Food Control 50 (2015) 548e553

Contents lists available at ScienceDirect

Food Control
journal homepage: www.elsevier.com/locate/foodcont

An improved method for microbiological testing of paper-based

laminates used in food packaging
Marlies Feichtinger a, Ulrike Zitz a, *, Helmut Fric b, Wolfgang Kneifel a, Konrad J. Domig a
a
BOKU e University of Natural Resources and Life Sciences, Vienna, Department of Food Sciences and Technology, Institute of Food Science, Muthgasse 18,
A-1190 Vienna, Austria
b
Constantia Teich GmbH, Quality Management, Mühlhofen 4, A-3205 Weinburg, Austria

a r t i c l e i n f o a b s t r a c t

Article history: Food packaging materials fundamentally contribute to food quality and safety, as they protect the
Received 30 July 2014 packaged food against external influences. In this context, the determination of the hygiene status of the
Received in revised form packaging material is of great importance. However, European legislation neither sets any microbio-
15 September 2014
logical criteria nor provides any approved standard for the microbiological testing of food packaging
Accepted 1 October 2014
Available online 13 October 2014
materials. Nevertheless, reliable routine control is essential for guaranteeing high hygienic quality of
packagings.
With the aim to achieve a maximum recovery rate at low contamination levels, an improved exper-
Keywords:
Food packaging
imental design was developed for the enumeration of the total colony count, yeasts and molds and
Hygiene Enterobacteriaceae on the surface of roll stock packaging materials. For this purpose, two different types
Paper laminates of paper laminates were selected and exemplarily used as objects of investigation. Moreover, the per-
Culture method formance of different growth media was compared for each microbiological parameter. This approach
was followed by method validation using a selection of quantitative reference materials of representative
microorganisms, including resistant forms of microbes such as bacterial endospores and fungal spores.
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction forms of microbes on packaging materials (Pirttij€arvi, Andersson, &

Packaging materials play a major role in ensuring microbiolog-
ical food safety as they pose a physical barrier against various
environmental contaminants. In addition, they also fulfill an
important function in protecting the packaged food from light,
oxygen and humidity, thus contributing to prolonged shelf-life
(Marsh & Bugusu, 2007). However, a high hygienic quality of the
packaging materials is required for excluding them as a potential
source of microbiological contamination for the filling goods.
Today, there exists neither a European regulation setting
microbiological threshold levels for food contact materials, nor any
internationally approved standard method for the microbiological
control of these materials. Although the microbial load of most
packaging materials is usually low, the predominance of resistant
Abbreviations: TSA, Tryptone Soya Agar; VRB, Violet Red Bile Agar; PCA, Plate
Count Agar; DRBC, Dichloran Rose Bengal Chloramphenicol Agar; SAB, modified
Sabouraud 1% Glucose 1% Maltose Agar; RM, reference material; FCS, food contact
side.
* Corresponding author. Tel.: þ43 1 47654 6634.
E-mail address: ulrike.zitz@boku.ac.at (U. Zitz).
Salkinoja-Salonen, 2000; Suihko & Stackebrandt, 2003; V€ aisa€nen,
Mentu, & Salkinoja-Salonen, 1991), such as bacterial endospores or
fungal spores at very low contamination levels, is very challenging
for the analytical surveillance. Therefore, under practical condi-
tions, an important focus should be set on the prevention of mi-
crobial entrance portals along the entire supply chain to avoid even
low microbiological contamination levels of the packaging mate-
rials. Thus it is of importance to strictly comply with good
manufacturing practices (GMP) (Raaska, 2005).
As far as packaging materials containing paper are considered,
such as paper-polyethylene terephthalate (PET) and paper-
aluminum laminates, the quality of the commodity paper is deci-
sive for the microbiological quality of the final product (Cerny &
Betz, 1999), since paper usually contains a certain microbial load
that is not completely eliminated during the production process.
Compared to other materials such as metals, glass or plastics, rather
low temperature treatment is applied to those laminates (Bergmair,
Washüttl, & Wepner, 2010). Experience has shown that undesired
exposure of the paper-layer, e.g. by faulty lamination, fosters the
release of embedded microbes and thus may pose a hazard for
spoilage of the packaged food. Moreover, bacterial and fungal

