Professional Documents
Culture Documents
AOAC Official Methods of AnalysisSM
AOAC Official Methods of AnalysisSM
AnalysisSM
Darryl Sullivan, Covance Laboratories and Past President, AOAC INTERNATIONAL
On March 28, 2011, the AOAC INTERNATIONAL
Board of Directors approved an alternative path to
achieve an Official Method (Official First Action status)
for methods selected and reviewed using the AOAC
volunteer consensus standards development
processes.
Thank you!
Standard Method Performance
Requirements [SMPR] Guideline
pendent
Inde-
Collaborative
Study
Classifications of Methods9
Qualitative Method
Qualitative Method
(trace or Identification Method
(main component1)
contaminant2)
AMDL
Collabora- Indepen-
dent
POD (0)
POD (0) POD (0)
Study
POD (c)
tive
Method B Total
(Reference)
Absent c d NA-
Result Total
Method A e f NA
Method B g h NB
Total N+ N- N
Estimators
Assume random samples from same
population, randomly assigned to A or B
Proportion Positive A = (e)/NA (= PODA)
Proportion Negative A = (f)/NA
Proportion Positive B = (g)/NB (= PODB)
Proportion Negative B = (h)/NB
Accuracy, sensitivity, specificity not
defined unless all results are confirmed
Chi-Square and POD
P+A - P+B = dPOD
Chi-Square test is the same as a Binomial
test of null hypothesis:
H0: dPOD = 0
Chi-Square Test
Checks only for differences between
observed and expected numbers of results
in each cell
“Expected” based on random assignment
of subjects to A or B, so expect same
proportion of positives in A and B (and
same proportion of negatives)
“Expected” calculated from marginal
frequencies
Estimators POD Concept
P+A: PODA = (e)/NA
P+B: PODB = (g)/NB
P+A - P+B = dPOD
Chi-Square test is the same as a Binomial
test of null hypothesis:
H0: dPOD = 0
Thank you
For the Validation of Qualitative
Methods
Paul Wehling
June 30, 2011
1
Qualitative (Binary) Methods
Methods that are restricted to 2 possible
outcomes:
Positive or Negative
Pass or Fail
Heads or Tails
1 or 0
Yes or No
Presence or Absence
Identified or Not Identified
2
POD
Parameter – Probability of Detection
General – Designed to be used by any Qualitative (Binary)
Method
Microbiological
Chemical
Bio Threat Agent Methods
Botanical Identification
Allergens
3
POD Parameter
Method Parameter that describes and
predicts method behavior
Probability of Detection or POD
The probability of getting a positive
result at a given concentration of
analyte.
POD is a function of concentration
4
POD Curve
5
POD Curve
6
POD
A simple descriptive statistic that describes the
method performance at a given concentration.
It is a calculation of proportion of observed positive
outcomes per total trials.
This simple statistic is inherent in all other systems,
such as Chi-Square, LOD, RLOD.
The “POD Concept” is only new in that it recognizes
the POD as a key parameter and plots a graph of POD
vs concentration.
7
WHY PLOT POD?
Plot of POD Curves are intended to assist
method users
To assist users in selecting best method for intended
use.
Understanding POD Curve is crucial for
interpretation of results.
The POD curve can be an indicator of the
“usefulness” of the method.
If POD were constant across all concentrations, the
method would not be useful.
8
VALIDATION
The task of validating a qualitative (binary)
method is characterizing the POD curve at
critical concentration points.
1. Make up a series of test materials at concentrations
of interest.
2. Analyze with replication
3. Calculate the proportion of positive responses at
the concentrations.
4. Plot observed proportions as POD curve by
concentration.
9
10
Example POD Response Curve
11
A BIT ABOUT POD
POD is a combination of sensitivity,
specificity, false positives, false
negatives.
Where did they all go?
12
“Where’s my False Negative?”
