at high speed on a vortex mixer for 1 min. Once the extraction is
complete, the tubes are spun in a table top centrifuge for 20 min at 10,000xg or until a
tight pellet and clear supernatant are obtained. Alternatively, ~1mL of vortexed sample
can be transferred to a microfuge tube and spun in a microfuge at 14,000xg for 5 minutes.
If the samples do not clear despite the spinning at high speed, the supernatant is passed
through a 0.45 µm syringe filter. The extract is then diluted 1:100 or 1:1000 into Sample
Dilution Buffer and is ready for analysis by ELISA. If necessary, the sample may be
stored at 4ºC for up to 24 h in a refrigerator prior to analysis.
Phytoplankton preparation: Approximately 0.1 to 1.0 L of cultured cells or sea water
samples are filtered onto a GF/F filter which can be immediately frozen at -80oC until the
filter can be processed or processed immediately. For processing, filters are placed in a
5mL conical tube and 3 mL of 20% methanol are added. The samples are sonicated until
the filter is completely homogenized. Care is needed to prevent the probe from rupturing
SLV Report & Results
11
the tube. The sonicator probe is cleaned carefully with 20% methanol between samples
to prevent cross-contamination. Next the homogenate is centrifuged at 3000xg for 10
minutes. The supernatant is passed through a 0.2 µm syringe filter. The extract is then
diluted into Sample Dilution Buffer and is ready for analysis by ELISA.
Storage of test kit: Any unused strips can be removed and stored in a desiccator pouch at
4oC for at least six months without compromising assay performance
Safety Requirements
General Good Laboratory Practices should be followed at all times.
Safety Glasses should be worn at all times.
The Stop solution in the assay contains 1 M hydrochloric acid. Care must be taken to
avoid skin or eye contact with the Stop solution.
Other Information (Cost of the Method, Special Technical Skills Required to Perform the
Method, Special Equipment Required and Associated Cost, Abbreviations and Acronyms
Defined and Details of Turn Around Times [Time Involved to Complete the Method])
Cost of the Method: The DAK 36 Domoic Acid Test Kit costs $200 and contains
sufficient assay reagents to perform 36 sample analyses (less than $6 per sample)
Special Technical Skills Required to Perform the Method: It is recommended that users
have prior experience performing ELISA assays or receive training from Mercury
Science Inc.
Special Equipment Required and Associated Cost (estimated):
 Microtiterplate orbital shaker
 Automated microtiterplate washer
 Multichannel pipette
 Pipetmen (P20, P200, P1000) (or equivalent)
 Microtiterplate reader (capable of reading at 450nm)
$500
$5,000
$700
$1,500
$6,500
This equipment is commonly available in most state laboratories.
Abbreviations and Acronyms Defined:
ELISA – Enzyme-Linked Immunosorbent Assay
HRP – Horseradish Peroxidase
TMB – Tetramethylbenzidine
DA – Domoic Acid
mAb – monoclonal Antibody
Details of Turn Around Times: As many as 36 sample extracts can be analyzed in <1.5
hours.
SLV Report & Results
12
Test Procedures, (Be Specific and Provide Easy-to-Follow Step-by-Step Procedures and
indicate critical steps)
The 96 well assay tray used in the assay contains 12 strips. Each strip of 8 wells can be
removed and stored until it is needed. The first two wells of each strip are used as a
control (no DA added). The remaining six wells are used to analyze 3 samples in
duplicate. This format provided the flexibility of running anywhere from 3 to 36
duplicate samples at a time.
1. For unknown sample analysis, extracts are diluted to a final concentration ranging
from 0.3 to 3 to ppb using the Sample Dilution Buffer [phosphate salt solution, pH
7.8, containing casein]. For clam tissues containing DA, sample dilutions of 1:100
and 1:1000 are typically used. (Preliminary tests with razor clam extracts showed
that a 25-fold dilution in sample dilution buffer eliminates matrix effects in ELISA
analysis.)
2. The immunoassay is started by adding 50 µl of the anti-DA antibody reagent to each
well using a multi-channel pipette.
3. Next, 50 µl of the Control solution (sample buffer without DA) is added to the first
two wells in each strip.
4. Duplicate 50 ul aliquots from the diluted DA extracts are then added to the remaining
wells in each strip and the plate is shaken at room temperature for 30 minutes on an
orbital shaker set to vigorously mix the solution in each well. Vigorous mixing is
key to reaching equilibrium in the allotted time and obtaining replicable results
from one run to the next. In this step, DA in the sample binds to available mAb in
proportion to [DA].
5. At the end of the incubation, 50 µl of DA HRP conjugate is added to each well and
the plate is shaken a second time for 30 min at room temperature on an orbital shaker.
The DA-HRP will bind to available mAb sites.
6. Following Step 5, the plate is washed three times with wash solution [Tris-HCl
buffered salt solution (pH 7.8) containing Tween 20 and sodium azide as a
preservative] using a commercial plate washer, making certain the fluid is completely
aspirated from all the wells. Alternatively, these washes can be done manually by
adding wash solution to wells using a multichannel pipettor and then flicking all fluid
from the wells. The manual method may result in slightly higher variability.
7. Next, 100 µL of SureBlue TMB substrate (5,5’-tetramethylbenzidine, kpl.com) is
added to each well.
8. The plate is placed on an orbital shaker for no more than 5 minutes, or until adequate
color development is observed.
9. Color development is terminated by adding 100 µL stop solution (1N hydrochloric
acid) to each well.
10. The absorbance in each well is measured at 450 nm using a plate reader.
11. The DA concentrations are determined using the sample (B) and control (Bo)
absorbances, the original tissue weights, and the volume of 20% or 50% methanol
used to extract each sample.
SLV Report & Results
13
The actual calculations are made using a Microsoft Excel work sheet that incorporates
the constants for a four parameter model (DA concentration =ED50(Bo/B -1)-slope).
This worksheet can be downloaded from:
http://www.mercuryscience.com/Domoic%20Acid%20Quantitation%208Well%20Strip.xls
Processing time for this assay is approximately 1.5 hours.
Quality Control (Provide Specific Steps)
Bo signals should be greater than 1.5 AU and less than 3.0 AU. When Bo values are
greater than 3.0, the user can remove 50 ul of the yellow solution from ALL wells on that
strip and re-read the signal.
Duplicate signals should be within 10% of their average value. For example: Two
duplicate wells having AU values of 1.500 and 1.600 are acceptable because the
difference between the values and their average (1.550) is less than 10%. If two
duplicate wells have AU values of 1.000 and 1.400, this result is invalid and should be
retested because the variation between the values is too great because: (1.200 1.000)/1.000 = 20%
Domoic Acid standard solutions can be run as needed to QC the accuracy of the assay.
QC protocols can be developed on a case-by-case basis with assistance provided by
Mercury Science Inc.
SLV Documents for Marine Biotoxin and Non-MPN Based Microbiological Methods
Validation Criteria (Include Accuracy / Trueness, Measurement Uncertainty, Precision
[Repeatability and Reproducibility], Recovery, Specificity, Working and Linear Ranges, Limit of
Detection, Limit of Quantitation / Sensitivity, Ruggedness, Matrix Effects and Comparability (if
intended as a substitute for an established method accepted by the NSSP).
Validation criteria were tested by following the protocols specified on the ISSC website for
Single Lab Validation for Laboratory Methods under section VII. SLV Documents for Marine
Biotoxin and Non-MPN Based Microbiological Methods (http://www.issc.org/lmrforms.aspx).
Seven different species of shellfish were tested for each criterion.
Validation criteria methods
Shellfish tissue sample collection
For the Beaufort NOAA validation trials, six species of shellfish were shipped from the
Northwest Fisheries Science Center in Washington to the NOAA CCFHR lab in Beaufort, NC.
These species included blue mussels (Mytilus edulis), butter clams (Saxidomus giganteus),
geoducks (Panopea generosa), manila clams (Venerupis philippinarum), Pacific oysters
(Crassostrea gigas), and razor clams (Siliqua patula). Each sample consisted of ~100g of
SLV Report & Results
14
homogenate made from blending 12 individuals. Geoduck samples were the only exception due
to logistical feasibility, and these homogenate samples consisted of 1-2 organisms. A seventh
species, Atlantic oysters (Crassostrea virginica), was obtained from a local vendor in Beaufort,
NC. Whole oysters were shucked and 12 individuals were blended to make the homogenate for
each sample. HPLC was performed at the Beaufort lab for the sample blanks of all species to
determine background levels of DA.
Domoic acid stock
The domoic acid stock was made up from 5mg Sigma domoic acid >90%. The DA powder was
added to 5ml 10% acetonitrile 90% dH2O solution. The stock was tested by HPLC and
compared to Canadian National research Council (NRC) (CRM DA-f) standards to determine the
exact concentration prior to spiking the samples.
Extraction
For each sample, a 1g aliquot of tissue homogenate was placed into a 15mL conical tube and
9mL 50% MeOH was added. For blank/unspiked samples, the tube was capped and vortexed for
1 minute. For spiked samples, domoic acid stock was added in an amount equal to the desired
concentration in tissue (in ppm). The sample was then capped and vortexed for 1 minute.
Approximately 1.2 ml of vortexed sample was transferred to a microfuge tube and centrifuged at
14000 rpm for 5 minutes. To be evaluated on the ELISA, spiked or natural samples must be
diluted into the working range of the assay of .3 to 3ppb. For consistency, the samples for all of
the tests were diluted at 1:1000. This dilution factor was sufficiently high to dilute out any matrix
effects from the different tissues or the methanol used for extraction, and it ensured accurate
detection for the relevant range of DA for regulatory purposes (able to detect low levels of
<5ppm, regulatory limit of 20ppm and high levels of >40ppm).
Validation Methods & Summary
Accuracy/Trueness & Measurement Uncertainty
For each sample, two 1g aliquots of tissue homogenate were tested. One aliquot was left blank
and one was spiked with a known concentration of DA. Both aliquots were processed on the
ELISA using a dilution factor of 1:1000. Twenty samples of each species were analyzed using a
range of DA concentrations equivalent to 5, 10, 20 and 40ppm in tissue.
Data summary
Percent accuracy/trueness was calculated by calculating the average concentration of the ELISA
results, dividing it by the average calculated spike value and multiplying by 100. To calculate
measurement uncertainty, each ELISA value was subtracted from the known spike value. The
two-sided 95% confidence interval was calculated for the differences, which represents the
measurement uncertainty. Percent accuracy/trueness ranged from 96to 108.6% with an average
of 100.8%. The measurement uncertainties ranged from -2.6 to 2ppm. This shows that the
SLV Report & Results
15
ELISA yields a very high level of accuracy with an average error within ±2ppm. See appendix
III for actual data/calculations.
Species
Butter Clam
Pacific Oyster
Geoduck
Razor clam
Blue mussel
Atlantic Oyster
Manila clam
% Accuracy/ Trueness Measurement uncertainty
108.6
-2.63 to -0.34
100.5
-0.99 to 0.81
103.7
-1.60 to 0.34
96.3
-0.55 to 2.01
97.5
-0.40 to 1.30
100.8
-1.63 to 1.33
98.4
-0.93 to 1.54
Ruggedness
For each sample, one aliquot of tissue homogenate was spiked with a known concentration of
DA. The spiked sample was split into two aliquots and run on two separate lots of ELISA test
kits. For each species, 10 samples were spiked, covering a range from 5 to 40ppm. For each
species the same two lots of kits were used as “Lot 1” and “Lot 2” for all 10 samples.
Data summary
To analyze the data, each lot was tested for symmetry and variance. The symmetry for all
species/lots was <2, indicating that the data were not skewed. Also, the ratios of the larger to
smaller of the variances for lot 1 and 2 were <2, showing that there was not a lack of
homogeneity. Since the symmetry tests and the variance ratios were all <2, a paired t-test was
used to test whether there was a significant difference between lots. None of the species showed
a significant difference between lots. See appendix IV for actual data/calculations.
Species
Sym
Lot 1
Butter Clam
1.58
Pacific Oyster
0.94
Geoduck
1.57
Razor clam
1.01
Blue mussel
1.37
Atlantic Oyster 1.66
Manila clam
1.11
Sym= Test of symmetry
Precision & Recovery
SLV Report & Results
Sym
Lot 2
Var
Lot 1
Var
Lot 2
1.18
139.10 130.10
0.62
81.76 76.03
1.80
98.78 109.25
1.08
147.30 217.94
1.60
126.83 122.78
1.40
242.18 315.12
1.24
142.72 141.34
Var= Test of Variance
Ratio
t-test
used
p
1.07
1.08
1.11
1.48
1.03
1.30
1.01
Paired
Paired
Paired
Paired
Paired
Paired
Paired
0.32
0.57
0.59
0.07
0.10
0.69
0.52
Significant
difference
(p<0.05)
NOT significant
NOT significant
NOT significant
NOT significant
NOT significant
NOT significant
NOT significant
16
Four aliquots of tissue were taken for each sample. One aliquot was set aside as the sample
blank. One of the aliquots was spiked with a low (but determinate at 1:1000 dilution) amount of
DA equivalent to 5ppm in tissue. One aliquot was spiked with a medium amount of DA,
equivalent to 20ppm, and the last was spiked with a high amount of DA at 40ppm. Two
replicates of each of the spiked samples were run on the ELISA. The sample blank was only run
once.
Data summary
For this analysis the ISSC protocol calls for a nested Analysis of Variance (ANOVA) and a one
way ANOVA to determine the precision and recovery of the test. However, for an ELISA,
performing an ANOVA is not particularly relevant or useful for determining the quality of the
test. Instead, we calculated the coefficients of variation (CV) for each spike level (low, medium,
high) and determined an average CV to demonstrate the reproducibility of the assay. The CVs
ranged from ~8-20%, which is to be expected for an ELISA test. See appendix V for actual data
and calculations.
Species
Butter Clam
Pacific Oyster
Geoduck
Razor clam
Blue mussel
Atlantic Oyster
Manila clam
Averages
%CV 5ppm
25.10
20.59
14.20
15.38
15.84
13.17
15.35
14.94
%CV 20ppm
18.57
11.36
10.81
10.16
12.71
11.46
12.09
10.69
%CV 40ppm
15.52
7.12
10.06
8.82
8.88
11.43
9.44
9.61
Average CV
19.73
13.02
11.69
11.45
12.48
12.02
12.29
11.74
Specificity
For each species, one sample free of natural DA (as determined by HPLC) was chosen. Three 1g
aliquots of this sample were placed into 15mL tubes with 8.5mL 50% MeOH. For one tube, an
additional 0.5mL 50% MeOH was added (for a total of 9ml) and the tube was capped and set
aside as the sample blank. The remaining two tubes were spiked with DA at a level of 5 ppm. To
one of these tubes, an additional 0.5mL 50% MeOH was added and this tube was set aside as the
DA-only spike. The interfering compounds (IC) tested were glutamic acid and glutamine. These
compounds are similar in structure to domoic acid and thus may have to potential to interact with
the ELISA antibodies giving false results. Glutamic acid and glutamine solutions were made up
at 2000 ppm in 100% DI water by adding 26.7 mg glutamic acid in 13.4 mL DI H2O and 26.7
mg glutamine in 13.4 mL DI H2O. For the third aliquot of tissue, 250uL of either the glutamic
acid or glutamine 2000 ppm solution was added, and then 250uL 100% MeOH was added for a
final volume of 9 mL 50% MeOH with 5 ppm DA and 50 ppm GA or glutamine. For each
potentially interfering compound, five replicates were run for the DA-only and the DA + ICspiked samples. The sample blank was run only once.
SLV Report & Results
17
Attempts to obtain epi-domoic acid and isodomoic acid reference materials were unsuccessful
and the cross reactivity of the assay with these compounds could not be determined during this
study.
Data summary
To determine whether the interfering compounds (IC) had an impact on the ELISA results, a
specificity index was calculated using the equation SI= (ELISA value DA-only)/(ELISA value
DA+IC). The SI was calculated for each replicate and an average SI was calculated. A two sided
t-test at the 0.05 significance level was used to determine whether the average SI value differed
from 1. Two of the seven species tested showed a significant effect in the presence of glutamic
acid, and only one showed a significant difference with glutamine. However, the margin of this
significance was generally very small, and usually within the inherent variance of the test of
~15%. The concentrations of the interfering compounds were also unnaturally high. This shows
that most of the tests were not affected by extreme levels of a potentially interfering compound.
See appendix VI for actual data/calculations.
Species
Interfering compound
SI avg
[SIavg-1]
u
Butter Clam
Pacific Oyster
Geoduck
Razor clam
Blue mussel
Atlantic Oyster
Manila clam
Glutamic Acid
Glutamic Acid
Glutamic Acid
Glutamic Acid
Glutamic Acid
Glutamic Acid
Glutamic Acid
1.41
1.185
1.11
1.084
0.837
0.795
1.031
0.41
0.19
0.11
0.08
0.16
0.21
0.03
0.43
0.51
0.17
0.35
0.12
0.16
0.16
Species
Interfering compound
SI avg
[SIavg-1]
u
Butter Clam
Pacific Oyster
Geoduck
Razor clam
Blue mussel
Atlantic Oyster
Manila clam
Glutamine
Glutamine
Glutamine
Glutamine
Glutamine
Glutamine
Glutamine
1.341
0.950
0.871
0.904
0.853
0.931
1.134
0.28
0.05
0.13
0.10
0.15
0.06
0.13
0.24
0.14
0.34
0.19
0.19
0.24
0.37
Sig diff from 1 ([SIavg –
mo]>u)
NOT significant
NOT significant
NOT significant
NOT significant
Significant
Significant
NOT significant
Sig diff from 1 ([SIavg
– mo]>u)
Significant
NOT significant
NOT significant
NOT significant
NOT Significant
NOT significant
NOT significant
Comparing methods
To compare the ELISA method to the currently accepted HPLC methods, four species with
samples from two different geographic locations were tested by each method. For each species,
20 samples from Washington and 20 samples from Alaska (for a total of 40) were analyzed. The
samples were spiked with a range of DA, corresponding to 2, 5, 10, 20 or 40 ppm in tissue. Once
spiked and vortexed, the samples were processed and run simultaneously by ELISA and by
HPLC.
SLV Report & Results
18
Data summary
To analyze the data, the HPLC and ELISA results were tested for symmetry and variance. The
symmetry for HPLC and ELISA values of all species was <2, indicating that the data were not
skewed. The ratios of the larger to smaller of the variances for HPLC and ELISA values were
also all <2, showing that there was not a lack of homogeneity. A paired t-test was used to test
whether there was a significant difference between HPLC and ELISA results. None of the
species tested showed a significant difference between the two methods. See appendix V for
actual data and calculations. Fig. 2 shows the correlations between the HPLC and ELISA values
for each species. See appendix VII for actual data and calculations.
Species
Sym
Sym
Var
HPLC
ELISA HPLC
Butter Clam
1.099
0.690
179.44
Pacific Oyster
1.063
0.915
172.35
Razor clam
0.932
0.856
172.07
Blue mussel
1.052
0.881
182.68
Sym= Test of symmetry Var= Test of Variance
Var
ELISA
131.02
140.33
133.81
144.56
Ratio
1.370
1.228
1.286
1.264
pvalue
0.576
0.837
0.215
0.699
Sig diff between
methods? (p<0.05)
NOT significant
NOT significant
NOT significant
NOT significant
Figure 2. Correlation between HPLC and ELISA values for four species of shellfish
SLV Report & Results
19
Limit of Detection, Limit of Quantitation/Sensitivity
The Limit of Quantitation and Limit of Detection were determined by testing 10 samples of each
species at 5 different concentrations, with 2 replicates at each concentration. The concentrations
used were 2, 5, 10, 20, and 40 ppm.
Data summary
To determine the limit of quantitation, the coefficient of variation (CV) was calculated for each
of the replicates at each concentration. The CVs were plotted against the spike concentration and
a linear regression was calculated. The equation of the line was used to calculate the value of the
concentration that would give rise to a 10% CV. This value was used as the limit of quantitation.
To find the limit of detection, the LOD was divided by 3.3. In three cases, the LOQ as
determined by this method was a negative number. For these species, an alternative method was
used to calculate the LOQ and LOD (see Appendix VI for data and calculations for both
methods). The limit of detection ranged from 0.2 to 2.77 ppm. At a 1:1000 dilution, based on the
linear range of the assay a detection limit of 3ppm is to be expected. These data show that the
limit of detection of the assay is sufficient for determining levels of DA relevant to management.
See appendix VIII for actual data and calculations.
Species
Butter clam
Pacific Oyster
Geoduck
Razor clam
Blue mussel
Atlantic Oyster
Manila Clam
LOQ
LOD
7.41
2.25
2.87
0.94
4.12
1.84
3.55
1.56
0.67
0.20
8.43
2.56
9.15
2.77
Linear range
During internal validation studies at Mercury Science, the assay was found to have an effective
quantitative range from approximately 0.3 to 3.0 ppb using domoic acid standard solutions. (See
Litaker et al, 2008).
SLV Report & Results
20
Comparability
As shown in the pre-proposal, the graph below shows the results of a year-long study done by the
Quinault Indian Nation (QIN) and the Washington Department of Health (WDOH) comparing
razor clam analysis performed by the Domoic Acid Test Kit versus HPLC analysis. One
hundred fifty six samples were compared. This independent study was planned and performed
without any input from Mercury Science or NOAA.
SLV Report & Results
21
ELISA Test Kit User Guide
Doc. 2009B
Domoic Acid
Screening
Test Kit
Colorimetric Immunoassay
for the detection of
Domoic Acid
in environmental samples
Instructions and User Guide
FOR SCIENTIFIC RESEARCH USE
Manufactured by
Mercury Science Inc.
www.mercuryscience.com
Tel: (866) 861-5836
DAK-36 User Guide and Supplemental Guide
22
Domoic Acid Screening Test Kit
For Scientific Research Use Only.
This product is not to be used for In Vitro or In Vivo
Diagnosis.
PRINCIPLES OF THE ASSAY
This product contains an antibody (Ab) that binds Domoic Acid and has been
developed for the semi-quantitative detection of Domoic Acid in sample
extracts. The signal of samples and a control are compared to determine the
amount of Domoic Acid present.
The Domoic Acid assay is a solid phase colorimetric immunoassay, based on
competition between Domoic Acid and enzyme-labeled Domoic Acid (DATracer) for anti-Domoic Acid antibody. Samples containing Domoic Acid
inhibit the binding of the DA-Tracer to the antibody molecules. Both the AbDomoic Acid and Ab-DA-Tracer complexes are captured on the surface of the
microtiter plate wells.
Following a wash step, the addition of an enzyme substrate (TMB) forms a
color proportional to the amount of DA-Tracer in the well. The amount of color
measured is inversely proportional to the concentration of Domoic Acid in the
sample.
Domoic Acid
Antibody
Y
Domoic Acid
Y
DA
Y
DA-Tracer
DA
DATracer
DA
DA
Solid phase
anti-mouse IgG
TMB
Y
Y
DA
DA-Tracer
TMB
Y
Y
DA
DA
DA
DATracer
DA
DA
TEST KIT CONTENTS Each Domoic Acid test kit contains reagents
for testing a maximum of 36 samples in duplicate.
The expiry date of the test kit is stated on the outer label.
Store the kit between 2oC and 8C.
DAK-36 User Guide and Supplemental Guide
Blue
23
Reagents
Store the reagents between 2oC and 8oC when not in use.
Component
Quantity
-------------------------------------------------------------------------------------------------------Control Solution
1 vial
2 mL
The control is a phosphate-buffered salt solution with casein.
Contains sodium azide as a preservative.
-------------------------------------------------------------------------------------------------------Sample Dilution Buffer
1 bottle 50 mL
Ready-to-use phosphate buffered (pH 7.8) salt solution with casein.