http://dx.doi.org/10.1016/j.foodcont.2014.10.002
0956-7135/© 2014 Elsevier Ltd. All rights reserved.
 M. Feichtinger et al. / Food Control 50 (2015) 548e553
spores, originating from the surrounding air or from cutting pro-
cedures of the laminates, may adhere to the packaging material
€rvi, Graeffe, & Salkinoja-Salonen, 1996). Hence not
surfaces (Pirttija
only quality losses of the product may result, but also some theo-
retical health risk to the consumer.
However, it should not be forgotten that usually process hygiene
criteria of food production possess some higher relevance in terms
of food safety than the contamination risk by paper-based pack-
aging materials (Ekman et al., 2009).
Microbial growth in foods strongly depends on the food in-
gredients and on environmental factors. Especially food products
with a high water activity (aw-value) and a high density of nutrients
like, e.g., yoghurts and dairy desserts, are prone to microbial spoilage.
Even if spores find suited environmental conditions, they are able to
germinate and to regrow upon getting contact with food of sufficient
aw-value. In contrast, dry foods with a low aw-value, such as choco-
late with an average aw of 0.4 (Copetti, Iamanaka, Frisvad, Pereira, &
Taniwaki, 2011), possess a significantly reduced risk to undergo mi-
crobial spoilage. Most gram-positive microorganisms require a
minimum aw-value of around 0.90 to grow, while gram-negative
bacteria prefer higher humidity (aw 0.97). Mold growth, however,
is already possible at an aw-value of 0.80 (Adams & Moss, 2008).
For the microbiological assessment, only a few official methods
exist. For example, DIN 10 050-3:1999 (Deutsches Institut für
Normung, 1999) describes the determination of the microbial
count of butter wrappers, and DIN 54 378:1993 (Deutsches Institut
für Normung, 1993) deals with the determination of the surface
colony count (yeasts and molds) of paper and board. The latter
method is also part of a recommended procedure outlined by the
German Industry Association for Food Technology and Packaging
for the microbiological testing of non-absorbent material surfaces
(Arbeitsgruppe “Lebensmittelerhaltung und Mikrobiologie,” 1974).
This method has been widely applied for routine purposes among
many packaging material producers.
In this document (German designation: “Merkblatt 21”) a cul-
tural technique for the determination of the total aerobic colony
count, of yeasts and molds and of Enterobacteriaceae is described.
According to this guideline, packaging materials first are cut into a
format that fits into standard petridishes (diameter 94 mm) by
using round templates. The petridishes are filled with a thin layer of
culture medium, which varies according to the microbiological
parameter. For the total colony count, Nutrient agar is used, while a
modified Sabouraud medium containing 1% Glucose and 1%
Maltose agar is used for yeasts and molds, and Violet Red Bile agar
for Enterobacteriaceae. After solidification of the medium, the cut
sample is put onto the agar and a second layer of medium is poured
over the sample.
This method shows some general as well as some specific de-
ficiencies regarding paper-based laminates: the procedure is time
consuming, and the assessment of the recommended sample area
of 100 cm2 cannot be achieved when using standard petridishes
with a diameter of 94 mm. Additionally, undesired growth of mi-
crobes originating from the cutting edges may hinder the unam-
biguous enumeration of the colonies. Also, the overlaying process
sometimes causes some curling of the laminate sample and hence
additional equipment is necessary to fix the sample into a correct
position. This again increases the risk of re-contamination and may
thus lead to false-positive results.
In order to overcome the above mentioned problems, this study
was undertaken to improve the cultural method for the microbio-
