1318 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO.

5, 2005

FOOD BIOLOGICAL CONTAMINANTS

Enumeration of Total Coliforms and E. coli in Foods by the

®
SimPlate Coliform and E. coli Color Indicator Method and
Conventional Culture Methods: Collaborative Study
PHILIP T. FELDSINE, ANDREW H. LIENAU, NERIE H. ROA, and SHANNON T. GREEN
BioControl Systems, Inc., 12822 SE 32nd St, Bellevue, WA 98005

Collaborators: J. Braut-Taormina; C. Braymen; C. Chavey; A. Cohen; A. Deans; S. Delancey; C. Elems; R. Forgey;

E. Gonzalez; V. Hall; M.C.J. Huang; L. Humes; F. Johnson; K. Kerdaji; J. King; L. Kipker; D. Lau; J. Nogle; C. Peters;
J. Pinkston; M. Porter; L. Potter; S. Rogers; J. Stephens; G. Skorupa; R. Taylor; E. Tuncan; K. Vought; D. Vrana

The relative effectiveness of the SimPlate® (SimPlate CEc-CI) method uses binary detection technology
Coliform and E. coli Color Indicator (CEc-CI) (BDT) to enumerate total coliforms and E. coli concentrations
method was compared to the AOAC 3-tube Most in foods after 24–28 h of incubation. This represents a
Probable Number (MPN) methods for enumerating significant time savings advantage over conventional
and confirming coliforms and Escherichia coli in enumeration methods which require 72–96 h (3–4 days) for
foods (966.23 and 966.24). In this study, test coliform confirmation and even longer for E. coli
portions were prepared and analyzed according to confirmation.
the conditions stated in both the AOAC methods The SimPlate CEc-CI method features a proprietary
and SimPlate directions for use. Six food types selective differential culture medium and a SimPlate device.
were artificially contaminated with coliform The medium is formulated with 2 indicators, one chromogenic
bacteria and E. coli: frozen burritos, frozen and one fluorogenic, to permit simultaneous detection of both
broccoli, fluid pasteurized milk, whole almond nut total coliforms and E. coli in a food suspension. Metabolic
meats, cheese, and powdered cake mix. Method activity by the indicator organisms results in a chromogenic
comparisons were conducted. Overall, the change in the presence of coliform bacteria and a fluorogenic
SimPlate method demonstrated <0.3 log difference reaction if E. coli is present in the test portion.
for total coliform and E. coli counts compared to The SimPlate device eliminates the uncertainty in
the AOAC reference methods for the majority of enumeration associated with the 3-tube Most Probable
food types and levels analyzed. In all cases, the Number (MPN) method because of factors such as food
repeatability and reproducibility of the SimPlate particulate interference and “the lack of precision in the MPN
CEc-CI method were not different from those of the tube method” (1). A test portion/medium homogenate is
reference methods and in certain cases, were dispensed into the test device and distributed into a fixed
statistically better than those of the AOAC 3-tube number of individual incubating wells: 84 for the normal
MPN methods. These results indicate that the counting range SimPlate test and 198 for the high counting
SimPlate CEc-CI method and the reference culture range. To enhance performance, medium supplements are
methods are comparable for enumeration of both provided for use with certain food matrixes. As such, discrete
total coliforms and E. coli in foods. aliquots are compartmentalized and isolated from each other
in the incubating wells where biochemical activities of viable
microorganisms are monitored in a fluid environment.
he demand for improved laboratory productivity Enumeration is measured by a simple binary reaction; each

T emphasizes faster turnaround times for quality control

testing in food companies. Total coliforms and
Escherichia coli are recognized as indicators of the quality of
well is either positive or negative. Any color change from the
original background color in each well or in the sponge is
interpreted as a positive reaction. For the SimPlate CEc-CI
food and the sanitation of the food processing environment. method, enumeration of total coliforms is achieved by
The SimPlate® for Coliform and E. coli Color Indicator counting the number of wells exhibiting a color change from
the original background color in each plate after incubation.
Submitted for publication June 2005. Colored wells are confirmed positive for coliform bacteria.
The recommendation was approved by the Methods Committee on Colored wells that also produce a blue fluorescence when
Microbiology and Extraneous Materials as First Action. See “Official exposed to a long-wavelength (366 nm) ultraviolet (UV) light
Methods Program Actions,” (2005) Inside Laboratory Management,
July/August issue. are confirmed positive for E. coli. The final count per plate is
Corresponding author’s e-mail: ptf@biocontrolsys.com. derived from the SimPlate Conversion Table (Table 1) that is
 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005 1319

based upon the principle of the Poisson distribution. The large Table 1. SimPlate Conversion Table
number of wells allows for an accurate measure of the true Number of positive wells = Population per platea
microbial population in a test sample.
A recent study compared the SimPlate CEc-CI method and
1=2 31 = 76 61 = 216
the AOAC 3-tube MPN methods (2, 3) for the enumeration of
total coliforms and E. coli in 21 different foods; 385 test 2=4 32 = 80 62 = 224
portions, including controls, were analyzed. Overall, the 3=6 33 = 84 63 = 232
results showed high correlation between the SimPlate CEc-CI 4=8 34 = 86 64 = 240
method and the AOAC methods for the detection of both total 5 = 10 35 = 90 65 = 248
coliforms and E. coli. The mean log counts produced by the 6 = 12 36 = 94 66 = 256
SimPlate CEc-CI method were not statistically different or
7 = 14 37 = 96 67 = 266
statistically higher for the majority of food types analyzed.
The repeatability standard deviation values of the SimPlate 8 = 16 38 = 100 68 = 276
CEc-CI method were generally lower than those of the AOAC 9 = 18 39 = 104 69 = 288
methods. A separate multilaboratory collaborative study 10 = 22 40 = 108 70 = 298
comparing the relative effectiveness of the SimPlate CEc-CI 11 = 24 41 = 112 71 = 312
method with the AOAC methods for the enumeration of total 12 = 26 42 = 116 72 = 324
coliforms and E. coli is reported here.
13 = 28 43 = 120 73 = 338
Collaborative Study 14 = 30 44 = 124 74 = 354
15 = 32 45 = 128 75 = 372
Design of Study
16 = 36 46 = 132 76 = 392
For this collaborative study, 6 food types representative of 17 = 38 47 = 136 77 = 414
a wide range of food categories were evaluated: frozen 18 = 40 48 = 142 78 = 440
burritos, frozen broccoli, fluid pasteurized whole milk,
19 = 42 49 = 146 79 = 470
almond nut meats, cheese, and powdered cake mix. All foods
20 = 46 50 = 150 80 = 508
were screened by AOAC reference Methods 966.23 (2) and
966.24 (3) prior to analysis and shown not to contain 21 = 48 51 = 156 81 = 556
significant numbers of coliform bacteria or E. coli. Foods 22 = 50 52 = 160 82 = 624
were artificially contaminated at a high inoculation level of 23 = 54 53 = 166 83 = 738
approximately 10 000 colony-forming units (CFU)/g, a 24 = 56 54 = 172 84 = >738
medium level of approximately 1000 CFU/g, and a low
25 = 58 55 = 178 If there are no positive wells and the sponge
inoculation level of approximately 100 CFU/g. All test is positive, then the population is 1
portions were tested by both the SimPlate CEc-CI method and
26 = 62 56 = 184
AOAC Official Methods (966.23 and 966.24). Duplicate test
27 = 64 57 = 190
portions representing each of the 3 inoculation levels and
uninoculated controls were tested for a total of 8 test portions 28 = 68 58 = 196 If there are no positive wells and the sponge
is negative, then the population is <1
per food type, with the exception of cheese, for which there
were 2 sets of duplicate test portions, with a total of 16 test 29 = 70 59 = 202
portions per analysis. 30 = 74 60 = 208

