Received: 23 January 2018 Revised: 6 June 2018 Accepted: 12 June 2018

DOI: 10.1111/jfs.12502

ORIGINAL ARTICLE

Most probable number technique in Escherichia coli count using

ISO 16649-3, ISO 7251, and rapid test enumeration device
(TEMPO EC) methods in milk and dairy products
Nuray Gamze Yörük

Kocaeli Food Control Laboratory Directorate,

Microbiology Department, Kocaeli, Turkey Abstract
Correspondence The aim of the study was to compare of International Standard Organization (ISO) 7251, ISO
Nuray Gamze Yörük, Kocaeli Food Control 16649-3 most probable number (MPN) assay for the enumeration of Escherichia coli group to
Laboratory Directorate, Microbiology
the Tempo EC automated method using a variety of spiked milk and milk product foods. Foods
Department, Kocaeli, Turkey.
Email: [email protected] evaluated in this study included wet pasta (crema), milk, ayran, and ice cream. The spiked food
samples were diluted at three inoculation levels such as low, medium, and high. A comparison of
the results indicated that automated system approximated the inoculation level to an automated
system. In this study, TEMPO EC, ISO 16649-3, and ISO 7251 methods are compliant for MPN
E. coli count in high, medium, and low contamination level studies. Nevertheless, for precise
results, performing primary validation studies by ISO, Association Française de Normalisation
(AFNOR), and/or Association Of Official Analytical Chemists (AOAC) will give more reliable
results.

Practical applications
In Turkish Food Codex Microbiological Criteria Communique, it is requested to perform studies
of E. coli analysis method in milk and milk products as per ISO 16649-3. However, in the revised
version of ISO 16649-3, it is stated in the scope of the standard that this method does not pro-
vide a full assessment for all products particularly for milk and milk products and that ISO 7251
method should be used in these samples. As it is required in new methods, it has been intended
to perform comparative analyses to low, medium, and high inoculation level with rapid the
TEMPO EC card and conventional test methods by ISO, AFNOR, and/or AOAC. If the use of
TEMPO EC, laboratory response time will improve with automation reducing the time for enu-
meration of E. coli from 4 days to 24 hr, especially critical for evaluation of perishable foods and
outbreak events.

1 | I N T RO D U C TI O N
Production of food materials in hygienic conditions and present-
ing those for consumption without disrupting the hygiene chain is an
All over the world, people need a sufficient and well-balanced nutri-
important criterion in nutrition. By contamination from different
tion supported by well-supplied and safe foods in order to be healthy,
sources, microorganisms reproduce rapidly upon favorable conditions
maintain their lives, and for physical development. Production and
consumption of safe and healthy foods are also compulsory for the in the chain of operations from food production to the end customer

maintenance of vital functions. Food problems are becoming gradually and may result a decline in sensory quality, economic losses, and food-
more complex by technological developments and the increase of borne diseases (Güner, Atasever, & Atasever, 2012). Personal hygiene
food varieties in the world. Inability to provide food safety is the most is one of the most important steps in the hygiene chain in this process.
important one among these problems. Cutting boards used in food processing, slicer, mixer and grinders,

J Food Saf. 2018;e12502. wileyonlinelibrary.com/journal/jfs © 2018 Wiley Periodicals, Inc 1 of 7

https://doi.org/10.1111/jfs.12502
2 of 7 YÖRÜK

process water, ambient air, waste kept in unsuitable conditions, pests, homogenization, homogeneous sample-media mixtures were spiked at
rodents, and pets are among other contamination sources. three different levels of bacteria using 1 mL E. coli National Collection
A variety of studies have been performed in this direction by of Type Cultures, England NCTC 12923 (550 cfu/g). Homogeneous
some investigators by comparing fast and conventional methods. For sample mixtures, which were diluted with Ringer's solution and con-
example, TEMPO Escherichia coli test (TEMPO EC, bioMérieux, taminated by microorganisms, were cultured using three different
Marcy-l'Étoile, France) and TBX chromogenic agar have been com- methods according to ISO 6887-1 and 6887-5 standards.
pared to determine E. coli number in cheese (Torlak, Akan, & Gökmen,
2008). Results obtained by both methods have been determined to be
2.1 | Horizontal method studies for ISO 16649-3
compatible by Torlak et al. Previous studies performed with TEMPO
β-glucuronidase positive E. coli count
system for different microorganisms have been determined to be pos-
itive compared to other reference methods. In this study, 10, 1, and 1 g/mL were taken (to provide 1, 0.1, and