logical assessment of laminated packaging materials. The particular
challenge was to simplify the working procedure and to enhance
method robustness considerably with respect to the observed
deficiencies of the currently used routine method, especially when
considering roll stocks of paper-based laminates.
549
2. Materials and methods
2.1. Materials
2.1.1. Quantitative reference materials
A selection of BioBall MultiShot-550 products (BTF, Sidney,
Australia), which are accredited reference materials (RM) under ISO
Guide 34 standard, was used for the microbiological spiking ex-
periments. According to the corresponding Certificates of Analysis
(BTF, n.d.), these quantitative reference materials contain a precise
number of colony-forming particles of a specified type culture
strain taken from the ATCC reference culture collection (ATCC,
American type culture collection), ranging between 500 and
600 cfu per BioBall. For method validation, the following microor-
ganisms were selected: Bacillus subtilis, ATCC 6633 (Lot no.: B1978
containing a stated reference value per BioBall of 568.7 ± 46.1 cfu
examined on Nutrient agar after 24 h aerobic incubation at 37  C);
Escherichia coli, ATCC 8739 (Lot no.: B1961 with a reference value of
588.5 ± 33.6 cfu on Nutrient agar after 24 h aerobic incubation at
37  C); Aspergillus brasiliensis, ATCC 16404 (Lot no.: B1982 having a
reference value of 548.5 ± 43.8 cfu on Dichloran Rose Bengal
Chloramphenicol (DRBC) agar after 48 h aerobic incubation at
37  C). A strict protocol (BTF, 2014) was followed for each BioBall re-
hydration procedure to ensure a homogenous suspension for
providing aliquots with a defined composition for each experiment.
2.1.2. Packaging material samples
Two different types of paper laminates, a paper-PET laminate
and a paper-aluminum laminate, served as sample material and
were examined for their microbial contamination prior to and after
spiking. The paper-PET laminate (termed as PaP in the following)
uses a metalized PET layer as food contact side (FCS) and is usually
intended for pot sealing of yoghurts and chilled dairy desserts. It is
delivered as large reels to the customers e usually food producers
or bottlers e who cut the material right after sealing it to the plastic
cups containing the food. The paper-aluminum laminate (termed as
PaA), which is used for wrapping chocolate bars, uses the paper
layer as FCS. Considering the risk of microbial growth, PaP is of
highest priority regarding food safety and was therefore selected
for the method development and validation with quantitative
reference materials.
2.2. Methods
2.2.1. Sampling
PaP samples were drawn right after the final cutting process in
the manufacturing site. Several layers of PaP were cut from the
large roll that is delivered to the customers, immediately wrapped
in polyethylene foil and transferred to the laboratory.
2.2.2. Method according to “Document 21”
The traditional method according to the guideline given in
“Document 21” is illustrated in Fig. 1 (on the left). For the deter-
mination of the total aerobic colony count, Nutrient agar was
replaced by Plate Count agar (PCA; Merck, Ref. 1.05463.0500) since
this culture medium has been recommended in several ISO stan-
dards, e.g., ISO 4833-1:2013 (International Organization for
Standardization, 2013). For the enumeration of yeasts and molds,
modified Sabouraud 1% Glucose 1% Maltose agar (SAB; Dinkelberg,
Ref. DB15970.0500) and for Enterobacteriaceae, Violet Red Bile agar
(VRB; Merck, Ref. 1.01406.0500) were used according to the given
guideline. In order to avoid a contamination of the samples, the
following steps were carried out in a laminar flow work bench.
Round samples of 40 cm2 e suitable to fit into a conventional
petridish e were cut using sterile scissors. To weigh down this
550 M. Feichtinger et al. / Food Control 50 (2015) 548e553

