Accepted Manuscript

Title: Foodborne pathogens in raw milk and cheese of sheep

and goat origin: a meta-analysis approach

Author: Ursula Gonzales-Barron Andiara Gonçalves-Tenório

Vânia Rodrigues Vasco Cadavez

PII: S2214-7993(17)30117-0
DOI: https://doi.org/doi:10.1016/j.cofs.2017.10.002
Reference: COFS 279

To appear in:

Received date: 2-9-2017

Revised date: 1-10-2017
Accepted date: 6-10-2017

Please cite this article as: Gonzales-Barron, U., Gonçalves-Tenório, A., Rodrigues, V.,
Cadavez, V.,Foodborne pathogens in raw milk and cheese of sheep and goat origin: a
meta-analysis approach, COFS (2017), https://doi.org/10.1016/j.cofs.2017.10.002

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Foodborne pathogens in raw milk and cheese of sheep and goat origin:
a meta-analysis approach

Ursula Gonzales-Barron, Andiara Gonçalves-Tenório, Vânia Rodrigues, Vasco

Cadavez*

t
ip
CIMO Mountain Research Centre, School of Agriculture, Polytechnic Institute of
Braganza Campus de Santa Apolónia, Apartado 1172, 5301-855 Braganza, Portugal

cr
(*) Corresponding author

us
ABSTRACT
an
This review compiles published information concerning the incidence of pathogenic
microorganisms – Listeria monocytogenes, Salmonella spp., Staphylococcus aureus and
M
shiga-toxin producing Escherichia coli (STEC) – in goat and sheep raw milk and
cheese. Meta-analytical data were extracted from 37 primary studies undertaken in
ed

Australia, Brazil, China, Colombia, Costa Rica, Czech Republic, Egypt, Germany,
Greece, Iran, Italy, Malaysia, Mexico, Norway, Poland, Portugal, Spain, Sweden,
Switzerland, Turkey, UK and USA. Pooled frequencies of detection of pathogens were
pt

found to be similar for sheep and goat raw milk: Salmonella (1.4 – 2.4%), L.
monocytogenes (2.9 – 3.6%), STEC (4.3 – 4.8%) and S. aureus (35 – 39%). Likewise,
ce

in goat cheeses, regardless of being made of raw or heat-treated milk, S. aureus has
been the most frequent contaminant (16.0%), whereas in raw milk cheeses, regardless of
Ac

origin, the pooled prevalence of S. aureus is equally high in hard (34.6%) and soft
cheeses (25.7%). L. monocytogenes is another important pathogen in sheep and goat
milk cheeses (3.6 – 12.8%) while E. coli O157 strains with virulence genes (4.3%) also
appear to persist during cheese manufacture. As expected, STEC has a higher pooled
incidence in raw milk cheeses (10.0%) than in pasteurised milk cheeses (4.7%). Thus,
the moderate contamination in raw milk and cheese of sheep and goat origin, revealed
by this meta-analysis, advocates the reinforcement of general prevention measures such
as close monitoring of hygiene on farms and eradication of disease by sheep and goat

Page 1 of 21
dairy farmers. Moreover, for the production of traditional cheeses made of raw milk,
preventive measures during processing, namely, regular sterilisation of dairy equipment,
process monitoring and hygiene of operators, should be even more stringent.

Keywords: Listeria monocytogenes, Staphylococcus aureus, Salmonella, Escherichia

coli, pasteurisation

t
ip
1. Introduction

cr
According to Eurostat [1], the milk production by the EU member states in 2016

us
was ~168.2 million tons. From this, only 1.78% represented the production of sheep
milk and 1.3% the production of goat milk. Interestingly, two-thirds of the world’s
production of sheep milk and over a quarter of the world’s production of goat milk are
an
concentrated in the Mediterranean region. Hence, Europe is positioned as the second
largest producer with about 3.0 and 2.5 million tons ewe and goat milk, respectively.
M
From these dairy herds, ~95% milk is transformed into dairy products [2]. The farming
of small ruminants in Europe has played an important role in rural sustainability,
allowing the exploitation of marginal areas of, otherwise, low productive potential.
ed