Contains sodium azide as a preservative.
-------------------------------------------------------------------------------------------------------Domoic Acid- Tracer
1 vial
7.5 mL
The tracer is in a MOPS-buffered solution containing bovine protein as a
stabilizer and methylisothiazolone, bromonitrodioxane, and Proclin 300 as
preservatives.
-------------------------------------------------------------------------------------------------------Domoic Acid Antibody
1 vial 7.5 mL
The antibody is in phosphate-buffered salt solution with casein.
Contains sodium azide as a preservative.
-------------------------------------------------------------------------------------------------------Wash Concentrate*
1 bottle 40 mL
A 25-fold concentration of Tris-HCI buffered (pH 7.8) salt solution with Tween
20. Contains sodium azide as a preservative.
*Prepare for use by mixing entire contents
with 960 mL of distilled water and placing in platewasher WASH Bottle.
-------------------------------------------------------------------------------------------------------Substrate Solution
1 bottle 15 mL
Tetramethylbenzidine (TMB) and H2O2
Keep away from direct sunlight.
-------------------------------------------------------------------------------------------------------Stop Solution
1 bottle 15 mL
1 N Hydrochloric Acid
-------------------------------------------------------------------------------------------------------Anti-Mouse IgG Microtitration Strips
1 plate (12 x 8 wells)
----------------------------------------------------------------------------------WARNINGS AND PRECAUTIONS
For research use only. Handle all samples as potentially hazardous.
Disposal of all waste should be in accordance with local regulations.
DAK-36 User Guide and Supplemental Guide
24
SCREENING ASSAY PROCEDURE
Perform each determination in duplicate for the Control and unknowns. All
sample extracts should be filtered prior to analysis. All reagents and samples
should be brought to room temperature prior to use. Use only the number of strips
needed. Keep unused strips stored in their aluminum foil pouch with the included
desiccant until needed.
1. Pipet 50 uL of the Domoic Acid Antibody solution into each well.
2. Pipet 50 uL of each Control solution or sample into a well using the sequence
shown in the table below. Always use wells A and B on each strip as Controls.
Always perform duplicate analyses of samples. Three samples can be tested per strip.
The example below shows the testing of eight samples.
1
2
3
4
5
6
7
8
9
10
11
12
A
Con- Con- Control trol
trol
B
Con- Con- Control trol
trol
C
1st
4th
Unk Unk
7th
Unk
D
1st
4th
Unk Unk
7th
Unk
E
2nd 5th
Unk Unk
8th
Unk
F
2nd 5th
Unk Unk
8th
Unk
G
3rd
6th
Unk Unk
H
3rd
6th
Unk Unk
3. Shake the wells for 30 minutes.
4. Pipet 50 uL of the Domoic Acid Tracer solution into each well.
5. Shake the wells for 30 minutes.
6. Wash the strips 3 times on the platewasher. Tap the strips upside-down firmly on a
paper towel to blot away any excess wash solution that may remain in the wells.
7. Add 100 uL of Substrate Solution to each well. Shake the plate for five minutes.
8. Add 100 uL of Stop Solution to each well. Shake the plate briefly.
9. Measure the absorbance in each well. Note: If Control absorbance is greater
than 3.0 AU, remove 50 uL from ALL WELLS and measure absorbance.
10. The data can be analyzed using the Excel worksheet available at the following
link: http://mercuryscience.com/Domoic Acid Quantitation 8Well Strip.xls
DAK-36 User Guide and Supplemental Guide
25
Additional Information
MATERIALS REQUIRED BUT NOT SUPPLIED WITH THE KIT
The Domoic Acid test kit is part of a complete system of immunodiagnostic
reagents and instrumentation. The system requires the following equipment.
1. Microtiterplate Reader able to measure Absorbance at 450 nm
2. Platewasher
3. Plate Shaker
4. 8 Channel pipet
5. Pipetmen (P10, P200 and P1000)
Other Notes:

Perform each Control and Sample in duplicate wells.
All sample extracts should be filtered prior to analysis.

All reagents and samples should be brought to room temperature
prior to use.

Use only the number of strips needed.

Keep unused strips stored in their aluminum foil pouch with the
included desiccant until needed.

If Control absorbance is greater than 3.0 AU, remove 50 uL from
ALL WELLS and repeat absorbance measurement.
An Excel worksheet has been developed to analyze results and quantitate the
amount of domoic acid in extracts.
Send your request for the “Domoic Acid Quantitation Worksheet - DAK-36” to:
info@mercuryscience.com
or the worksheet can be downloaded at:
www.mercuryscience.com/Domoic Acid Quantitation 8Well Strip.xls
Structure of Domoic Acid
COOH
COOH
COOH
N
H
DAK-36 User Guide and Supplemental Guide
26
PERFORMANCE CHARACTERISTICS
Reproducibility
Inter-Assay Standard Curve
The average values and standard deviation of 5 separate standard curves is
shown below.
100
90
Relative Signal
80
70
standards data
60
best fit curve
50
40
30
20
10
0
0.1
1
10
100
ppb DA in Standard
Intra-assay Signal Precision
Analysis of 12 replicates for five different samples
A
B
Signal (% of Control)
99.5
76.5
Standard Deviation
1.4
1.2
% Coeff. Var.
1.4
1.6
C
47.5
2.0
4.2
D
23.5
2.3
9.8
Intra-assay Concentration Precision
Analysis of 3 different samples measured in 6 separate quantitative assays.
A
B
C
Average Conc. (ppb)
0.56
1.54
3.66
Standard Deviation (ppb)
0.01
0.13
0.19
% Coeff. Var.
2.1
8.6
5.3
DAK-36 User Guide and Supplemental Guide
E
10.4
1.1
10.9
27
PERFORMANCE CHARACTERISTICS (Cont.)
Detection Limit
The detection limit is defined as the minimum concentration of Domoic Acid
that can be distinguished from a blank standard with 95% confidence. A
detection limit of 0.1 ppb Domoic Acid in extraction buffer has been
demonstrated with this assay. The quantitative range of the assay is from 0.3
ppb to 6 ppb Domoic Acid.
Cross Reactivity
This assay is specific for the detection of domoic acid. The ability of the assay
to detect structurally related compounds is shown in the following table.
Analyte
Domoic Acid
Kainic Acid
Glutamic Acid
Glutamine
% Reactivity
100
0.3
less than 0.1
less than 0.1
PROCEDURAL NOTES
Please read all instructions thoroughly before using this kit. Do not mix
reagents from kits having different lot numbers. Do not use kits after the
expiration date printed on the kit label.
Reagents should be at room temperature when used.
During washing steps, check that each well is completely filled during wash
solution additions. After washing is complete, invert the wells and tap them
firmly against a paper towel to remove excess liquid.
The platewasher should be rinsed with distilled water at the end of each day of
use to prevent clogging of the dispensing and aspirating ports. Prime the
platewasher with wash solution before the first wash each day.
Care must be taken during each step to prevent contamination of reagents and
equipment. Do not use the same pipet tip in two different reagents.
For Technical Assistance, contact Mercury Science Inc: (866) 861-5836.
DAK-36 User Guide and Supplemental Guide
28
Domoic Acid Test Kit
Summary Protocol Sheet
Add
Antibody
50 uL
Add Control
and Samples
50 uL
Incubate
Shake for 30 minutes
Add
Tracer
50 uL
Incubate
Shake for 30 minutes
Wash
“3 WASHES”
program
Add
Substrate
100 uL,
Shake for 5 minutes
Add
Stop
100uL
Measure
Absorbance at 450 nm
Note: If Control absorbance is greater than 3.0 AU, remove 50 uL from ALL
WELLS and repeat absorbance measurement.
DAK-36 User Guide and Supplemental Guide
29
DAK-36
Domoic Acid Test Kit
Supplemental Guide
Version 2010A
This guide contains advice for obtaining optimal assay performance and answers
frequently asked questions. Please contact Mercury Science for technical support and
additional help.
KIT EXPIRATION DATE
The kit expiration date applies to kits that are refrigerated (40o C) when not in use. After
opening, any unused strips of goat anti-mouse wells must be stored sealed with
desiccant and should be used within the kit expiration date.
NOTES ON EQUIPMENT
1. Plate Shaking: The incubation steps should be performed using a shaker designed
specifically for 96-well microtiter (ELISA) plates.
The rotary motion of these devices mixes the samples so that optimal equilibrium
conditions are reached during the shake times specified in the assay. The rotational
speed varies between devices depending on their orbital motion, but is often in the
range of 800 to 1000 rpm.
2. Plate Washing: The washing step should be performed using a plate washer
designed specifically for 96-well microtiter (ELISA) plates.
If wells must be washed manually, consistency of washes within each strip of wells is
important.
3. Pipettes: The volatility of 50% methanol solutions requires the use of positive
placement pipettes for precision in pipetting aliquots of shellfish extracts. Mercury
Science recommends using a Gilson Microman (or similar product) when diluting
shellfish extracts or spiked control. Air displacement pipetters are suitable for use
with the Antibody, Tracer, Control and Sample Dilution Buffer provided with the kit.
4. Pipette Calibration: Check the accuracy and precision of your pipettes.
Assay results are dependent on consistent sample dilution and addition of reagents
to all wells. Continually monitor pipette operation.
DAK-36 User Guide and Supplemental Guide
30
NOTES ON OPTIMIZING THE CONTROL SIGNAL
The optimal Control values should be between 1.8 and 3.0 absorbance units (AU)
What to do if the Control signal is low (<1.8 AU)
1. Change shake speed:
The optimal shake speed is critical for obtaining recommended Control values in the
assay. Too slow and the binding kinetics do not reach equilibrium. Too fast and
evaporation and aerosol formation can lead to erratic results. A good starting point
is between 800 – 1000 rpm on standard ELISA shakers. New kit users may want to
run a single strip of wells at their regular shake speed to ensure good performance. If
the Control signal is 2.5 AU or higher the mixing is ideal. If the signal is low, try
increasing the speed with another strip and compare the results. It usually only takes
a few tries to get the shaking step optimized.
2. Check your water source:
Water sources can affect the assay. Some reverse osmosis water systems have
been found to produce water that inhibits assay signal. The actual cause is unknown.
In this event, switching to an alternative source of commercially available water has
been successful. Bottled distilled water found at the local grocery store is an
inexpensive alternative that usually works well with the assay. This water can be
used for both the preparation of the 1 x Wash Solution and preparation of shellfish
extraction solvent.
3. Change the Substrate development time:
The standard color development time is 5 minutes. However, depending on the
temperature of the room and other factors this can vary slightly. Optimal
development can sometimes occur in only 3 minutes or in up to 7 minutes. If the
signal is consistently low after stopping development at 5 minutes, try letting the
assay develop an additional 30 seconds to 1 min longer before adding the Stop
solution. If it is still too low, increase the time by an additional minute. The assay can
develop for up to 7 minutes without interfering with the results. After running a few
assays you will also be able to tell visually how the development is progressing.
4. Check or change your reagent trays and plastics:
Another cause of low signal as well as higher variability among replicate samples is
the leaching of inhibitory compounds from certain types of plastic ware such as
reagent trays. All reagent trays should be rinsed with deionized water and completely
dried before each use.
DAK-36 User Guide and Supplemental Guide
31
5. Check the control solution
Contamination of the reagents can cause low signal. Care should be taken to
prevent contamination of the Control solution as well as other reagents. If a problem
with the Control solution is suspected, the Sample Dilution Buffer may be substituted
for the Control solution.
If you are not running the entire plate at once, the control solution should always be
transferred into a separate container (such as a microfuge tube) so that the glass vial
is not contaminated during pipetting. If the glass vial containing the control solution
does become contaminated or if you run out of control solution, the sample dilution
buffer can be used for the control wells.
What to do if the Control signal is too high (>3.0 AU)
1. Remove 50 l from each well and read again
If the Control values are above 3.0 AU, accuracy can be affected. Test strips with
high Control values should have 50 l removed from each well in that strip and read
again. If the Bo values are below 3.0 then you can successfully read the assay. If still
too high then remove another 50 l and read again. The assay is most accurate
when Control values are in the 1.8 to 3.0 AU range. Avoid removing too much fluid
initially and dropping the Bo values below 1.8. This can lead to an underestimate of
the actual Bo values.
2. Change the development time
As mentioned before, although the recommended development time is 5 minutes,
under certain conditions the assay may take as little as 3 or as many as 7 minutes to
fully develop. If the readings are consistently over 3.0, you can try decreasing the
development time. While the assay is developing, the color can be visually inspected
for how quickly it is developing. If the wells are a dark blue after only 3 or 4 minutes,
the Stop solution can be added to stop the color development.
NOTES ON USING THE EXCEL WORKSHEET
The DAK-36 test kit uses a monoclonal antibody for detection of domoic acid. This
antibody provides a consistent assay response when used with the standard protocol
described in the user guide. A standard curve is not required to obtain quantitative
results. Instead, unknown samples are compared to a Control (blank) response.
Control and samples are run in duplicate to aide in quality control. It should be noted
that Control signal may vary between strips.
DAK-36 User Guide and Supplemental Guide
32
As stated in the User Guide, it is important that Controls be run on every strip of
wells.
Mercury Science provides an Excel sheet to assist in converting assay results to
concentrations. This spreadsheet is available for download at:
http://www.mercuryscience.com/Domoic%20Acid%20Quantitation%208Well%20Strip.xls
1. Check your dilution factors and extraction values:
Since different extraction methods can be used to prepare samples, always check to
make sure that the worksheet reflects the method you used. This information should
be in columns F (Dilution Factor), J (Shellfish Weight, grams), and K (Extraction
Volume, mL). These values are critical for being able to calculate the amount of DA in
the original tissue.
Some of my data are red, what does this mean?
Samples are always tested in duplicate wells to provide an internal quality control.
Duplicate signals should be within 10% of their average value. If the values of the two
wells have a difference of greater than 10% of their average, the values will be ‘redflagged’ on the Excel sheet. This means that the difference is too large to provide an
accurate estimate of the DA concentration and the sample should be re-tested. If the
control wells are red-flagged, the entire strip should be re-tested.
For example: Two duplicate wells having AU values of 1.500 and 1.600 are
acceptable because the difference between the values and their average (1.550) is
less than 10%. If two duplicate wells have AU values of 1.000 and 1.400, this result
is invalid and should be retested because the variation between the values is too
great: (1.200 -1.000)/1.000 = 20%.
At what level do you get LOW and HIGH as opposed to a numerical estimate?
The worksheet will give values of “LOW” or “HIGH” instead of a numerical value if the
B/B0 is out of the linear range of the assay. For “LOW” values, this is at a B/B0 of
about 85%. For “HIGH” values this is at about 15%. At a 1:1000 dilution, this means
that concentrations of <2.7ppm will be reported as “LOW” and concentrations >56
ppm will be reported as “HIGH”. If you get a LOW or HIGH result that you want an
actual value for, you can try testing the sample again using a different dilution factor.
For example, you can try 1:200 or 1:500 for samples with low levels of DA, or 1:2000
for samples with very high levels of DA.
ASSAY SPECIFICITY
DAK-36 User Guide and Supplemental Guide
33
Cross Reactivity with epi-Domoic Acid and isoDomoic Acids:
Naturally occurring samples of domoic acid may also contain small quantities of epidomoic acid and various isodomoic acids. Analysis of samples containing epi-DA
and iso-DA suggests the cross reactivity is less than 6%. (Litaker, et al. 2008
available online at: http://mercuryscience.com/LitakerStewartDec2008.pdf)
In correlation studies of DAK-36 analysis and HPLC analysis of contaminated
shellfish, no appreciable difference in the quantitation of domoic acid has been seen.
DAK-36 User Guide and Supplemental Guide
34
SHELLFISH TISSUE EXTRACTION
SUGGESTED METHOD
While several different methods can be used, the following method provides an
example of how a shellfish extraction can be performed to prepare a sample for use
on the ELISA.
1. Place 10-12 shucked and rinsed organisms into a commercial blender and blend until
a uniform homogenate is formed.
2. Weigh out 1.0g of homogenate into a 15 mL plastic tube
3. Add 9 mL of 50% MeOH (Methanol:Water 50:50, v:v)
4. Cap and vortex the tube on high speed using a vortex mixer for at least 1 minute
5. Transfer approximately 1 mL of the vortexed solution to a microfuge tube
6. Spin the tube in a microfuge at 14,000 rpm for 5 minutes
7. The resulting clear supernatant is diluted into Sample Dilution Buffer immediately
prior to analysis.
8. Extracts should be run on the ELISA the same day
Alternative methods
1. If a vortex mixer is unavailable (step 4 above), samples can be hand shaken for 2
minutes.
2. Instead of spinning in a microfuge (steps 5-6 above), the 15mL tube can be spun in a
large table top centrifuge at 3,000 rpm for 20 minutes. Depending on the centrifuge
available, spin speeds and times can be adjusted as long as a general rule is
followed of using shorter times for higher speeds and longer spin times for slower
speeds.
3. Instead of spinning samples in a microfuge or centrifuge (step 5-6 above), the sample
may be filtered using a 0.2 to 0.45um GF/F syringe tip filter. The sample should be
poured into the syringe barrel and filtered into a clean 15 mL tube. If the filter tip
clogs, the tip can be replaced and the rest of the sample pushed through. For
shellfish species with very fatty tissues, such as oysters, a couple of mL of glass fiber
can be placed into the syringe prior to filtering to help prevent clogging.
4. Instead of using 1g of tissue in 9mL 50% MeOH, 2g of tissue can be used in 18 mL
50% MeOH.
SUGGESTED DILUTION SERIES
Before being run on the ELISA, a shellfish extract must be diluted so that the
concentration of DA is within the working range of the assay (0.3 to 3ppb). A 1:1000
dilution ensures that the assay will detect levels relevant to the regulatory limit of 20
ppm. Note: A 1:1000 dilution is used assuming the initial extraction was a 1:10
dilution of tissue into solvent (1g tissue in 9mL 50% MeOH). If a different extraction is
used then a different dilution series may have to be calculated.
DAK-36 User Guide and Supplemental Guide
35
To perform a 1:1000 dilution, we have found a two step dilution series to work best.
Dilution 1: Using a positive displacement pipette, add 13.1 L of extract to 400 L
Sample Dilution Duffer. Vortex thoroughly.
Dilution 2: Using either a positive displacement or air pipette, add 13.1 L of Dilution
1 to 400 l Sample Dilution Buffer. Vortex thoroughly. This dilution is ready to be
added to the assay wells.
DAK-36 User Guide and Supplemental Guide
36
Appendix I. DAK-36 Kit Stability
Six Month Product Shelf life:
The DAK-36 Domoic Acid Test Kit has a stated shelf life of six months from the date of kit
assembly when stored at 40o F.
During product development, stability studies were begun on each DAK-36 kit component.
Most of the components of the kit have retained greater than 90% of their original activity after
more than four years at 40o F. The limiting reagent of the kit shelf life is the Substrate solution
which causes assay signal to gradually decline.
A full kit stability study is currently ongoing. This study uses complete kits that have been
manufactured and kept for archival purposes. During this validation study, spiked samples were
occasionally run with old DAK-36 kits. The oldest kit tested had an age of 48 months. A graph
showing a compilation of some of the data obtained from kits of varying age is shown below. In
general, the Control signal produced by the kits had decreased approximately 50% after 3 years.
The response of the assay to domoic acid, measured as the ratio of the sample to the Control
signal (B/Bo), has not been altered even after four years.
In light of the current available evidence and in the interest of ensuring optimal kit performance,
Mercury Science Inc suggests that the DAK-36 kit continue to be shipped with a six month shelf
life.
Appendix I. DAK-36 Kit Stability
37
Appendix II. Mercury Science Lab Validation Results
A concurrent SLV was performed at Mercury Science Inc. in Raleigh, NC testing 8 species of
shellfish on four of the validation criteria. Six species of shellfish were received from the
University of Alaska-Fairbanks School of Fisheries and Ocean Science, including butter clams
(Saxidomus giganteus), blue mussels (Mytilus edulis), cockles (Clinocardium nuttalli), California
mussels (Mytilus californianus), littleneck clams (Protothaca staminea), and Pacific oysters
(Crassostrea gigas). The two additional species tested, geoducks (Panopea generosa) and razor
clams (Siliqua patula), were obtained from the Washington samples tested by NOAA.
Accuracy/Trueness & Measurement Uncertainty
Accuracy and trueness and measurement uncertainty were calculated by spiking 20 samples for
each species over a range of DA concentrations ranging from 2 to 40 ppm. Percent
accuracy/trueness ranged from 73 to 106%. Measurement uncertainty calculations indicated that
for most species the uncertainty was within ± 2ppm.
Species
Blue mussel
Butter clam
California mussel
Cockles
Geoduck
Littleneck clam
Pacific oyster
Razor Clam
% accuracy/ Measurement uncertainty
trueness
101.54
-1.12 to 0.59
85.39
1.11 to 3.93
93.91
-0.09 to 2.19
105.71
-2.09 to 0.12
76.50
2.80 to 6.93
100.78
-0.75 to 0.48
95.48
-0.45 to 2.01
72.66
3.16 to 8.16
Ruggedness
To test ruggedness, 10 split samples for each species were run on two different lots of test kits.
Four of the eight species showed a significant difference between lots of tests.
Species
Sym
Lot 1
Sym
Lot 2
Var
Lot 1
Var
Lot 2
Ratio
t-test
used
p
Blue mussel
Butter clam
California mussel
Cockles
Geoduck
Littleneck clam
Pacific oyster
Razor clam
1.35
0.97
1.45
1.34
1.43
1.58
1.82
1.24
1.28
0.98
1.70
1.44
2.09
1.30
1.71
1.46
60.0
69.7
115.5
126.4
81.8
67.6
93.7
88.6
59.3
82.5
117.3
120.2
212.6
82.1
137.1
178.0
1.01
1.18
1.02
1.05
2.60
1.21
1.46
2.01
Paired
Paired
Paired
Paired
Welch’s
Paired
Paired
Welch’s
0.72
0.0001
0.06
0.01
0.51
0.001
0.04
0.44
Appendix II. Mercury Science Data
Significant
difference
(p<0.05)
Not significant
Significant
Not significant
Significant
Not significant
Significant
Significant
Not significant
38
Precision & Recovery
To test precision and recovery, 10 samples were tested for each species with three spike levels
(low, medium, and high) for each sample and two reps at each spike level. Instead of using an
ANOVA to calculate the precision and recovery, the coefficients of variation at each spike level
were calculated to test reproducibility, since that is more relevant to an ELISA test. The CVs
ranged from 5 to 22%, which is typical for an ELISA, with an average CV of 10.5% CV which is
very good for an ELISA. The CVs decreased slightly in the higher concentrations, but overall all
of the CVs were fairly consistent.
Species
Blue mussel
Butter Clam
California mussel
Cockles
Geoduck
Littleneck clam
Pacific oyster
Razor clam
Averages
%CV 5ppm
11.54
19.36
7.22
12.44
12.04
10.96
21.65
9.21
13.05
%CV 20ppm
5.04
12.73
8.58
12.21
6.98
6.16
12.82
8.61
9.14
%CV 40ppm
6.46
8.41
11.92
10.06
7.71
8.91
13.73
5.63
9.10
Average CV
7.68
13.50
9.24
11.57
8.91
8.68
16.07
7.82
10.43
Specificity
To test the specificity of the assay, two potentially interfering compounds (IC), glutamic acid and
glutamine, were tested. One sample from each species was split into two aliquots where one was
spiked with only 5ppm DA and the other was spiked with both 5ppm DA and 50ppm of either
glutamic acid or glutamine. The results show that none of the species showed a significant
difference between the DA only spike and the DA+IC spike when the IC was glutamic acid.
When glutamine was tested as the IC, four of the eight species showed a significant difference.