ISO 16649-3 β-glucuronidase positive E. coli count method spec- 0.01 g/mL dilutions) from the samples that were contaminated for

ifies a horizontal method for the detection and enumeration of E. coli count with MPN method under aseptic conditions according to

β-glucuronidase positive E. coli, by means of the liquid-medium culture the three-tube method; and 10 mL baseline dilution was inoculated to

technique and calculation of the most probable number (MPN) after the first three tubes (10−1) containing sterile selective enrichment

incubation at (37  1) C, then at (44  1) C. This method has not medium (mineral modified glutamate medium [LAB M- LAB080A,
been fully evaluated for all matrices (e.g., for milk and milk products). A Neogen Company, UK], sodium glutamate [LAB M-LAB080B, A Neo-
This method is applicable to the products intended for human con- gen Company, UK], ammonium chloride [Sigma-Aldrich, Riedel-de Haën,
sumption and the feeding of animals or environmental samples in the Steinheim, Germany]) that were prepared as double-power (10 mL);
area of food production and food handling. 1 mL baseline dilution (10−1) was inoculated to the second three tubes
ISO 7251 E. coli method gives general guidelines for the detection consisting of single-power selective enrichment media; and finally 9 mL
and enumeration of presumptive E. coli by means of the liquid-medium Ringer's solution was analyzed diluted to inoculate 1 mL suspension
culture technique and calculation of the MPN after incubation at 37 C, diluted with 1 mL Ringer's solution in the third of the three-tubes (10−2)
then at 44 C. This International Standard is applicable to products containing selective enrichment media prepared as a single power, and
intended for human consumption and the feeding of animals and envi- then the analysis was completed (ISO 16649-3:2015).
ronmental samples in the area of food production and food handling. The tubes were incubated at 37  C for 24 hr. The tubes that were
MPN E. coli count with rapid test enumeration device (TEMPO observed to have acid formation and turn to yellow (from violet) were
EC, bioMérieux, Marcy-l'Étoile, France) unit is signifying that the cards inoculated at 44 C for 24 hr by streaking tryptone bile glucuronide
have completed incubation; the cards were taken from the incubator agar (TBX; LAB M Harlequin TBGA HAL 003, A Neogen Company,
with their panels, placed in reader unit as per 16-layered three-tube UK) with a sterile disposable loop. After the incubation, blue–green
MPN system with first 225 μL, second 22. 5 μL, and third 2.25 μL vol- colored colonies in Petri plates were recognized to be β-glucuronidase
umes and read from the screen according to the number determined positive E. coli, and according to the positivity of the tubes, MPN/g-ml
with fluorescent radiation of microorganisms as per the three-order number was determined using the MPN table, the rating scale of ISO
system consisting of 16 wells each. 7218:2014 three-tube method.
The most important advantage of TEMPO method compared to
conventional methods is that it gives results in a shorter time. In addi- 2.2 | Horizontal method studies for ISO 7251 MPN
tion, the time spent for processing the samples by the staff is signifi- E. coli count
cantly shorter compared to that of the MPN method.
For E. coli count with the MPN method, 10, 1, and 1 g/mL contaminated
The hypothesis of this study is to present the sensitivity of three
samples (1, 0.1, and 0.01 g/mL) were taken according to three-tube
different methods for the precise identification of E. coli in milk, milk
method; and 10 mL baseline dilution (10−1) was inoculated to the first
products, and various types of that are consumed by all parts of soci-
three-tubes containing Durham tubes in sterile lauryl tryptose broth
ety and to present labor and time-consuming, and economic benefits
(LAB M-LAB 196, A Neogen Company, UK) that were prepared as dou-
of rapid test methods.
ble power (10 mL); 1 mL baseline dilution (10−1) was inoculated to sec-
ond three tubes consisting of single-power lauryl tryptose broth, and
2 | MATERIAL AND METHODS 9 mL Ringer's solution was diluted; and then the tubes were incubated
at 37 C for 24–48 hr. In the result of incubation, a passage was per-
In this study, 120 samples that does not contain E. coli (n = 30 wet cake, formed from the tubes with gas production and turbidity in Durham
n = 30 milk, n = 30 ice cream, and n = 30 ayran) were contaminated with tubes to sterile EC medium containing Durham tubes (selective medium;
NCTC 12923 strain of E. coli National Collection of Type Cultures, England, LAB M-LAB 171, A Neogen Company, UK) with sterile disposable loops
in higher, moderate, and lower levels by powers of 10, and E. coli count in the amount of one loop, and it was incubated at 44 C for 24–48 hr.
was compared using ISO 16649-3, ISO 7251, and TEMPO EC methods. At the end of this period, a passage was performed from the tubes with
On different days, 10 g/mL samples were weighed in TEMPO gas production to indole free peptone water media-tryptone water (LAB
stomacher bag under aseptic conditions, and they were homogenized M-LAB 129, A Neogen Company, UK) broths without Durham tubes
with 90 mL sterile peptone water for 1 min in the homogenizer. After using sterile disposable loops in the amount of one loop, and it was
YÖRÜK 3 of 7