Although sheep and goat milk are products of high nutritional value, in
microbiological terms, raw milk is a rich medium for the development of pathogenic
pt

microorganisms responsible for several foodborne diseases [3]. Milk contamination can
occur in two ways: by endogenous transfer (i.e., direct transfer of blood or by infection
ce

in the udder), or by exogenous transfer (i.e., during or after the milking process) [4].
The main microorganisms associated with the consumption of raw milk from cows,
Ac

goats and sheep are: Bacillus cereus, Campylobacter spp., Staphylococcus aureus,
Helicobacter pylori, shigatoxin-producing Escherichia coli (STEC), Listeria
monocytogenes, Salmonella spp., Clostridium botulinum, Brucella spp., Mycobacterium
bovis, Leptospira, Cryptosporidium parvum and Toxoplasma gondii [4-6]. The main use
of sheep and goat milk is the production of traditional cheeses, at farm level or by small
local dairies or by cheese industries working at regional level. These traditional cheeses
are commonly produced from raw milk because of tradition and the enhanced
organoleptic properties attributed. In detriment, pathogenic microorganisms, when

Page 2 of 21
present in raw milk, can remain viable in cheese even when they are ripened for several
weeks.
Since the frequency of outbreaks attributed to milk and dairy products has been
considerable in the past years, the presence of pathogens in raw milk and thereof
products seems to remain a public health hazard. An exhaustive literature search
conducted by the authors indicated that the pathogenic microorganisms most frequently

t
searched for in microbiological surveys in goat and sheep milk/cheese are: S. aureus, E.

ip
coli O157, L. monocytogenes and Salmonella spp. Therefore, the objective of this work

cr
was to estimate the overall incidences of the aforementioned pathogens in raw milk and
cheese of sheep and goat origin. Furthermore, the analysis of contamination in cheeses

us
was broken down by type of cheese (soft and hard) and type of milk (raw and
pasteurised). In order to do so, separate multilevel meta-analysis models were adjusted
to up-to-date information extracted from the literature.
an
2. Meta-analysis
M
For this meta-analysis, the population was defined as milk and cheeses surveyed
from bulk tanks and at retail anywhere in the world, while the measured outcome was
ed

the detection of pathogens. Electronic searches, using the Scopus and Scielo databases,
were carried out to find articles and official reports published since the year 2000,
pt

summarising the occurrence of microbiological hazards in these products. The

bibliographic searches were undertaken using the formula:
ce

[(occurrence) OR (prevalence) OR (incidence) OR (concentration) OR

(contamination) OR (count) OR (survey) OR (sampling)] AND [(Listeria) OR (STEC)
Ac

OR (VTEC) OR (O157) OR (Escherichia coli) OR (Staphylococcus aureus) OR

(Salmonella)] AND [(goat milk) OR (sheep milk) OR (goat cheese) OR (sheep
cheese)] AND NOT [(artificial) OR (meta-analysis) OR (“systematic review”) OR
(inocul*) OR (spiked) OR (feed) OR (livestock) OR (cow)].

After assessing all the information presented in every study, a total of 37

primary studies passed the two quality criteria of having used approved microbiological
methods and presenting sufficient data in extractable form. The meta-analyses for goat
raw milk and cheese were based on 35 primary studies [3, 7-40] while those for sheep

Page 3 of 21
raw milk and cheese were based on 16 studies [3, 12, 18-20, 22, 24, 26-27, 30, 32, 34,
47, 40-42].
The occurrence of microbial hazards in goat and sheep raw milk and cheese is a
binary trait (i.e., a sample tests either positive or negative for the pathogen), thus the
parameter to measure the effect size is the raw proportion p (calculated as the number of
successes, i.e., positive samples, s, divided by the total sample size, n). From every

t
primary study, the values of s, n and p were extracted. In addition, study characteristics

ip
such as country, pasteurisation (yes/no), cheese type (soft/hard) and survey’s year were

cr
annotated.

Several multilevel random-effect meta-analysis regressions were fitted to

us
appropriate data partitions in order to pool incidences of: (i) pathogens in sheep and
goat raw milk; (ii) pathogens in sheep and goat milk cheeses; (iii) S. aureus and STEC
an
in raw and pasteurised cheeses; and (iv) S. aureus in hard and soft cheeses made of raw
milk. For a detailed explanation on multilevel meta-analysis modelling for prevalence
data, refer to [43] and [44]. Meta-analysis models and graphs were built in R Studio
M
version 1.0.136 using the ‘metafor’ package.
ed

3. Incidence of foodborne pathogens in sheep and goat raw milk

In spite of the broad knowledge concerning the huge negative impact that
pt

foodborne diseases can represent for sheep’ and goats’ small producers, data concerning
microbial pathogens occurrence in sheep and goat milk and thereof products are scarce.
ce

Only 150 incidence observations for the four pathogens were retrieved from a total of
37 primary studies published since 2000.
Ac