POD Response vs Concentration
1
“False Negative at 1 ppm”
0.9
1-POD(1 ppm)
0.8
0.7 “Specificity”
1-POD(0)
0.6
POD
“Sensitivity at 1 ppm”
0.5
POD(1 ppm)
0.4
0.3
0.2
“False Positive”
0.1
POD(0)
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Concentration (ppm)
13
Some Statistics
To do Classical Collab Statistics, Code Results
“Positive” = 1
“Negative” = 0
14
POD = Mean
LAB1 LAB2 LAB3
Trial1 1 1 0
Trial2 0 1 1
Trial3 1 1 1
Trial4 0 0 0
Trial5 1 0 0
Trial6 1 0 1
Trial7 0 0 1
Trial8 0 1 0
Trial9 1 1 0
Trial10 0 1 0
Mean 0.5 0.6 0.4 LPOD = 0.50
15
Analogous Parameters
Quantitative Quantitative Qualitative
Method Attribute Qualitative Estimate
Parameter Estimate Parameter
General Mean or
Expectation
Mean, μ Mean, x POD POD or LPOD
Repeatability
Variance
r2 sr2 r2 sr2
Reproducibility
Variance
R2 sR2 R2 sR2
Laboratory Variance
L2 sL2 L2 sL2
Expected difference
between Two Methods*
Bias, B x1 x2 dPOD POD1 POD2
16
Difference Between Methods -
dPOD
Compare any two methods by comparing
POD values at a given concentration.
Difference by subtraction
17
dPOD(3.5) = -0.10
18
dPOD Curve vs Concentration
19
Big Ideas
Combine sensitivity, specificity, false
positive, false negative into 1 parameter –
Probability of Detection or ‘POD’
Graph POD vs. Concentration with
Confidence Intervals
Compare methods by difference of POD at
same concentration
Use the classic statistical model and
descriptive stats for quantitative methods as
the tool for calculating qualitative stats.
20
POD Concept
Works for single lab and Multilab experiments.
Works for paired and unpaired designs.
Provides harmonization across
qualitative/quantitative methodologies.
Does comparisons and hypothesis tests via confidence
interval analysis – equivalent to chi-squared tests.
POD Curve plots mean response and uncertainty on
the same graph.
21
Qualitative Method Validation
Studies for Quantal Data:
LOD, dPOD, PRE = RLOD and ω
E. coli O157:H7 Raw ground beef Unpaired H 3.18 0.72 74.41 11.80
L 0.84 0.46 11.61 7.93
Pre-
Collaborative
Phase(s)
-Reference -Defined in ISO -Can be various pre- -Acceptable Ref published ISO, CEN, NMKL, -Must be BAM, unless there For FSIS regulated products,
Method 16140-1 existing recognized by HC (Part 1) BAM, etc. It is up to is no BAM reference the current FSIS method,
-1st priority is ISO analytical methods -May include any methods the applicant; method. which is found in the
method, 2nd priority is e.g. AOAC OMA, from methods organizations, however, as the EU -If these is no BAM Microbiology Laboratory
CEN method, if ISO, FDA BAM, such as AOAC, BAM, regulation in EC reference method, but if Guidebook (MLG), is the
neither exists, then 3rd FSIS MLG and APHA, ICMSF, IDF, ISO 2073/2005 there is a most appropriate reference
priority is other Health Canada etc. Microbiological nationally/internationally cultural method for validating
recognized methods -If no appropriate Ref -Where no Ref exists, MMC criteria states EN ISO recognized reference methods used by FSIS-
can indicate “NA” in assess on case by case basis methods, these are method, then FSIS MLG, regulated establishments.
Note: definition still summary tables for most frequently used. AOAC, ISO, and Health FDA BAM, or methods
under discussion at POD Canada are all potential referenced by ISO or Codex
ISO level to open up reference methods. APHA, Alimentarius may be
for non ISO/CEN ICMSF, and IDF methods appropriate. Non-cultural
methods (PIV) also may be used as methods applicable in some
reference methods. circumstances.