incubated at 44 C for 24–48 hr. After dropping 0.5 mL Kovac's reagent
(indole reagent;Merck, Germany) to all tubes that have completed incu-
bation period, formation of red rings on the surface was assessed as
indole positive (ISO 7251:2015); according to the positivity of the tubes,
MPN/g-mL number was determined using the MPN table, the rating
scale of ISO 7218:2014 three-tube method.

2.3 | MPN E. coli count with rapid test enumeration

device (TEMPO EC, bioMérieux, Marcy-
l'Étoile, France)
For E. coli count by Tempo device using MPN method, 1 mL was taken
FIGURE 1 Comparison of E. coli results (log count per gram) obtained
from the contaminated samples, 3 mL sterile bidistilled water was added
by TEMPO EC (y axis) using (log count per gram) MPN method ISO
to pH = 7.4 media bottles containing sterile TEMPO EC (bioMérieux,
16649-3 (x axis) (•) spiked samples of medium and low contamination
Marcy-l'Étoile, France) media, the theoretical formula of which is theo-
retical formula in g/L = 9 g Bio-Soyase and food + 0.25 g growth fac- recoveries were 3.23 log cfu/g, 3.04 log cfu/g, 3 log cfu/mL, and 3.20
tors + 20.8 g MOPS (3-[N-morpholino] propanesulphonic) sodium salt + log cfu/mL, respectively. Recoveries in tests using low contamination
12.6 g MOPS acid (*) + 0.7 g sodium deoxycholate (bovine and sheep) level of 2 × 104 cfu/g were 2.34 log cfu/g, 2.18 log cfu/g, 2.18 log
+ 0.19 g substrate and enzyme regulators + 0.4 g antifoaming agent; cfu/mL, and 2.38 log cfu/mL, respectively.
they were cultured after adding 1 mL homogeneous spiked sample. The A total of 120 samples were found in the scale according to ISO
analysis type to be inoculated was introduced by lot to the preparation 7218:2014 method, and they were assessed as shown in the tables.
unit of TEMPO device, and sample name was entered. Cultured media MPN/g-mL values were converted to log base 10 before statistical
bottle was vortexed for approximately 3 s to become homogeneous. analysis. The significance level was determined as p < .05.
TEMPO EC (bioMérieux Marcy-l'Étoile, France) Kit Package Insert REF Samples were studied by three different methods, and the results
80004 cards were read after entering the sample name and by entering were assessed with linear regression analysis (R2) with regard to E. coli
dilution rate as 1/40 and 1/4, it was intended for the device to count number. The correlation between methods was demonstrated with
microorganisms even at low levels of 1 cfu/g and 1 MPN/g-mL. The linear regression graphs for wet cake in Figures 1–3, ice cream in Fig-
samples corresponding to placed cards were fed into the bottles in the ures 4–6, milk in Figures 7–9, and ayran in Figures 10–12.
preparation unit and waited for 3 min, whereas the contaminated For the comparison of TEMPO EC (bioMérieux, Marcy-
sample-media mixture was drawn from the preparation unit into wells l'Étoile, France) and ISO 16649-3 methods, 30 cake samples were stud-
by the card to cut-off transfer ends. Afterward, the cards were placed ied with higher, moderate, and lower levels by powers of 10; higher
into panels and incubated at 37 C for 24–27 hr. When a green square, values gave identical results for three methods. As a result of the study
the warning light, appears in TEMPO reader unit signifying that the
performed with lower and moderate levels, R2 0.84 value analysis results
cards have completed incubation; the cards were taken from the incuba-
are in compliance with each other as shown in Figure 1.
tor with their panels, placed in reader unit as per 16-layered three-tube
For the comparison of TEMPO EC, (bioMérieux, Marcy-
MPN system with first 225 μL, second 22. 5 μL, and third 2.25 μL vol-
l'Étoile, France) and ISO 7251 methods, 30 wet cake samples were
umes, and read from the screen according to the number determined
studied with higher, moderate, and lower levels of contamination;
with fluorescent radiation of microorganisms as per the three-order sys-
higher values gave identical results for three methods. As a result of
tem consisting of 16 wells each.
the study performed with lower and moderate levels, R2 0.76 value