The results of the meta-analysis on the incidence of L. monocytogenes,

Salmonella spp., S. aureus and STEC in goat and sheep raw milk are presented in Table
1. Pooled prevalence, 95% confidence intervals, number of extracted observations (n),

between-study variability (t2), residuals variance (σ2) and test for publication bias (viz.
effect of study size) are compiled. The pooled incidences of pathogens in sheep raw
milk are comparable to those of goat raw milk (Table 1). Among the pathogens,
Salmonella spp. presented the lowest overall incidence in both sheep (1.37%; 95%: 0.27
– 6.62%) and goat raw milk (2.36%; 95%: 0.87 – 6.26%), being the incidence in the

Page 4 of 21
former lower than that of the latter. To this respect, in [45] it was stated that higher fat
content in ewe milk, as compared with cow and goat milk’s, may confer a protective
effect on Gram-negative bacteria.
At a significantly higher level of contamination lies L. monocytogenes in raw
milk of both sheep (3.56%; 95%: 1.53 – 8.11%) and goat (2.92%; 95%: 1.49 – 5.66%)
origin. The contamination of goat and ewe milk with Listeria spp. is assumed to occur

t
by the same contamination routes as for cow milk; this is, through silage, other feeds,

ip
cross-contamination of feeds by manure, poor-quality bedding, contaminated milking

cr
equipment or mastitis infection attributable to Listeria spp. [45]. Infection of the
mammary gland by L monocytogenes (which is the most prevalent pathogenic strain, as

us
a source of milk contamination) is more common in sheep, and presumably in goats,
than in cows [46]. Sheep and goat suffering from listerial mastitis can contaminate milk
over long periods of time without clinical signs of listeriosis [47], hence contaminating
an
all the milk production chain and processing environment.
Following L. monocytogenes, shigatoxin-producing E. coli is the next pathogen
M
of high prevalence in raw milk (Table 1). Its pooled incidence in sheep raw milk
(4.83%; 95% CI: 2.18 – 10.39%) is not significantly different from that of goat raw
milk (4.33%; 95% CI: 2.15 – 8.55%). Preventing animals’ faecal material from
ed

contaminating the milk is an important step in reducing the prevalence of STEC in raw
milk. The implementation of effective cleaning procedures to remove faecal material
pt

from udders prior to milking can reduce the risk yet not eliminate it entirely.
The pathogen of highest incidence is S. aureus in both sheep raw milk (39.4%;
ce

95% CI: 22.7 – 58.9%) and goat raw milk (35.2%; 95% CI: 23.2 – 49.3%; Table 1). The
presence of this pathogen in milk does not only indicates the poor hygienic quality
Ac

under which such milk was produced, but also may indicate udder inflammation, since
Staphylococcus spp. are the main etiological agents of small ruminant’s intra-mammary
infections, with the most frequent isolates belonging to S. aureus. In fact, it is common
that the presence of S. aureus is due to post-secretory contamination [45].
Recommended methods for reducing the risk of S. aureus infection of the mammary
gland include regular monitoring of udders, teats, milking equipment and milk, and the
maintenance of scrupulous hygiene practices, including the isolation of animals with
mastitis infection.

Page 5 of 21
 The overall prevalence of the four pathogens in raw milk of sheep (7.45%; 95%
CI: 3.89 – 13.8%) and goat (6.63%; 95% CI: 4.15 – 10.4%) origin can be read off from
the Galbraith plots [43]. These plots also suggest the presence of heterogeneity among
the observations extracted from the literature, which was also corroborated by the
significant residual between-study heterogeneity values (τ2) from both meta-analyses
(Table 1). For the meta-analysis on goat raw milk data, no evidence of publication bias

t
was found (p=0.762); by contrast, for the sheep milk meta-analysis, there was strong

ip
evidence of publication bias (p<.0001; Table 1). This means that it is very likely that

cr
survey studies of small sample size which failed to detect any contamination in raw
milk may have remained unpublished.

us
4. Incidence of foodborne pathogens in sheep and goat cheese

an
Pathogens – taken as a whole – are more prevalent in cheeses made of goat milk
(9.83%; 95% CI: 5.82 – 16.1%) than in those of sheep milk (2.94%; 1.32 – 6.45%;
M
Figure 2). STEC has been more frequently detected in goat milk cheeses (4.31%; 95%
CI: 1.78 – 10.0%) than in sheep milk cheeses (2.76%; 95% CI: 1.09 – 6.79%; Table
2).When artificially-contaminated in low numbers, E. coli O157 was capable of
ed

surviving the cheese manufacture process, ripening and storage period of soft Hispanic,
Paneer, Camembert, feta and Cheddar cheeses [48]. Similarly, L. monocytogenes
pt

presented, on meta-analysis, a significantly higher pooled prevalence in goat milk

cheeses (12.8%; 95% CI: 4.49 – 31.6%) than in sheep milk cheeses (3.61%; 95% CI:
ce