sample after placing it onto the first layer of the culture medium
with sterile tweezers, a sterile glass ring was used and further
covered by a second layer of medium. According to the given
guideline, the obtained colony counts were finally multiplied
with a factor of 2.5 to refer them to a sample area of 100 cm2.

2.2.3. Improved method

Several laminate sheets (PaP) with an optimum size of
30  50 cm were cut out of the PaP layers taken in the production
site. One sheet originating from the middle of the sample stack was
taken out with sterile tweezers and placed onto a stainless steel
tray (Rist PROFI, Ref. 3178102; 530  325  20 mm) with the FCS up
in a laminar flow work bench. Six sterile purpose-built stainless
steel frames (110  110  7 mm) were placed onto this sheet on the
tray, and subsequently the frames were filled with culture medium.
To avoid unwanted microbial growth from the cutting edges, a
proper distance between the steel frames and the end of the sample
was kept. Then the sample areas were covered with either the
bottom or the lid of a petridish of square format (Greiner bio-one,
Ref. 688161; 120  120  17 mm) and incubated under defined
conditions (PCA: 30  C for 3 days; SAB: 25  C for 5 days) after so-
lidification of the medium. For the enumeration, all single colonies
were counted. Spreading colonies were also registered as single
colonies (according to Commission Decision 91/180/EEC
(Anonymous, 1991)). If the obtained result has to be referred to an
area of 100 cm2, square templates (10  10 cm) may be applied. The
individual steps of the procedure of the improved method are also
illustrated in Fig. 1 (on the right). Furthermore, the image in Fig. 2
visualizes the assembly used for the optimized microbiological
examination of paper-based laminates.
Fig. 2. Image of the final assembly used for the improved microbiological assessment
of paper-based laminates.
2.2.4. Assessment of the microbiological status of PaP and PaA
Initially, the hygiene status of four PaP sample batches had to be
examined in order to be able to select those lots for the spiking
experiments showing no detectable microbial contamination. For
the total aerobic colony counts, PCA was used as culture medium,
and samples prepared as described above were incubated at 30  C
for 3 days. Likewise, yeasts and molds were cultured on SAB at 25  C
for 5 days, and E. coli was examined using VRB incubated at 37  C
for 24 h, but only in case of any growth detected on PCA. In total,
twelve samples per batch (PaP-1, PaP-2, PaP-3, PaP-4) and per
growth medium (PCA, SAB) were investigated. Additionally, two
batches of PaA were tested for the same parameters.
2.2.5. Method validation using quantitative reference materials
For simulating a microbiological contamination of PaP with
microorganisms from different possible sources, spiking experi-
ments were performed using BioBall MultiShot-550 reference

Fig. 1. Flow diagram describing the steps of the traditional (according to Arbeitsgruppe “Lebensmittelerhaltung und Mikrobiologie,” 1974) and the improved test method for the
microbiological assessment of paper-based laminates.
 M. Feichtinger et al. / Food Control 50 (2015) 548e553 551