a
Preparation of Inocula and Test Portions The population reflects the number of microorganisms per plate.
To determine the number of microorganisms per g (mL) food
product, refer to text, I(c), Reading and Interpretation of Results.
Test portions of frozen burritos, frozen broccoli, fluid
pasteurized whole milk, almond nut meats, cheese, and
powdered cake mix were inoculated with one of the following
coliforms: Klebsiella pneumoniae, Enterobacter cloacae, food products. These test portions were then inoculated with
Enterobacter spp., Citrobacter freundii, Citrobacter spp., and an E. coli isolate so that each inoculated test portion contained
Klebsiella spp., respectively, for method comparison of total one non-E. coli coliform and one E. coli coliform type.
coliforms. Test cultures were propagated from either Test cultures were grown in Brain Heart Infusion broth
American Type Culture Collection strains or natural food (BHI) for 18–24 h at 35°C. All foods were inoculated in bulk.
isolates maintained in the in-house culture collection at Low-moisture foods were inoculated with a culture
BioControl Systems, Inc. All test cultures were cultured, lyophilized in 10% nonfat dry milk. High-moisture foods
streaked onto differential or selective agar media, and their were inoculated with a wet suspension of stationary phase
identities were verified by typical colony morphology and cells. All foods were mixed thoroughly to ensure content
biochemical analysis using the appropriate API™ test by uniformity. Frozen burritos and frozen broccoli were stored at
BioMérieux, Inc. These bacteria were chosen because they are –20°C until use. Fluid pasteurized milk, almond nut meats,
representative of those coliforms commonly isolated from cheese, and powdered cake mix were maintained in a chilled
1320 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005

condition (2–8°C). Inoculated test portions were prepared at Using MPN Table 966.24A, the MPN of E. coli was computed
least 1 week prior to analysis to allow populations to stabilize. based on the number of EC broth tubes containing
Test Sample Distribution biochemically confirmed E. coli isolates from streaked plates.
For the SimPlate CEc-CI method, aliquots from each test
For every food type except cheese, each collaborator
portion were analyzed for total coliforms and E. coli using
received a total of 8 test portions. For cheese, each
SimPlate CEc-CI according to the product directions for use,
collaborator received 2 sets of 8 test portions each. Each test
portion contained approximately 50 g or 50 mL of test which are provided in the OMA procedures described below.
material. Test portions were distributed by an overnight The number of positive wells containing coliforms and
delivery service to arrive not later than the week preceding the E. coli were converted to microorganisms/g using the
Monday that analyses were initiated. Cheese and fluid SimPlate Conversion Table (Table 1). If the original food test
pasteurized milk were shipped with ice packs to maintain a portion was diluted before inoculation of the SimPlate device,
chilled temperature during shipment. Frozen burritos and then the CFU/g of total coliforms or E. coli was calculated by
frozen broccoli were shipped with dry ice; nut meats and cake multiplying the population from the Conversion Table
mix were shipped at room temperature. (Table 1) by the dilution factor. Data were recorded onto the
appropriate worksheets.
Analysis
Collaborators analyzed each test portion by both the Statistical Analysis
AOAC culture Methods (966.23 and 966.24) and the
SimPlate CEc-CI method. Test portions were prepared as The number of microorganisms/g generated by both the
stated in the Official Methods of Analysis (OMA), 966.23 (2), SimPlate CEc-CI and AOAC methods was derived separately
for both the AOAC and SimPlate CEc-CI methods for for both total coliforms and E. coli. For the AOAC methods,
analysis. For all food types, except cheese and nut meats, 50 g the number of gas-positive tubes was converted to MPN/g.
test portions were added to 450 mL Butterfield’s Phosphate For the SimPlate CEc-CI method, microorganisms/g was
Buffered Diluent (BPBD) and homogenized by blending for determined by using the SimPlate Conversion Table (Table 1).
2 min. The frozen test portions were added directly and The base 10 logarithms (log10) of SimPlate counts and the
blended to obtain a fluid homogenate. After homogenization, MPN index were used for statistical analysis. There were
the suspensions were serially diluted in BPBD. The number of instances where indefinite values were reported. To perform
dilutions necessary to perform the test was dependent upon statistical analysis on these particular data values, the
the contamination level of food type being analyzed. indefinite value signs on these data were truncated before
Subsequent 10-fold dilutions were added to 90 mL sterile statistical analysis. This treatment of out-of-range data is
diluent and shaken 25 times in a 30 cm arc. BioControl well-established and is reported in previous AOAC
Systems, Inc. provided collaborators with a suggested serial collaborative studies (4–6). Repeatability (sr) and
dilution scheme for each food type to ensure that collaborators reproducibility (sR), standard deviations, relative standard
had a general guideline for diluting inoculated test portions to deviations of repeatability (RSDr) and reproducibility
achieve countable SimPlates and readable MPN tubes. (RSDR), and repeatability (r) and reproducibility (R; 7) values
For one set of cheese test portions, 50 g portions were were calculated according to the methods of Youden and
added to 450 mL BPBD and macerated for 2 min. These test Steiner (8) after the Cochran and Grubb’s tests were applied to
portion homogenates were analyzed by the SimPlate CEc-CI eliminate suspect outliers. Repeatability variances were
method. It was determined that use of a blender for compared by using an F statistic that computes the ratio of the
homogenization of the cheese matrix was unsuitable for the 2 variances. Mean responses for the 2 methods were
SimPlate method because of its detrimental effect on the compared by using a 2 sample (paired) t-test for all foods
readability of the test. For the other set of duplicate cheese test except cheese. Cheese was analyzed by using a 2 sample
portions, 50 g test portions were added to 450 mL BPBD and (unpaired) t-test.
homogenized by blending for 2 min. These test portion
homogenates were analyzed by the AOAC 3-tube MPN AOAC Official Method 2005.03
Method (966.23) and confirmed by AOAC Method 966.24. Detection and Confirmed Quantitation of Coliforms
For almond nut meats, 50 g test portions were added to 50 mL and E. coli in Foods
BPBD in a sterile jar and shaken vigorously (50 times through SimPlate Coliform and E. coli Color Indicator
a 30 cm arc) to obtain a 10r dilution. The homogenate was First Action 2005
allowed to stand for 3–5 min before serial dilutions were (Applicable to detection and quantitation of confirmed
made. total coliforms and E. coli in cake mix, chocolate, condiments,
For the AOAC methods, total coliforms and E. coli were dairy foods, egg products, flour, frozen prepared meals, frozen
analyzed and confirmed as stated in AOAC Method 966.24. vegetables, fruits, fruit juices, mushrooms, nut meats, pasta,
The MPN of coliforms was computed on the basis of number of poultry meats, red meats, seafood, and spices.)
tubes of Brilliant Green Lactose Broth (BGLB) producing gas See Tables 2005.03A and 2005.03B for the results of the
at the end of the incubation period using MPN Table 966.24A. interlaboratory study supporting acceptance of the method.
Table 2005.03A. Interlaboratory study results for total coliforms by AOAC and SimPlate CEc-CI methods

Meanb sre RSDr, %f rg sRh RSDR, %i Rj

Food group Lot/level Na SIMc AOACd SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC

Frozen burrito Low 8 2.28 2.14 0.22 0.22 9.54 10.63 0.61 0.64 0.22 0.33 9.54 15.40 0.61 0.92
Medium 9 2.71 2.71 0.17 0.25 6.33 9.33 0.48 0.71 0.20 0.27 7.50 10.00 0.57 0.76
High 9 3.63 3.70 0.22 0.21 6.18 5.75 0.63 0.60 0.22 0.32 6.18 8.58 0.63 0.89
k l
Frozen broccoli Low 8 2.53 2.16 0.11 0.55 4.43 25.37 0.31 1.54 0.27 0.72 10.69 33.22 0.76 2.01
Medium 8 3.03 2.75 0.14k 0.36 4.75 13.08 0.40 1.01 0.32l 0.42 10.58 15.33 0.89 1.18
k l
High 8 3.57 3.72 0.11 0.39 2.99 10.86 0.31 1.09 0.12 0.39 3.20 10.86 0.33 1.09
Nutmeats Low 10 1.70m 1.41 0.39 0.40 22.76 28.60 1.08 1.13 0.65 0.62 38.17 43.91 1.81 1.73
Medium 10 1.91 1.64 0.42 0.44 21.70 27.16 1.16 1.24 0.56 0.67 29.23 41.12 1.57 1.88
High 10 3.33 3.12 0.37 0.48 11.15 15.51 1.04 1.36 0.51 0.71 15.22 22.73 1.42 1.99
m n
Fluid milk Low 9 2.28 2.05 0.15 0.35 6.65 16.90 0.42 0.97 0.35 0.50 15.41 24.32 0.98 1.40
Medium 9 2.73m 2.60 0.13k 0.41 4.75 15.74 0.37 1.15 0.29l 0.53 10.48 20.48 0.81 1.49
k
High 9 3.24 3.33 0.05 0.28 1.68 8.47 0.15 0.79 0.32 0.40 10.03 12.03 0.91 1.12
Cheese Low 9 2.96 2.77 0.07k 0.26 2.48 9.59 0.20 0.74 0.11l 0.42 3.64 15.35 0.30 1.18
k l
Medium 10 3.44 3.28 0.07 0.36 2.06 10.91 0.20 1.00 0.12 0.36 3.48 10.91 0.33 1.00
High 8 3.96 4.09 0.13n 0.26 3.36 6.26 0.37 0.72 0.13l 0.40 3.36 9.89 0.37 1.13
Cake mix Low 9 1.78 1.64 0.45 0.48 25.28 28.95 1.26 1.33 0.61 0.63 34.04 38.30 1.70 1.76
Medium 9 2.71 2.49 0.46 0.63 16.90 28.16 1.28 1.76 0.60 0.83 22.21 37.28 1.69 2.33
High 9 3.65 3.53 0.32 0.47 8.83 13.44 0.90 1.33 0.45 0.68 12.41 19.16 1.27 1.89
a
N = Number of laboratories with valid data.
b
Mean log total coliforms/gram.
c
SIM = SimPlate method.
d
AOAC = AOAC culture method.
e
sr = Repeatability standard deviation.
f
RSDr = Repeatability relative standard deviation.
g
r = Repeatability values, 2.8 ´ sr.
h
sR = Reproducibility standard deviation.
i
RSDR = Reproducibility relative standard deviation.
j
R = Reproducibility values, 2.8 ´ sR.
k
Significantly better repeatability p < 0.01.
l
Significantly better reproducibility p < 0.01.
m
Significantly better mean log counts p < 0.05; statistically not different p < 0.01.
n
Significantly better repeatability p < 0.05; statistically not different p < 0.01.
FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005 1321
Table 2005.03B. Interlaboratory study results for E. coli by AOAC and SimPlate CEc-CI methods

Meanb sre RSDr, %f rg sRh RSDR, %i Rj

Food group Lot/level Na SIMc AOACd SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC

Frozen burrito Low 9 1.96 1.96 0.28 0.40 14.17 20.32 0.78 1.11 0.27 0.40 14.17 20.32 0.78 1.11
Medium 9 2.37 2.42 0.20 0.35 8.27 14.30 0.55 0.97 0.21 0.38 9.09 15.54 0.60 1.05
k l
High 9 3.30 3.17 0.15 0.35 4.47 11.22 0.43 0.98 0.17 0.35 5.09 11.22 0.49 0.98
Frozen broccoli Low 8 2.20 2.21 0.26 0.32 11.87 14.72 0.73 0.91 0.30 0.53 13.87 23.99 0.85 1.48
Medium 8 2.59 2.64 0.23 0.24 8.86 9.23 0.64 0.68 0.31 0.35 12.05 13.38 0.87 0.99
High 8 3.17 3.06 0.10k 0.40 3.16 13.09 0.28 1.12 0.15m 0.48 4.68 15.72 0.42 1.35
Nut meats Low 10 0.88 0.84 0.49 0.36 55.39 42.97 1.36 1.01 0.49 0.69 55.39 82.11 1.36 1.93
Medium 10 1.10 0.90 0.33 0.33 29.21 36.62 0.91 0.92 0.49 0.70 43.71 77.52 1.37 1.95
High 10 2.65 2.52 0.28 0.42 10.75 16.64 0.80 1.18 0.44 0.70 16.58 27.72 1.23 1.96
Fluid milk Low 9 1.08 1.34 0.39 0.28 36.00 21.13 1.09 0.79 0.57 0.54 52.78 40.68 1.59 1.52
Medium 9 1.79 1.88 0.44 0.43 24.41 22.75 1.22 1.20 0.51 0.48 28.39 25.34 1.42 1.34
High 9 2.19 2.34 0.19 0.32 8.71 13.67 0.54 0.90 0.46 0.47 20.80 20.26 1.30 1.33
n k m
Cheese Low 10 2.35 2.04 0.12 0.37 5.26 18.31 0.35 1.05 0.12 0.44 5.26 21.36 0.35 1.22
Medium 10 2.73 2.65 0.07k 0.27 2.62 10.10 0.20 0.75 0.07m 0.32 2.62 11.90 0.20 0.88
k m
High 10 3.16 2.93 0.17 0.65 5.26 22.08 0.47 1.81 0.17 0.65 5.26 22.08 0.47 1.81
Cake mix Low 9 0.82 0.54 0.40 0.54 48.84 100.94 1.13 1.53 0.40 0.66 48.84 122.77 1.13 1.86
1322 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005

o
Medium 9 1.70 1.40 0.76 0.79 43.25 52.53 2.12 2.22 0.96 1.14 54.77 74.74 2.69 3.21
High 8 3.19n 2.77 0.52 0.54 16.23 19.39 1.45 1.51 0.60 0.87 18.89 31.23 1.69 2.42

a
N = Number of laboratories with valid data.
b
Mean log E. coli/gram.
c
SIM = SimPlate method.
d
AOAC = AOAC culture method.
e
sr = Repeatability standard deviation.
f
RSDr = Repeatability relative standard deviation.
g
r = Repeatability values, 2.8 ´ sr.
h
sR = Reproducibility standard deviation.
i
RSDR = Reproducibility relative standard deviation.
j
R = Reproducibility values, 2.8 ´ SR.
k
Significantly better repeatability p < 0.01.
l
Significantly better reproducibilty p < 0.05; not statistically different p < 0.01.
m
Significantly better reproducibility p < 0.01.
n
Significantly better mean log counts p < 0.01.
o
Significantly better mean log counts p < 0.05; not statistically different p < 0.01.
 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005 1323