Page 1 of 8
ISO 16140 AOAC OMA Health Canada NordVal FDA Draft USDA/FSIS
-Selection of RA SLV RA, SE, SP, Kappa The selection of foods is Matrices commonly sampled
food -5 categories for all -All claimed matrices -5 categories for all foods -5 categories for all determined by FDA’s in FSIS regulated
foods applications, 3 must be included in applications, 3 food types foods applications, 3 regulatory needs. establishments: meat,
food types per the study, in other per category (Table 1). food types per poultry, and egg products,
category (see below) words, no defined -Environmental samples is category (see below) and environmental samples
-Feed, environmental categories, and “all additional category -Feed, environmental (sponges, swabs, brines)
samples and primary foods” claim not samples are
production samples applicable additional categories All claimed matrices must be
(PIV)are additional -Environmental LOD included in the study.
categories surfaces claim Same, except 1 food Contains proposal to create
RLOD require 3-7 different type per category (if matrix categories based on
Same, except 1 food surfaces (# possible) a different intrinsic properties. “All
type per category (if required is under food type Foods” claim not applicable
possible) a different review (RF)
food type IV
At least 1 matrix that
was tested in the
SLV. For every 5
foods claimed, 1 food
matrix must be
included
Page 2 of 8
ISO 16140 AOAC OMA Health Canada NordVal FDA Draft USDA/FSIS
- Food Each Food type can be Only one single food Each food type can be made Each Food type can Currently, foods are
category/type/ made of various item is accepted to of various relevant food be made of various validated individually and
item relevant food items. meet the sample size items. Table 1. relevant food items. there are no category
Annex B provides requirement of a food At NordVal’s claims. There are no “All
guidance ( not type, i.e. 20 CLARIFICATION homepage Foods” claims.
mandatory) replicates. NEEDED (www.nmkl.org)
These are then Can these be grouped provides a list of food
grouped together to together to meet the sample categories
meet the sample number requirement of a These are then
number requirement of food type, in this case 20 grouped together to
a food type, i.e. 20 samples? meet the sample
samples. Yes, they can be group number requirement
together to meet the sample of a food type, i.e. 20
This is allow for the number requirement. This samples.
use of naturally notion has been introduce to
contaminated samples allow for heterogeneity with
(BL) in a food type. Products in a
type may vary greatly in
origin, composition,
preparation processes,
natural background; all
those small variabilities
could have an influence on
the detectability of the target
organism. (I.I.)
Page 3 of 8
ISO 16140 AOAC OMA Health Canada NordVal FDA Draft USDA/FSIS
-No. of RA SLV and IV 3 levels: RA Level 1, 6 replicates/level, For each matrix and analyte:
levels/samples 20 samples per food 3 levels: -negative controls =5 20 samples per food single level 1) minimum 60 samples
type or 60 samples per -negative controls =5 samples type or 60 samples Level 2, 6 replicates/level, 1 inoculated at fractional
category samples -1 level with fractional per category inoculated level + 1 recovery level per
RLOD -1 level with positive results = 20 LOD uninoculated level (5 alternative and reference
3 levels fractional positive samples 3 levels replicates) method
-negative controls =5 results = 20 samples -Another level up to 1 log -negative controls =5 Level 3, 10 replicates/level, 2) 5-10 uninoculated
samples -Another (high) level higher= 20 samples samples 1 inoculated level + 1 samples per alternative
-1 level ( theoretical = 5 or 20 samples -1 level ( theoretical uninoculated level (5 and reference method
LOD, with fractional (under review (RF) LOD) = 20 samples replicates)
positive results (BL)) -Another level = at Level 4, 20 replicates/level,
= 20 samples least 5 samples 1 inoculated level + 1
-Another level = at uninoculated level (5
least 5 samples replicates)
It is proposed that each of
the 4 levels use 20 replicate
test portions and that all
levels have a negative
control.
-Sample size Undefined Standard is 25 g or 25 -25 g, but larger sample Undefined 25 g unless otherwise Application dependent.
MicroVal: Is specified mL, unless Ref sizes are permitted specified. Portions should not be made
in the reference method specified -Sample size must be the larger without validation.
method, other (larger) larger sample size same for alternate and Ref Validation study conclusions
samples size is methods, consult MMC if from larger portions
allowed but specified testing composite samples applicable to smaller
in the certificate. portions.