This difference may be due to the fact that the samples were spiked with unnaturally high levels
of glutamine. In nature, any IC would be found in much lower concentrations and would most
likely not affect the effectiveness of the test. More tests may be done to show that lower levels
do not contribute to a significant difference.
Species
Blue mussel
Butter Clam
California mussel
Cockles
Geoduck
Interfering
compound
Glutamic Acid
Glutamic Acid
Glutamic Acid
Glutamic Acid
Glutamic Acid
Appendix II. Mercury Science Data
SI avg
[SIavg-1]
mo
1.049
0.983
1.056
1.024
1.075
0.049
0.017
0.056
0.024
0.075
0.14
0.11
0.12
0.12
0.13
Sig diff from 1
([SIavg – mo]>u)
Not significant
Not significant
Not significant
Not significant
Not significant
39
Littleneck clam
Pacific oyster
Razor clam
Glutamic Acid
Glutamic Acid
Glutamic Acid
1.212
0.969
1.028
0.212
0.031
0.028
0.24 Not significant
0.32 Not significant
0.13 Not significant
Species
Interfering
compound
Glutamine
Glutamine
Glutamine
Glutamine
Glutamine
Glutamine
Glutamine
Glutamine
SI avg
[SIavg-1]
mo
1.104
1.004
1.143
0.999
1.166
1.046
1.028
0.907
0.104
0.004
0.143
0.001
0.166
0.046
0.028
0.093
0.05
0.13
0.13
0.22
0.15
0.08
0.38
0.91
Blue mussel
Butter Clam
California mussel
Cockles
Geoduck
Littleneck clam
Pacific oyster
Razor clam
Appendix II. Mercury Science Data
Sig diff from 1
([SIavg – mo]>u)
Significant
Not significant
Significant
Not significant
Significant
Not significant
Not significant
Significant
40
Appendix III. Accuracy and Trueness Data and Calculations
VII. #1 Marine Biotoxin and Non-MPN Based Microbiological Methods SOP – Accuracy/Trueness
& Measurement Uncertainty
VALIDATION CRITERIA
Accuracy/Trueness is the closeness of agreement between test results and the accepted reference value.
To determine method accuracy/trueness, the concentration of the targeted analyte/measurand/organism of
interest as measured by the analytical method under study is compared to a reference concentration.
Measurement uncertainty is a single parameter (usually a standard deviation or confidence interval)
expressing the possible range of values around the measured result within which the true value is
expected to be with a stated degree of probability. It takes into account all recognized effects operating on
the result including: overall precision of the complete method, the method and laboratory bias and matrix
effects.
Procedure: This procedure is applicable for use with either growing waters or shellfish tissues. Make
every effort to use samples free of the target analyte/measurand/organism of interest. For each shellfish
type of interest use a minimum of 10-12 animals per sample. For each sample take two (2) aliquots of
either the homogenate or growing water sample appropriately sized for your work and spike one(1) of the
two (2) aliquots with a suitable known concentration of the target analyte/measurand/organism of interest.
Do not spike the second aliquot. This is the sample blank. For microbiological methods determine the
concentration of the target organism of interest used to spike each sample by plating on/in appropriate
agar. Process both aliquots of sample as usual to determine the method concentration for the target
analyte/measurand/organism of interest. For growing waters do twenty (20) samples collected from a
variety of growing areas. For shellfish do twenty (20) samples for each shellfish tissue type of interest
collected from a variety of growing areas, the same growing area harvested on different days or from
different process lots. Use a variety of concentrations spanning the range of concentrations of
importance in the application of the method to spike sample homogenates or growing water
samples. Both the low and high level spike concentrations must yield determinate values when analyzed
by the method under study.
Data:
Working Range ___________
Sample Type _____________
Agar used to determine spike concentration _____________________
Organism used for spiking _________________________________
Sample Spike conc/plate count Sample blank conc Spiked sample conc from analysis
1
2
3
4
5
6
7
8
9
10
Sample Spike conc/plate count Sample blank conc Spiked sample conc from analysis
11
12
13
14
Appendix III. Accuracy & Trueness Data & Calculations
A. Butter clams
41
15
16
17
18
19
20
For shellfish samples, repeat for each tissue type of interest.
DATA HANDLING
Accuracy/Trueness
The accuracy/trueness of a method consists of two distinct components, the portion due to the method
itself regardless of the laboratory performing it and the portion contributed by the laboratory’s
performance. In a single laboratory method validation, it is impossible to distinguish the contribution of
each to the overall accuracy/trueness of the method. Consequently, what is being estimated is the
accuracy/trueness of the method as implemented by the laboratory performing the analysis. Good
accuracy/trueness suggests the appropriateness of the method and the laboratory’s performance of it for
the intended work. Poor accuracy/trueness on the other hand indicates the potential unsuitability of the
method and/or the laboratory’s performance of it for the intended work.
Accuracy /trueness will be determined by calculating the closeness of agreement between the test results
and either a known reference value or a reference value obtained by plate count for microbiological
methods.
To determine the accuracy/trueness of the method as implemented by the laboratory over the range in
concentrations important to the intended application of the method, the data is worked-up in the following
manner.
1. Convert plate counts to logs.
2. If necessary use the sample blank (converted to logs for microbiological methods) to correct the
results from the spiked samples for matrix effects.
3. Calculate the average reference concentration of the analyte/measurand used to spike the
samples; or, for microbiological methods calculate the average plate count of the data in logs.
The average plate count represents the average reference concentration for the microbiological
method.
4. Calculate the average concentration of the analyte/measurand/organism of interest in the spiked
samples. For microbiological methods log transformed data is used for this calculation.
5. Divide the average concentration calculated from the spiked samples by the average reference
concentration.
6. Multiply the quotient by 100. This provides an estimate in percent of the accuracy/trueness of
the method as implemented by the laboratory over the range in concentrations of importance to
the intended application of the method.
Measurement uncertainty
Measurement uncertainty can be determined by subtracting the results for each spiked sample from the
reference value for the sample and calculating the 95% confidence interval of these differences. The
confidence interval of these differences represents the range in values within which the true measurement
uncertainty lies. A narrow range in values indicates that the method as implemented by the laboratory
produces reliable results.
Use the log transformed data for both the plate count and the microbial results obtained from the spiked
samples. If necessary use the sample blank (converted to logs for microbiological methods) to correct the
spiked sample for matrix effects and calculate the two-sided, 95% confidence interval for the difference in
concentrations between the reference and the spiked samples. This range in counts represents the
measurement uncertainty of the method as implemented by the laboratory.
Data Summary:
Appendix III. Accuracy & Trueness Data & Calculations
A. Butter clams
42
Calculated % accuracy/trueness ____________
Calculated measurement uncertainty ________________
A. Butter clams
Data
Working Range 2 to 40ppm
Sample Type Shellfish tissue
Agar used to determine spike concentration Not applicable
Organism used for spiking Butter clams (Saxidomus giganteus)
Count
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Averages
St dev
Sample # Date
Kit
Spike
Blank
ELISA
1543 3/10/2010
100304R
5.0
0.0
1418 3/17/2010
100304R
20.0
0.0
1932 3/17/2010
100304R
5.0
0.0
1508
3/9/2010
100304R
40.0
0.0
1660 3/18/2010
100304R
40.0
0.0
1832 3/18/2010
100304R
5.0
0.0
1371 3/19/2010
100304R
20.0
0.0
1521 3/19/2010
100304R
20.0
0.0
1949 3/19/2010
100304R
40.0
0.0
1364 3/22/2010
100304R
5.0
0.0
1694 3/23/2010
100304R
5.0
0.0
1645 3/23/2010
100304R
10.0
0.0
1419 3/23/2010
100304R
20.0
0.0
1390 3/24/2010
100302
5.0
0.0
1776 3/24/2010
100302
10.0
0.0
1496 3/24/2010
100302
20.0
0.0
1634 3/22/2010
100304R
5.0
0.0
1348 3/22/2010
100304R
20.0
0.0
1563 3/25/2010
100302
10.0
0.0
1736 3/25/2010
100302
40.0
0.0
17.25
18.74
3.5
24.9
5.8
42.1
45.5
5.1
24.1
22.2
35.2
6.3
3.9
12.9
22.5
5.9
13.0
26.2
5.5
19.8
11.3
39.0
Difference
1.5
-4.9
-0.8
-2.1
-5.5
-0.1
-4.1
-2.2
4.8
-1.3
1.1
-2.9
-2.5
-0.9
-3.0
-6.2
-0.5
0.2
-1.3
1.0
-1.49
2.62
Accuracy/Trueness calculations
1. Convert plate counts to logs: N/A
2. If necessary use the sample blank (converted to logs for microbiological methods) to correct the
results from the spiked samples for matrix effects: N/A
3. Calculate the average reference concentration of the analyte/measurand used to spike the samples;
or, for microbiological methods calculate the average plate count of the data in logs. The average
plate count represents the average reference concentration for the microbiological method: 17.25
4. Calculate the average concentration of the analyte/measurand/organism of interest in the spiked
samples. For microbiological methods log transformed data is used for this calculation: 18.74
5. Divide the average concentration calculated from the spiked samples by the average reference
concentration: (18.74)/(17.25)=1.086
Appendix III. Accuracy & Trueness Data & Calculations
A. Butter clams
43
6. Multiply the quotient by 100. This provides an estimate in percent of the accuracy/trueness of the
method as implemented by the laboratory over the range in concentrations of importance to the
intended application of the method. 1.086*100= 108.6%
Measurement uncertainty calculation
1. Calculate the two-sided, 95% confidence interval for the difference in concentrations between the
reference and the spiked samples: 95% CI calculated using the formula
=1.96*(2.62/√20)= 1.15
x=-1.49
-1.49 - 1.15= -2.63
-1.49 + 1.15= -0.34
Data Summary:
Calculated % accuracy/trueness 108.6%
Calculated measurement uncertainty -2.63 - -0.34
Appendix III. Accuracy & Trueness Data & Calculations
A. Butter clams
44
B. Pacific Oyster
Data
Working Range 2 to 40 ppm
Sample Type Shellfish tissue
Organism used for spiking Pacific oyster (Crassostrea gigas)
Count
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Averages
St dev
Sample #
2115
2116
2117
2118
2119
1816
1817
2165
2175
2176
2207
2208
2209
2211
2330
2275
2181
2329
1566
2110
Date
3/26/2010
3/30/2010
3/30/2010
3/31/2010
3/31/2010
3/31/2010
3/31/2010
4/1/2010
4/1/2010
4/1/2010
4/5/2010
4/5/2010
4/6/2010
4/6/2010
4/6/2010
4/6/2010
4/6/2010
4/6/2010
4/6/2010
4/6/2010
Kit
100302-2
100302-2
100302-2
100302-3
100302-3
100302-3
100302-3
100302-4
100302-4
100302-4
100302-5
100302-5
100302-6
100302-5
100302-5
100302-5
100302-5
100302-5
100302-6
100302-6
Spike
40.0
5.0
20.0
40.0
20.0
5.0
5.0
20.0
40.0
20.0
5.0
10.0
20.0
5.0
10.0
20.0
5.0
10.0
20.0
40.0
18.0
Blank
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.6
0.0
ELISA
Difference
(corrected for
blank)
37.7
2.3
5.2
-0.2
22.6
-2.6
43.0
-3.0
20.6
-0.6
3.9
1.1
4.5
0.5
24.0
-4.0
38.7
1.4
17.7
2.3
5.3
-0.3
13.2
-3.2
22.2
-2.2
4.8
0.2
10.2
-0.2
18.1
1.9
5.2
-0.2
9.6
0.4
18.9
1.1
36.3
3.7
18.1
-0.09
2.05
Accuracy/Trueness calculations
1. Convert plate counts to logs: N/A
2. Use sample blank to correct the results from the spiked samples for matrix effects: Done in table
3. Calculate the average reference concentration of the analyte used to spike the samples: 18.00
4. Calculate the average concentration of the analyte in the spiked samples: 18.09
5. Divide the average concentration calculated from the spiked samples by the average reference
concentration: (18.1)/(18.0)=1.005
6. Multiply the quotient by 100. 1.005*100= 100.5%
Measurement uncertainty calculation
1. 95% CI calculated using the formula
=1.96*(2.05/√20)= 0.90
x=-0.09; -0.09-0.9= -0.99; -0.09+0.90= 0.81
Data Summary:
Calculated % accuracy/trueness 100.5%
Calculated measurement uncertainty -0.99 - 0.81
Appendix III. Accuracy & Trueness Data & Calculations
B. Pacific oyster
45
C. Geoduck
Data
Working Range 2 to 40 ppm
Sample Type Shellfish tissue
Organism used for spiking Geoduck (Panopea generosa)
Count
Sample
Date
Kit
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Average
St dev
1288
1289
1342
1358
1425
1428
1430
1485
1487
1566
1577
1569
1574
1575
1629
1661
1711
1770
1852
1856
4/7/2010
4/9/2010
4/9/2010
4/9/2010
4/9/2010
4/12/2010
4/12/2010
4/12/2010
4/12/2010
4/12/2010
4/13/2010
4/13/2010
4/13/2010
4/14/2010
4/14/2010
4/14/2010
4/15/2010
4/15/2010
4/15/2010
4/15/2010
100302-6
100302-7
100302-7
100302-7
100302-7
100302-8
100302-8
100302-8
100302-8
100302-9
100323-1
100323-1
100323-1
100323-1
100323-1
100323-1
100323-2
100323-2
100323-2
100323-2
Spike
5.0
20.0
40.0
5.0
20.0
40.0
5.0
20.0
40.0
5.0
5.0
10.0
20.0
5.0
10.0
20.0
5.0
10.0
20.0
40.0
17.3
Blank
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
ELISA
Difference
5.4
22.1
44.5
5.0
20.5
41.5
5.2
21.8
41.2
3.3
6.4
10.8
13.4
6.1
10.8
18.9
7.8
12.0
22.1
39.0
17.9
-0.4
-2.1
-4.5
0.0
-0.5
-1.5
-0.2
-1.8
-1.2
1.7
-1.4
-0.8
6.6
-1.1
-0.8
1.1
-2.8
-2.0
-2.1
1.0
-0.63
2.22
Accuracy/Trueness calculations
1. Convert plate counts to logs: N/A
2. Use the sample blank to correct the results from the spiked samples for matrix effects: N/A
3. Calculate the average reference concentration of the analyte used to spike the samples: 17.25
4. Calculate the average concentration of the analyte in the spiked samples: 17.88
5. Divide the average concentration calculated from the spiked samples by the average reference
concentration: (17.9)/(17.3)=1.037
6. Multiply the quotient by 100: 1.037*100= 103.7%
Measurement uncertainty calculation
1. 95% CI calculated using the formula
=1.96*(2.22/√20)= 0.97
x=-0.63; -0.63-0.97= -1.60; -0.63+0.97= 0.34
Data Summary:
Calculated % accuracy/trueness 103.7%
Calculated measurement uncertainty -1.60 - 0.34
Appendix III. Accuracy & Trueness Data & Calculations
C. Geoduck
46
D. Razor clam
Data
Working Range 2 to 40 ppm
Sample Type Shellfish Tissue
Organism used for spiking Razor clam (Siliqua patula)
Count
Sample Date
Kit
Spike
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Average
St dev
302
344
345
375
413
412
423
377
424
459
323
342
487
376
468
493
494
449
491
492
5/3/2010
5/3/2010
5/4/2010
5/4/2010
5/7/2010
5/7/2010
5/7/2010
5/13/2010
5/11/2010
5/11/2010
5/12/2010
5/12/2010
5/13/2010
5/13/2010
5/13/2010
5/5/2010
5/5/2010
5/4/2010
5/4/2010
5/4/2010
100323-7
100323-7
100323-8
100323-8
100323-9
100323-9
100323-9
100323-11
100323-9
100323-10
100323-11
100323-11
100323-11
100323-11
100323-11
100323-8
100323-8
100323-8
100323-8
100323-8
5.0
20.0
40.0
5.0
40.0
40.0
20.0
40.0
5.0
20.0
5.0
10.0
10.0
20.0
40.0
5.0
10.0
5.0
20.0
40.0
20.0
Blank
0.3
0.0
0.3
0.6
0.9
0.7
0.7
0.3
0.0
0.4
0.0
0.0
0.2
0.0
0.0
0.9
0.7
1.3
0.3
0.6
ELISA
Difference
(corrected
for blank)
6.0
-1.0
19.8
0.2
35.0
5.0
6.2
-1.2
37.1
2.9
36.3
3.7
17.0
3.1
41.7
-1.7
5.6
-0.6
19.1
0.9
5.0
0.0
12.0
-2.0
11.0
-1.0
24.4
-4.4
33.4
6.6
5.1
-0.1
11.8
-1.8
5.8
-0.8
19.2
0.9
33.9
6.1
19.27
0.73
2.92
Accuracy/Trueness calculations
1. Convert plate counts to logs: N/A
2. Use sample blank to correct the results from the spiked samples for matrix effects: Done in table
3. Calculate the average reference concentration of the analyte used to spike the samples: 20.00
4. Calculate the average concentration of the analyte in the spiked samples:: 19.27
5. Divide the average concentration calculated from the spiked samples by the average reference
concentration: (19.27)/(20.00)=0.9635
6. Multiply the quotient by 100: 0.9635*100= 96.35%
Measurement uncertainty calculation
1. 95% CI calculated using the formula
x=0.73; 0.73-1.28= -0.55; 0.73+1.28= 2.01
=1.96*(2.92/√20)= 1.28
Data Summary:
Calculated % accuracy/trueness 96.35%
Calculated measurement uncertainty -0.55 - 2.01
Appendix III. Accuracy & Trueness Data & Calculations
D. Razor Clam
47
E. Blue mussel
Data
Working Range 2 to 40 ppm
Sample Type Shellfish Tissue
Organism used for spiking Blue mussel (Mytilus edulis)
Count
Sample
Date
Kit
Spike
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Average
St dev
804
857
854
873
874
875
876
884
886
897
898
899
924
928
942
945
946
957
513
557
4/19/2010
4/23/2010
4/23/2010
4/23/2010
4/23/2010
4/26/2010
4/26/2010
4/26/2010
4/27/2010
4/27/2010
4/29/2010
4/29/2010
4/29/2010
4/30/2010
4/30/2010
4/30/2010
4/30/2010
4/30/2010
4/30/2010
4/30/2010
100323-3
100323-3
100323-3
100323-4
100323-4
100323-4
100323-4
100323-5
100323-5
100323-5
100323-6
100323-6
100323-6
100323-6
100323-6
100323-6
100323-6
100323-6
100323-6
100323-6
5.0
20.0
40.0
20.0
40.0
5.0
20.0
20.0
40.0
5.0
5.0
10.0
20.0
5.0
10.0
20.0
5.0
10.0
20.0
40.0
18.0
Blank
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.9
0.0
ELISA
Difference
5.3
19.1
36.8
17.3
39.0
5.5
20.3
15.7
35.6
6.0
7.1
10.9
22.0
5.1
10.2
21.3
5.6
10.6
20.1
37.7
17.55
0.45
1.94
-0.3
0.9
3.3
2.7
1.0
-0.5
-0.3
4.3
4.4
-1.0
-2.1
-0.9
-2.0
-0.1
-0.2
-1.3
-0.6
-0.6
-0.1
2.3
Accuracy/Trueness calculations
1. Convert plate counts to logs: N/A
2. Use sample blank to correct the results from the spiked samples for matrix effects: Done in table
Calculate the average reference concentration of the analyte used to spike the samples: 18.0
3. Calculate the average concentration of the analyte in the spiked samples: 17.55
4. Divide the average concentration calculated from the spiked samples by the average reference
concentration: (17.55)/(18.0)=.957
5. Multiply the quotient by 100: 0.957*100= 95.7%
Measurement uncertainty calculation
1. 95% CI calculated using the formula
x=0.45; 0.45-0.85= -0.40; 0.45+0.85= 1.30
=1.96*(1.94/√20)= 0.85
Data Summary:
Calculated % accuracy/trueness 95.7%
Calculated measurement uncertainty -0.40 - 1.30
Appendix III. Accuracy & Trueness Data & Calculations
E. Blue Mussel
48
F. Atlantic Oyster
Data
Working Range 2 to 40 ppm
Sample Type Shellfish Tissue
Organism used for spiking Atlantic oyster (Crassostrea virginica)
Count
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Average
Stdev
Sample
3
13
19
20
21
22
12
25
26
30
35
33
34
31
32
37
38
39
36
11
Date
5/14/2010
5/17/2010
5/18/2010
5/18/2010
5/18/2010
5/18/2010
5/18/2010
5/19/2010
5/19/2010
5/19/2010
5/24/2010
5/24/2010
5/24/2010
5/24/2010
5/24/2010
5/24/2010
5/24/2010
5/24/2010
5/24/2010
5/24/2010
Kit
100323-12
100323-13
100323-14
100323-14
100323-14
100323-14
100323-14
100323-15
100323-15
100323-15
100503-1
100503-1
100503-1
100323-15
100503-1
100323-15
100503-1
100323-15
100503-1
100503-2
Spike
Blank
ELISA
Difference
5.0
0.0
4.5
0.5
5.0
0.0
5.0
0.0
20.0
0.0
17.2
2.8
40.0
0.0
34.8
5.2
20.0
0.0
18.5
1.5
40.0
0.0
39.6
0.4
20.0
0.0
19.5
0.5
5.0
0.0
5.1
-0.1
20.0
0.0
21.8
-1.8
40.0
0.0
40.4
-0.4
5.0
0.0
5.5
-0.5
20.0
0.0
24.6
-4.6
40.0
0.0
52.0
-12.0
5.0
0.0
4.8
0.2
10.0
0.0
8.3
1.7
20.0
0.0
17.9
2.1
40.0
0.0
39.4
0.6
10.0
0.0
10.4
-0.4
10.0
0.0
8.3
1.7
5.0
0.0
5.4
-0.4
19.0
19.15
-0.15
3.38
Accuracy/Trueness calculations
1. Convert plate counts to logs: N/A
2. Use the sample blank to correct results from the spiked samples for matrix effects N/A
3. Calculate the average reference concentration of analyte used to spike the samples: 19.0
4. Calculate the average concentration of the analyte in the spiked samples: 19.15
5. Divide the average concentration calculated from the spiked samples by the average reference
concentration: (19.15)/(19.0)=1.008
6. Multiply the quotient by 100. 1.008*100= 100.8%
Measurement uncertainty calculation
1. 95% CI calculated using the formula
x=-0.15; -0.15-1.48= -1.63; -0.15+1.48= 1.33
=1.96*(3.38/√20)= 1.48
Data Summary:
Calculated % accuracy/trueness 100.8%
Calculated measurement uncertainty -1.63 - 1.33
Appendix III. Accuracy & Trueness Data & Calculations
F. Atlantic Oyster
49
G. Manila clam
Data
Working Range Samples 2 to 40 ppm
Sample Type Shellfish Tissue
Organism used for spiking Manila clam (Venerupis philippinarum)
Count
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Average
St dev
Sample
329
390
582
630
2111
2332
1052
2258
2523
2395
410
597
618
285
2246
2333
2396
2437
2536
2559
Date
5/20/2010
5/20/2010
5/20/2010
5/20/2010
5/20/2010
5/20/2010
5/20/2010
5/20/2010
5/20/2010
5/27/2010
5/26/2010
5/26/2010
5/26/2010
5/26/2010
5/26/2010
5/26/2010
5/26/2010
5/26/2010
5/27/2010
5/27/2010
Kit
100323-16
100323-16
100323-16
100323-17
100323-17
100323-17
100323-17
100323-17
100323-17
100503-4
100503-2
100503-2
100503-2
100503-3
100503-3
100503-3
100503-3
100503-3
100503-3
100503-3
Spike
Blank
ELSA
5.0
0.0
20.0
0.0
40.0
0.0
5.0
0.0
20.0
0.0
40.0
0.0
5.0
0.8
20.0
0.0
40.0
0.0
20.0
0.0
5.0
0.0
10.0
0.0
20.0
0.0
40.0
0.0
5.0
0.0
10.0
0.0
20.0
0.0
40.0
0.0
10.0
0.0
5.0
0.0
19.0
18.7
5.5
21.2
39.6
5.5
20.4
42.9
4.7
20.8
37.7
14.6
4.8
10.8
17.8
45.0
5.8
10.2
20.2
31.3
10.4
4.9
Difference
-0.5
-1.2
0.4
-0.5
-0.4
-2.9
0.3
-0.8
2.4
5.4
0.2
-0.8
2.2
-5.0
-0.8
-0.2
-0.2
8.7
-0.4
0.1
0.30
2.82
Accuracy/Trueness calculations
1. Convert plate counts to logs: N/A
2. Use the sample blank to correct results from the spiked samples for matrix effects Done in table
3. Calculate the average reference concentration of analyte used to spike the samples: 19.0
4. Calculate the average concentration of the analyte in the spiked samples: 18.7
5. Divide the average concentration calculated from the spiked samples by the average reference
concentration: (18.7)/(19.0)=0.984
6. Multiply the quotient by 100. 0.984*100= 98.4%
Measurement uncertainty calculation
1. 95% CI calculated using the formula
x=0.30; 0.30-1.24= -0.93; 0.30+1.24= 1.54
=1.96*(2.82/√20)= 1.24
Data Summary:
Calculated % accuracy/trueness 98.4%
Calculated measurement uncertainty -0.93 - 1.54
Appendix III. Accuracy & Trueness Data & Calculations
G. Manila clam
50
Appendix IV. Ruggedness Data and Calculations
VII. #2 Marine Biotoxin and Non-MPN Based Microbiological Methods SOP – Ruggedness
VALIDATION CRITERIA
Ruggedness is the ability of a particular method to withstand relatively minor changes in analytical
technique, reagents or environmental factors likely to arise in different test environments.