A. Principle (b) If an alternative test portion size is specified in testing

procedure or standard, prepare a 10% weight-to-volume
The SimPlate Coliform and E. coli Color Indicator suspension.
(CEc-CI) method is used for detection and confirmed
(c) If necessary, prepare 10-fold serial dilutions
quantitation of total coliform and E. coli populations. The
appropriate for the anticipated population of the product.
CEc-CI medium and test portion mixture is dispensed into a
SimPlate device and incubated for a minimum of 24–28 h. F. CEc-CI Test Procedure, Single Test Medium
The medium changes color in the presence of total coliforms.
(a) For 1.0 mL test suspension size.—Resuspend
E. coli is detected by the presence of a blue fluorescent color
powdered medium with 9.0 mL sterile distilled or deionized
change in the medium when examined under ultraviolet (UV)
water. Add 1.0 mL prepared test suspension and mix well. Do
light (366 mm). The total coliform and E. coli counts are
Not count this reconstitution as a dilution.
determined by counting the wells with changed color and
(b) For 0.1 mL test suspension size.—Resuspend
those with fluorescent wells, respectively, referring to the
powdered medium with 9.9 mL sterile deionized water. Add
SimPlate Conversion Table (Table 1).
0.1 mL prepared test suspension and mix well. This is an
B. Media and Reagents additional 1:10 dilution from step E. The final volume of test
portion/medium mixture in the container should be
Items (a) and (b) are available from BioControl Systems 10 ± 0.2 mL.
(12822 SE 32nd St, Bellevue, WA 98005, USA; Note: For select raw foods (meats, seafood, and
www.biocontrolsys.com). vegetables), add 0.1 mL Supplement R to the rehydrated
(a) Dehydrated CEc-CI medium.—Individually packaged medium. Supplements are available from BioControl
single test or multiple test format with supplement(s). Systems, Inc. The final volume of suspension medium
(b) SimPlate devices.—In packs of 20. mixture into the container should be 10 ± 0.2 mL.
(c) Butterfield’s phosphate buffer.—(1) Stock Remove lid from the SimPlate device and transfer test
solution.—Dissolve 34.0 g KH2PO4 in 500 mL H2O, adjust to suspension/medium mixture onto center of plate. Immediately
pH 7.2 with ca 175 mL 1M NaOH, and dilute to 1 L with H2O. replace lid. Continue with step H.
Store in the refrigerator. (2) Diluent.—Dilute 1.25 mL stock
solution to 1 L with H2O. Prepare dilution blanks with this G. CEc-CI Test Procedure, Multiple Test Medium
solution. Autoclave 15 min at 121°C. (a) Empty contents of one multidose container into
100 mL sterile distilled or deionized water. Shake to
C. Apparatus completely dissolve.
(a) Incubator.—Maintaining 35°–37°C in a dark (b) Remove lid from SimPlate device. Pipet prepared test
environment. suspension onto center of plate. If the prepared test suspension
(b) Micropipettor.—Accurately dispensing 0.1 and size is 1.0 mL, overlay the test portion with 9.0 mL medium.
1.0 mL. Do not count this media addition as a dilution.
(c) Pipets.—Glass or plastic sterile, 1 mL with 0.01 mL (c) For 0.1 mL prepared test suspension, overlay with
graduations, and 10 mL with 0.1 mL graduations. 9.9 mL medium: this is an additional 1:10 dilution of test
(d) Stomacher/masticator.—IUL Instruments (Cincinnati, suspension from step E. The final volume of test
OH, USA; www.iul-inst.com), or equivalent, for macerating suspenson/medium mixture on the plate should be
test portions. 10 ± 0.2 mL.
Note: For select raw foods (meats, seafood, and vegetables),
(e) Long wavelength ultraviolet light source (366 nm).
add 1.0 mL Supplement R to the rehydrated medium.
D. General Instructions Supplements are available from BioControl Systems, Inc. The
final volume of suspension medium mixture into the container
Discard the SimPlate device if the wells are crushed or should be 10 ± 0.2 mL.
pinched. Do not use expired medium. Store reconstituted (d) Immediately replace the lid. Continue with step H.
medium between 2° and 25°C in the dark and use within 12 h.
Dispose of used medium and SimPlate devices in a H. Test Procedure for Single and Multiple Tests
decontamination container and sterilize according to Good (a) Gently swirl to distribute the test suspension into all
Laboratory Practices (U.S. Food and Drug Administration wells. Hold plate with both hands tilted slightly to help
(1997) Code of Federal Regulations, Title 21, Chapter 1, distribute medium into wells.
Subchapter A, part 58, U.S. Government Printing Office, (b) Pour off excess medium by holding lid against plate on
Washington, DC, USA). either side of sponge cavity. Tip plate toward operator to allow
fluid to drain into sponge. Observe the background color of
E. Preparation of Test Suspensions
the wells. Background is defined as the color of the test
(a) Weigh 50 g test portion into 450 mL sterile diluent, i.e., suspension inside the wells.
Butterfield’s phosphate buffer. This is a 1:10 dilution. (c) Do not invert the SimPlate device. Incubate at
Macerate test portion. 35°–37°C for 24–28 h in the dark.
1324 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005

Table 2. Collaborator participation for the SimPlate CEc-CI collaborative study by food typea

Lab Frozen burritos Frozen broccoli Milk 1 Milk 2 Almond nut meats Cheese 1 Cheese 2 Powdered cake mix

1 Y N Y Y Y N Y Y
b
2 Y Y Y Y Y Y Y Y
3 Y Y Y Y Y Y Y Y
c c c c c c c
4 Y Y Y Y Y Y Y Yc
d d d d d d d
5 Y Y Y Y Y Y Y Yd
6 Ye N Ye N N N N N
7 Y N Y Y Y Y Y Y
b
8 N N N N Y Y N N
9 Y Y Yd Y N N Y Y
10 Y Y N N Y N N N
11 Y Y N Y N Y Y N
e
12 Y Y Y Y Y Y Y Y
13 Y N Ye Y Y Yb Y Y
e b
14 Y Y Y Y Y Y Y Y
15 N Y N N Y N Y Y
f
Total 13 10 11 11 12 10 12 11

a
Y = Collaborator analyzed this food type; N = collaborator did not analyze this food type.
b
Laboratory reported data that is out of range. Results not included in the statistical analysis for the designated food types.
c
Laboratory did not submit data for this analysis. Results not included in the statistical analysis for the designated food types.
d
Laboratory did not follow study instructions. Results not included in the statistical analysis for the designated food types.
e
Laboratory invalidated due to contamination of diluent. Results not included in the statistical analysis for the designated food types.
f
Total number of laboratories participating in the analysis of this food type.

I. Reading and Interpretation of Results study, test portions were prepared and incubated according to
the conditions stated in AOAC Method 966.23. Fifteen
(a) After incubation, observe color change of fluid in the
laboratories participated in the study (Table 2). Six
wells. Disregard particulate matter if present. Count the
laboratories analyzed all 6 food types, 4 laboratories analyzed
number of wells showing any color change from the
5 food types, 2 laboratories analyzed 4 food types,
background color. Wells with changed color correspond to the
1 laboratory analyzed 3 food types, and 2 laboratories
presence of total coliforms. Count the number of colored
analyzed 2 food types (Table 2). The numbers of total
wells showing white to blue fluorescence by holding a UV
coliforms and E. coli (reported in log10 values) recovered from
light (366 nm) 15–30 cm (approximately 6 in.) above the
individual test portions are presented in Tables 3–8.
SimPlate device. These wells correspond to the presence of
Interlaboatory study results as well as repeatability and
E. coli.
reproducibility values are presented in Tables 2005.03A and
(b) To determine the population, perform the following
2005.03B.
calculations: (1) Count the number of positive wells on the
plate; (2) use the SimPlate Conversion Table (Table 1) to Frozen Burritos
determine the total population per plate.
(c) To calculate the number of microorganisms/g of food, Frozen burrito test portions inoculated at a low, medium,
multiply the count in I(b)(2) by the appropriate dilution factor and high level were analyzed (Table 3). Uninoculated controls
(see steps E and F for single test or steps E and G for multiple were included in each analysis. Thirteen laboratories
test). participated in the analysis of frozen burritos.
Reference: J. AOAC Int. 88, 1318 (2005). Laboratories 5–7 did not follow study instructions. The data
from these laboratories were excluded from statistical
Results and Discussion analysis. Laboratory 4 did not submit data for this analysis.
Data generated by the AOAC methods and the SimPlate
The SimPlate Coliform and E. coli-Color Indicator Method CEc-CI method for total coliforms in frozen burritos were
(CEc-CI) was compared to the AOAC 3-tube MPN method compared statistically. Eight laboratories submitted valid data
for the enumeration of coliforms and E. coli in foods. For this for the low level and 9 laboratories submitted valid data for the
Table 3. Total coliforms and E. coli counts for frozen prepared meals test portions (log10 CFU/g) by SimPlate CEc-CI and AOAC methods