materials. For this, B. subtilis ATCC 6633 was chosen as represen-
tative for gram-positive spore-formers originating from paper
€rvi et al., 1996, 2000; Va
(Pirttija €isa
€nen et al., 1991; Va €isa
€nen,
Nurmiaho-Lassila, Marmo, & Salkinoja-Salonen, 1994), Asp. brasi-
liensis ATCC 16404 from the processing environment air and E. coli
ATCC 8739, a quality control strain, as hygiene indicator of human
origin.
Reflecting the usually low contamination rate of PaP and in or-
der to provide low numbers of microorganisms (approximately
5 cfu), aliquots of 10 ml of the RM e suspension were used. For this
purpose, one BioBall was suspended in 1.1 ml of Re-Hydration Fluid
(bioMe rieux, Ref. 410386, Lot no.: B124) as described in the in-
structions for use (BTF, 2014). A piston stroke pipette (Gilson
Pipetman pipette, variable volume of 1e10 ml, Gilson Inc., USA) with
sterile pipette tips (10 ml, Biozym) was used for all spiking experi-
ments, in which the 10 ml aliquots were applied as small dots over
the sample area of 100 cm2 and subsequently covered with growth
medium within 2 min.
The validity of the culture media and incubation conditions
prescribed in “Document 21” has to be proven in comparison to the
Certificate of Analysis of the BioBalls (BTF, n.d.), since only under
these cultivation conditions the assigned cfu reference values are
guaranteed.
TSA and PCA were compared regarding the growth behavior of
B. subtilis and incubated aerobically at 30  C for 20e28 h,
regarding E. coli (aerobic incubation at 37  C for 20e28 h), PCA and
VRB were evaluated, and for Asp. brasiliensis (aerobic incubation at
25  C for 3e5 days), DRBC and SAB were compared. For that
purpose, aliquots of 10 ml of the RM e suspension were transferred
into a standard petridish (94 mm diameter). The experiment was
performed with 20 samples per microorganism and culture me-
dium. To ensure a constant cfu value per aliquot, the RM e sus-
pension was vortexed for 5 s (2000 rpm/min) before each step.
The inoculated suspension was then covered with approximately
20 ml of culture medium poured into the petridishes, mixed and
incubated.
For the spiking experiments with selected PaP samples, the RM
e suspensions were prepared as described above and using the
method with the steel frames. Five of the defined sample areas on
the sheet were used for spiking and one served as negative control.
Before filling the frames with the media, each inoculated portion of
the RM e suspension was allowed to dry for 2e3 min before being
filled with the culture medium and covered with a petridish of
square format. The trays with the spiked samples were incubated
for 20e28 h at 30  C (B. subtilis) and 37  C (E. coli), respectively, and
for 3e5 days at 25  C (Asp. brasiliensis). Colonies were counted as
described in Section 2.2.3.
2.2.6. Statistical evaluation of the results
Both a non-parametric test method (Wilcoxon matched-pairs
test) and a parametric test method (Bland-Altman method com-
parison) were used for paired sample comparison in order to decide
if there are any significant differences between the media tested for
each microorganism. Method comparison is usually based on the
differences between observations made on the same subjects by
performing the microbiological methods in parallel, where the
differences are considered as being normally distributed (Altman &
Bland, 1983; Bland & Altman, 1986). Furthermore, statistical eval-
uation was done on the basis of the decadic logarithms of the
colony counts by BiAS for Windows, Version 9.12-05/2011, ©
epsilon 2011.
3. Results and discussion
3.1. Experimental arrangement
In order to avoid cutting of samples, a large sheet of the pack-
aging material was directly put onto a sterile stainless steel tray,
whereas sample areas could be defined by using stainless steel
frames of a given size. Subsequently, the culture medium was
poured into the frame and the assembly was covered with a square
petridish.
3.2. Method validation
Basically, the hygiene status of the selected PaP samples from
four different batches displayed very low microbial growth. Neither
total aerobic bacterial colonies nor yeasts and molds were detected
on PaP-1 and PaP-2. Hence these batches were selected for the
subsequent spiking experiments. The average contamination level
of the four batches was 0.02 cfu per 100 cm2 (total colony count)
and 0.2 cfu per 100 cm2 for yeasts and molds. However, the mold
counts varied significantly within the batches.
The two batches of PaA had an average aerobic colony count of
31 cfu per 100 cm2 on the paper side, but showed almost no growth
(0.25 cfu per 100 cm2) on the aluminum layer. In analogy, yeasts
and molds were only found on the paper side (non-FCS) of PaA-2,
with a mean contamination level of 0.92 cfu per 100 cm2.
All organisms found within the group of total colony counts,
both for PaP and PaA, appeared to belong to the gram-positive
spore-formers. This finding is also in agreement with the litera-
ture (V€ais€
anen et al., 1991). Moreover, as another previous study
has shown, the transfer rate of Bacillus endospores from paper
packaging to dry foods like chocolate or rice is usually very low and
thus need not be considered as food safety risk (Ekman et al., 2009).

Table 1
Recovery results of the reference material-suspension aliquots, each containing a precise number of colony-forming particles of one of the reference strains of B. subtilis, E. coli
and Asp. brasiliensis, when applied to the test system as artificial contaminants (by spiking the foil or without the foil) and using different culture media.