2.4 | Statistical analyses

Results converted to logarithmic form. Homogeneity of variances con-
trolled with Levene’s statistic test. Pearson correlation coefficient and
linear regression tests performed using IBM® SPSS® (V. 21, Armonk,
NY, USA) (Torlak et al., 2008).

3 | RESULTS

In this study, wet cake, ice cream, ayran, and milk samples were con-
taminated with 6.5 × 106 cfu/g E. coli as a high level of contamination
in microbiological analyses, and recovery was determined as 4.98 log
FIGURE 2 Comparison of E. coli results (log count per gram) obtained
cfu/g, 4.73 log cfu/g, 4.70 log cfu/mL, and 4.83 log cfu/mL, respec- by TEMPO EC (y axis) using (log count per gram) MPN method ISO
tively. In tests using moderate contamination level of 3 × 105 cfu/g, 7251 (x axis) (•) spiked samples of medium and low contamination
4 of 7 YÖRÜK

FIGURE 5 Comparison of E. coli results (log count per gram) obtained

FIGURE 3 Comparison of E. coli results (log count per gram) obtained by TEMPO EC (y axis) using (log count per gram) MPN method ISO
by MPN method ISO 16649-3 (y axis) and (log count per gram) MPN 7251 (x axis) (•) spiked samples of medium and low contamination
method ISO 7251 (x axis) (•) spiked samples of medium and low
contamination levels, R2 0.90 value analysis results are in compliance with each other
as shown in Figure 6.

analysis results are in compliance with each other as shown in For the comparison of TEMPO EC and ISO 16649-3 methods,

Figure 2. 30 milk samples were studied with higher, moderate, and lower levels

For the comparison of ISO 16649-3 and ISO 7251 methods, by powers of 10; higher values gave identical results for three

30 wet cake samples were studied with lower, moderate, and higher methods. As a result of the study performed with lower and moderate

levels of contamination; higher values gave identical results for three levels, R2 0.91 value analysis results are in compliance with each other

methods. As a result of the study performed with lower and moderate as shown in Figure 7.
For the comparison of TEMPO EC and ISO 7251 methods, 30 milk
levels, R2 0.79 value analysis results are in compliance with each other
samples were studied with higher, moderate, and lower levels of con-
as shown in Figure 3.
For the comparison of TEMPO EC and ISO 16649-3 methods, tamination; and higher values gave identical results for all the three