Uninoculated Low

Total coliforms E. coli Total coliforms E. coli

1 2 1 2 1 2 1 2

Lab SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC

1 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00a 1.88 <1.00a 1.63 1.00a 2.18 1.00a 2.18
2 <1.00 0.56 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.45 1.97 2.00 1.97 2.58 2.36 2.15 2.63
3 <1.00 <0.48 <1.00 0.56 <1.00 <0.48 <1.00 <0.48 2.08 1.32 1.90 2.66 2.15 1.63 2.20 1.88
9 <1.00 0.96 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.15 2.18 1.60 1.63 2.38 1.88 2.15 2.18
10 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.26 1.63 1.90 1.63 2.45 2.18 2.15 2.18
11 1.30 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.00 2.18 2.00 1.97 2.15 2.18 2.42 1.97
12 1.30 0.56 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.00 1.97 1.30 2.38 2.45 2.66 2.15 2.38
13 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 1.78 1.63 2.08 1.88 2.08 1.88 2.38 2.46
14 <1.00 <0.48 <1.00 0.56 <1.00 <0.48 <1.00 <0.48 1.90 2.38 1.90 1.97 2.15 2.38 2.56 1.97
Medium High

1 2.58 2.66 2.75 2.66 1.90 1.97 2.23 1.97 3.71 3.38 3.67 4.04 3.45 3.04 3.69 3.66
2 2.85 2.97 2.48 2.97 2.58 2.18 2.00 1.97 3.70 3.97 3.64 3.97 3.38 2.88 3.31 2.88
3 2.87 2.38 2.92 2.66 2.48 2.38 2.56 2.66 3.53 3.38 3.51 3.32 3.11 3.38 3.29 3.32
9 2.75 2.38 2.20 2.38 2.56 2.66 2.15 2.08 3.62 3.66 3.79 3.66 3.66 3.66 3.45 3.04
10 2.76 3.04 2.73 2.38 2.45 1.97 2.38 2.38 3.90 3.66 3.48 3.32 3.44 3.66 3.38 2.88
a a
11 2.56 2.66 2.51 2.66 2.26 2.66 2.26 2.66 3.75 4.38 2.95 4.04 3.55 4.38 2.56 4.04
12 2.88 3.18 2.93 2.38 2.56 3.18 2.56 1.97 3.70 3.38 3.48 3.38 3.51 3.04 3.08 3.04
13 2.97 3.04 2.81 3.04 2.60 2.66 2.60 2.88 3.68 3.66 3.76 3.66 3.45 2.88 3.38 2.97
14 2.51 2.66 2.68 2.67 2.15 2.66 2.15 2.66 3.70 3.66 3.81 4.04 3.66 3.18 3.45 2.43

a
Outlier, data not used for analysis in method comparison.
FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005 1325
Table 4. Total coliforms and E. coli counts for frozen vegetables test portions (log10 CFU/g) by SimPlate CEc-CI and AOAC methods

Uninoculated Low

Total coliforms E. coli Total coliforms E. coli

1 2 1 2 1 2 1 2

Lab SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC

2 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.38 1.32 2.34 2.38 2.20 1.46 2.15 2.38
3 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.38 3.38 2.58 2.38 1.90 3.04 2.41 2.38
9 <1.00 <0.48 <1.00 1.63 <1.00 <0.48 <1.00 <0.48 2.00 1.63 2.34 <0.48 1.30 1.63 2.15 1.63
10 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 3.13 2.18 2.98 1.97 2.62 2.88 2.51 2.97
11 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.48 2.88 2.48 2.38 2.00 2.88 2.26 2.38
12 1.60 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.58 2.38 2.56 1.97 2.20 1.97 2.08 1.97
13 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.62 2.18 2.76 2.38 2.34 1.97 2.42 1.63
15 1.60 0.96 —a 1.18 <1.00 <0.48 1.90 0.86 2.48 2.66 2.51 2.38 2.34 2.176 2.26 1.97
Medium High

2 3.02 2.97 2.95 2.36 2.60 2.97 2.79 2.88 3.70 3.36 3.60 3.36 3.20 3.36 3.18 3.63
3 3.08 3.38 3.05 2.66 2.58 3.38 2.66 2.66 3.62 3.38 3.57 3.38 3.24 3.66 2.92 3.38
9 3.16 3.04 3.02 3.04 2.79 3.04 2.76 2.66 3.57 3.18 3.80 3.66 3.13 3.18 3.29 2.66
10 2.15 2.18 2.48 1.97 1.78 2.18 2.20 1.97 3.80 3.66 3.71 2.88 2.97 2.88 3.09 2.38
1326 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005

11 3.02 2.38 3.37 3.32 2.20 2.38 2.70 2.66 3.71 3.38 3.62 4.38 2.98 3.38 3.05 3.66
12 3.15 2.63 3.19 2.63 2.34 2.63 2.87 2.36 3.95 3.97 3.76 3.63 3.38 2.63 3.38 3.18
—a
13 3.25 3.04 3.32 3.18 2.75 2.66 2.87 2.88 3.79 3.80 3.66 3.12 2.45 3.18 3.18
15 3.02 2.38 3.28 2.88 2.60 2.38 2.88 2.59 3.68 4.04 3.92 3.66 3.26 3.18 3.38 2.18

a
— = Out of range, data not used for analysis in methods comparison.
Table 5. Total coliforms and E. coli counts for milk test portions (log10 CFU/g) by SimPlate CEc-CI and AOAC methods

Uninoculated Low

Total coliforms E. coli Total coliforms E. coli

1 2 1 2 1 2 1 2

Lab SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC

1 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.38 2.38 1.90 1.97 <1.00 1.63 <1.00 0.96
2 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.87 2.97 2.85 2.97 <1.00 1.36 <1.00 0.56
3 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.08 1.97 1.78 1.88 1.90 1.97 <1.00 1.88
7 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.25 1.97 2.08 1.88 1.30 1.97 <1.00 1.88
9 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.83 2.66 2.79 1.63 <1.00 0.56 <1.00 <0.48
11 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.00 2.18 2.20 1.63 <1.00 1.63 <1.00 1.63
12 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.38 1.97 2.38 1.97 1.90 1.36 2.08 1.63
13 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.08 2.38 2.08 1.97 1.00 1.32 1.30 1.63
14 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.15 1.63 2.00 0.96 2.08 0.96 1.30 0.96
Medium High