Bacillus subtilis ATCC 6633 Escherichia coli ATCC 8739 Aspergillus brasiliensis ATCC 16404

PCA PCA þ foil TSA TSA þ foil TSA PCA þ foil VRB VRB þ foil DRBC DRBC þ foil SAB SAB þ foil

Reference 5.17 5.17 5.17 5.17 5.35 5.35 5.35 5.35 4.99 4.99 4.99 4.99
valuea
Sample size, n 20 20 20 20 20 20 20 20 20 20 20 20
Average 4.30 ± 1.34 3.75 ± 1.73 5.50 ± 2.67 5.20 ± 2.29 4.75 ± 1.62 3.50 ± 2.10 0 0.35 ± 1.98 5.00 ± 2.25 5.20 ± 1.98 3.70 ± 2.23 4.35 ± 1.98
value ± SDb
Recovery 83 73 106 101 89 65 0 7 100 104 74 87
rate [%]

Abbreviations: PCA, Plate Count Agar; TSA, Tryptone Soya Agar; VRB, Violet Red Bile Agar; DRBC, Dichloran Rose Bengal Chloramphenicol Agar; SAB, modified Sabouraud 1%
Glucose 1% Maltose Agar.
a
The reference value of each BioBall MultiShot-550 reference material suspension (RM-suspension) is expressed as cfu per aliquot according to the corresponding Cer-
tificate of Analysis (BTF, n.d.).
b
The average value of colony counts obtained with all RM-suspension aliquots with respect to the culture media tested with and without foil (selected PaP samples) and
corresponding standard deviation SD [cfu per aliquot].
552 M. Feichtinger et al. / Food Control 50 (2015) 548e553
Table 2
Results of the statistical evaluation (Bland Altman and Wilcoxon matched pairs test) of the comparison experiments performed to assess the growth performance of the culture
media.
Escherichia coli ATCC 8739c
Culture medium (sample size, n) TSA (26) VRB (26) PCA (20)
Bland Altman method comparison
Growth [%]a 100 50 (38e64) 100
Significanceb 0.000007 0.029
Wilcoxon matched pairs test
Growth [%]a 100 55 (42e67) 100
Significanceb 0.000000 0.008
Abbreviations: TSA, Tryptone Soya Agar; VRB, Violet Red Bile Agar; PCA, Plate Count Agar; DRBC, Dichloran Rose Bengal Chloramphenicol Agar; SAB, modified Sabouraud 1%
Glucose 1% Maltose Agar.
a
The growth performance is expressed as proportion of microbial growth on the medium showing lower cfu results and corresponding 95% confidence interval (within
brackets) with respect to the medium to be compared.
b
The significance of the difference between the cfu results (expressed as logarithms) with the media that are compared, is given with 95% confidence, p (95%).
c
For statistical evaluation only cfu results with resuscitated overnight cultures of E. coli RM-suspensions were used.
In order to check whether PCA or TSA is better suited for
culturing bacterial spore-formers, these two media were compared.
Corresponding results are shown in Table 1 and indicate that both
roughly yield the same recovery of B. subtilis as specified for the
microbial reference material in the Certificate of Analysis (BTF, n.d.).
However, the use of TSA resulted in a slightly higher recovery rate.
Asp. brasiliensis tended to produce higher colony count numbers on
DRBC than on SAB. E. coli showed good growth performance on TSA,
but interestingly no growth on VRB. Obviously, the selective com-
ponents of this medium may exert inhibitory effects on sub-lethally
damaged E. coli cells, at least as the control strain used is concerned.
Alternatively, we rather recommend to use a non-selective growth
medium like PCA or TSA for the assessment of the total colony
count and then to sub-culture suspected colonies on VRB agar in
order to verify them as members of the Enterobacteriaceae family.
Anyway, the levels of a contamination with gram-negatives on
packaging materials can, in general, be regarded as low, since these
bacteria show little resistance against typical conditions applied
during packaging material manufacture.
In accordance with the recovery results obtained with the
different media presented in Table 1, the spiking experiments car-
ried out with selected PaP samples were similar to the comparison
experiments without any samples and also the same influences of
the media on the individual growth performance of the test strains
became evident.