(PIV)
-Fractional Can be achieved by Can be achieved by Can be achieved by either Can be achieved by Yes, one or both methods defined as a range of 20-80%
positive either alternate or Ref. either alternate or alternate or Ref. either alternate or must give 40 – 90% positive confirmed positive results
- All samples should Ref. -proportion of positives Ref. results. It is proposed that using reference method
not be all positive or -proportion of 25% to 75%, - All samples should the percentage positive
all negative. positives 25% to not be all positive or results be changed to 25 –
-Ideal is 10 positive 75%, ideal is approx all negative. 75%.
and 10 negative (50%) 50% -Ideal is 10 positive
but any fractional (10% to 90% is under and 10 negative
results is acceptable review (RF) (50%) but any
fractional results is
acceptable
Page 4 of 8
ISO 16140 AOAC OMA Health Canada NordVal FDA Draft USDA/FSIS
-Results analysis RA - by level and by Method Equivalence: RA By level/individual Performed for each matrix.
and criteria -By type and by matrix -POD -By type and by experiment for each matrix. Unpaired study: One sided
category - by POD Probability -one-tailed POD 95% category Per AOAC Microbiology chi-square test with alpha =
-Relative accuracy of Detection 95% confidence interval (I.I.) -Relative accuracy guidelines, McNemar Chi 0.05. Criterion:
AC, relative confidence interval AC, Square statistics are used. indistinguishable or better
specificity SP, relative for the alternate the Performance parameters: -Relative specificity performance than reference
sensitivity SE Ref and presumptive - by level and by food, but SP, method. Paired study:
-First by unconfirmed and confirmed results only calculated for those -Relative sensitivity Evaluate sensitivity with
results, again by -then by difference that passed POD SE minimum 29 confirmed
confirmed results between POD successfully - Kappa positive results. Zero false
-McNemar test as alternate and POD For Unpaired : -First by unconfirmed negative results from 29
criteria, (for paired Ref, confidence level -Performance parameters is results, again by confirmed positives would be
and unpaired) with must contain zero for the comparison of confirmed results consistent with a test having
caveats i.e. really not method to be presumptive vs. confirmed a sensitivity that met or
suitable for unpaired considered not results of the alternate Criteria: exceeded 90% and zero
and “never be different at 95% method ( not the Ref SE ≥ 95% negative results from 50
interpreted by only confidence method results) Kappa ≤ 0.80 confirmed positives would be
the McNemar test” - Chi Square is not -Specificity is based on LOD: fit for purpose consistent with a test with a
RLOD required but presumptive results sensitivity that met or
-by category “interesting” -Sensitivity is based on final exceeded 94%. Criterion:
-LOD of alternate ( confirmed) results none proposed
method divided by - Equivalence of alternate
LOD of Ref method and Ref can only be
For paired, no lower determined by the number
limit, but LOD of true positives in both sets,
alternate might not be done by POD method
> 2 times the LOD Ref
For unpaired samples, For Paired:
no lower limit, the Use “absolute” results
LOD alternate might where Ref can have FN
not be >3 times the
LOD Ref (In the Criteria:
ISO/CD 16140-2 Sensitivity 98%
version, I don’t find Specificity 90.4%
any acceptability False negative rate < 2%
limit settled for False positive rate ≤ 9.6%
unpaired samples, Efficacy 94%
only specified for LOD must be comparable or
paired samples BL) exceed the lower LOD of
the Ref
The values of 2
(paired) and 3
(unpaired) are still
tentative values!!!!!
(PIV)
Page 5 of 8
ISO 16140 AOAC OMA Health Canada NordVal FDA Draft USDA/FSIS
Inter- Applicable to alternative
laboratory methods with a major
Study modification, defined as any
significant change in the
design or the component
reagents for a screening test,
for example, the introduction
of a new antibody or
oligonucleotide primer.