1 2.58 2.38 2.83 2.66 1.78 1.63 2.15 2.18 3.38 3.66 3.37 3.38 1.78 2.38 1.78 2.66
2 3.35 2.97 3.46 4.04 1.30 1.63 <1.00 1.15 4.02 4.08 3.92 3.97 1.30 2.36 1.30 1.63
3 2.60 2.18 2.92 2.66 2.00 2.18 2.34 2.66 2.87 2.66 3.00 3.04 2.34 2.66 2.66 3.04
7 2.58 2.08 2.60 2.38 <1.00 2.08 2.15 2.38 3.03 3.66 3.11 3.04 2.26 3.38 2.62 2.66
9 2.93 3.04 3.13 3.04 1.78 1.97 1.30 0.56 3.38 3.38 3.41 3.66 1.60 1.97 2.08 1.97
11 2.79 2.66 2.51 1.63 1.78 1.97 2.08 1.63 3.22 3.66 3.24 3.04 2.45 2.38 2.00 1.57
12 2.60 2.18 2.60 2.38 2.08 1.97 1.90 1.97 3.11 3.38 3.08 2.97 2.70 2.32 2.66 2.38
13 2.60 2.18 2.51 2.63 1.90 1.63 1.90 2.18 3.15 2.97 3.20 3.38 2.45 2.63 2.62 2.18
14 2.60 2.66 2.66 3.04 2.41 2.18 2.00 1.97 2.85 2.97 2.97 2.97 2.34 1.96 2.48 1.56
FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005 1327
Table 6. Total coliforms and E. coli counts for nutmeats test portions (log10 CFU/g) by SimPlate CEc-CI and AOAC methods

Uninoculated Low

Total coliforms E. coli Total coliforms E. coli

1 2 1 2 1 2 1 2

Lab SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC

1 <1.00 <0.48 1.00 0.96 <1.00 <0.48 <1.00 <0.48 1.00 0.96 1.00 1.18 <1.00 0.56 <1.00 <0.48
2 1.30 <0.48 2.20 <0.48 <1.00 <0.48 <1.00 <0.48 1.30 0.96 0.65 1.36 <1.00 0.56 <1.00 <0.48
3 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 1.90 1.63 1.60 1.63 1.30 1.36 <1.00 0.56
7 <1.00 1.36 1.60 1.36 <1.00 <0.48 <1.00 1.36 1.30 1.36 1.30 1.32 <1.00 1.36 <1.00 1.04
8 <1.00 <0.48 <1.00 0.96 <1.00 <0.48 <1.00 <0.48 2.00 0.96 2.00 2.38 <1.00 <0.48 <1.00 1.18
9 <1.00 <0.48 1.60 <0.48 <1.00 <0.48 <1.00 <0.48 1.60 1.18 1.78 1.36 <1.00 <0.48 1.30 <0.48
11 2.56 0.56 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 1.60 <0.48 <1.00 <0.48 1.60 <0.48 <1.00 <0.48
12 1.30 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 1.90 1.63 2.34 2.38 <1.00 1.36 1.30 1.97
13 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 3.20 1.97 2.15 2.18 2.26 1.97 <1.00 2.18
14 <1.00 0.48 1.30 <0.48 <1.00 <0.48 <1.00 <0.48 2.56 1.97 2.08 1.36 <1.00 0.96 <1.00 0.56
Medium High
a a a a
1 1.78 1.88 1.30 1.63 <1.00 0.56 <1.00 0.56 <1.00 <0.48 3.18 3.04 <1.00a <0.48a <1.00a 3.04a
2 1.78 1.97 1.90 0.96 <1.00 0.96 <1.00 0.56 3.68 3.36 3.48 3.63 2.85 2.63 2.76 1.97
1328 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005

3 1.78 1.63 2.34 2.66 1.30 0.96 2.00 0.96 3.40 3.04 4.57 4.38 2.76 2.38 3.53 3.66
7 1.30 1.18 1.30 0.96 1.30 1.18 <1.00 0.56 2.34 2.66 2.83 3.04 1.78 2.66 2.20 3.04
8 1.78 1.63 0.65 1.63 0.65 0.56 <1.00 0.96 3.71 3.63 3.18 2.36 3.09 2.63 2.51 2.36
9 2.15 0.86 1.60 1.63 1.30 0.56 <1.00 <0.48 2.68 1.88 3.02 1.63 2.15 1.36 2.15 1.63
11 1.30 0.56 2.41 0.56 1.30 <0.48 1.30 <0.48 <1.00a 1.18 <1.00a 1.36 <1.00a <0.48a <1.00a <0.48a
12 2.73 1.97 2.48 2.38 2.00 1.97 1.60 2.38 3.51 3.38 3.19 3.18 2.87 2.97 2.62 2.97
13 2.56 3.04 2.45 2.18 1.78 1.63 1.30 2.18 3.41 3.32 3.28 3.38 2.67 3.18 2.59 3.38
14 2.30 1.46 2.38 1.97 1.30 <0.48 <1.00 1.18 3.41 3.46 3.59 3.63 3.09 2.08 2.76 1.45

a
Outlier, data not used for analysis in method comparison.
Table 7. Total coliforms and E. coli counts for cheese test portions (log10 CFU/g) by SimPlate CEc-CI and AOAC methods (run 2)

Uninoculated Low

Total coliforms E. coli Total coliforms E. coli

1 2 1 2 1 2 1 2

Lab SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC

1 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.95 2.38 3.09 2.66 2.38 1.97 2.45 2.38
2 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 3.05 2.66 3.06 3.18 2.38 1.63 2.60 2.32
3 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.88 2.38 2.98 2.38 2.48 2.38 2.34 2.38
a a a a a a a
7 <1.00 <0.48 <1.00 <0.48 — — — — 2.88 2.66 2.75 3.04 — — — —a
9 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.88 2.66 2.92 2.38 2.26 1.97 2.38 1.63
b b
11 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.68 3.04 3.33 3.04 2.15 2.38 2.41 1.36
12 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.90 2.38 2.87 2.38 2.45 1.97 2.26 2.38
13 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.87 2.66 3.05 2.38 2.34 2.38 2.26 2.38
14 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 3.09 3.38 3.13 2.66 2.34 1.18 2.08 1.45
15 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.85 3.66 2.95 3.38 2.38 2.66 2.34 1.97
Medium High

1 3.51 3.38 3.41 2.88 2.75 2.38 2.73 2.38 4.03 3.66 3.88 3.66 3.19 2.66 2.95 2.66
2 3.41 4.04 3.35 3.18 2.70 2.97 2.81 2.36 3.83 4.18 4.03 4.38 3.08 3.88 3.41 2.63
b b
3 3.53 3.38 3.47 3.04 2.79 3.38 2.62 3.04 3.00 2.38 2.78 3.18 3.26 2.38 2.95 3.18
7 3.28 3.38 3.32 3.66 —a —a —a —a 4.02 4.04 3.98 4.38 —a —a —a —a
9 3.28 3.18 3.42 3.66 2.66 2.38 2.68 2.66 3.85 3.38 3.87 4.04 3.08 3.38 3.26 3.38
11 3.33 3.38 3.19 3.04 2.73 2.38 2.87 3.04 3.51b 3.38 3.09b 3.04 3.33 2.38 3.06 2.38
12 3.42 3.38 3.53 3.32 2.73 2.63 2.81 2.38 4.06 4.63 3.93 4.38 3.20 3.86 3.38 3.18
13 3.46 3.04 3.55 3.04 2.73 2.38 2.60 2.66 3.95 4.66 4.00 4.04 3.15 3.18 2.90 2.88
b b
14 3.53 2.38 3.67 3.38 <1.00 1.46 1.78 1.45 4.28 4.38 3.83 4.38 3.26 1.18 3.08 3.32
15 3.51 3.38 3.55 3.38 2.75 2.66 2.75 2.66 3.87 3.66 3.87 3.59 3.25 3.38 3.16 2.86

a
— = Reference E. coli broth tubes enriched incorrectly; data not used in analysis.
b
Outlier, data not used for analysis in method comparison.
FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005 1329
Table 8. Total coliforms and E. coli counts for cake mix test portions (log10 CFU/g) by SimPlate CEc-CI and AOAC methods