3.3. Statistical evaluation
The data sets were merged for statistical analysis where
appropriate in order to enhance the statistical power of the tests by
increasing the sample number. As a prerequisite, all comparison
experiments were based on the same study design. All colony count
results were transferred into their logarithms to provide normally
distributed values.
Regarding the media-specific growth performance of B. subtilis
as a representative of gram-positive bacteria producing endo-
spores, the two media, TSA and PCA, were compared with both the
parametric Bland-Altman method comparison and the non-
parametric Wilcoxon matched pairs test. As shown in Table 2,
both showed a statistically significant difference in favor of TSA.
However, the lower recovery rate obtained with PCA could be the
result of better visibility of colonies on TSA than on PCA. According
to additional observations, colonies on PCA more often appeared as
spreading colonies than on TSA.
The results of the comparison experiments carried out with the
media used for the detection and enumeration of molds based on
Bacillus subtilis ATCC 6633 Aspergillus brasiliensis ATCC 16404
VRB (20) TSA (40) PCA (40) DRBC (60) SAB (60)
69 (50e96) 100 79 (64e97) 100 82 (70e97)
0.024 0.023
71 (57e89) 100 79 (63e100) 100 82 (71e99)
0.036 0.019
the use of Asp. brasiliensis as a test strain, are also shown in Table 2.
Both statistical assessments clearly showed a significant difference
in favor of DRBC. Besides the higher recovery rate, DRBC has several
additional advantages: this growth medium is internationally
approved and widely accepted by several standards for the inves-
tigation of different foodstuffs for yeasts and molds (e.g. ISO 21 527-
1:2008 (International Organization for Standardization, 2008)).
Furthermore, the pink coloration of DRBC facilitates a better
detection and enumeration of mostly white and light-yellow
molds.
Regarding the influence of the culture medium on the growth
characteristics of E. coli, it could be shown that there is a significant
difference in favor of the non-selective medium, since the selective
VRB medium showed almost no recovery of this microorganism
(Table 1). However, for the statistical evaluation, cfu counts unequal
to zero are required and thus the experiments had to be repeated
with resuscitated overnight cultures of E. coli RM (Table 2).
4. Conclusions
There is some considerable need for having methods available
for the improved microbiological examination of paper-PET lami-
nates and other roll stock packaging materials. Based on the new
experimental design and on the validation using representative
quantitative reference materials a powerful procedure could be
developed. In particular, the problem of microbial growth origi-
nating from the exposed paper layer resulting from cutting the
laminate is avoided. In addition, the risk of re-contaminating the
samples during analytical preparation is significantly reduced.
Furthermore, the simplified working procedure is time saving and
considerably enhances method robustness.
In conclusion, the developed analytical approach was shown to
be capable of achieving a maximum recovery rate at low contam-
ination levels with the microorganisms tested, reflecting the
generally low contamination rate of food packaging materials.
References
Adams, M. R., & Moss, M. O. (2008). Food microbiology (3rd ed.). RSC Publishing.
http://dx.doi.org/10.1039/9781847557940-00447.
Altman, D. G., & Bland, J. M. (1983). Measurement in medicine: the analysis of
method comparison studies. The Statistician, 32, 307e317.
Anonymous. (1991). Commission Decision of 14 February 1991 laying down certain
methods of analysis and testing of raw milk and heat-treated milk (91/180/
EEC).
Arbeitsgruppe “Lebensmittelerhaltung und Mikrobiologie”. (1974). Merkblatt 21:
Bestimmung der Oberfla €chenkeimzahl (Bakterien, Schimmelpilze, Hefen und