Page 6 of 8
ISO 16140 AOAC OMA Health Canada NordVal FDA Draft USDA/FSIS
- number of 1; relevant food item, 1 At least 1 1; relevant food One or more.
foods inoculated with target, item, inoculated with
using a challenging target, using a
enrichment protocol challenging
enrichment protocol
- number of 3; negative control, 3; negative control, 3; negative control, one 3; negative control, 2 for a Level 2 study (1
levels one level which one level which level which produce one level which inoculated and 1
produce fractional produce fractional fractional positive and produce fractional uninoculated.
positive and another positive and another another level about 10 times positive and another 3 for Levels 3 & 4 (high,
level level greater than the detection level low, and uninoculated.
level
-Confirmation for Paired, only Matched or Confirm all samples for Paired, only Yes.
confirm the + Alt/- unmatched, confirm confirm the + Alt/-
Ref, for Unpaired, all samples Ref, for Unpaired,
confirm all confirm all
enrichments enrichments
Page 7 of 8
ISO 16140 AOAC OMA Health Canada NordVal FDA Draft USDA/FSIS
- Comparisons Analyzed two ways: By level and by CLARIFICATION Alternative to reference
1.Unconfirmed matrix analyzed and NEEDED method (if available).
Alternate method reported separately Consistent with Pre-
results vs. confirmed collaborative?
Ref By level and by matrix, all
2. Confirmed result confirmed
Alternate method Confirmed alternate method
results vs. confirmed results vs reference (I.I.)
Ref
-Parameters Specificity ( only for Cross Lab Probability CLARIFICATION Rel Specificity Per AOAC guidelines,
Calculated Neg controls) of Detection (LPOD) NEEDED Rel Sensitivity Sensitivity, Specificity,
Sensitivity ( only for Difference between Consistent with Pre- Rel Accuracy False Negative, and False
inoculated levels ) Alternate LPOD and collaborative? Kappa Positive Rates.
Relative Accuracy Ref LPOD Yes,
(%of agreements ) POD , dPOD determined
RLOD of the for each matrix-level . All
different participants dPOD data is then used to
(BL) assess the comparative
performance of both
methods
All 5 method
parameter(specificity,
selectivity, FP, FN and
method efficacy) calculated
in one of two ways,,
depending if sample is
paired or un paired. (I.I.)
- Interpretation
McNemar test (chi If confidence interval CLARIFICATION Criteria: Per AOAC guidelines,
square) of dLPOD does not NEEDED SE ≥ 95% McNemar Chi Square
RLOD is for contain zero, then the Consistent with Pre- Kappa ≤ 0.80 statistics.
information only diff is statistically collaborative? [LOD: fit for its
: analysis of deviance significant Yes , dPOD one-tailed and purpose]
test to assess the method parameter
laboratory effect on requirement must be met.
RLOD then (I.I.)
acceptability of
RLOD global value
(BL)
Page 8 of 8
International Stakeholder Panel on Alternative Methods
Microbiology Working Group for Harmonized Matrix Comparison
Twinbrook
HILTON WASHINGTON D.C./ROCKVILLE EXECUTIVE MEETING CENTER
ISPAM
1) EN ISO 16140:2008 E
AOAC ISPAM Small Group Micro Working Groups Meeting Agenda nlm PRE-DECISIONAL Page 1
EN ISO 16140:2008 (E) nlm
EN ISO 16140:2008 (E) nlm
Matrix for Salmonella (NORDVAL)
Matrix group
9. Animal faeces
10. Miscellaneous
NORDVALnlm
Matrix for Listeria (NORDVAL)
Matrix group Matrix Food Examples
1. Meat 1.1 Raw red meat Minced meat, (tatar – type)
1.3 Raw smoked salted Bacon, smoked filet
meat-products
1.4 Heat treated products Sliced meat and poultry products
1.5 Fermented products Salami
2. Fish 2.1 Raw fish, shelfish and Cold smoked salmon
Fish products
2.3 Heat treated fish products Heat treated shrimps
3. Milk 3.1 Milk Raw milk
3.4.1 Firm cheese Yellow cheese