30 ice cream samples were studied with higher, moderate, and lower methods. As a result of the study performed with lower and moderate

levels by powers of 10; higher values gave identical results for three levels, R2 0.93 value analysis results are in compliance with each other

methods. As a result of the study performed with lower and moderate as shown in Figure 8.
For the comparison of ISO 7251 and ISO 16649-3 methods,
levels, R2 0.92 value analysis results are in compliance with each other
30 milk samples were studied with higher, moderate and lower levels
as shown in Figure 4.
of contamination; and higher values gave identical results for three
For the comparison of TEMPO EC and ISO 7251 methods, 30 ice
methods. As a result of the study performed with lower and moderate
cream samples were studied with higher, moderate, and lower levels
levels, R2 0.91 value analysis results are in compliance with each other
of contamination; higher values gave identical results for three
as shown in Figure 9.
methods. As a result of the study performed with lower and moderate
For the comparison of ISO 16649-3 and TEMPO EC methods,
levels, R2 0.89 value analysis results are in compliance with each other
30 ayran samples were studied with higher, moderate, and lower
as shown in Figure 5.
levels by powers of 10; and higher values gave identical results for
For the comparison of ISO 16649-3 and ISO 7251 methods,
three methods. As a result of the study performed with lower and
30 ice cream samples were studied with higher, moderate, and lower
moderate levels, R2 0.85 value analysis results are in compliance with
levels of contamination; higher values gave identical results for three
each other as shown in Figure 10.
methods. As a result of the study performed with lower and moderate

FIGURE 6 Comparison of E. coli results (log count per gram) obtained

FIGURE 4 Comparison of E. coli results (log count per gram) obtained by MPN method ISO 16649-3 (y axis) and (log count per gram) MPN
by TEMPO EC (y axis) using (log count per gram) MPN method ISO method ISO 7251 (x axis) (•) spiked samples of medium and low
16649-3 (x axis) (•) spiked samples of medium and low contamination contamination
YÖRÜK 5 of 7

FIGURE 9 Comparison of E. coli results (log count per milliliter)

FIGURE 7 Comparison of E. coli results (log count per gram) obtained obtained by MPN method ISO 7251 (y axis) using (log count per
by TEMPO EC (y axis) using (log count per gram) MPN method ISO milliliter) MPN method ISO16649-3 (x axis) (•) spiked samples of
16649-3 (x axis) (•) spiked samples of medium and low contamination medium and low contamination

For the comparison of ISO 7251 and TEMPO EC methods, levels, it was observed that methods were in compliance with each
30 ayran samples were studied with higher, moderate, and lower other according to R2 analysis results as shown in the figures.
levels of contamination; and higher values gave identical results for Tekinşen, Nizamlıog
lu, Bayar, Telli, and Köseog
lu (2008) studied
three methods. As a result of the study performed with lower and on 45 Greek yogurt samples of nine different brands manufactured in
moderate levels, R2 0.83 value analysis results are in compliance with Konya with regard to E. coli count, and they have determined that
each other as shown in Figure 11. results varied between <3–11 MPN/g and that two samples were not
For the comparison of ISO 7251 and ISO 16649-3 methods, in compliance with Turkish Food Codex Fermented Milk Products
30 ayran samples were studied with higher, moderate, and lower Communique.
levels of contamination; and higher values gave identical results for Çalışkan and Törnük (2016) have analyzed 100 vanilla, chocolate,
three methods. As a result of the study performed with lower and and fruit cake samples they had collected from 10 different districts
moderate levels, R2 0.80 value analysis results are in compliance with
of Istanbul. In the result, an average level of 15 × 101 MPN/g E. coli
each other as shown in Figure 12.
has been determined in a total of 18 samples (18%).
Önganer and Kırba
g (2009) have determined the presence of coli-
form and E. coli of fecal origin in fresh cottage cheese based on
4 | DISCUSSION
“MPN” for “coliform and E. coli of fecal origin.” To determine the