Procedure: This procedure is applicable for use with either growing waters or shellfish tissue. Make
every effort to use samples free of the target analyte/measurand/organism of interest. For each shellfish
type of interest use a minimum of 10 – 12 animals. For each sample take two (2) aliquots of either the
growing water sample or shellfish homogenate appropriately sized for your work. Spike both aliquots
with a suitable concentration of the target analyte/measurand/organism of interest. Process both aliquots
of the sample as usual to determine method concentration for the target analyte/measurand/organism of
interest. For the second aliquot of each sample, however, use a different batch or lot of culture media
and/or test reagents as appropriate to process this aliquot. For growing waters, do ten (10) samples
collected from a variety of growing waters. For shellfish do ten (10) samples for each shellfish tissue type
of interest collected from a variety of growing areas, the same growing area harvested on different days or
from different process lots. Use the same two batches or lots of culture media and/or test reagents to
process each sample such that “batch or lot 1” is used to process the first aliquot of each sample and
“batch or lot 2” is used to process the second aliquot of each sample. Use a range of concentrations which
spans the range of the method’s intended application to spike the sample aliquots. However both aliquots
of the same sample must be spiked with the same concentration of the target analyte/measurand/organism
of interest. Process samples over a period of several days.
Data:
Sample type ____________
Sample Conc “Batch or Lot 1” Conc “Batch or Lot 2”
Media and/or Reagents Media and/or Reagents
1
2
3
4
5
6
7
8
9
10
For shellfish samples, repeat for each tissue type of interest.
DATA HANDLING
Ruggedness
In the day to day operations of the laboratory there will be changes in the batches/lots of culture media
and/or test reagents used to process samples. Environmental factors are also likely to change over time.
None of these factors, however, should adversely impact test results if the method as implemented is
sufficiently rugged to be used routinely for regulatory monitoring.
Procedure: To determine whether the method as implemented is sufficiently rugged to withstand the
types of changes anticipated to occur in routine use, a two-sided t-test at a significance level (α) of .05
will be used on the data to ascertain if results obtained using different culture media and/or test reagent
batches/lots under slightly varying environmental conditions are significantly affected by such minor
changes. Either a paired t-test or Welch’s t-test will be used depending upon the shape of the distribution
produced by the data for each batch/lot and their respective variances. Use log transformed data for the
Appendix IV. Ruggedness Data & Calculations
A. Butter clams
51
results obtained from microbiological methods. The appropriate t-test to be used for the analysis is
determined in the following manner.
1. Test the symmetry of the distribution of results from both batch/lot 1 and batch/lot 2.
2. Calculate the variance of both batch/lot 1 and batch/lot 2 data.
3. Values for the test of symmetry for either batch/lot 1 or batch/lot 2 outside the range of -2 to +2
indicate a significant degree of skewness in the distribution.
4. A ratio of the larger of the variances of either batch/lot 1 or batch/lot 2 to the smaller of the
variances of either batch/lot 1 or batch/lot 2 >2 indicates a lack of homogeneity of variance.
5. Use either the paired t-test or Welch’s t-test for the analysis based on the following
considerations.
If the distributions of the data from batch/lot 1 and batch/lot 2 are symmetric
(within the range of -2 to +2) and there is homogeneity of variance, use a
paired t-test for the analysis.
If the distributions of the data from batch/lot 1 and batch/lot 2 are symmetric
(within the range of -2 to +2) but there is a lack of homogeneity of variance in
the data, use Welch’s t-test for the analysis.
If the distribution of the data from batch/lot 1 and batch/lot 2 are skewed
(outside the range of -2 to +2) and the skewness for both groups is either
positive for both or negative for both and there is homogeneity of variance in
the data, use the paired t-test for the analysis.
If the distributions of the data from batch/lot 1 and batch/lot 2 are skewed and
the skewness for both groups is either positive for both or negative for both
but the data lacks homogeneity of variance, use Welch’s t-test to analyze the
data.
Data Summary:
Value for the test of symmetry of the distribution of batch/lot 1 data _____
Value for the test of symmetry of the distribution of batch/lot 2 data ______
Variance of batch/lot 1 data ______
Variance of batch/lot 2 data _______
Ratio of the larger to the smaller of the variances of batch/lot 1 and batch/lot 2 ______
Is there a significant difference between batch/lot 1 samples and batch/lot 2 samples ____Y/N_______
A. Butter clam
Data
Sample type: Butter clam tissue
Count
Sample #
Date
1
1694
3/23/2010
2
1645
3/23/2010
3
1419
3/23/2010
4
1390
3/24/2010
5
1776
3/24/2010
6
1496
3/24/2010
7
1634
3/24/2010
8
1348
3/24/2010
9
1563
2/25/2010
Lot 1 (100304R)
3.5
11.8
19.9
7.3
11.5
22.3
6.1
21.4
10.7
Appendix IV. Ruggedness Data & Calculations
Lot 2 (100302)
3.8
10.1
20.3
9.0
14.1
25.1
5.5
23.4
11.4
Lot 1-Lot 2
-0.3
1.7
-0.4
-1.7
-2.6
-2.8
0.6
-2.0
-0.7
A. Butter clams
52
10
1736
2/25/2010
43.8
41.5
Average
St dev
2.3
-0.6
1.7
Ruggedness calculations
1. Test the symmetry of the distribution of results from both batch/lot 1 and batch/lot 2.
Symmetry was calculated using the formula
Lot 1 Symmetry: 1.577
Lot 2 Symmetry: 1.182
2. Calculate the variance of both batch/lot 1 and batch/lot 2 data.
Variance was calculated by using the formula
Lot 1 Variance = 139.1
Lot 2 Variance = 130.1
3. Values for the test of symmetry for either batch/lot 1 or batch/lot 2 outside the range of -2 to +2
indicate a significant degree of skewness in the distribution. N/A
4. A ratio of the larger of the variances of either batch/lot 1 or batch/lot 2 to the smaller of the variances
of either batch/lot 1 or batch/lot 2 >2 indicates a lack of homogeneity of variance.
Variance Lot 1/Variance Lot 2= 139.1/130.1= 1.069
5. Use either the paired t-test or Welch’s t-test for the analysis based on the following considerations.
If the distributions of the data from batch/lot 1 and batch/lot 2 are symmetric (within the range of
-2 to +2) and there is homogeneity of variance, use a paired t-test for the analysis.
If the distributions of the data from batch/lot 1 and batch/lot 2 are symmetric (within the range of
-2 to +2) but there is a lack of homogeneity of variance in the data, use Welch’s t-test for the analysis.
If the distribution of the data from batch/lot 1 and batch/lot 2 are skewed (outside the range of -2
to +2) and the skewness for both groups is either positive for both or negative for both and there is
homogeneity of variance in the data, use the paired t-test for the analysis.
If the distributions of the data from batch/lot 1 and batch/lot 2 are skewed and the skewness for
both groups is either positive for both or negative for both but the data lacks homogeneity of variance,
use Welch’s t-test to analyze the data.
A paired t-test was performed using the formula
lots
where d is the average difference between the
t=│ (-0.06)/ (√1.72/10)│=1.052
For a two-tailed p value of 0.05 with 9 degrees of freedom, the critical value is 2.262.
Appendix IV. Ruggedness Data & Calculations
A. Butter clams
53
Since t< 2.262 the two lots are NOT significantly different.
Data Summary:
Value for the test of symmetry of the distribution of batch/lot 1 data 1.577
Value for the test of symmetry of the distribution of batch/lot 2 data 1.182
Variance of batch/lot 1 data 139.1
Variance of batch/lot 2 data 130.1
Ratio of the larger to the smaller of the variances of batch/lot 1 and batch/lot 2 1.069
Is there a significant difference between batch/lot 1 samples and batch/lot 2 samples? No
Appendix IV. Ruggedness Data & Calculations
A. Butter clams
54
B. Pacific oyster
Data
Sample type: Pacific oyster tissue
Count
Sample Date
1
2207
4/5/2010
2
2208
4/5/2010
3
2209
4/5/2010
4
2211
4/6/2010
5
2330
4/6/2010
6
2275
4/6/2010
7
2181
4/6/2010
8
2329
4/6/2010
9
1566
4/6/2010
Lot 1 (100302)
5.1
12.8
22.4
5.4
10.5
19.0
5.8
11.0
19.4
Lot 2 (100323)
5.9
12.6
21.6
5.0
10.5
20.9
5.3
13.4
20.3
10
33.2
31.5
Average
St deviation
2110
4/6/2010
Ruggedness calculations
1. Test the symmetry of the distribution
Lot 1 Symmetry: 0.936
Lot 1-Lot 2
-0.8
0.2
0.8
0.4
0.0
-1.9
0.5
-2.4
-0.9
1.7
-0.2
1.3
Lot 2 Symmetry: 0.624
2. Calculate the variance
Lot 1 Variance = 81.76
Lot 2 Variance = 76.03
3. Values outside the range of -2 to +2 indicate a significant degree of skewness: N/A
4. Ratio of the larger to the smaller of the variances of >2 indicates a lack of homogeneity of variance.
Variance Lot 1/Variance Lot 2= 81.76/76.03= 1.075
5. Use either the paired t-test or Welch’s t-test for the analysis
If the distributions of the data from batch/lot 1 and batch/lot 2 are symmetric (within the range of
-2 to +2) and there is homogeneity of variance, use a paired t-test for the analysis.
A paired t-test was performed using the formula
where d is the average difference between lots
t=│ (-0.2)/ (√1.32/10)│=0.582
For a two-tailed p value of 0.05 with 9 degrees of freedom, the critical value is 2.262.
Since t< 2.262 the two lots are NOT significantly different.
Data Summary:
Value for the test of symmetry of the distribution of batch/lot 1 data 0.936
Value for the test of symmetry of the distribution of batch/lot 2 data 0.624
Variance of batch/lot 1 data 81.76
Variance of batch/lot 2 data 76.03
Ratio of the larger to the smaller of the variances of batch/lot 1 and batch/lot 2 1.075
Is there a significant difference between batch/lot 1 samples and batch/lot 2 samples? No
Appendix IV. Ruggedness Data & Calculations
B. Pacific oyster
55
C. Geoduck
Data
Sample type: Geoduck tissue
Count Sample
Date
1
1577
4/13/2010
Lot 1 (100302)
Lot 2 (100323)
5.9
10.6
12.8
5.2
10.4
18.2
7.1
12.1
2
3
4
5
6
7
8
1569
1574
1575
1629
1661
1711
1770
4/13/2010
4/13/2010
4/14/2010
4/14/2010
4/14/2010
4/15/2010
4/15/2010
5.7
10.8
16.2
5.8
10.7
19.0
6.4
12.8
9
10
1852
1856
4/15/2010
4/15/2010
22.1
38.2
Ruggedness calculations
1. Test the symmetry of the distribution
Lot 1 Symmetry: 1.573
23.2
39.8
Average
St deviation
Lot 1-Lot 2
-0.2
0.2
3.4
0.6
0.4
0.8
-0.7
0.7
-1.1
-1.6
0.2
1.4
Lot 2 Symmetry: 1.801
2. Calculate the variance
Lot 1 Variance = 98.78
Lot 2 Variance = 109.25
3. Values outside the range of -2 to +2 indicate a significant degree of skewness: N/A
4. Ratio of the larger to the smaller of the variances of >2 indicates a lack of homogeneity of variance.
Variance Lot 2/Variance Lot 1= 109.25/98.78= 1.106
5. Use either the paired t-test or Welch’s t-test for the analysis
If the distributions of the data from batch/lot 1 and batch/lot 2 are symmetric (within the range of
-2 to +2) and there is homogeneity of variance, use a paired t-test for the analysis.
A paired t-test was performed using the formula
where d is the average difference between lots
t=│ (0.2)/ (√1.42/10)│= 0.559
For a two-tailed p value of 0.05 with 9 degrees of freedom, the critical value is 2.262.
Since t< 2.262 the two lots are NOT significantly different.
Data Summary:
Value for the test of symmetry of the distribution of batch/lot 1 data 1.573
Value for the test of symmetry of the distribution of batch/lot 2 data 1.801
Variance of batch/lot 1 data 98.78
Variance of batch/lot 2 data 109.25
Ratio of the larger to the smaller of the variances of batch/lot 1 and batch/lot 2 1.106
Is there a significant difference between batch/lot 1 samples and batch/lot 2 samples? No
Appendix IV. Ruggedness Data & Calculations
C. Geoduck
56
D. Razor clam
Data
Sample type: Razor clam tissue
Count Sample
Date
Lot 1 (100323)
Lot 2 (100302)
Lot 1-Lot 2
1
449
5/4/2010
6.3
6.4
0.0
2
3
4
5
6
7
8
491
492
493
494
323
342
487
5/4/2010
5/4/2010
5/5/2010
5/5/2010
5/12/2010
5/12/2010
5/13/2010
18.6
35.7
5.0
11.4
5.1
9.6
11.9
20.4
36.0
4.2
10.2
6.6
12.7
13.4
-1.8
-0.4
0.9
1.2
-1.5
-3.2
-1.5
9
10
376
468
5/13/2010
5/13/2010
22.2
38.0
29.5
48.6
-7.3
-10.6
Ruggedness calculations
1. Test the symmetry of the distribution
Lot 1 Symmetry: 1.014
Average
St deviation
-2.4
3.8
Lot 2 Symmetry: 1.080
2. Calculate the variance
Lot 1 Variance = 147.30
Lot 2 Variance = 217.94
3. Values outside the range of -2 to +2 indicate a significant degree of skewness: N/A
4. Ratio of the larger to the smaller of the variances of >2 indicates a lack of homogeneity of variance.
Variance Lot 2/Variance Lot 1= 217.94/147.30= 1.48
5. Use either the paired t-test or Welch’s t-test for the analysis
If the distributions of the data from batch/lot 1 and batch/lot 2 are symmetric (within the range of
-2 to +2) and there is homogeneity of variance, use a paired t-test for the analysis.
A paired t-test was performed using the formula
where d is the average difference between lots
t=│ (-2.4)/ (√3.82/10)│= 2.03
For a two-tailed p value of 0.05 with 9 degrees of freedom, the critical value is 2.262.
Since t< 2.262 the two lots are NOT significantly different.
Data Summary:
Value for the test of symmetry of the distribution of batch/lot 1 data 1.014
Value for the test of symmetry of the distribution of batch/lot 2 data 1.080
Variance of batch/lot 1 data 147.30
Variance of batch/lot 2 data 217.94
Ratio of the larger to the smaller of the variances of batch/lot 1 and batch/lot 2 1.48
Is there a significant difference between batch/lot 1 samples and batch/lot 2 samples? No
Appendix IV. Ruggedness Data & Calculations
D. Razor clam
57
E. Blue mussel
Data
Sample type: Blue mussel tissue
Count Sample
Date
Lot 1 (100323)
Lot 2 (100302)
Lot 1-Lot 2
1
898
4/29/2010
8.0
6.5
1.5
2
3
4
5
6
7
8
899
924
928
942
945
946
957
4/29/2010
4/29/2010
4/30/2010
4/30/2010
4/30/2010
4/30/2010
4/30/2010
12.6
22.5
4.9
11.6
22.2
5.1
12.2
12.3
22.3
5.4
12.1
18.4
5.3
10.6
0.3
0.2
-0.5
-0.5
3.8
-0.2
1.6
9
10
513
557
4/30/2010
4/30/2010
21.6
42.1
20.5
41.8
1.1
0.2
0.8
1.3
Ruggedness calculations
1. Test the symmetry of the distribution
Lot 1 Symmetry: 1.373
Average
St deviation
Lot 2 Symmetry: 1.601
2. Calculate the variance
Lot 1 Variance = 126.83
Lot 2 Variance = 122.78
3. Values outside the range of -2 to +2 indicate a significant degree of skewness: N/A
4. Ratio of the larger to the smaller of the variances of >2 indicates a lack of homogeneity of variance.
Variance Lot 1/Variance Lot 2= 126.83/122.78= 1.03
5. Use either the paired t-test or Welch’s t-test for the analysis
If the distributions of the data from batch/lot 1 and batch/lot 2 are symmetric (within the range of
-2 to +2) and there is homogeneity of variance, use a paired t-test for the analysis.
A paired t-test was performed using the formula
where d is the average difference between lots
t=│ (0.8)/ (√1.32/10)│= 1.81
For a two-tailed p value of 0.05 with 9 degrees of freedom, the critical value is 2.262.
Since t< 2.262 the two lots are NOT significantly different.
Data Summary:
Value for the test of symmetry of the distribution of batch/lot 1 data 1.373
Value for the test of symmetry of the distribution of batch/lot 2 data 1.601
Variance of batch/lot 1 data 126.83
Variance of batch/lot 2 data 122.78
Ratio of the larger to the smaller of the variances of batch/lot 1 and batch/lot 2 1.03
Is there a significant difference between batch/lot 1 samples and batch/lot 2 samples? No
Appendix IV. Ruggedness Data & Calculations
E. Blue mussel
58
F. Atlantic Oyster
Data
Sample type: Atlantic Oyster tissue
Count Sample
Date
Lot 1 (100323)
Lot 2 (100302)
Lot 1-Lot 2
1
35
5/24/2010
5.8
5.6
0.2
2
3
4
5
6
7
8
33
34
31
32
37
38
39
5/24/2010
5/24/2010
5/24/2010
5/24/2010
5/24/2010
5/24/2010
5/24/2010
21.5
54.2
4.3
10.9
21.6
32.0
9.5
23.3
53.7
4.9
8.0
18.5
46.3
7.5
-1.7
0.6
-0.6
2.9
3.0
-14.2
2.0
9
10
36
11
5/24/2010
5/26/2010
11.2
6.1
9.1
6.8
2.1
-0.7
-0.6
5.0
Ruggedness calculations
1. Test the symmetry of the distribution
Lot 1 Symmetry: 1.655
Average
St deviation
Lot 2 Symmetry: 1.405
2. Calculate the variance
Lot 1 Variance = 242.18
Lot 2 Variance = 315.12
3. Values outside the range of -2 to +2 indicate a significant degree of skewness: N/A
4. Ratio of the larger to the smaller of the variances of >2 indicates a lack of homogeneity of variance.
Variance Lot 2/Variance Lot 1= 315.12/242.18= 1.301
5. Use either the paired t-test or Welch’s t-test for the analysis
If the distributions of the data from batch/lot 1 and batch/lot 2 are symmetric (within the range of
-2 to +2) and there is homogeneity of variance, use a paired t-test for the analysis.
A paired t-test was performed using the formula
where d is the average difference between lots
t=│ (-0.6)/ (√5.02/10)│= 0.407
For a two-tailed p value of 0.05 with 9 degrees of freedom, the critical value is 2.262.
Since t< 2.262 the two lots are NOT significantly different.
Data Summary:
Value for the test of symmetry of the distribution of batch/lot 1 data 1.655
Value for the test of symmetry of the distribution of batch/lot 2 data 1.405
Variance of batch/lot 1 data 242.18
Variance of batch/lot 2 data 315.12
Ratio of the larger to the smaller of the variances of batch/lot 1 and batch/lot 2 1.301
Is there a significant difference between batch/lot 1 samples and batch/lot 2 samples? No
Appendix IV. Ruggedness Data & Calculations
F. Atlantic oyster
59
G. Manila clam
Data
Sample type: Manila clam tissue
Count Sample
Date
Lot 1 (100323)
Lot 2 (100302)
Lot 1-Lot 2
1
410
5/26/2010
5.7
5.6
0.1
2
3
4
5
6
7
8
597
618
285
2246
2333
2396
2437
5/26/2010
5/26/2010
5/26/2010
5/26/2010
5/26/2010
5/26/2010
5/26/2010
11.1
19.8
37.8
7.1
9.4
20.3
37.1
11.3
20.2
41.5
7.3
11.4
18.5
30.7
-0.2
-0.3
-3.7
-0.1
-1.9
1.9
6.3
9
10
2536
2559
5/27/2010
5/27/2010
13.1
7.0
12.3
4.0
0.8
3.0
0.6
2.7
Ruggedness calculations
1. Test the symmetry of the distribution
Lot 1 Symmetry: 1.105
Average
St deviation
Lot 2 Symmetry: 1.238
2. Calculate the variance
Lot 1 Variance = 142.72
Lot 2 Variance = 141.34
3. Values outside the range of -2 to +2 indicate a significant degree of skewness: N/A
4. Ratio of the larger to the smaller of the variances of >2 indicates a lack of homogeneity of variance.
Variance Lot 1/Variance Lot 2= 142.72/141.34= 1.01
5. Use either the paired t-test or Welch’s t-test for the analysis
If the distributions of the data from batch/lot 1 and batch/lot 2 are symmetric (within the range of
-2 to +2) and there is homogeneity of variance, use a paired t-test for the analysis.
A paired t-test was performed using the formula
where d is the average difference between lots
t=│ (0.6)/ (√2.72/10)│= 0.669
For a two-tailed p value of 0.05 with 9 degrees of freedom, the critical value is 2.262.
Since t< 2.262 the two lots are NOT significantly different.
Data Summary:
Value for the test of symmetry of the distribution of batch/lot 1 data 1.105
Value for the test of symmetry of the distribution of batch/lot 2 data 1.238
Variance of batch/lot 1 data 142.72
Variance of batch/lot 2 data 141.34
Ratio of the larger to the smaller of the variances of batch/lot 1 and batch/lot 2 1.01
Is there a significant difference between batch/lot 1 samples and batch/lot 2 samples? No
Appendix IV. Ruggedness Data & Calculations
G. Manila clam
60
Appendix V. Precision & Recovery Data and Calculations
VII. #3 Marine Biotoxin and Non-MPN Based Microbiological Methods SOP – Precision &
Recovery
VALIDATION CRITERIA
Precision is the closeness of agreement between independent test results obtained under stipulated
conditions.