Uninoculated Low

Total coliforms E. coli Total coliforms E. coli

1 2 1 2 1 2 1 2

Lab SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC SIM AOAC

1 2.00 0.56 1.78 0.56 <1.00 <0.48 <1.00 <0.48 1.30 2.38 1.30 1.88 <1.00 <0.48 <1.00 <0.48
2 1.30 <0.48 1.30 <0.48 <1.00 <0.48 <1.00 <0.48 1.30 <0.48 1.30 1.36 <1.00 0.56 <1.00 <0.48
a a
3 <1.00 0.97 <1.00 2.46 <1.00 <0.48 <1.00 <0.48 2.45 1.88 2.34 2.18 2.45 2.38 2.34 2.56
7 1.30 <0.48 1.90 0.56 <1.00 <0.48 <1.00 <0.48 2.81 2.38 2.56 2.38 <1.00 2.38 1.30 1.04
9 <1.00 0.56 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.08 0.96 <1.00 1.63 1.60 <0.48 <1.00 <0.48
12 1.30 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 1.30 1.36 2.51 2.66 <1.00 <0.48 1.78 <0.48
13 1.30 <0.48 <1.00 <0.48 <1.00 <0.48 <1.00 <0.48 2.00 1.36 2.08 1.36 <1.00 1.36 <1.00 <0.48
14 1.60 <0.48 1.60 <0.48 1.30 <0.48 1.90 <0.48 1.30 0.96 1.30 1.36 <1.00 <0.48 <1.00 1.36
15 1.90 0.96 1.00 0.48 <1.00 <0.48 <1.00 <0.48 1.60 1.63 1.90 1.63 <1.00 <0.48 <1.00 <0.48
Medium High

1 2.81 2.66 1.60 1.36 1.30 1.36 <1.00 0.56 4.37 4.18 3.98 4.38 4.28 4.18 3.90 3.97
2 2.56 1.04 2.34 1.97 1.30 0.56 1.78 <0.48 2.62 2.18 3.51 3.04 2.26 1.97 3.44 2.66
b b
3 3.18 1.30 3.22 1.30 3.19 3.38 3.22 — 4.39 2.45 4.03 4.04 4.39 — 3.85 —b
1330 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005

7 3.87 4.04 3.41 2.46 3.33 4.04 2.26 2.46 4.00 4.38 3.95 4.66 3.45 4.38 3.20 3.20
9 2.66 3.66 2.38 2.38 2.26 1.36 2.15 1.36 3.29 3.38 3.35 3.18 3.11 2.66 3.03 2.38
12 2.26 1.97 2.08 2.38 1.30 0.56 1.78 1.63 3.64 3.97 3.33 3.38 3.06 2.32 2.15 1.63
13 2.73 1.63 2.26 1.63 <1.00 <0.48 <1.00 0.96 3.80 3.38 3.11 2.88 3.37 2.63 2.15 1.36
14 3.31 3.04 3.09 2.66 2.00 1.63 <1.00 0.96 3.81 3.66 3.38 3.66 3.62 2.97 3.05 2.63
15 1.90 1.97 3.20 2.59 <1.00 <0.48 3.05 2.38 3.66 3.38 3.45 3.38 3.56 3.04 3.42 2.38

a
Outlier, data not used for analysis in methods comparison.
b
— = Out of range, data not used for analysis in methods comparison.
 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005 1331

medium and high levels. For the low level, Laboratory 1 data showed higher mean log counts for the low and medium levels
were determined to be an outlier by the Grubb’s test, and than did AOAC; there was no statistical difference between
therefore the paired data were not included for analyses in the the mean log counts for the high level for the 2 methods. Data
method comparison. Mean log counts recovered from test from the SimPlate CEc-CI method demonstrated lower Sr for
portions for all 3 levels were not statistically different between all 3 levels, and a lower SR value for the medium level. There
the 2 methods. The Sr and SR values of the 2 methods were was no statistical difference in SR for the other 2 levels (Table
statistically analyzed (Table 2005.03A). The Sr and SR values 2005.03A).
were also not statistically different for all 3 levels between the Data generated by the AOAC and the SimPlate CEc-CI
2 methods (Table 2005.03A). method were also compared statistically for enumeration of
For E. coli in frozen burritos, 9 laboratories submitted valid E. coli in milk. There was no statistical difference between the
data for all 3 levels of inoculation. Laboratory 1 was 2 methods in mean log counts, Sr and SR values for all 3 levels
determined to be an outlier for the low level by the Grubb’s (Table 2005.03B).
test, and Laboratory 11 was determined to be an outlier for the
high level by the Cochran method. Paired data from both Almond Nut Meats
laboratories were not included in the statistical analysis. Mean Almond nut meats were analyzed at 3 different levels
log counts recovered from test portions for all 3 levels were (Table 6). Twelve laboratorys participated in this analysis.
not statistically different between the 2 methods. The Sr and Laboratory 5 did not follow study directions. Laboratory 4 did
SR values were not statistically different between the not submit data for this analysis. These laboratories were
2 methods for low and medium levels. For the high level, excluded from statistical analysis.
SimPlate CEc-CI method had lower Sr and SR values than Data generated by the AOAC methods and the SimPlate
those of the AOAC method (Table 2005.03B). CEc-CI method were compared statistically for both total
Frozen Broccoli coliforms and E. coli. Ten laboratories submitted valid data for
all 3 levels. For total coliforms, the mean log counts for
Ten laboratories participated in the analysis of frozen almond nut meats were not statistically different for the
broccoli. Laboratory 5 was excluded because it did not follow medium and high level test portions analyzed; the mean log
study directions. Laboratory 4 did not submit data for this counts for the SimPlate CEc-CI method were significantly
analysis. higher than those of AOAC for the low level. The Sr and SR
Data generated by the AOAC methods and the SimPlate values for all 3 levels for both methods were not statistically
CEc-CI method for frozen broccoli were compared different (Table 2005.03A).
statistically (Table 4). Eight laboratories submitted valid data. For enumeration of E. coli, mean log counts, Sr and SR
For total coliforms, the mean log counts from all 3 levels values generated by AOAC and SimPlate CEc-CI methods
between the 2 methods were not statistically different. The were not statistically different for all 3 levels
Sr and SR values for the SimPlate CEc-CI method were (Table 2005.03B).
significantly lower than those of the AOAC methods for all
3 levels of inoculation (Table 2005.03A). Cheese
For E. coli in frozen broccoli, the mean log counts for all Cheese was analyzed twice. Ten laboratories participated
3 levels between the 2 methods were not statistically different. in the first run. Laboratory 5 did not follow study directions.
The SimPlate CEc-CI method had lower Sr and SR values than Laboratory 4 did not submit its study data for this analysis.
the AOAC method for the high level of inoculation. The Laboratories 2, 8, 13, and 14 reported data with no end points.
Sr and SR values were not statistically different between the As a result, there were valid data from only 4 laboratories,
2 methods for low and medium levels (Table 2005.03B). insufficient to meet the minimum laboratory requirement.
Twelve laboratories participated in the rerun of cheese.
Fluid Pasteurized Milk
Laboratory 5 did not follow study instructions. Laboratory 4
Fluid pasteurized milk was run twice. Eleven laboratories did not submit data for this analysis. Nine laboratories
participated in the first run. Laboratories 5 and 6 did not submitted valid data for the low level, 10 laboratories
follow study directions. Laboratories 12–14 were invalidated submitted valid data for the medium level, and 8 laboratories
because of contaminated diluent. Laboratory 4 did not submit submitted valid data for the high level cheese test portions.
data for this analysis. There were valid data from only Cheese test portions, inoculated at 3 different levels, were
6 laboratories, insufficient to meet the minimum laboratory analyzed (Table 7).
requirement. As a result, milk was rerun. During the second Data generated by the AOAC and SimPlate CEc-CI
run, 11 laboratories participated. Laboratory 5 did not follow methods were analyzed for statistical equivalence.
study directions. Laboratory 4 did not submit data for this Laboratory 11 was an outlier for the low level based on the
analysis. There were valid data from 9 laboratories for the Cochran test, and Laboratories 3 and 11 were outliers based
second run of milk. Data generated by the AOAC methods on the Grubb’s test for the high level. Paired data for the test
and the SimPlate method for the enumeration of total portions were not included for analyses. For total coliforms,
coliforms and E. coli in milk were compared statistically there was no difference in mean log counts between the
(Table 5). For total coliforms, the SimPlate CEc-CI method 2 methods for all 3 levels. The SimPlate CEc-CI method data
1332 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005