In this study, 120 samples that does not contain E. coli (n = 30 wet microbiological quality of unpackaged cottage cheese released for sale
cake, n = 30 milk, n = 30 ice cream, and n = 30 ayran) were contami- in Diyarbakır, they have found that seven of these products (23.3%)
nated with NCTC 12923 strain of E. coli National Collection of Type Cul- had MPN E. coli and coliform bacteria.
tures, England, in higher, moderate, and lower levels, and E. coli count Among 100 water buffalo yogurt products provided from bazaars
was compared using ISO 16649-3, ISO 7251 and TEMPO EC, bioMér- (open-air markets) of Kayseri province and its districts, it was deter-
ieux, Marcy-l'Étoile, France methods. mined that nine (9%) had 1.85  0.18 MPN/g E. coli as a result of ana-
According to study results, 30 wet cake, ice cream, milk, and lyses (Ertaş, Al, Karadal, & Gönülalan, 2014).
ayran samples were studied with low, moderate, and high level of con- Tempo device was used in another study performed in 2007, and
tamination for the comparison of TEMPO EC, ISO 16649-3 and ISO it has been stated in the analysis of food samples that TEMPO
7251 methods; and high values gave identical results for three shortens analysis period and provides saving from time and labor of
methods. As a result of the study performed with low and medium

FIGURE 10 Comparison of E. coli results (log count per milliliter)

FIGURE 8 Comparison of E. coli results (log count per gram) obtained obtained by ISO 16649-3 (y axis) using (log count per milliliter) MPN
by TEMPO EC (y axis) using (log count per gram) MPN method ISO method TEMPO EC (x axis) (•) spiked samples of medium and low
7251 (x axis) (•) spiked samples of medium and low contamination contamination
6 of 7 YÖRÜK

equivalent results with BS ISO 21528-2:2004 (Owen, Willis, &

Lamph, 2010).
Bacterial contamination of raw milk and milk products may occur
on nipples and milking machines and may be caused due to lack of
hygiene on instruments and equipment, insufficient pasteurization,
contamination after pasteurization, quality of used raw materials and
excipients, applied process, and inadequate hygiene in the workplace;
frequent outbreaks connected with the use of these products have
been observed in recent years (Fidan & Ag
ao
glu, 2004; Karagözlü,
2011; Temelli, 2002).
FIGURE 11 Comparison of E. coli results (log count per milliliter) E. coli is one of the important indicators of potential hygiene
obtained by ISO 7251 (y axis) using (log count per milliliter) MPN problems in food processing plants. As it is a hygiene indicator, it has
method TEMPO EC (x axis) (•) spiked samples of medium and low types that can be derived from the intestinal tract of animals, and for
contamination
this reason, they are used as an evidence of cross-contamination from
laboratory staff as a result of 96–99% high correlation values raw meat or food contact surfaces and/or undercooking, apart from
observed in the comparative analysis between TEMPO E. coli and TBX pointing out businesses with poor hygienic structure. In addition, they
agar performed on naturally and artificially contaminated feta cheese, might show uncontrolled time and temperature in food production
uncooked minced meat, and frozen vegetables (Kunicka, 2007). Simi- process (Erol, 2007; Health Protection Agency, 2009).
larly, in a study performed in 2007 comparing TEMPO system and Any test designed for providing rapid response to hygiene prob-
Petri plate method for the determination of E. coli, coliform group and lems and determining intended microorganisms should ideally give a
total aerobic plate count (TAPC) in chicken samples (Cosby & Bailey, fast and safe result. Time to obtain results is important for the release
2007), it has been revealed by the authors that TAPC results are com- of products in the shortest time possible after manufacturing and
pliant between TEMPO and Petri plate methods. It was stated that should especially gain the attention of manufacturers of products with

there is a correlation between TEMPO and cultural MPN method for short shelf lives.