Recovery is the fraction or percentage of an analyte/measurand/organism of interest recovered following
sample analysis.
Procedure: This procedure is applicable for use with either growing waters or shellfish tissue. Make
every effort to use samples free of the target analyte/measurand/organism of interest. For each shellfish
type of interest use a minimum of 10-12 animals per sample. For each sample take four (4) aliquots of
either the shellfish homogenate or growing water sample appropriately sized for the work. Spike one of
the four aliquots with a low (but determinable by the method under study) concentration of the target
analyte/measurand/organism of interest. Spike the second aliquot of the growing water sample or shellfish
homogenate with a medium concentration of the target analyte/measurand/organism of interest. Spike the
third aliquot of the growing water sample or shellfish homogenate with a high (but determinable by the
method under study) concentration of the target analyte/measurand/organism of interest. Do not spike the
fourth aliquot of the growing water sample or shellfish homogenate. This is the sample blank. Spiking
levels must cover the range in concentrations important to the application of the method (working range).
For microbiological methods determine the concentration of the target organism of interest used to spike
each aliquot by plating in/on appropriate agar. Process each aliquot including the sample blank as usual to
determine the method concentration for the target analyte/measurand/organism of interest. Do two (2)
replicates for each of the three (3) spiked aliquots. Replicate analysis is unnecessary for the sample blank.
Do only one sample blank per sample. For growing waters, do ten (10) samples collected from a variety
of growing areas. For shellfish, do ten (10) samples for each shellfish tissue type of interest collected
from a variety of growing areas, the same growing area harvested on different days or from different
process lots. Use the same spiking levels for each of the ten (10) samples analyzed in this exercise (i.e.
1
3
5
10 , 10 and 10 ).
Data:
Working Range _____________
Sample Type ________________
Agar used to determine spike concentration _____________________
Organism used for spiking _________________________________
Sample Spike conc/Plate count/Conc of blank Conc in spiked sample from analysis
1L 1La 1Lb
1M 1Ma 1Mb
1H 1Ha 1Hb
1B
2L 2La 2 Lb
2M 2Ma 2Mb
2H 2Ha 2Hb
2B
““““““
““
10L 10La 10Lb
10M 10Ma 10Mb
AppendixV. Precision & Recovery Data & Calculations
A. Butter clams
61
10H 10Ha 10Hb
10B
L, M and H refer to low, medium and high concentrations respectively. La, Lb, Ma, Mb, Ha and Hb refer to
the replicate determinations of the sample aliquots spiked with low (L), medium (M) and high (H)
concentrations of the target analyte/measurand/organism of interest. B refers to the sample blank.
For shellfish samples, repeat for each tissue type of interest.
DATA HANDLING
To determine the precision and recovery, the NSSP method calls for the use of a nested and one way
ANOVA. However, determining recovery is not particularly relevant to an ELISA. Instead,
reproducibility was determined by finding the coefficient of variation at each concentrations and then
averaging those CVs in order to give an overall indication of the reproducibility of the ELISA over the
working range of the test.
A. Butter clams
Data
Working Range 2 to 40 ppm
Sample Type Shellfish tissue
Agar used to determine spike concentration N/A
Organism used for spiking Butter clams_
1
2
3
4
5
6
7
8
9
10
Samp
#
1543
1418
1508
1932
1660
1832
1371
1521
1949
1364
Date
Kit
3/10/10
3/17/10
3/9/10
3/17/10
3/18/10
3/18/10
3/19/10
3/19/10
3/19/10
3/22/10
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
Avg
Stdev
Blank
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
L
spike
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
La
Lb
4.2
3.4
6.5
6.1
5.6
6.0
5.5
5.5
4.4
4.8
4.0
9.7
5.4
5.7
4.9
5.5
3.9
4.3
5.8
5.2
5.31
1.33
M
spike
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
Ma
Mb
11.5
14.0
25.9
23.8
21.5
22.4
19.5
19.9
23.1
20.0
21.3
19.3
26.9
25.2
26.6
20.0
23.3
20.2
27.0
20.6
21.59
4.01
H
spike
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
Ha
28.5
29.7
53.3
51.7
43.0
42.3
42.6
47.2
43.1
46.5
38.3
39.3
54.3
51.4
43.5
43.9
38.2
44.4
42.5
46.7
43.51
6.76
Reproducibility calculations
1. Calculate the coefficient of variation (CV) for low spikes using the formula CV=(stdev/avg)*100
=1.33/5.31*100=25.1%
2. Calculate the CV for medium spikes: CV= 4.01/21.59*100= 18.57%
3. Calculate the CV for high spikes: CV= 6.76/43.51*100= 15.52%
4. Calculate the average CV for all spikes: (25.1+18.57+15.52)/3= 19.73%
Data summary
AppendixV. Precision & Recovery Data & Calculations
Hb
A. Butter clams
62
CV for low concentrations 25.1%
CV for medium concentrations 18.57%
CV for high concentrations 15.52%
Average CV for all concentrations: 19.73%
AppendixV. Precision & Recovery Data & Calculations
A. Butter clams
63
B. Pacific oysters
Data
Working Range 2 to 40 ppm
Sample Type Shellfish tissue
Agar used to determine spike concentration N/A
Organism used for spiking Pacific oysters
1
2
3
4
5
6
7
8
9
10
Samp
#
2115
2116
2117
2118
2119
1816
1817
2165
2175
2176
Date
Kit
3/26/2010
3/30/2010
3/30/2010
3/31/2010
3/31/2010
3/31/2010
4/1/2010
4/1/2010
4/1/2010
4/1/2010
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
Blank
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Avg
Stdev
L
spike
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
La
Lb
4.8
6.8
5.7
5.8
3.8
5.0
6.0
6.8
5.5
4.7
4.0
4.3
5.9
6.1
6.4
6.2
3.6
3.9
4.0
4.1
5.16
1.06
M
spike
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
Ma
Mb
23.6
22.0
20.0
20.9
22.9
24.9
20.4
21.0
21.3
21.0
18.8
17.6
24.4
22.3
22.6
21.4
20.0
18.9
15.7
17.3
20.85
2.37
H
spike
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
Ha
34.5
42.8
45.2
42.9
39.3
38.0
44.6
40.7
39.9
41.2
34.7
38.6
38.2
40.6
39.0
41.5
38.2
40.9
36.8
40.4
39.90
2.84
Reproducibility calculations
1.
2.
3.
4.
Calculate the CV for low spikes using the formula CV=(stdev/avg)*100 =1.06/5.16*100= 20.59
Calculate the CV for medium spikes: CV= 2.37/20.85*100= 11.36
Calculate the CV for high spikes: CV= 2.84/39.90*100= 7.12
Calculate the average CV for all spikes: (20.59+11.36+7.12)/3= 13.02
Data summary
CV for low concentrations 20.59%
CV for medium concentrations 11.36%
CV for high concentrations 7.12%
Average CV for all concentrations: 13.02%
Appendix V. Precision & Recovery Data & Calculations
B. Pacific Oyster
Hb
64
C. Geoducks
Data
Working Range 2 to 40 ppm
Sample Type Shellfish tissue
Agar used to determine spike concentration N/A
Organism used for spiking Geoducks
1
2
3
4
5
6
7
8
9
10
Samp
#
1288
1289
1342
1358
1425
1428
1430
1485
1487
1566
Date
Kit
4/7/2010
4/9/2010
4/9/2010
4/9/2010
4/9/2010
4/12/2010
4/12/2010
4/12/2010
4/12/2010
4/12/2010
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
Blank
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Avg
Stdev
L
spike
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
La
Lb
4.8
4.7
5.5
6.1
4.4
6.1
4.6
5.7
5.0
5.5
4.4
4.4
5.3
6.8
4.5
5.1
4.2
5.4
4.4
4.2
5.06
0.72
M
spike
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
Ma
Mb
18.9
17.6
21.3
21.2
23.5
20.2
18.2
18.7
21.6
18.5
22.4
20.7
21.6
16.0
22.5
18.0
19.7
15.7
19.0
18.7
19.69
2.13
H
spike
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
Ha
35.9
35.2
43.8
47.0
34.1
40.1
40.6
41.2
37.2
38.4
40.9
42.3
30.2
37.3
37.9
40.7
36.0
38.7
33.2
40.1
38.53
3.88
Reproducibility calculations
1.
2.
3.
4.
Calculate the CV for low spikes using the formula CV=(stdev/avg)*100 =0.72/5.06*100= 14.20
Calculate the CV for medium spikes: CV= 2.13/19.69*100= 10.81
Calculate the CV for high spikes: CV= 3.88/38.53*100= 10.06
Calculate the average CV for all spikes: (14.2+10.81+10.06)/3= 11.69
Data summary
CV for low concentrations 14.20%
CV for medium concentrations 10.81%
CV for high concentrations 10.06%
Average CV for all concentrations: 11.69%
Appendix V. Precision & Recovery Data & Calculations
C. Geoducks
Hb
65
D. Razor clams
Data
Working Range 2 to 40 ppm
Sample Type Shellfish tissue
Agar used to determine spike concentration N/A
Organism used for spiking Razor clams
1
2
3
4
5
6
7
8
9
10
Samp
#
302
344
345
375
377
412
413
423
424
459
Date
Kit
5/3/2010
5/3/2010
5/4/2010
5/4/2010
5/5/2010
5/5/2010
5/7/2010
5/7/2010
5/11/2010
5/11/2010
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
Blank*
0.3
0.0
0.3
0.6
0.7
0.7
0.9
0.7
0.0
0.4
Avg
Stdev
*All values are corrected for blanks
Reproducibility calculations
1.
2.
3.
4.
L
spike
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
La
Lb
6.0
6.1
4.3
4.8
5.4
5.6
5.8
6.5
4.5
5.3
4.7
4.5
5.7
6.3
3.8
4.9
6.0
6.9
4.6
5.5
5.36
0.82
M
spike
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
Ma
Mb
21.2
19.3
20.3
18.2
17.9
16.5
17.4
15.8
16.9
16.9
19.5
20.0
22.2
18.4
20.8
21.4
22.5
19.6
21.2
19.3
19.25
1.96
H
spike
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
Ha
34.6
36.4
32.8
38.0
31.9
35.7
32.4
35.4
34.1
39.8
37.2
35.8
37.9
43.9
31.9
33.7
36.6
37.5
36.3
42.1
36.20
3.19
Calculate the CV for low spikes using the formula CV=(stdev/avg)*100 =0.82/5.36*100= 15.38
Calculate the CV for medium spikes: CV= 1.96/19.25*100= 10.16
Calculate the CV for high spikes: CV= 3.19/36.20*100= 8.82
Calculate the average CV for all spikes: (15.38+10.16+8.82)/3= 11.45
Data summary
CV for low concentrations 15.38%
CV for medium concentrations 10.16%
CV for high concentrations 8.82%
Average CV for all concentrations: 11.45 %
Appendix V. Precision & Recovery Data & Calculations
D. Razor clams
Hb
66
E. Blue mussels
Data
Working Range 2 to 40 ppm
Sample Type Shellfish tissue
Agar used to determine spike concentration N/A
Organism used for spiking Blue mussels
1
2
3
4
5
6
7
8
9
10
Samp
#
804
854
857
873
874
875
876
884
886
897
Date
Kit
4/19/2010
4/23/2010
4/23/2010
4/23/2010
4/23/2010
4/26/2010
4/26/2010
4/26/2010
4/27/2010
4/27/2010
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
Blank
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Avg
Stdev
L
spike
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
La
Lb
6.6
6.7
4.8
4.4
3.9
4.2
5.3
5.5
6.2
5.9
4.4
4.7
5.4
5.5
4.7
5.9
5.4
5.0
4.0
5.6
5.20
0.82
M
spike
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
Ma
Mb
16.8
15.5
19.9
22.4
21.5
21.1
17.8
17.0
17.3
15.2
19.1
16.8
19.6
18.1
15.1
14.0
20.7
18.5
18.8
19.6
18.23
2.32
H
spike
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
Ha
41.1
41.3
38.0
40.0
36.9
35.8
33.7
33.5
35.1
37.7
30.5
32.4
33.3
35.0
33.1
31.4
33.9
35.2
32.2
33.0
35.14
3.12
Reproducibility calculations
1.
2.
3.
4.
Calculate the CV for low spikes using the formula CV=(stdev/avg)*100 =0.82/5.20*100= 15.84
Calculate the CV for medium spikes: CV= 2.32/18.23*100= 12.71
Calculate the CV for high spikes: CV= 3.12/35.14*100= 8.88
Calculate the average CV for all spikes: (15.84+12.71+8.88)/3= 12.48
Data summary
CV for low concentrations 15.84%
CV for medium concentrations 12.71%
CV for high concentrations 8.88%
Average CV for all concentrations: 12.48%
Appendix V. Precision & Recovery Data & Calculations
E. Blue mussels
Hb
67
F. Atlantic Oyster
Data
Working Range 2 to 40 ppm
Sample Type Shellfish tissue
Agar used to determine spike concentration N/A
Organism used for spiking Atlantic oysters
1
2
3
4
5
6
7
8
9
10
Samp
#
3
12
13
19
20
21
22
25
26
30
Date
Kit
5/14/2010
5/14/2010
5/17/2010
5/17/2010
5/18/2010
5/18/2010
5/18/2010
5/19/2010
5/19/2010
5/19/2010
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
Blank
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Avg
Stdev
L
spike
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
La
4.3
5.2
5.3
5.3
6.3
6.1
5.1
5.0
4.0
5.2
Lb
4.9
5.0
6.1
5.7
6.1
4.7
5.0
4.9
3.8
5.2
5.14
0.68
M
spike
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
Ma
17.5
19.9
25.2
19.1
20.0
18.5
19.9
24.0
18.1
17.6
Mb
19.1
21.6
19.7
20.7
20.5
17.2
17.1
23.5
18.8
18.6
19.81
2.27
H
spike
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
Ha
41.9
33.4
38.8
36.7
33.6
43.1
37.4
44.3
48.4
39.6
Reproducibility calculations
1.
2.
3.
4.
Calculate the CV for low spikes using the formula CV=(stdev/avg)*100 =0.68/5.14*100= 13.17
Calculate the CV for medium spikes: CV= 2.27/19.81*100= 11.46
Calculate the CV for high spikes: CV= 4.53/39.67*100= 11.43
Calculate the average CV for all spikes: (13.17+11.46+11.43)/3= 12.02
Data summary
CV for low concentrations 13.17%
CV for medium concentrations 11.46%
CV for high concentrations 11.43%
Average CV for all concentrations 12.02 %
Appendix V. Precision & Recovery Data & Calculations
F. Atlantic Oyster
Hb
34.7
37.8
39.7
33.6
35.2
43.1
40.2
47.4
45.1
39.5
39.67
4.53
68
G. Manila clam
Data
Working Range 2 to 40 ppm
Sample Type Shellfish tissue
Agar used to determine spike concentration N/A
Organism used for spiking Manila clams
1
2
3
4
5
6
7
8
9
10
Samp
#
329
390
582
630
2111
2332
1052
2258
2523
2395
Date
Kit
5/20/2010
5/20/2010
5/20/2010
5/20/2010
5/20/2010
5/20/2010
5/20/2010
5/20/2010
5/20/2010
5/20/2010
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
Blank
0.0
0.0
0.0
0.0
0.0
0.0
0.8
0.0
0.0
0.0
Avg
Stdev
L
spike
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
La
4.9
4.5
6.8
4.3
6.0
5.1
5.9
5.8
5.2
4.6
Lb
5.4
4.7
7.8
5.1
6.2
5.8
5.3
6.2
5.5
5.5
5.52
0.85
M
spike
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
Ma
Mb
25.2
20.7
22.6
24.3
20.0
20.1
24.2
23.7
24.3
21.0
20.2
19.4
18.7
26.2
17.3
18.5
19.3
21.9
19.6
18.8
21.30
2.57
H
spike
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
Ha
39.6
43.7
39.0
46.3
39.3
40.5
41.4
46.2
38.2
37.4
Reproducibility calculations
1.
2.
3.
4.
Calculate the CV for low spikes using the formula CV=(stdev/avg)*100 =0.85/5.52*100= 15.35
Calculate the CV for medium spikes: CV= 2.57/21.30*100= 12.09
Calculate the CV for high spikes: CV= 3.98/42.16*100= 9.44
Calculate the average CV for all spikes: (15.35+12.09+9.44)/3= 12.29
Data summary
CV for low concentrations 15.35%
CV for medium concentrations 12.09%
CV for high concentrations 9.44%
Average CV for all concentrations 12.29%
Appendix V. Precision & Recovery Data & Calculations
G. Manila clam
Hb
42.7
51.7
39.0
42.9
39.6
40.8
44.3
50.3
42.2
38.1
42.16
3.98
69
Appendix VI. Specificity Data and Calculations
VII. #4 Marine Biotoxin and Non-MPN Based Microbiological Methods SOP – Specificity
VALIDATION CRITERIA
Specificity is the ability of the method to measure only what it is intended to measure. To determine
method specificity samples containing suspected interferences (interfering organisms/compounds/toxins)
are analyzed in the presence of the analyte/measurand/targeted organism of interest.
Procedure: This procedure is applicable for use with either growing waters or shellfish tissue. Make
every effort to use samples free of the targeted analyte/measurand/organism of interest. For each shellfish
tissue type of interest use a minimum of 10-12 animals per sample. For each sample take three (3)
aliquots of either the shellfish homogenate or growing water sample appropriately sized for the work and
spike two (2) of the three (3) with a low but determinate level (by the method under study) of the targeted
analyte/measurand/ organism of interest. Take one of these two (2) aliquots and also spike it with a
moderate to high level of a suspected interfering organism/compound/toxin if not naturally incurred. Do
not spike the third aliquot. This is the sample blank. Process each aliquot, the sample blank, the aliquot
spiked with the targeted analyte/measurand/organism of interest and the aliquot spiked with the targeted
analyte/measurand/organism of interest in the presence of the suspected interfering
organism/compound/toxin as usual to determine the method concentration for the targeted
analyte/measurand/organism of interest. Do five (5) replicates for each aliquot excluding the sample
blank. Do one sample blank per analysis. Repeat this process for all suspected interfering
organisms/compounds/toxins.
Data:
Name of suspected interfering organism/compound/toxin #1 ______________________
Sample type ________________
Sample blank concentration for the targeted analyte/measurand/organism of interest ____
Concentration of aliquot spiked Concentration of aliquot spiked
with targeted analyte/measurand/ with targeted analyte/measurand/
organism of interest organism of interest and
suspected interfering organism/
compound/toxin
Replicate 1
2
3
4
5
Repeat for each suspected interfering organism tested.
DATA HANDLING
The Specificity index will be used to test the specificity of the method in the presence of suspected
interfering organisms/compounds/toxins. The Specificity index (SI) is calculated as indicated below:
Specificity index (SI) = Sample spiked with target of interest only
Sample spiked with both target and suspected interferences
All microbiological count data must be converted to logs before analysis. Samples spiked with both the
targeted analyte/measurand/organism of interest and the targeted analyte/measurand/organism of interest
in the presence of a suspected interfering organism/compound/toxin may have to be corrected for matrix
effects before determining the Specificity index (SI). The sample blank accompanying the analysis is used
for this purpose. Any corrections that may be necessary to microbiological data for matrix effects are
done using log transformed data.
AppendixVI. Specificity Data & Calculations
A. Butter clams
70
The Specificity index should equal one (1) in the absence of interferences. To test the significance of a
Specificity index other than one (1) for any suspected interfering organism/compound/toxin, a two-sided
t-test is used. For each suspected interfering organism/compound/toxin calculate the average Specificity
Index (SI) for the 5 replicates analyzed for each sample by obtaining the average concentration for both
the aliquot containing the targeted analyte/measurand/organism of interest only and the aliquot containing
the targeted analyte/measurand/organism of interest in the presence of suspected interfering
organisms/compounds/toxins and using the formula below.
SIavg = Avg concentration of sample spiked with target of interest only
Avg concentration of sample spiked with both target and suspected interferences
Perform a two-sided t-test at the .05 significance level to determine if the average Specificity index (SI)
obtained from the 5 replicates of each analysis differs from one (1).
Repeat for all interfering organisms/compounds/toxins tested.
Note: The calculation steps for performing the t-test were adapted from the ISSC 2007 Summary of
Actions p317.
Data Summary:
Interfering organism/compound/toxin #1 ___________________ SIavg______
Significant difference from 1 _____
Interfering organism/compound/toxin #2 ___________________ SIavg_______ Significant difference
from 1 _____
Interfering organism/compound/toxin #3 ____________________ SIavg ___________
Significant difference from 1 _____
Interfering organism/compound/toxin #n ___________________ SIavg _______ Significant difference
from 1 _____
A. Butter Clams
Data:
Name of suspected interfering organism/compound/toxin #1 Glutamic Acid
Sample type Butter clam tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc of spike Conc of Spike + IC
Specificity Index
1
5.0
3.5
1.42
2
5.4
2.9
1.86
3
5.6
3.6
1.56
4
4.4
3.4
1.30
5
5.2
5.7
0.92
Avg
5.12
3.82
1.41
Perform the two-sided t-test to determine if the average specificity index (SI) obtained from the 5 replicates
of each analysis differs from one (1).
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t. 2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 1.41
4. Calculate the standard deviation (s). 0.346
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.346/ √5=0.430
6. If [SIavg – mo] > u, where mo = 1 the average Specificity index (SIavg) differs from one (1) and that the
method may not be specific for the analyte [SIavg – mo] = 1.41-1 = 0.41 < 0.43 = u
AppendixVI. Specificity Data & Calculations
A. Butter clams
71
Name of suspected interfering organism/compound/toxin #2 Glutamine
Sample type Butter clam tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc of spike Conc of Spike + Specificity Index
IC
1
4.9
4.4
1.11
2
5.8
4.5
1.30
3
6.0
3.8
1.59
4
5.2
4.6
1.14
5
5.2
4.2
1.24
Avg
5.45
4.30
1.28
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t. 2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 1.28
4. Calculate the standard deviation (s). 0.191
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.191/ √5=0.237
6. If [SIavg – mo] > u, where mo = 1 decide that the average Specificity index (SIavg) differs from
one (1)
and that the method may not be specific for the analyte [SIavg – mo] = 1.28-1= 0.28>0.24=u
Data Summary
Interfering compound #1: Glutamic Acid
Interfering compound #2: Glutamine
SIavg: 1.40
SIavg: 1.28
AppendixVI. Specificity Data & Calculations
Significant difference from 1: NO
Significant difference from 1: YES
A. Butter clams
72
B. Pacific oysters
Data:
Name of suspected interfering organism/compound/toxin #1 Glutamic Acid
Sample type Pacific oyster tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc
of Conc of Spike + Specificity Index
spike
IC
1
5.0
2.8
1.78
2
2.7
3.8
0.72
3
4.2
3.1
1.36
4
4.9
3.5
1.39
5
4.5
4.8
0.95
Avg
4.28
3.61
1.185
Perform the two-sided t-test to determine if the average specificity index (SI) obtained from the 5
replicates
of each analysis differs from one (1).
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t. 2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 1.185
4. Calculate the standard deviation (s). 0.413
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.413/ √5=0.513
6. If [SIavg – mo] > u, where mo = 1 decide that SIavg differs from 1 and that the method may not
be specific for the analyte [SIavg – mo] = [1.185-1]= 0.185 < 0.513 = u
Name of suspected interfering organism/compound/toxin #2 Glutamine
Sample type Pacific oyster tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc of spike Conc of Spike + IC
Specificity Index
1
4.9
4.7
1.03
2
4.7
4.4
1.08
3
4.2
4.7
0.90
4
4.6
4.7
0.97
5
4.3
5.4
0.79
Avg
4.5
4.8
0.96
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t.