showed lower Sr and SR values than did those of the AOAC because an examination of the data revealed no reason to
method for all 3 levels (Table 2005.03A). exclude either data set.
For E. coli in cheese, Laboratory 7 incubated the EC broth It is generally recognized that there is a lack of precision
tubes incorrectly for all 3 levels. Data for this laboratory were inherent to the AOAC 3-tube MPN method (1). Consequently,
excluded from analysis. For the medium level, it is difficult to compare Sr values and mean log counts. This
Laboratory 14 was an outlier based on Grubb’s test and limitation in the reference method may have contributed to the
therefore paired data for these test portions were not included higher incidence of the SimPlate CEc-CI method having
for analysis. The SimPlate CEc-CI method reported a higher better Sr and SR values.
mean log count for the low level; those for the medium and
high levels were statistically not different. The SimPlate Conclusions
CEc-CI method data showed better repeatability and
reproducibility than did the AOAC method for all 3 levels In general, there was a <0.3 log count difference between
(Table 2005.03B). the AOAC method and the corresponding SimPlate CEc-CI
method for the detection and quantitation of total coliforms
Powdered Cake Mix and E. coli. For total coliforms, significantly higher mean log
counts (95% confidence level) were recovered by the
Powdered cake mix test portions inoculated at a low, SimPlate CEc-CI method for the nut meats at the low level and
medium, and high level were analyzed (Table 8). Eleven milk at the low and medium level test portions when
laboratories participated in this analysis. Laboratories 4 and 5 compared to the AOAC method. The Sr and SR values were
did not submit data. Nine laboratories submitted valid data for similar between the 2 methods, although SimPlate produced
powdered cake mix test portions. lower standard deviations in general.
Data generated by the AOAC method and the SimPlate For E. coli, the SimPlate CEc-CI method recovered higher
CEc-CI method were compared statistically. For total mean log counts for powdered cake mix at the medium and
coliforms, the difference in mean log counts for cake mix high levels, and for cheese at the low level. The Sr and SR
between the AOAC method and the SimPlate CEc-CI method values were similar between the 2 methods for E. coli,
were not statistically significant for the low, medium, and high although SimPlate produced lower standard deviations in
test portions. There was also no statistical difference in general.
repeatability and reproducibility between the 2 methods
(Table 2005.03A). Recommendations
For E. coli in powdered cake mix, Laboratory 3 was an
outlier for the low level based on Grubb’s test, and therefore The data generated in this collaborative indicate that the
paired data from both methods, for these test portions were not SimPlate Coliform and E. coli Color Indicator method is
included for analysis in method comparison. In general, the comparable to the AOAC 3-tube MPN method. It is
SimPlate CEc-CI method recovered higher mean log counts recommended that the SimPlate Coliform and E. coli Color
than did the AOAC method for the medium and high levels of Indicator method be adopted as First Action for confirmed
inoculation; there was no difference between the 2 methods for enumeration of total coliforms and E. coli in foods.
the low level. The Sr and SR values for all 3 levels were not
statistically different between the 2 methods (Table 2005.03B). Acknowledgments
In this collaborative study, the SimPlate CEc-CI method
was compared to the AOAC 3-tube MPN method for the The participation of the following collaborators is
enumeration of total coliforms and E. coli. Six food types acknowledged with appreciation:
were evaluated by the 2 methods in this study. Three levels of Jasna Braut-Taormina, Alice Cohen, and Eufemia Gonzalez,
contamination and uninoculated controls were analyzed in U.S. Food and Drug Administration, Jamaica Bay, NY
replicates. The SimPlate CEc-CI method and the AOAC Craig Braymen and Kevin Vought, Minnesota State
method for recovery of total coliforms were statistically Department of Agriculture, Minneapolis, MN
different for 3 of the 18 contaminated food lots analyzed, Cathy Chavez, Goldkist Poultry, Sumter, SC
specifically almond nut meats for the low level and fluid Amy Deans, Sarah DeLancey, and James Stephens,
pasteurized milk for the low and medium levels BioControl Systems, Inc., Bellevue, WA
(Table 2005.03A). The SimPlate CEc-CI method recovered Carol Elems, Mai-Chiung J. Huang, Lorraine M. Humes,
E. coli at a significantly higher rate than did the AOAC Judith Nogle, and Christopher Peters, U.S. Food and Drug
methods for 3 of the 18 lots analyzed, specifically for cheese Administration, Alameda, CA
at the low level and powdered cake mix for both the medium Robin Forgey, Costco Wholesale, Issaquah, WA
and high levels (Table 2005.03B). There were occurrences in Venus Hall, Miller Laboratories, Ogden, UT
which collaborators reported E. coli levels higher than the Khalil F. Kerdahi, Frederick D. Johnson, Mike Porter,
coliform bacteria levels for individual test portions by the Lonna Potter, and Greg Skorupa, U.S. Food and Drug
AOAC method only. Although E. coli levels cannot be higher Administration, Jefferson, AR
than coliform levels, the data were included as reported Jerry King, Midwest Laboratories, Omaha, NE
 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 88, NO. 5, 2005 1333

Lisa Kipker, Jessica Pinkston, and Shannon Rogers, (3) Official Methods of Analysis (2005) 18th Ed., AOAC
Goldkist Poultry, Boaz, AL INTERNATIONAL, Gaithersburg, MD, Method 966.24
Bob Taylor, Biotec Inc., Hillsville, PA
(4) Kinneberg, K.M., & Lindberg, K.G. (2002) J. AOAC Int. 85,
Erdal Tuncan and David Vrana, ConAgra Foods Inc., 56–71
Columbia, MO
(5) Silbernagel, K.M., & Lindberg, K.G. (2001) J. AOAC Int. 84,
References 1431–1443
(6) Entis, P. (1996) J. AOAC Int. 79, 1069–1082
(1) Jay, J.M. (2000) Modern Food Microbiology, 6th Ed., Aspen
(7) McClure, F.D. (2001) J. AOAC Int. 84, 301–302
Publishers, Inc., Gaithersburg, MD, Chapter 10, p. 186
(2) Official Methods of Analysis (2005) 18th Ed., AOAC (8) Youden, W.J., & Steiner, E.H. (1975) Statistical Manual of
INTERNATIONAL, Gaithersburg, MD, Method 966.23 the AOAC, AOAC INTERNATIONAL, Gaithersburg, MD