coliforme Keime) auf nicht saugfa €higen Packstoffen. Verpackungs-Rundschau,
Technisch-Wissenschaftliche Beilage. Technische Universit€
at München.
 M. Feichtinger et al. / Food Control 50 (2015) 548e553
Bergmair, J., Washüttl, M., & Wepner, B. (2010). Prüfpraxis für Kunststoffverpackungen:
Lebensmittel-, Pharma- und Kosmetikverpackungen (2nd ed.). Behr's Verlag.
Bland, J. M., & Altman, D. G. (1986). Statistical methods for assessing agreement
between two methods of clinical measurement. Lancet, 1(8476), 307e310.
Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2868172.
BTF. (2014). BioBall Multishot-550-instructions for use. Retrieved 03.09.14, from
http://bioball.com/instruction_for_use.php?nav¼BioBall.
BTF. (n.d.). Certificate of analysis. Retrieved 03.09.14, from http://bioball.com/cofa.
php?nav¼BioBall.
Cerny, G., & Betz, M. (1999). Einfluss von Altpapier und Altpapier-
Aufbereitungstechnologien auf die Verkeimungsrate von Sekund€ arfaser-
stoffen. Das Papier, 10A, V42eV45.
Copetti, M. V., Iamanaka, B. T., Frisvad, J. C., Pereira, J. L., & Taniwaki, M. H. (2011).
Mycobiota of cocoa: from farm to chocolate. Food Microbiology, 28(8),
1499e1504. http://dx.doi.org/10.1016/j.fm.2011.08.005.
Deutsches Institut für Normung. (1993). DIN 54378: Prüfung von Papier, Karton und
Pappe e Bestimmung der Oberfla €chenkoloniezahl (OKZs).
Deutsches Institut für Normung. (1999). DIN 10050-3:1999-09: Prüfung von But-
tereinwicklern e Teil 3: Keimzahlbestimmung.
Ekman, J., Tsitko, I., Weber, A., Nielsen-LeRoux, C., Lereclus, & Salkinoja-Salonen, M.
(2009). Transfer of Bacillus cereus spores from packaging paper into food.
Journal of Food Protection, 72(11), 2236e2242.
International Organization for Standardization. (2008). ISO 21527-1:2008 e
microbiology of food and animal feeding stuffs e horizontal method for the
enumeration of yeasts and moulds e part 1: colony count technique in products
with water activity greater than 0.95.
International Organization for Standardization. (2013). EN ISO 4833-1:2013:
microbiology of the food chain e horizontal method for the enumeration of
553
microorganisms e part 1: colony count at 30 degrees C by the pour plate
technique.
Marsh, K., & Bugusu, B. (2007). Food packaging e roles, materials, and environ-
mental issues. Journal of Food Science, 72(3), R39eR55. http://dx.doi.org/10.1111/
j.1750-3841.2007.00301.x.
€rvi, T. S. M., Andersson, M. A., & Salkinoja-Salonen, M. S. (2000). Properties of
Pirttija
Bacillus cereus and other bacilli contaminating biomaterial-based industrial
processes. International Journal of Food Microbiology, 60(2e3), 231e239. http://
dx.doi.org/10.1016/S0168-1605(00)00313-5.
€rvi, T. S. M., Graeffe, T. H., & Salkinoja-Salonen, M. S. (1996). Bacterial con-
Pirttija
taminants in liquid packaging boards: assessment of potential for food spoilage.
Journal of Applied Bacteriology, 81(4), 445e458.
Raaska, L. (2005). Managing contamination risks from packaging materials. In
H. L. M. Lelieveld, M. A. Mostert, & J. Holah (Eds.), Handbook of hygiene control in
the food industry (pp. 378e395). Woodhead Publishing.
Suihko, M.-L., & Stackebrandt, E. (2003). Identification of aerobic mesophilic bacilli
isolated from board and paper products containing recycled fibres. Journal of
Applied Microbiology, 94(1), 25e34. http://dx.doi.org/10.1046/j.1365-
2672.2003.01803.x.
€isa
Va €nen, O. M., Mentu, J., & Salkinoja-Salonen, M. S. (1991). Bacteria in food
packaging paper and board. The Journal of Applied Bacteriology, 71(2), 130e133.
Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1917722.
€isa
Va €nen, O. M., Nurmiaho-Lassila, E. L., Marmo, S. A., & Salkinoja-Salonen, M. S.
(1994). Structure and composition of biological slimes on paper and board ma-
chines. Applied and Environmental Microbiology, 60(2), 641e653. Retrieved from
http://www.pubmedcentral.nih.gov/articlerender.fcgi?
artid¼201361&tool¼pmcentrez&rendertype¼abstract.