2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 0.96
4. Calculate the standard deviation (s). 0.113
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.113/
√5=0.14
6. If [SIavg – mo] > u, where mo = 1 decide that SIavg differs from 1 and that the method
may not be specific for the analyte [SIavg – mo] = [0.96-1] = 0.04< 0.14 = u
Data Summary
Interfering compound #1: Glutamic Acid
Interfering compound #2: Glutamine
SIavg: 1.185
SIavg: 0.96
Appendix VI. Specificity Data & Calculations
Significant difference from 1: NO
Significant difference from 1: NO
B. Pacific Oysters
73
Appendix VI. Specificity Data & Calculations
B. Pacific Oysters
74
C. Geoducks
Data:
Name of suspected interfering organism/compound/toxin #1 Glutamic Acid
Sample type Geoduck tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc
of Conc of Spike + Specificity Index
spike
IC
1
5.1
5.2
0.98
2
5.2
5.2
1.00
3
5.6
5.1
1.10
4
5.9
4.6
1.29
5
6.1
5.0
1.22
Avg
5.59
5.02
1.11
Perform the two-sided t-test to determine if the average specificity index (SI) obtained from the 5
replicates
of each analysis differs from one (1).
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t. 2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 1.11
4. Calculate the standard deviation (s). 0.136
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.083/ √5=0.168
6. If [SIavg – mo] > u, where mo = 1 decide that SIavg differs from 1 and that the method may
not be specific for the analyte [SIavg – mo] = [1.11-1]= 0.11 > 0.17 = u
Name of suspected interfering organism/compound/toxin #2 Glutamine
Sample type Geoduck tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc of spike Conc of Spike + IC
Specificity Index
1
5.3
6.0
0.87
2
5.0
6.5
0.77
3
6.0
4.7
1.26
4
3.0
5.6
0.53
5
4.9
4.8
1.01
Avg
4.82
5.53
0.89
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t.
2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 0.89
4. Calculate the standard deviation (s). 0.270
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.270/
√5=0.335
6. If [SIavg – mo] > u, where mo = 1 decide that SIavg differs from 1 and that the method
may not be specific for the analyte [SIavg – mo] = [0.89-1] = 0.11 < 0.335 = u
Data Summary
Interfering compound #1: Glutamic Acid
Interfering compound #2: Glutamine
SIavg: 1.11
SIavg: 0.89
Appendix VI. Specificity Data & Calculations
Significant difference from 1: NO
Significant difference from 1: NO
C. Geoducks
75
Appendix VI. Specificity Data & Calculations
C. Geoducks
76
D. Razor clams
Data:
Name of suspected interfering organism/compound/toxin #1 Glutamic Acid
Sample type Razor clam tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc
of Conc of Spike + Specificity Index
spike
IC
1
4.8
4.7
1.01
2
7.7
4.9
1.58
3
4.5
4.6
0.98
4
4.6
5.3
0.88
5
5.4
5.5
0.99
Avg
5.41
4.99
1.08
Perform the two-sided t-test to determine if the average specificity index (SI) obtained from the 5
replicates
of each analysis differs from one (1).
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t. 2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 1.08
4. Calculate the standard deviation (s). 0.278
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.278/ √5=0.345
6. If [SIavg – mo] > u, where mo = 1 decide that SIavg differs from 1 and that the method may not
be specific for the analyte [SIavg – mo] = [1.08-1]= 0.08 < 0.345 = u
Name of suspected interfering organism/compound/toxin #2 Glutamine
Sample type Razor clam tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc of spike Conc of Spike + IC
Specificity Index
1
5.0
6.0
0.85
2
4.5
6.3
0.71
3
5.3
4.9
1.09
4
5.9
5.7
1.04
5
5.4
6.1
0.88
Avg
5.23
5.78
0.91
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t.
2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 0.91
4. Calculate the standard deviation (s). 0.153
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.153/
√5=0.190
6. If [SIavg – mo] > u, where mo = 1 decide that SIavg differs from 1 and that the method
may not be specific for the analyte [SIavg – mo] = [0.91-1] = 0.09 < 0.19 = u
Data Summary
Interfering compound #1: Glutamic Acid
Interfering compound #2: Glutamine
SIavg: 1.08
SIavg: 0.91
Appendix VI. Specificity Data & Calculations
Significant difference from 1: NO
Significant difference from 1: NO
E. Blue mussels
77
E. Blue mussels
Data:
Name of suspected interfering organism/compound/toxin #1 Glutamic Acid
Sample type Blue mussel tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc
of Conc of Spike + Specificity Index
spike
IC
1
5.6
6.9
0.80
2
5.5
6.5
0.84
3
5.3
7.1
0.75
4
6.0
7.5
0.81
5
6.6
6.6
1.01
Avg
5.79
6.92
0.84
Perform the two-sided t-test to determine if the average specificity index (SI) obtained from the 5
replicates
of each analysis differs from one (1).
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t. 2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 0.84
4. Calculate the standard deviation (s). 0.100
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.100/ √5=0.12
6. If [SIavg – mo] > u, where mo = 1 decide that SIavg differs from 1 and that the method may not
be specific for the analyte [SIavg – mo] = [0.84-1]= 0.16 > 0.12 = u
Name of suspected interfering organism/compound/toxin #2 Glutamine
Sample type Blue mussel tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc of spike Conc of Spike + IC
Specificity Index
1
5.6
6.8
0.81
2
5.5
7.0
0.78
3
5.3
7.4
0.71
4
6.0
7.0
0.86
5
6.6
6.0
1.11
Avg
5.79
6.86
0.85
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t.
2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 0.85
4. Calculate the standard deviation (s). 0.15
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.15/ √5=0.19
6. If [SIavg – mo] > u, where mo = 1 decide that SIavg differs from 1 and that the method
may not be specific for the analyte [SIavg – mo] = [0.85-1] = 0.15 < 0.19 = u
Data Summary
Interfering compound #1: Glutamic Acid
Interfering compound #2: Glutamine
SIavg: 0.84
SIavg: 0.85
Appendix VI. Specificity Data & Calculations
Significant difference from 1: YES
Significant difference from 1: NO
E. Blue mussels
78
F. Atlantic oysters
Data:
Name of suspected interfering organism/compound/toxin #1 Glutamic Acid
Sample type Atlantic oyster tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc
of Conc of Spike + Specificity Index
spike
IC
1
3.8
5.3
0.72
2
4.6
5.0
0.91
3
4.6
5.3
0.86
4
4.0
6.4
0.62
5
4.8
5.3
0.90
Avg
4.34
5.46
0.79
Perform the two-sided t-test to determine if the average specificity index (SI) obtained from the 5
replicates
of each analysis differs from one (1).
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t. 2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 0.79
4. Calculate the standard deviation (s). 0.12
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.12/ √5=0.16
6. If [SIavg – mo] > u, where mo = 1 decide that SIavg differs from 1 and that the method may not
be specific for the analyte [SIavg – mo] = [0.79-1]= 0.21 > 0.16 = u
Name of suspected interfering organism/compound/toxin #2 Glutamine
Sample type Atlantic oyster tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc of spike Conc of Spike + IC
Specificity Index
1
5.7
5.3
1.08
2
4.2
5.4
0.77
3
5.4
4.7
1.14
4
5.7
5.6
1.02
5
4.6
6.5
0.72
Avg
5.11
5.49
0.94
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t.
2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 0.94
4. Calculate the standard deviation (s). 0.19
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.19/ √5=0.24
6. If [SIavg – mo] > u, where mo = 1 decide that SIavg differs from 1 and that the method
may not be specific for the analyte [SIavg – mo] = [0.94-1] = 0.06 < 0.24 = u
Data Summary
Interfering compound #1: Glutamic Acid
Interfering compound #2: Glutamine
SIavg: 0.79
SIavg: 0.94
Appendix VI. Specificity Data & Calculations
Significant difference from 1: YES
Significant difference from 1: NO
F. Atlantic oyster
79
G. Manila clams
Data:
Name of suspected interfering organism/compound/toxin #1 Glutamic Acid
Sample type Manila clam tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc
of Conc of Spike + Specificity Index
spike
IC
1
4.8
4.3
1.11
2
4.6
4.1
1.12
3
4.8
4.1
1.16
4
4.1
4.7
0.86
5
4.1
4.4
0.93
Avg
4.47
4.33
1.03
Perform the two-sided t-test to determine if the average specificity index (SI) obtained from the 5
replicates
of each analysis differs from one (1).
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t. 2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 1.03
4. Calculate the standard deviation (s). 0.13
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.13/ √5=0.16
6. If [SIavg – mo] > u, where mo = 1 decide that SIavg differs from 1 and that the method may not
be specific for the analyte [SIavg – mo] = [1.03-1]= 0.03 < 0.16 = u
Name of suspected interfering organism/compound/toxin #2 Glutamine
Sample type Manila clam tissue
Sample blank concentration for the targeted analyte of interest 0.0 ppm
Replicate
Conc of spike Conc of Spike + IC
Specificity Index
1
4.6
5.0
0.91
2
4.2
4.7
0.90
3
4.9
3.2
1.52
4
5.3
3.6
1.46
5
4.7
4.3
1.09
Avg
4.73
4.17
1.18
1. Let the significance level, α = .05
2. Look up t1-α/2 for n-1 degrees of freedom (n = 4) in the Table of the distribution of t.
2.776
3. Calculate the average Specificity index (SIavg) from the data for each analysis. 1.18
4. Calculate the standard deviation (s). 0.30
5. Calculate u, u = t1-α/2 s/n^0.5, n = # of replicates per analysis. u= 2.776 * 0.30/ √5=0.37
6. If [SIavg – mo] > u, where mo = 1 decide that SIavg differs from 1 and that the method
may not be specific for the analyte [SIavg – mo] = [1.18-1] = 0.18 < 0.37 = u
Data Summary
Interfering compound #1: Glutamic Acid
Interfering compound #2: Glutamine
SIavg: 1.03
SIavg: 1.18
Appendix VI. Specificity Data & Calculations
Significant difference from 1: NO
Significant difference from 1: NO
G. Manila clams
80
Appendix VII. Comparing Methods Data and Calculations
IX. #1 SOP for the Single Laboratory Validation of New or Modified Analytical Methods Intended
as Alternatives to Officially Recognized NSSP Methods – Comparing Methods
VALIDATION CRITERIA
Comparability is the acceptability of a new or modified analytical method as a substitute for an
established method in the NSSP. To be acceptable the new or modified method must not produce a
significant difference in results when compared to the officially recognized method. Comparability must
be demonstrated for each substrate or tissue type of interest by season and geographic area if applicable.
Comparison of Methods:
New or modified methods demonstrating comparability to officially recognized methods must not
produce significantly different results when compared
Procedure to compare the new or modified method to the officially recognized method: This
procedure is applicable for use with either growing waters or shellfish tissue. For each shellfish type of
interest use a minimum of 10-12 animals per sample. For each sample take two (2) aliquots and analyze
one by the officially recognized method and the other by the alternative method. Actual samples are
preferable; but, in cases where the occurrence of the analyte/measurand/organism of interest is
intermittent (such as marine biotoxins), spiked samples can be used. Samples having a variety of
concentrations which span the range of the method’s intended application should be used in the
comparison. Analyze a minimum of thirty (30) paired samples for each season from a variety of growing
areas for a total of at least 120 samples over the period of a year for naturally incurred samples. For
spiked samples analyze a minimum of ten (10) samples for each season from a variety of growing areas
for a total of at least 40 samples over the period of a year.
Note: Instead of testing samples over the range in seasons, samples were tested across a geographic range.
Twenty (20) samples from two different locations (Alaska and Washington) were tested simultaneously
by HPLC and ELISA for a total of 40 samples.
Data:
Sample type ______________
Date Sample/Station # Conc. Recognized method Conc. Alternative Method
1
2
3
4
5
6
7
8
9
10
n
n is the last sample in the comparison
For shellfish samples, repeat for each tissue type of interest
Data handling to compare the new or modified method to the officially recognized
Two methods of analysis are considered to be comparable when no significant difference can be
demonstrated in their results. To determine whether comparability in methods exists, a two-sided t-test at
a significance level (α) of .05 will be used to test the data. Either a paired t-test or Welch’s t-test will be
Appendix VII. Comparing Methods Data & Calculations
A. Butter clams
81
used depending upon the shape of the distributions produced by the data for each method and their
respective variances. Use log transformed data for the results obtained from microbiological methods.
The appropriate t-test to be used for the analysis is determined in the following manner.
1. Test the symmetry for the distribution of results from both the officially recognized analytical
method and the proposed alternative analytical method.
2. Calculate the variance of the data for both the officially recognized analytical method and the
proposed alternative analytical method.
3. Values for the test of symmetry for either method outside the range of -2 to +2 indicate a
significant degree of skewness in the distribution.
4. A ratio of the larger of the variances of either method to the smaller of the variances of either
method >2 indicates a lack of homogeneity of variance.
5. Use either the paired t-test or Welch’s t-test for the analysis of the data based on the following
considerations.
• If the distribution of the data from the officially recognized analytical method and the
proposed alternative analytical method are symmetric (within the range of -2 to +2) and
there is homogeneity of variance use a paired t-test for the data analysis.
• If the distributions of the data for both analytical methods are symmetric (within the range
-2 to +2) but there is a lack of homogeneity of variance in the data, use Welch’s t-test
for the analysis of the data.
• If the distributions of the data from the officially recognized and proposed alternative
analytical methods are skewed (outside the range -2 to +2) and the skewness for both
methods is either positive for both or negative for both and there is homogeneity of
variance in the data, use the paired t-test for the analysis of the data.
• If the distributions of the data from the officially recognized and the proposed alternative
analytical methods are skewed and the skewness for both analytical methods is either
positive or negative for both but the data lacks homogeneity of variance, use Welch’s ttest to analyze the data.
Data summary for the comparison of the new or modified method to the officially recognized
method:
Value for the test of symmetry for the distribution of the data generated by the officially recognized
method _______________
Value for the test of symmetry for the distribution of the data generated by the proposed alternative
method ________________
Variance of the data generated from the officially recognized analytical method _______
Variance of the data generated from the proposed alternative analytical method _______
Ratio of the larger to the smaller of the variances generated by the officially recognized and proposed
analytical methods ________________
Is there a significant difference between the analytical methods Y/N
A. Butter clams
Data
Sample type Butter clam tissue
Count Location
1
WA
2
WA
3
WA
4
WA
5
WA
Sample
1694
1645
1419
1390
1776
Date
Kit
3/23/2010 100304R
3/22/2010 100304R
3/23/2010 100304R
3/24/2010
100302
3/24/2010
100302
Spike
5
10
20
5
10
Appendix VII. Comparing Methods Data & Calculations
HPLC Area
58
168.8
227
74.3
141.8
HPLC ppm
3.6
10.4
13.9
4.6
8.7
ELISA ppm
4.1
11.2
21.8
7.2
11.1
A. Butter clams
82
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
WA
WA
WA
WA
WA
WA
WA
WA
WA
WA
WA
WA
WA
WA
WA
AK
AK
AK
AK
AK
AK
AK
AK
AK
AK
AK
AK
AK
AK
AK
AK
AK
AK
AK
AK
1496
1634
1348
1563
1736
1543
1418
1508
1932
1660
1830
1371
1521
1949
1364
BC1
BC2
BC3
BC4
BC5
BC6
BC7
BC8
BC9
BC10
BC11
BC12
BC13
BC14
BC15
BC16
BC17
BC18
BC19
BC20
3/24/2010
100302
3/22/2010 100304R
3/22/2010 100304R
3/25/2010
100302
3/25/2010
100302
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
9/29/2010
100916
3/14/2010 Cayman
3/14/2010 Cayman
3/14/2010 Cayman
3/14/2010 Cayman
3/14/2010 Cayman
3/14/2010 Cayman
3/14/2010 Cayman
3/14/2010 Cayman
3/14/2010 Cayman
3/14/2010 Cayman
20
5
20
10
40
2
2
2
2
5
10
20
40
40
40
2
2
2
2
5
10
20
40
40
40
5
10
20
5
10
20
5
10
20
40
Appendix VII. Comparing Methods Data & Calculations
301.3
70.9
302.9
127.3
450
35.1
29.4
33.3
28.7
88
167.4
284.8
675.5
633.4
594.5
34.5
36.7
33.5
34.9
86.2
161.2
315.3
723.7
706.6
612
75.6
153.2
295.8
72.4
151.9
282.5
73
144.7
288.1
571.2
18.5
4.3
18.6
7.8
27.6
2.2
1.8
2.0
1.8
5.4
10.3
17.5
41.4
38.9
36.5
2.1
2.3
2.1
2.1
5.3
9.9
19.3
44.4
43.3
37.5
4.6
9.4
18.1
4.4
9.3
17.3
4.5
8.9
17.7
35.0
A. Butter clams
22.8
5.6
23.7
10.0
38.1
3.4
3.7
2.7
3.0
6.2
9.6
17.0
35.5
31.4
30.8
2.5
2.4
2.5
2.1
5.2
8.2
15.2
32
34.7
28.8
5.9
11.8
20.4
5.1
14.3
22.7
7.1
13.1
20.4
36.1
83
1. Test the symmetry for the distribution of results from both the officially recognized analytical method
(HPLC) and the proposed alternative analytical method (ELISA).
Symmetry was calculated using the formula:
HPLC symmetry: 1.099
ELISA symmetry: 0.690
2. Calculate the variance of the data for both the officially recognized analytical method and the proposed
alternative analytical method.
Variance was calculated using the formula:
HPLC variance: 179.44
ELISA variance: 131.02
3. Values for the test of symmetry for either method outside the range of -2 to +2 indicate a significant
degree of skewness in the distribution. N/A
4. A ratio of the larger of the variances of either method to the smaller of the variances of either method
>2 indicates a lack of homogeneity of variance. N/A
Variance HPLC/ELISA: 179.44/131.02= 1.37
5. Use either the paired t-test or Welch’s t-test for the analysis of the data based on the following
considerations.
• If the distribution of the data from the officially recognized analytical method and the
proposed alternative analytical method are symmetric (within the range of -2 to +2) and
there is homogeneity of variance use a paired t-test for the data analysis.
Appendix VII. Comparing Methods Data & Calculations
A. Butter clams
84
• If the distributions of the data for both analytical methods are symmetric (within the range
-2 to +2) but there is a lack of homogeneity of variance in the data, use Welch’s t-test
for the analysis of the data.
• If the distributions of the data from the officially recognized and proposed alternative
analytical methods are skewed (outside the range -2 to +2) and the skewness for both
methods is either positive for both or negative for both and there is homogeneity of
variance in the data, use the paired t-test for the analysis of the data.
• If the distributions of the data from the officially recognized and the proposed alternative
analytical methods are skewed and the skewness for both analytical methods is either
positive or negative for both but the data lacks homogeneity of variance, use Welch’s ttest to analyze the data.
A paired t-test was performed using the Excel formula =TTEST(array1, array2, tails, type) where array1=
HPLC data; array2= ELISA data; tails= 2; type= 1(paired)
p= 0.576
Since p > 0.05, the HPLC and ELISA values are NOT significantly different.
Data summary for the comparison of the new or modified method to the officially recognized
method:
Value for the test of symmetry for the distribution of the data generated by the officially recognized
method 1.099
Value for the test of symmetry for the distribution of the data generated by the proposed alternative
method 0.690
Variance of the data generated from the officially recognized analytical method 179.44
Variance of the data generated from the proposed alternative analytical method 131.02
Ratio of the larger to the smaller of the variances generated by the officially recognized and proposed
analytical methods 1.37
Is there a significant difference between the analytical methods? No
Appendix VII. Comparing Methods Data & Calculations
A. Butter clams
85
B. Pacific Oyster
Data
Sample type Pacific oyster tissue
Count
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Location Sample
Date
Kit Spike
WA
2207 4/5/2010 100302
5
WA
2208 4/5/2010 100302
10
WA
2209 4/5/2010 100302
20
WA
2211 4/6/2010 100302
5
WA
2330 4/6/2010 100302
10
WA
2275 4/6/2010 100302
20
WA
2181 4/6/2010 100302
5
WA
2329 4/6/2010 100302
10
WA
1566 4/6/2010 100302
20
WA
2110 4/6/2010 100302
40
WA
2115 9/30/2010 100916
2
WA
2116 9/30/2010 100916
2
WA
2117 9/30/2010 100916
2
WA
2118 9/30/2010 100916
2
WA
2119 9/30/2010 100916
5
WA
1816 9/30/2010 100916
10
WA
1817 9/30/2010 100916
20
WA
2165 9/30/2010 100916
40
WA
2175 9/30/2010 100916
40
WA
2176 9/30/2010 100916
40
AK
PO11 3/22/2010 100301
5
AK
PO12 3/22/2010 100301
10
AK
PO13 3/22/2010 100301
20
AK
PO14 3/22/2010 100301
5
AK
PO15 3/22/2010 100301
10
AK
PO16 3/22/2010 100301
20
AK
PO17 3/22/2010 100301
5
AK
PO18 3/22/2010 100301
10
AK
PO19 3/22/2010 100301
20
AK
PO20 3/22/2010 100301
40
AK
PO1 10/4/2010 100916
2
AK
PO2 10/4/2010 100916
2
AK
PO3 10/4/2010 100916
2
AK
PO4 10/4/2010 100916
2
AK
PO5 10/4/2010 100916
5
AK
PO6 10/4/2010 100916
10
AK
PO7 10/4/2010 100916
20
AK
PO8 10/4/2010 100916
40
AK
PO9 10/4/2010 100916
40
AK
PO10 10/4/2010 100916
40
Appendix VII. Comparing Methods Data & Calculations
HPLC Area
70.8
162.4
315.6
79.7
150.9
300.3
78.3
150.4
320.7
622.5
25.4
29.1
28.6
30.4
76.4
158
294.7
633.1
516.5
642.4
66
124.3
237.7
64.7
120.1
224.6
56.9
115.3
240.3
470.2
30.5
28.2
27.5
29.2
88.5
141.9
311.6
622.3
660.6
676.9
HPLC ppm
4.3
10.0
19.4
4.9
9.3
18.4
4.8
9.2
19.7
38.2
1.6
1.8
1.8
1.9
4.7
9.7
18.1
38.8
31.7
39.4
4.0
7.6
14.6
4.0
7.4
13.8
3.5
7.1
14.7
28.8
1.9
1.7
1.7
1.8
5.4
8.7
19.1
38.2
40.5
41.5
ELISA ppm
4.0
8.7
21.8
5.1
9.7
19.4
5.6
9.8
20.4
35.7
2.1
2.5
2.0
2.0
5.5
9.5
16.7
34.7
26.9
30.4
4.7
10.1
20.9
4.8
10.7
21.0
5.3
10.2
19.3
45.6
2.47
2
2.4
2.4
4.6
8.5
17
32.3
29.8
33.1
B. Pacific oysters
86
1. Calculate the symmetry
HPLC symmetry: 1.063
ELISA symmetry: 0.915
2. Calculate the variance
HPLC variance: 172.35
ELISA variance: 140.33
3. Values for the test of symmetry outside the range of -2 to +2 indicate skewness in the distribution. N/A
4. A ratio of the larger to smaller of the variances >2 indicates a lack of homogeneity of variance. N/A
Variance HPLC/ELISA: 172.35/140.33= 1.228
5. Use either the paired t-test or Welch’s t-test for the analysis of the data based on the following
considerations.
• If the distribution of the data are symmetric (within the range of -2 to +2) and there is homogeneity
of variance use a paired t-test for the data analysis.
A paired t-test was performed using the Excel formula =TTEST(array1, array2, tails, type) where array1=
HPLC data; array2= ELISA data; tails= 2; type= 1(paired)
Paired t-test: p= 0.873
Since p > 0.05, the HPLC and ELISA values are NOT significantly different.