E. coli and coliform group microorganisms; however, there is no such

conformity between TEMPO and Petri film methods. The authors
5 | CONC LU SIONS
have attributed this to insufficiency of Petri plates to distinguish low-
level background flora.
Upon the review of all studies performed for providing safe food pro-
Moreover, Enterobacteriaceae was compared with TEMPO and
duction, it is observed that the levels determined for E. coli in milk and
conventional method in artificially and naturally contaminated sam-
milk products are high and this brings to mind that they are not safe
ples, R2 0.78 value was determined as a result of the regression analy-
for human health. It has been suggested that these food products may
sis on 67 samples using TEMPO and MPN tube method (HPA method
have negative effects on health because this problem involves milk,
F18); R2 0.75 value was determined as a result of regression analysis
milk products, and various types of which are consumed by nearly all
on 47 samples using TEMPO and pour plate method (HPA method
parts of society. As a fecal contamination criterion, E. coli reflects inad-
F23); and R2 0.78 value was determined as a result of regression anal-
equate hygiene and sanitation conditions for public health in milk and
ysis on 84 samples using TEMPO and spread plate method. In this milk products. For this reason, studies on MPN E. coli should be per-
sense, studies have demonstrated that there is no statistical difference formed more frequently with sensitive methods.
between TEMPO results and three alternative methods (MPN Turkish Food Codex Microbiological Criteria Communique (Anonim,
method, BS ISO 21528:1:2004, pour plate method, BS ISO 2011) also indicates the right to take samples and perform analysis in a
21528:2:2004 and spread plate method based on BS ISO more detailed manner for microorganisms that are not stated in Annex-
21528:2:2004). The studies have revealed that TEMPO EB gives or risk analysis on food products with suspicious safety. In this respect,
MPN method and analysis are also important in the microbiological ana-
lyses of high-risk foods, apart from cultural methods. Literature studies
on E. coli count with MPN using ISO 16649-3, ISO 7251 and rapid test
Eenumeration methods (TEMPO EC) in milk and milk products have
been reviewed, and no study was found up to this date that compares
these three methods for MPN E. coli count.
In E. coli count analyses with MPN Method, the conformity of results
have been statistically studied between conventional methods ISO
16649-3, ISO 7251, and TEMPO EC used as rapid test method. Although
it was stated in TEMPO EC Kit Package Insert REF 80004, summary and
FIGURE 12 Comparison of E. coli results (log count per gram)
obtained by MPN method ISO 7251 (y axis) using (log count per gram) remarks that the kit has been developed to provide equivalent perfor-
MPN method ISO16649-3 (x axis) (•) spiked samples of medium and mance as per AOAC Official Methods 966.23 and 966.24 standards and
low contamination works in compliance with ISO 16649-2 method (E. coli colony count in
YÖRÜK 7 of 7