Data summary for the comparison of the new or modified method to the officially recognized
method:
Value for the test of symmetry for the distribution of the data generated by the HPLC method 1.063
Value for the test of symmetry for the distribution of the data generated by the ELISA method 0.915
Variance of the data generated from the HPLC method 172.35
Variance of the data generated from the ELISA method 140.33
Ratio of the larger to the smaller of the variances 1.228
Is there a significant difference between the analytical methods? No
Appendix VII. Comparing Methods Data & Calculations
B. Pacific oysters
87
C. Razor clams
Data
Sample type Razor clam tissue
Count
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Location Sample
Date
Kit Spike
WA
493 5/5/2010 100302
5
WA
494 5/5/2010 100323
10
WA
449 5/12/2010 100323
5
WA
323 5/12/2010 100323
5
WA
342 5/12/2010 100323
10
WA
492 5/12/2010 100302
40
WA
491 5/13/2010 100323
20
WA
487 5/13/2010 100323
10
WA
376 5/13/2010 100323
20
WA
468 5/13/2010 100323
40
WA
302 9/27/2010 100916
5
WA
344 9/27/2010 100916
10
WA
345 9/27/2010 100916
20
WA
375 9/27/2010 100916
2
WA
377 9/27/2010 100916
2
WA
412 9/27/2010 100916
2
WA
413 9/27/2010 100916
2
WA
423 9/27/2010 100916
40
WA
424 9/27/2010 100916
40
WA
459 9/27/2010 100916
40
AK
RC1 9/28/2010 100916
2
AK
RC2 9/28/2010 100916
5
AK
RC3 9/28/2010 100916
10
AK
RC4 9/28/2010 100916
20
AK
RC5 9/28/2010 100916
40
AK
RC6 9/28/2010 100916
2
AK
RC7 9/28/2010 100916
5
AK
RC8 9/28/2010 100916
10
AK
RC9 9/28/2010 100916
20
AK
RC10 9/28/2010 100916
40
AK
RC11 9/28/2010 100916
2
AK
RC12 9/28/2010 100916
5
AK
RC13 9/28/2010 100916
10
AK
RC14 9/28/2010 100916
20
AK
RC15 9/28/2010 100916
40
AK
RC16 9/28/2010 100916
2
AK
RC17 9/28/2010 100916
5
AK
RC18 9/28/2010 100916
10
AK
RC19 9/28/2010 100916
20
AK
RC20 9/28/2010 100916
40
Appendix VII. Comparing Methods Data & Calculations
HPLC Area
90.4
167.8
102.5
79.2
141.2
668.1
315.1
172.1
315.8
706.1
75
151.4
246.6
36
33.5
42.3
37.4
629.7
444.6
527.6
28.5
71.1
145.9
302.6
546
30.3
69.7
156.9
292.6
482.4
25.2
75.8
146.1
309.3
617.4
28.1
78.7
155.9
313.6
638.4
HPLC ppm
4.8
9.6
5.7
4.9
8.7
41.0
19.1
10.4
19.4
43.3
4.3
9.3
14.8
1.6
1.4
1.9
1.4
37.9
27.3
32.0
1.7
4.4
9.0
18.6
33.5
1.9
4.3
9.6
18.0
29.6
1.5
4.7
9.0
19.0
37.9
1.7
4.8
9.6
19.2
39.2
ELISA ppm
4.8
10.0
5.6
5.8
9.8
39.3
17.7
12.3
23.5
38.1
4.8
9.7
19.5
2.7
3.2
3.3
2.3
32.2
25.8
26.8
2.5
5.0
9.0
17.5
34.1
1.9
4.8
8.6
16.9
28.3
1.7
3.9
9.5
16.4
36.1
2.1
4.4
10.2
15.8
29.2
C. Razor clams
88
6. Calculate the symmetry
HPLC symmetry: 0.932
ELISA symmetry: 0.856
7. Calculate the variance
HPLC variance: 172.07
ELISA variance: 133.81
8. Values for the test of symmetry outside the range of -2 to +2 indicate skewness in the distribution. N/A
9. A ratio of the larger to smaller of the variances >2 indicates a lack of homogeneity of variance. N/A
Variance HPLC/ELISA: 172.07/133.81= 1.286
10. Use either the paired t-test or Welch’s t-test for the analysis of the data based on the following
considerations.
• If the distribution of the data are symmetric (within the range of -2 to +2) and there is homogeneity
of variance use a paired t-test for the data analysis.
A paired t-test was performed using the Excel formula =TTEST(array1, array2, tails, type) where array1=
HPLC data; array2= ELISA data; tails= 2; type= 1(paired)
Paired t-test: p= 0.215
Since p > 0.05, the HPLC and ELISA values are NOT significantly different.
Data summary for the comparison of the new or modified method to the officially recognized
method:
Value for the test of symmetry for the distribution of the data generated by the HPLC method 0.932
Value for the test of symmetry for the distribution of the data generated by the ELISA method 0.856
Variance of the data generated from the HPLC method 172.07
Variance of the data generated from the ELISA method 133.81
Ratio of the larger to the smaller of the variances 1.286
Is there a significant difference between the analytical methods? No
Appendix VII. Comparing Methods Data & Calculations
C. Razor clams
89
D. Blue mussels
Data
Sample type Butter clam tissue
Count Location Sample
Date
Kit Spike
1
WA
898 4/29/2010 100302
5
2
WA
899 4/29/2010 100302
10
3
WA
924 4/29/2010 100302
20
4
WA
928 4/30/2010 100302
5
5
WA
942 4/30/2010 100302
10
6
WA
945 4/30/2010 100302
20
7
WA
946 4/30/2010 100302
5
8
WA
957 4/30/2010 100302
10
9
WA
513 4/30/2010 100302
20
10
WA
557 4/30/2010 100302
40
11
WA
804 9/30/2010 100916
2
12
WA
854 9/30/2010 100916
2
13
WA
857 9/30/2010 100916
2
14
WA
873 9/30/2010 100916
2
15
WA
874 9/30/2010 100916
5
16
WA
875 9/30/2010 100916
10
17
WA
876 9/30/2010 100916
20
18
WA
884 9/30/2010 100916
40
19
WA
886 9/30/2010 100916
40
20
WA
897 9/30/2010 100916
40
1
AK BM11 4/21/2010 100302
5
2
AK BM12 4/21/2010 100302
10
3
AK BM13 4/21/2010 100302
20
4
AK BM14 4/21/2010 100302
5
5
AK BM15 4/21/2010 100302
10
6
AK BM16 4/21/2010 100302
20
7
AK BM17 4/21/2010 100302
5
8
AK BM18 4/21/2010 100302
10
9
AK BM19 4/21/2010 100302
20
10
AK BM20 4/21/2010 100302
40
11
AK
BM1 10/4/2010 100916
2
12
AK
BM2 10/4/2010 100916
2
13
AK
BM3 10/4/2010 100916
2
14
AK
BM4 10/4/2010 100916
2
15
AK
BM5 10/4/2010 100916
5
16
AK
BM6 10/4/2010 100916
10
17
AK
BM7 10/4/2010 100916
20
18
AK
BM8 10/4/2010 100916
40
19
AK
BM9 10/4/2010 100916
40
20
AK BM10 10/4/2010 100916
40
Appendix VII. Comparing Methods Data & Calculations
HPLC Area
83.8
157.5
298.1
65.7
164.8
315.1
67.6
167.8
332.5
652.3
32.9
35.2
32.6
32.2
87.9
164.9
306.0
655.1
656.4
512.2
72.5
137.2
232.1
66.3
132.4
258.5
74.6
125
255
527.1
33.1
32.5
31.9
32.7
81.6
162.6
330.1
681.7
653.8
685.3
HPLC ppm
5.1
9.7
18.3
4.0
10.1
19.3
4.1
10.3
20.4
40.0
2.0
2.2
2.0
2.0
5.4
10.1
18.8
40.2
40.3
31.4
4.4
8.4
14.2
4.1
8.1
15.9
4.6
7.7
15.6
32.3
2.0
2.0
2.0
2.0
5.0
10.0
20.3
41.8
40.1
42.0
ELISA ppm
6.3
10.7
21.0
4.9
10.6
17.2
4.9
8.6
16.5
38.0
2.2
2.4
2.9
2.7
4.7
10
18.8
32.2
33.6
27.7
6.5
13.9
23.3
6.9
13.3
24.6
7.2
12.9
24.5
47.2
2.2
2.4
2.3
2.2
5.0
8.4
14.6
31.6
28.8
29.5
D. Blue mussels
90
1. Calculate the symmetry
HPLC symmetry: 1.052
ELISA symmetry: 0.881
2. Calculate the variance
HPLC variance: 182.68
ELISA variance: 144.56
3. Values for the test of symmetry outside the range of -2 to +2 indicate skewness in the distribution. N/A
4. A ratio of the larger to smaller of the variances >2 indicates a lack of homogeneity of variance. N/A
Variance HPLC/ELISA: 182.68/144.56= 1.264
5. Use either the paired t-test or Welch’s t-test for the analysis of the data based on the following
considerations.
• If the distribution of the data are symmetric (within the range of -2 to +2) and there is homogeneity
of variance use a paired t-test for the data analysis.
A paired t-test was performed using the Excel formula =TTEST(array1, array2, tails, type) where array1=
HPLC data; array2= ELISA data; tails= 2; type= 1(paired)
Paired t-test: p= 0.699
Since p > 0.05, the HPLC and ELISA values are NOT significantly different.
Data summary for the comparison of the new or modified method to the officially recognized
method:
Value for the test of symmetry for the distribution of the data generated by the HPLC method 1.052
Value for the test of symmetry for the distribution of the data generated by the ELISA method 0.881
Variance of the data generated from the HPLC method 182.68
Variance of the data generated from the ELISA method 144.56
Ratio of the larger to the smaller of the variances 1.264
Is there a significant difference between the analytical methods? No
Appendix VII. Comparing Methods Data & Calculations
D. Blue mussels
91
Appendix VIII. Limit of Quantitation/Limit of Detection Data and Calculations
VII. #5 Marine Biotoxin and Non-MPN Based Microbiological Methods SOP – Linear Range, Limit of
Detection, Limit of Quantitation/Sensitivity
VALIDATION CRITERIA
Linear Range is the range within the working range where the results are proportional to the
concentration of the analyte/measurand/organism of interest present in the sample.
Limit of Detection is the minimum concentration at which the analyte/measurand/organism of interest
can be identified.
Limit of Quantitation/Sensitivity is the minimum concentration of the analyte/measurand/organism of
interest that can be quantified with an acceptable level of precision and accuracy under the conditions of
the test.
Procedure: This procedure is applicable for use with either growing waters or shellfish tissue. Make
every effort to use samples free of the target analyte/measurand/organism of interest. For each shellfish
type of interest use a minimum of 10-12 animals per sample. For each sample take at least six (6) aliquots
of either the growing water sample or shellfish homogenate appropriately sized for your work and spike
a
b
n
five (5) of the six (6) aliquots with five (5) different concentrations (i.e. 10 , 10 …10 ) of the target
analyte/measurand/organism of interest spanning 50 – 150% of the working range/range of interest for the
method under study. Do not spike the sixth or last aliquot of each sample. This is the sample blank. For
microbiological methods determine the concentration of the target analyte/measurand/organism of interest
used to spike each aliquot of each sample by plating in/on appropriate agar. Do not use aliquots of the
same master solution/culture to spike all the samples in this exercise. A separate master solution /culture
should be used for each sample. Process each aliquot including the sample blank as usual to determine
method concentration for the target analyte/measurand/organism of interest. Do three (3) replicates for
each aliquot excluding the sample blank. Do only one blank per sample. For growing waters do ten (10)
samples collected from a variety of growing areas. For shellfish do ten (10) samples for each shellfish
tissue type of interest collected from a variety of growing areas, the same growing area harvested on
different days or from different process lots. Use the same spiking levels for each of the ten (10) samples
a
b
n
analyzed (10 , 10 …10 ).
Data:
Sample type _________
Working range/Range of interest ____________
Range in spiking levels used __________________
Agar used to determine spike concentration _____________________
Organism used for spiking _________________________________
Aliquot 0* 1 2 3 4 5
Sample 1
Spike conc. /plate count
Response, replicate 1
Response, replicate 2
Response, replicate 3
Aliquot 0* 1 2 3 4 5
Sample 2
Spike conc. /plate count
Response, replicate 1
Response, replicate 2
Response, replicate 3
Sample 3
Spike conc. /plate count
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations
A. Butter clams
92
Response, replicate 1
Response, replicate 2
Response, replicate 3
Sample 4
Spike conc. /plate count
Response, replicate 1
Response, replicate 2
Response, replicate 3
Sample 5
Spike conc. /plate count
Response, replicate 1
Response, replicate 2
Response, replicate 3
Sample 6
Spike conc. /plate count
Response, replicate 1
Response, replicate 2
Response, replicate 3
Sample 7
Spike conc. /plate count
Response, replicate 1
Response, replicate 2
Response, replicate 3
Sample 8
Spike conc. /plate count
Response, replicate 1
Response, replicate 2
Response, replicate 3
*
Aliquot 0 1 2 3 4 5
Sample 9
Spike conc. /plate count
Response, replicate 1
Response, replicate 2
Response, replicate 3
Sample 10
Spike conc. /plate count
Response, replicate 1
Response, replicate 2
Response, replicate 3
* Unspiked sample blank
Response is the signal data (absorbance, florescence, Ct value), colonies, plaques, etc resulting from
the analysis.
For shellfish samples repeat for each tissue type of interest.
DATA HANDLING
Limit of Detection and Limit of Quantitation/Sensitivity
To determine the minimum concentration at which the analyte/measurand/organism of interest can be
identified and subsequently quantified with an acceptable level of precision and accuracy under the
conditions of the test, the data is manipulated in the following manner.
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations
A. Butter clams
93
1. Calculate the coefficient of variation or relative standard deviation for each concentration of
analyte/measurand/organism of interest spiked into the samples. Use the log transformed data
for manipulating microbiological results.
2. Plot the coefficient of variation/relative standard deviation on the y-axis for each concentration
of analyte/measurand/organism of interest spiked into the samples and plotted on the x-axis.
Use log transformed concentration values for the microbiological data.
3. Fit the curve and determine from the graph the concentration of analyte/measurand/organism of
interest which gave rise to a coefficient of variation/relative standard deviation of 10%. This
is the limit of quantitation/sensitivity of the method as implemented by the laboratory.
4. Divide the value for the limit of quantitation/sensitivity obtained from step 3 above by 3.3 or
determine the concentration of analyte/measurand/organism of interest that gave rise to a
coefficient of variation/relative standard deviation of 33%. This value is the limit of detection
of the method as implemented by the laboratory.
For single laboratory validation, the concepts of “blank + 3σ” and “blank + 10σ” generally suffice for
determining the limit of detection and the limit of quantitation/sensitivity. Since the blank is in theory
zero (0), then the limit of detection and the limit of quantitation /sensitivity become 3σ and 10σ
respectively. An absolute standard deviation of 3 and 10 equates to a coefficient of variation/relative
standard deviation of 33% and 10% respectively. Accordingly the limit of detection and the limit of
quantitation/sensitivity become the concentration of analyte/measurand/organism of interest which give
rise to these values.
ALTERNATIVE DATA HANDLING METHOD
For several species, the LOQ and LOD that were calculated based on the curve were negative numbers.
For these cases, an alternative method was used to calculate the LOD and LOQ. For that method, the
values for each concentration were plotted on the y axis with the spike value on the x axis. A linear
regression was performed to find the equation of the line. The standard error of the y intercept was
calculated using Excel. This value represents sigma. The limit of quantitation was determined by using
the y intercept of equation of the line and adding 10 times sigma (If the equation of the line is y=mx+b,
LOQ= b+10*sigma). The limit of detection was determined by adding 3 times sigma to the y intercept.
Data Summary:
The limit of detection of the method as implemented ______________
The limit of quantitation/sensitivity of the method as implemented ____________
A. Butter Clams
Data
Sample type Shellfish tissue
Working range/Range of interest 2 to 40 ppm
Range in spiking levels used 2 to 40 ppm
Agar used to determine spike concentration Not applicable
Organism used for spiking Butter clams
Count Sample # Date
Kit
Spike Rep 1 Rep 2
1
1543 9/15/2010
100323
2.0
3.5
4.0
2
1418 9/15/2010
100323
2.0
2.8
2.5
3
1508 9/15/2010
100323
2.0
3.3
2.3
4
1932 9/15/2010
100323
2.0
2.5
2.9
5
1660 9/15/2010
100323
2.0
2.1
3.1
CV
0.090
0.085
0.248
0.116
0.259
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations
A. Butter clams
94
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1830
1371
1521
1949
1364
1543
1418
1508
1932
1660
1830
1371
1521
1949
1364
1543
1418
1508
1932
1660
1830
1371
1521
1949
1364
1543
1418
1508
1932
1660
1830
1371
1521
1949
1364
1543
1418
1508
1932
1660
1830
1371
1521
1949
1364
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
3/10/2010
3/17/2010
3/9/2010
3/17/2010
3/18/2010
3/18/2010
3/19/2010
3/19/2010
3/19/2010
3/22/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
3/10/2010
3/17/2010
3/9/2010
3/17/2010
3/18/2010
3/18/2010
3/19/2010
3/19/2010
3/19/2010
3/22/2010
3/10/2010
3/17/2010
3/9/2010
3/17/2010
3/18/2010
3/18/2010
3/19/2010
3/19/2010
3/19/2010
3/22/2010
100323
100323
100323
100323
100323
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
100304R
2.0
2.0
2.0
2.0
2.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
20.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
2.5
2.1
2.5
2.0
1.9
4.2
6.5
5.6
5.5
4.4
4.0
5.4
4.9
3.9
5.8
14.7
12.0
11.4
12.7
11.1
12.8
10.4
9.5
9.4
10.3
11.5
25.9
21.5
19.5
23.1
21.3
26.9
26.6
23.3
27.0
28.5
53.3
43.0
42.6
43.1
38.3
54.3
43.5
38.2
42.5
2.2
1.9
2.1
2.1
2.1
3.4
6.1
6.0
5.5
4.8
9.7
5.7
5.5
4.3
5.2
11.5
11.6
10.9
12.2
12.4
13.1
11.6
10.0
9.3
11.2
14.0
23.8
22.4
19.9
20.0
19.3
25.2
20.0
20.2
20.6
29.7
51.7
42.3
47.2
46.5
39.3
51.4
43.9
44.4
46.7
0.089
0.081
0.114
0.033
0.080
0.152
0.044
0.050
0.001
0.069
0.598
0.042
0.078
0.074
0.079
0.174
0.023
0.027
0.030
0.079
0.020
0.077
0.032
0.012
0.059
0.134
0.058
0.030
0.015
0.102
0.068
0.046
0.201
0.104
0.192
0.029
0.022
0.012
0.072
0.053
0.019
0.039
0.007
0.106
0.066
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations
A. Butter clams
95
Equation of line: y= -0.0017x+0.1126
Solve equation for x where y=0.1 (10% CV)
0.1 = -0.0017x+0.1126
-0.0126 = -0.0017 x
7.41 = x = Limit of Quantitation (LOQ)
Divide LOQ by 3.3 to calculate Limit of Detection (LOD)
7.41/3.3= 2.25
Data Summary:
The limit of detection of the method as implemented 2.25
The limit of quantitation/sensitivity of the method as implemented 7.41
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations
A. Butter clams
96
B. Pacific Oyster
Data
Sample type Shellfish tissue
Working range/Range of interest 2 to 40 ppm
Range in spiking levels used 2 to 40 ppm
Agar used to determine spike concentration Not applicable
Organism used for spiking Pacific oyster
Count Sample # Date
Kit
Spike Rep 1
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
2115
2116
2117
2118
2119
1816
1817
2165
2175
2176
2115
2116
2117
2118
2119
1816
1817
2165
2175
2176
2115
2116
2117
2118
2119
1816
1817
2165
2175
2176
2115
2116
2117
2118
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
3/26/2010
3/30/2010
3/30/2010
3/31/2010
3/31/2010
3/31/2010
4/1/2010
4/1/2010
4/1/2010
4/1/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
3/26/2010
3/30/2010
3/30/2010
3/31/2010
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100302
100302
100302
100302
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
20.0
20.0
20.0
20.0
2.2
2.3
2.8
2.4
2.7
2.4
2.2
2.2
2.6
2.2
4.8
5.7
3.8
6.0
5.5
4.0
5.9
6.4
3.6
4.0
9.4
12.2
7.7
8.0
10.4
9.4
9.8
10.1
9.0
8.3
23.6
20.0
22.9
20.4
Rep 2
2.2
2.1
2.1
3.0
1.2
2.6
2.0
2.1
2.1
2.1
6.8
5.8
5.0
6.8
4.7
4.3
6.1
6.2
3.9
4.1
10.0
8.9
8.3
8.3
9.7
9.5
9.8
10.7
9.3
7.9
22.0
20.9
24.9
21.0
CV
0.004
0.048
0.217
0.145
0.544
0.043
0.062
0.047
0.135
0.035
0.244
0.010
0.190
0.087
0.124
0.063
0.024
0.016
0.051
0.016
0.046
0.221
0.049
0.023
0.047
0.009
0.002
0.042
0.025
0.034
0.051
0.033
0.061
0.020
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations B. Pacific Oyster
97
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
2119
1816
1817
2165
2175
2176
2115
2116
2117
2118
2119
1816
1817
2165
2175
2176
3/31/2010
3/31/2010
4/1/2010
4/1/2010
4/1/2010
4/1/2010
3/26/2010
3/30/2010
3/30/2010
3/31/2010
3/31/2010
3/31/2010
4/1/2010
4/1/2010
4/1/2010
4/1/2010
Coefficients
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
Standard Error
20.0
20.0
20.0
20.0
20.0
20.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
t Stat
21.3
18.8
24.4
22.6
20.0
15.7
34.5
45.2
39.3
44.6
39.9
34.7
38.2
39.0
38.2
36.8
21.0
17.6
22.3
21.4
18.9
17.3
42.8
42.9
38.0
40.7
41.2
38.6
40.6
41.5
40.9
40.4
P-value
0.009
0.046
0.062
0.038
0.038
0.069
0.152
0.038
0.024
0.065
0.023
0.076
0.042
0.043
0.048
0.067
Lower
95%
Intercept
0.1090
0.2764
0.3945
0.6941 -0.4395
X
Variable
0.9997
0.0134
74.6343
0.0000
0.9732
Equation of line: y= -0.997x+0.109
y intercept= 0.109
Sigma= Standard Error of the y intercept= 0.2764
Limit of Quantitation = y intercept + 10*sigma = 0.109+10*0.2764 = 2.87