TBX agar) by being ISO-approved; it was determined in this study that Ertaş, N., Al, S., Karadal, F., & Gönülalan, Z. (2014). Kayseri İlinde Satışa
TEMPO EC, ISO 16649-3, and ISO 7251 methods are compliant for MPN Sunulan Manda Yog urtlarının Mikrobiyolojik Kalitesi. Journal of Faculty
Veterinary Medicine Istanbul University, 40(1), 83–89.
E. coli count in high, moderate, and low contamination level studies. Nev- Fidan, F., & Ag aoglu, S. (2004). Agrı Bölgesinde Bulunan Lokantaların
ertheless, for precise results, performing primary validation studies by Hijyenik Durumu Üzerine Araştırmalar. Journal of Yüzüncü Yıl University
ISO, AFNOR, and/or AOAC will give more reliable results. Veterinary Faculty, 15(1–2), 107–114.
Güner, A., Atasever, M., & Atasever, A. M. (2012). Yeni Ortaya Çıkan ve
The most important advantage of TEMPO method compared to con-
Tekrar Önem Kazanan Gıda Kaynaklı Bakteriyel Patojenler. Journal of
ventional methods is that it gives results in a shorter time. In addition, the Kafkas University Veterinary Faculty, 18(5), 889–898.
time spent for processing the samples by the staff is significantly shorter Health Protection Agency. (2009). Guidelines for assessing the microbiologi-
cal safety of ready-to-eat foods placed on the market. Retrieved from
compared to that of MPN method. The lack of approval steps required by
http://www.hpa.org.uk/webw/EPAweb& EPAwebStandard/EPAweb
TEMPO provides saving on the time of staff and the cost of media. In addi- Karagözlü, N. (2011). Gıda Kaynaklı Toksikoenfeksiyonlar. Gıda Mikrobiyo-
tion to delivery costs of large liquid media in vials from the manufacturing lojisi, Erkmen O. Eflatun Basım Da gıtım Yayıncılık Danışmanlık Yatırım
place to the laboratory, media preparation time is also shorter. ve tic. Ltd. Şti. Ankara, 153–171.
Kunicka, A. (2007). Evaluation of the TEMPO system: An automated
method for food microbiological quality control. Journal Biotechnology,
131, S69–S72.
Önganer, A. N., & Kırba g, S. (2009). Diyarbakır'da Taze Olarak Tüketilen
ACKNOWLEDGMENT Çökelek Peynirlerinin Mikrobiyolojik Kalitesi. Erciyes University Fen
The author would like to express his gratitude for Kocaeli Food Con- Bilimleri Enstitüsü Dergisi, 25(1–2), 24–33.
Owen, M., Willis, C., & Lamph, D. (2010). Evaluation of the TEMPO most
trol Laboratory Directorate Manager Veterinary Medicine Hasan
probable number technique for the enumeration of Enterobacteriaceae
AKYILDIZ and also for Dr. Erdem Danyer. in food and dairy products. Journal of Applied Microbiology, 109,
1810–1816. https://doi.org/10.1111/j.1365-2672.2010.04810
Tekinşen, K. K., Nizamlıo glu, M., Bayar, N., Telli, N., & Köseo glu, İ. E.
(2008). Konya'da Üretilen Süzme (Torba) Yog urtların Bazı Mikrobiyo-
ORCID
lojik ve Kimyasal Özellikleri. Journal of Veterinary Science, 24(1),
Nuray Gamze Yörük http://orcid.org/0000-0003-0867-4141 69–75.
Temelli, S. (2002). Gıda Zehirlenmesine Neden Olan E. coli O157:H7 ve
Önemi. Uludag University Journal of The Faculty of Veterinary Medicine,
21, 133–138.
RE FE R ENC E S Torlak, E., Akan, I. M., & Gökmen, M. (2008). Comparison of TEMPO EC
Anonim. (2011). Turkish Food Codex Microbiological Criteria Communi- and TBX medium for the enumeration of Escherichia coli in cheese.
que, Official Newspaper (29/12/2011), 3rd place, 28157. additional Letter Application Microbiology, 47, 566–570.
page 1.6.9.2 (tarts and wet cakes).
Çalışkan, E., & Törnük, F. (2016). İstanbul İlinde Satışa Sunulan Pastalarda
Patojen Bakteri Varlıgının İncelenmesi. Yıldız Teknik University Fen
Bilimleri Enstitüsü Gıda Mühendislig i Yüksek Lisans Tezi. How to cite this article: Yörük NG. Most probable number
Cosby, D. E., & Bailey, J. S. (2007). Comparison of the TEMPO system, Pet-
technique in Escherichia coli count using ISO 16649-3, ISO
rifilm and cultural MPN procedure for enumeration of E. coli, coliforms
and total aerobic plate counts. International Association for Food Protec- 7251, and rapid test enumeration device (TEMPO EC)
tion, 161, 2–58. methods in milk and dairy products. J Food Saf. 2018;e12502.
Erol, İ. (2007). Gıda Hijyeni ve Mikrobiyolojisi. Baskı. Ankara: Pozitif Matbaa- https://doi.org/10.1111/jfs.12502
cılık ltd. Şti.