Limit of Detection = y intercept + 3*sigma= 0.109+3*0.2764 = 0.94
Upper
95%
Lower
95.0%
Upper
95.0%
0.6576
-0.4395
0.6576
1.0263
0.9732
1.0263
Data Summary:
The limit of detection of the method as implemented 0.94
The limit of quantitation/sensitivity of the method as implemented 2.87
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations B. Pacific Oyster
98
C. Geoduck
Data
Sample type Shellfish tissue
Working range/Range of interest 2 to 40 ppm
Range in spiking levels used 2 to 40 ppm
Agar used to determine spike concentration Not applicable
Organism used for spiking Geoduck
Count Sample # Date
Kit
Spike Rep 1 Rep 2
1
1288 10/1/2010 100916
2.0
3.0
3.3
2
1289 10/1/2010 100916
2.0
3.3
3.3
3
1342 10/1/2010 100916
2.0
3.8
3.3
4
1358 10/1/2010 100916
2.0
3.0
3.2
5
1425 10/1/2010 100916
2.0
3.0
2.8
6
1428 10/1/2010 100916
2.0
2.9
2.6
7
1430 10/1/2010 100916
2.0
3.9
4.0
8
1485 10/1/2010 100916
2.0
3.5
3.4
9
1487 10/1/2010 100916
2.0
2.4
2.1
10
1566 10/1/2010 100916
2.0
2.8
1.6
1
1288 4/7/2010 100302
5.0
4.8
4.7
2
1289 4/9/2010 100302
5.0
5.5
6.1
3
1342 4/9/2010 100302
5.0
4.4
6.1
4
1358 4/9/2010 100302
5.0
4.6
5.7
5
1425 4/9/2010 100302
5.0
5.0
5.5
6
1428 4/12/2010 100302
5.0
4.4
4.4
7
1430 4/12/2010 100302
5.0
5.3
6.8
8
1485 4/12/2010 100302
5.0
4.5
5.1
9
1487 4/12/2010 100302
5.0
4.2
5.4
10
1566 4/12/2010 100302
5.0
4.4
4.2
1
1288 10/4/2010 100916
10.0
11.6
11.0
2
1289 10/4/2010 100916
10.0
8.9
8.5
3
1342 10/4/2010 100916
10.0
12.1
12.3
4
1358 10/4/2010 100916
10.0
11.2
11.5
5
1425 10/4/2010 100916
10.0
10.4
11.0
6
1428 10/4/2010 100916
10.0
9.9
10.9
7
1430 10/4/2010 100916
10.0
14.5
13.4
8
1485 10/4/2010 100916
10.0
10.0
8.3
9
1487 10/4/2010 100916
10.0
7.9
8.6
10
1566 10/4/2010 100916
10.0
8.6
9.2
1
1288 4/7/2010 100302
20.0
18.9
17.6
2
1289 4/9/2010 100302
20.0
21.3
21.2
3
1342 4/9/2010 100302
20.0
23.5
20.2
4
1358 4/9/2010 100302
20.0
18.2
18.7
CV
0.078
0.004
0.089
0.066
0.060
0.061
0.024
0.017
0.084
0.377
0.021
0.082
0.228
0.144
0.064
0.010
0.174
0.086
0.180
0.025
0.036
0.034
0.014
0.018
0.041
0.069
0.056
0.126
0.053
0.054
0.047
0.001
0.105
0.016
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations
C. Geoduck
99
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1425
1428
1430
1485
1487
1566
1288
1289
1342
1358
1425
1428
1430
1485
1487
1566
4/9/2010
4/12/2010
4/12/2010
4/12/2010
4/12/2010
4/12/2010
4/7/2010
4/9/2010
4/9/2010
4/9/2010
4/9/2010
4/12/2010
4/12/2010
4/12/2010
4/12/2010
4/12/2010
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
100302
20.0
20.0
20.0
20.0
20.0
20.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
21.6
22.4
21.6
22.5
19.7
19.0
35.9
43.8
34.1
40.6
37.2
40.9
30.2
37.9
36.0
33.2
18.5
20.7
16.0
18.0
15.7
18.7
35.2
47.0
40.1
41.2
38.4
42.3
37.3
40.7
38.7
40.1
Standard
Error
0.3246
0.111
0.056
0.210
0.157
0.160
0.013
0.016
0.050
0.113
0.011
0.023
0.024
0.149
0.051
0.051
0.134
Coefficients
t Stat
P-value
Lower 95%
Intercept
0.8689
2.6764
0.0087
0.2246
X
Variable
0.9415
0.0157 59.8415
0.0000
0.9103
Equation of line: y= 0.9415x+ 0.8689
y intercept= 0.8689
Sigma= Standard Error of the y intercept= 0.3246
Limit of Quantitation = y intercept + 10*sigma = 0.8689+10*0.3246 = 4.12
Limit of Detection = y intercept + 3*sigma= 0. 8689+3*0.3246 = 1.84
Upper
95%
1.5131
Lower
95.0%
0.2246
Upper
95.0%
1.5131
0.9727
0.9103
0.9727
Data Summary:
The limit of detection of the method as implemented 1.84
The limit of quantitation/sensitivity of the method as implemented 4.12
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations
C. Geoduck
100
D. Razor clam
Data
Sample type Shellfish tissue
Working range/Range of interest 2 to 40 ppm
Range in spiking levels used 2 to 40 ppm
Agar used to determine spike concentration Not applicable
Organism used for spiking Razor clams
Count Sample # Date
Kit
Spike Rep 1
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
302
344
345
375
377
412
413
423
424
459
302
344
345
375
377
412
413
423
424
459
302
344
345
375
377
412
413
423
424
459
302
344
345
9/27/2010
9/27/2010
9/27/2010
9/27/2010
9/27/2010
9/27/2010
9/27/2010
9/27/2010
9/27/2010
9/27/2010
5/3/2010
5/3/2010
5/4/2010
5/4/2010
5/5/2010
5/5/2010
5/7/2010
5/7/2010
5/11/2010
5/11/2010
9/27/2010
9/27/2010
9/27/2010
9/27/2010
9/27/2010
9/27/2010
9/27/2010
9/27/2010
9/27/2010
9/27/2010
5/3/2010
5/3/2010
5/4/2010
100916
100916
100916
100916
100916
100916
100916
100916
100916
100916
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100916
100916
100916
100916
100916
100916
100916
100916
100916
100916
100323
100323
100323
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
20.0
20.0
20.0
1.7
1.5
2.4
3.0
2.7
3.3
2.2
1.9
3.1
1.5
6.0
4.3
5.4
5.8
4.5
4.7
5.7
3.8
6.0
4.6
12.9
7.4
8.3
10.3
9.0
8.7
8.5
9.7
8.8
8.1
21.2
20.3
17.9
Rep 2
1.9
1.9
2.7
3.0
3.4
2.4
2.7
2.3
2.0
1.1
6.1
4.8
5.6
6.5
5.3
4.5
6.3
4.9
6.9
5.5
13.6
7.2
8.7
9.5
9.1
8.9
8.2
9.5
8.8
8.2
19.3
18.2
16.5
CV
0.075
0.154
0.096
0.019
0.166
0.236
0.160
0.135
0.318
0.204
0.012
0.086
0.032
0.080
0.120
0.031
0.068
0.166
0.096
0.123
0.037
0.013
0.037
0.055
0.006
0.012
0.026
0.017
0.002
0.007
0.068
0.080
0.059
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations
D. Razor clams
101
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
375
377
412
413
423
424
459
302
344
345
375
377
412
413
423
424
459
5/4/2010
5/5/2010
5/5/2010
5/7/2010
5/7/2010
5/11/2010
5/11/2010
5/3/2010
5/3/2010
5/4/2010
5/4/2010
5/5/2010
5/5/2010
5/7/2010
5/7/2010
5/11/2010
5/11/2010
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
20.0
20.0
20.0
20.0
20.0
20.0
20.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
17.4
16.9
19.5
22.2
20.8
22.5
21.2
34.6
32.8
31.9
32.4
34.1
37.2
37.9
31.9
36.6
36.3
15.8
16.9
20.0
18.4
21.4
19.6
19.3
36.4
38.0
35.7
35.4
39.8
35.8
43.9
33.7
37.5
42.1
0.069
0.000
0.020
0.130
0.021
0.096
0.066
0.036
0.103
0.079
0.064
0.109
0.028
0.105
0.037
0.017
0.105
Lower
95%
Upper
95%
Lower
95.0%
Upper
95.0%
0.1406
1.2682
0.1406
1.2682
X Variable
0.8934
0.0138 64.8915
0.0000
0.8661
Equation of line: y= 0.8934x + 0.7044
y intercept= 0.7044
Sigma= Standard Error of the y intercept= 0.2841
Limit of Quantitation = y intercept + 10*sigma = 0.7044+10*0.2841 =3.55
Limit of Detection = y intercept + 3*sigma= 0.7044+3*0.2841 = 1.56
0.9207
0.8661
0.9207
Intercept
Coefficients
Standard Error
0.7044
0.2841
t Stat
2.4795
P-value
0.0149
Data Summary:
The limit of detection of the method as implemented 1.56
The limit of quantitation/sensitivity of the method as implemented 3.55
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations
D. Razor clams
102
E. Blue mussel
Data
Sample type Shellfish tissue
Working range/Range of interest 2 to 40 ppm
Range in spiking levels used 2 to 40 ppm
Agar used to determine spike concentration Not applicable
Organism used for spiking Blue mussel
Count Sample # Date
Kit
Spike Rep 1
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
804
854
857
873
874
875
876
884
886
897
804
854
857
873
874
875
876
884
886
897
804
854
857
873
874
875
876
884
886
897
804
854
857
873
9/7/2010
9/7/2010
9/7/2010
9/8/2010
9/8/2010
9/8/2010
9/8/2010
9/8/2010
9/8/2010
9/8/2010
4/19/2010
4/23/2010
4/23/2010
4/23/2010
4/23/2010
4/26/2010
4/26/2010
4/26/2010
4/27/2010
4/27/2010
9/7/2010
9/7/2010
9/7/2010
9/8/2010
9/8/2010
9/8/2010
9/8/2010
9/8/2010
9/8/2010
9/8/2010
4/19/2010
4/23/2010
4/23/2010
4/23/2010
100503
100503
100503
100503
100503
100503
100503
100503
100503
100503
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100503
100503
100503
100503
100503
100503
100503
100503
100503
100503
100323
100323
100323
100323
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
20.0
20.0
20.0
20.0
1.4
2.4
2.5
1.8
1.8
1.8
0.9
1.8
1.1
0.9
6.6
4.8
3.9
5.3
6.2
4.4
5.4
4.7
5.4
4.0
11.2
11.6
12.0
10.0
9.6
10.5
11.0
11.0
10.1
10.8
16.8
19.9
21.5
17.8
Rep 2
1.7
2.9
1.9
1.5
1.7
1.7
1.1
2.3
2.0
0.9
6.7
4.4
4.2
5.5
5.9
4.7
5.5
5.9
5.0
5.6
12.4
12.3
12.4
9.9
11.2
11.1
10.0
12.2
10.5
10.9
15.5
22.4
21.1
17.0
CV
0.145
0.143
0.207
0.114
0.052
0.042
0.129
0.162
0.433
0.043
0.014
0.054
0.048
0.026
0.034
0.039
0.021
0.152
0.046
0.238
0.068
0.040
0.022
0.009
0.105
0.039
0.065
0.073
0.029
0.005
0.058
0.082
0.013
0.036
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations
E. Blue mussel
103
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
874
875
876
884
886
897
804
854
857
873
874
875
876
884
886
897
4/23/2010
4/26/2010
4/26/2010
4/26/2010
4/27/2010
4/27/2010
4/19/2010
4/23/2010
4/23/2010
4/23/2010
4/23/2010
4/26/2010
4/26/2010
4/26/2010
4/27/2010
4/27/2010
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
20.0
20.0
20.0
20.0
20.0
20.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
17.3
19.1
19.6
15.1
20.7
18.8
41.1
38.0
36.9
33.7
35.1
30.5
33.3
33.1
33.9
32.2
15.2
16.8
18.1
14.0
18.5
19.6
41.3
40.0
35.8
33.5
37.7
32.4
35.0
31.4
35.2
33.0
0.092
0.089
0.059
0.054
0.082
0.027
0.004
0.037
0.020
0.003
0.050
0.044
0.035
0.039
0.027
0.017
Equation of line: y= -0.0021x+0.1014
Solve equation for x where y=0.1 (10% CV)
0.1 = -0.0021x+0.1014
-0.0014 = -0.0021 x
0.67 = x = Limit of Quantitation (LOQ)
Divide LOQ by 3.3 to calculate Limit of Detection (LOD)
0.67/3.3= 0.202
Data Summary:
The limit of detection of the method as implemented 0.20
The limit of quantitation/sensitivity of the method as implemented 0.67
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations
E. Blue mussel
104
F. Atlantic Oyster
Data
Sample type Shellfish tissue
Working range/Range of interest 2 to 40 ppm
Range in spiking levels used 2 to 40 ppm
Agar used to determine spike concentration Not applicable
Organism used for spiking Atlantic oyster
Count Sample # Date
Kit
Spike Rep 1
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
3
12
13
19
20
21
22
25
26
30
3
12
13
19
20
21
22
25
26
30
3
12
13
19
20
21
22
25
26
30
3
12
13
19
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
5/14/2010
5/14/2010
5/17/2010
5/17/2010
5/18/2010
5/18/2010
5/18/2010
5/19/2010
5/19/2010
5/19/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
9/15/2010
5/14/2010
5/14/2010
5/17/2010
5/17/2010
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
20.0
20.0
20.0
20.0
1.5
1.1
0.9
1.8
3.2
1.9
1.7
2.1
1.5
2.7
4.3
5.2
5.3
5.3
6.3
6.1
5.1
5.0
4.0
5.2
8.4
8.1
10.0
10.1
8.9
10.1
9.6
8.1
11.5
10.7
17.5
19.9
25.2
19.1
Rep 2
1.0
0.7
1.7
2.5
2.7
1.6
2.4
1.6
1.3
3.1
4.9
5.0
6.1
5.7
6.1
4.7
5.0
4.9
3.8
5.2
9.8
8.8
9.1
10.3
9.0
9.8
9.7
8.7
11.3
9.8
19.1
21.6
19.7
20.7
CV
0.247
0.324
0.455
0.246
0.103
0.147
0.235
0.200
0.091
0.110
0.097
0.024
0.102
0.044
0.017
0.178
0.017
0.010
0.029
0.004
0.108
0.056
0.066
0.015
0.014
0.023
0.006
0.048
0.010
0.062
0.062
0.060
0.175
0.056
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations F. Atlantic Oyster
105
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
20
21
22
25
26
30
3
12
13
19
20
21
22
25
26
30
5/18/2010
5/18/2010
5/18/2010
5/19/2010
5/19/2010
5/19/2010
5/14/2010
5/14/2010
5/17/2010
5/17/2010
5/18/2010
5/18/2010
5/18/2010
5/19/2010
5/19/2010
5/19/2010
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
100323
20.0
20.0
20.0
20.0
20.0
20.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
40.0
20.0
18.5
19.9
24.0
18.1
17.6
41.9
33.4
38.8
36.7
33.6
43.1
37.4
44.3
48.4
39.6
20.5
17.2
17.1
23.5
18.8
18.6
34.7
37.8
39.7
33.6
35.2
43.1
40.2
47.4
45.1
39.5
0.019
0.052
0.106
0.015
0.029
0.039
0.131
0.089
0.016
0.062
0.034
0.001
0.051
0.048
0.049
0.001
Equation of line: y= -0.0023x+0.1194
Solve equation for x where y=0.1 (10% CV)
0.1 = -0.0023x+0.1194
-0.0194 = -0.0023 x
8.43 = x = Limit of Quantitation (LOQ)
Divide LOQ by 3.3 to calculate Limit of Detection (LOD)
8.43/3.3= 2.56
Data Summary:
The limit of detection of the method as implemented 2.56
The limit of quantitation/sensitivity of the method as implemented 8.43
Appendix VIII. Limit of Quantitation /Limit of Detection Data & Calculations F. Atlantic Oyster
106
G. Manila Clams
Data
Sample type Shellfish tissue
Working range/Range of interest 2 to 40 ppm
Range in spiking levels used 2 to 40 ppm
Agar used to determine spike concentration Not applicable
Organism used for spiking Manila clams
Count Sample #
Date
Kit
Spike Rep 1 Rep 2
1
329
5/20/2010 100323 5.0
4.9
5.4
2
390
5/20/2010 100323 5.0
4.5
4.7
3
582
5/20/2010 100323 5.0
6.8
7.8
4
630
5/20/2010 100323 5.0
4.3
5.1
5
2111
5/20/2010 100323 5.0
6.0
6.2
6
2332
5/20/2010 100323 5.0
5.1
5.8
7
1052
5/20/2010 100323 5.0
5.9
5.3
8
2258
5/20/2010 100323 5.0
5.8
6.2
9
2523
5/20/2010 100323 5.0
5.2
5.5
10
2395
5/20/2010 100323 5.0
4.6
5.5
1
329
5/20/2010 100323 20.0
25.2
20.2
2
390
5/20/2010 100323 20.0
20.7
19.4
3
582
5/20/2010 100323 20.0
22.6
18.7
4
630
5/20/2010 100323 20.0
24.3
26.2
5
2111
5/20/2010 100323 20.0
20.0
17.3
6
2332
5/20/2010 100323 20.0
20.1
18.5
7
1052
5/20/2010 100323 20.0
24.2
19.3
8
2258
5/20/2010 100323 20.0
23.7
21.9
9
2523
5/20/2010 100323 20.0
24.3
19.6
10
2395
5/20/2010 100323 20.0
21.0
18.8
1
329
5/20/2010 100323 40.0
39.6
42.7
2
390
5/20/2010 100323 40.0
43.7
51.7
3
582
5/20/2010 100323 40.0
39.0
39.0
4
630
5/20/2010 100323 40.0
46.3
42.9
5
2111
5/20/2010 100323 40.0
39.3
39.6
6
2332
5/20/2010 100323 40.0
40.5
40.8
7
1052
5/20/2010 100323 40.0
41.4
44.3
8
2258
5/20/2010 100323 40.0
46.2
50.3
9
2523
5/20/2010 100323 40.0
38.2
42.2
10
2395
5/20/2010 100323 40.0
37.4
38.1
1
329
9/9/2010 100503 2.0
1.6
1.7
2
390
9/8/2010 100503 2.0
2.2
1.7
3
582
9/8/2010 100503 2.0
2.2
1.9
4
630
9/8/2010 100503 2.0
5.1
2.1
5
2111
9/8/2010 100503 2.0
2.7
1.8
Appendix IX. Definitions
CV
0.071
0.028
0.097
0.119
0.016
0.084
0.071
0.047
0.050
0.128
0.156
0.045
0.134
0.054
0.100
0.059
0.161
0.057
0.149
0.076
0.052
0.119
0.000
0.053
0.005
0.005
0.048
0.060
0.071
0.013
0.036
0.167
0.114
0.579
0.300
107
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
2332
1052
2258
2523
2395
329
390
582
630
2111
2332
1052
2258
2523
2395
9/8/2010
9/8/2010
9/9/2010
9/8/2010
9/8/2010
9/9/2010
9/8/2010
9/8/2010
9/8/2010
9/8/2010
9/8/2010
9/8/2010
9/9/2010
9/8/2010
9/8/2010
100503
100503
100503
100503
100503
100503
100503
100503
100503
100503
100503
100503
100503
100503
100503
2.0
2.0
2.0
2.0
2.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10.0
1.7
3.0
1.3
2.5
2.0
11.3
10.4
11.9
12.0
12.8
11.8
12.1
10.3
11.8
11.3
1.8
2.5
1.5
4.0
1.9
11.0
10.2
11.4
12.9
12.0
10.9
11.7
10.2
11.9
13.4
0.048
0.115
0.103
0.336
0.056
0.017
0.012
0.029
0.047
0.043
0.056
0.022
0.009
0.006
0.122
Equation of line: y=-0.002x+0.1183
Solve equation for x where y=0.1 (10% CV)
0.1 = -0.002 x + 0.1183
-0.0183 = -0.002 x
9.15 = x = Limit of Quantitation (LOQ)
Divide LOQ by 3.3 to calculate Limit of Detection (LOD)
9.15/3.3= 2.77
Data Summary:
The limit of detection of the method as implemented 2.77
The limit of quantitation/sensitivity of the method as implemented 9.15
Appendix IX. Definitions
108
Appendix IX. Definitions
(http://www.issc.org/client_resources/lmr%20documents/ii.%20definitions.pdf)
1.
2.
3.
Accuracy/Trueness - Closeness of agreement between a test result and the accepted reference value.
Analyte/measurand - The specific organism or chemical substance sought or determined in a sample.
Blank - Sample material containing no detectable level of the analyte or measurand of interest that is subjected to the
analytical process and monitors contamination during analysis.
4. Comparability – The acceptability of a new or modified method as a substitute for an established method in the NSSP.
Comparability must be demonstrated for each substrate or tissue type by season and geographic area if
applicable.
5. Fit for purpose – The analytical method is appropriate to the purpose for which the results are likely to be used.
6. HORRAT value – HORRAT values give a measure of the acceptability of the precision characteristics of a method. 4
7. Limit of Detection – the minimum concentration at which the analyte or measurand can be identified. Limit of detection
is matrix and analyte/measurand dependent.4
8. Limit of Quantitation/Sensitivity – the minimum concentration of the analyte or measurand that can be quantified with an
acceptable level of precision and accuracy under the conditions of the test.
9. Linear Range – the range within the working range where the results are proportional to the concentration of the analyte or
measurand present in the sample.
10. Measurement Uncertainty – A single parameter (usually a standard deviation or confidence interval) expressing the
possible range of values around the measured result within which the true value is expected to be with a stated degree
of probability. It takes into account all recognized effects operating on the result including: overall precision of the
complete method, the method and laboratory bias and matrix effects. .
11. Matrix – The component or substrate of a test sample.
12. Method Validation – The process of verifying that a method is fit for purpose.1
13. Precision – the closeness of agreement between independent test results obtained under stipulated conditions.1, 2 There are
two components of precision:
a. Repeatability – the measure of agreement of replicate tests carried out on the same sample in the same
laboratory by the same analyst within short intervals of time.
b. Reproducibility – the measure of agreement between tests carried out in different laboratories. In single laboratory
validation studies reproducibility is the closeness of agreement between results obtained with the same method on
replicate analytical portions with different analysts or with the same analyst on different days.
14. Quality System - The laboratory’s quality system is the process by which the laboratory conducts its activities so as to
provide data of known and documented quality with which to demonstrate regulatory compliance and for other decision–
making purposes. This system includes a process by which appropriate analytical methods are selected, their capability is
evaluated, and their performance is documented. The quality system shall be documented in the laboratory’s quality
manual.
15. Recovery – The fraction or percentage of an analyte or measurand recovered following sample analysis.
16. Ruggedness – the ability of a particular method to withstand relatively minor changes in analytical technique,
reagents,
or environmental factors likely to arise in different test environments.4
17. Specificity – the ability of a method to measure only what it is intended to measure. 1
18. Working Range – the range of analyte or measurand concentration over which the method is applied.
REFERENCES:
1. Eurachem Guide, 1998. The Fitness for Purpose of Analytical Methods. A Laboratory Guide to Method Validation
and Related Topics. LGC Ltd. Teddington, Middlesex, United Kingdom.
2. IUPAC Technical Report, 2002. Harmonized Guidelines for Single-Laboratory Validation of Methods of Analysis,
Pure Appl. Chem., Vol. 74, (5): 835-855.
3. Joint FAO/IAEA Expert Consultation, 1999. Guidelines for Single-Laboratory Validation of Analytical Methods for
Trace-Level Concentrations of Organic Chemicals.
4. MAF Food Assurance Authority, 2002. A Guide for the Validation and Approval of New Marine Biotoxin Test
Methods. Wellington, New Zealand.
5. National Environmental Laboratory Accreditation. , 2003. Standards. June 5.
6. EPA. 2004. EPA Microbiological Alternate Procedure Test Procedure (ATP) Protocol
for Drinking Water,
Ambient Water, and Wastewater Monitoring Methods: Guidance. U.S. Environmental Protection Agency (EPA),
Office of Water Engineering and Analysis Division, 1200 Pennsylvania Avenue, NW, (4303T), Washington, DC
20460. April.
Appendix IX. Definitions
109
Acknowledgements
We would like to sincerely thank Linda Chandler for her help and guidance on completing the
statistics and calculations for this report. We would also like to thank the Northwest Fisheries
Science Center in Washington and the University of Alaska-Fairbanks School of Fisheries and
Ocean Sciences for providing shellfish samples.
Acknowledgements