0.70 – 16.4%; Table 2). The higher levels of these two pathogens in goat milk cheeses
may arise from the goat production itself which, in regions of the Mediterranean and
Ac

Middle East, are linked to more artisanal systems. In developing regions, the equipment
of farm may be rudimentary including poor milking facilities, hand-milking and
consequently long milking times. Hence, goat cheeses produced in those on-farm
facilities, traditionally manufactured without thermal treatment, are expected to bear
more contamination.
While the overall occurrence of S. aureus in goat milk cheeses was found to be
considerably high (16.0%; 7.92 – 29.8%; Table 2), cheeses made of raw milk,
regardless of origin, presented an even higher prevalence of S. aureus (38.7%; 95% CI:

Page 6 of 21
9.28 – 79.6%; Table 3). S aureus, besides being present in the animal's udder, is also
present in humans. Thus, contamination by this microorganism can occur from the time
of milking to the processing of milk [49]. The presence of S. aureus in cheeses is
relevant because it is responsible for the production of thermostatic enterotoxins, which
may survive the pasteurisation process. Furthermore, even if the milk is pasteurised,
cross-contamination with S. aureus could still occur during cheese manufacture [15].

t
This affirmation is validated by this meta-analysis, where the cheeses prepared with

ip
pasteurised milk, although, at a lower mean occurrence (4.20%; 95% CI: 0.63 – 23.3%;

cr
Table 3), still bear S. aureus. Generally speaking, the occurrence of S. aureus in cheeses
produced with both raw and pasteurised milk is an indicator of hygiene deficiency

us
during milking and processing.
In cheeses made of raw milk either of sheep or goat origin, longer ripening
periods, overall, did not appear to decrease the viability of S. aureus: the pooled
an
prevalence in hard cheeses (34.6%; 95% CI: 6.27 – 80.7%) was not significantly
different from that of soft cheeses (25.7%; 95% CI: 5.0 – 69.3%; Table 3). On the other
M
hand, in the production of cheese, pasteurisation of milk is an effective control measure
to reduce the exposure of consumers to E. coli O157; as shown by the significant
difference in STEC prevalence between raw milk cheeses (10.0%; 95% CI: 3.25 –
ed

26.9%) and pasteurised milk cheeses (4.70%; 95% CI: 2.00 – 10.7%; Table 3).
pt

5. Conclusions
ce

This systematic review indicated that up-to-date data on incidence of pathogens in goat
and sheep raw milk and cheeses are limited, particularly for Salmonella and L.
Ac

monocytogenes in raw cheeses. However, meta-analysis suggested that these raw milks
have an important pathogenic contamination. This is in accordance with the EFSA’s
report on public health risks associated to raw milk [50], where Salmonella spp., STEC
and L. monocytogenes were considered to have an important impact on human health, in
the first pathogen because of its high incidence rate while in the other two because of
the high severity of their infections. Moreover, the meta-analyses revealed that the
pathogens under study followed the same orderly frequency of detection in both sheep
and goat raw milk: Salmonella (1.4 – 2.4%), L. monocytogenes (2.9 – 3.6%), E. coli

Page 7 of 21
O157 (4.3 – 4.8%) and S. aureus (35 – 39%). In goat milk cheeses, the pooled
prevalence of pathogens (9.8%) was greater than in sheep milk cheeses (2.9%), being
particularly high for S. aureus (16.0%) in goat milk cheeses. In raw milk cheeses
regardless of origin, the pooled prevalence of S. aureus was equally high in hard
cheeses (34.6%) and soft cheeses (25.7%). On the other hand, in the production of
cheese, pasteurisation of milk appears as an effective control measure against

t
pathogens, as shown by the significant lower incidence of S. aureus (4.2%) and STEC

ip
(4.7%) in pasteurised milk cheeses as compared to those of raw milk cheeses (38.7%

cr
and 10.0%, respectively). Nonetheless, since the manufacture of sheep and goat cheeses
are traditionally made of raw milk – and hence likely to be associated with more

us
contamination levels – it is important to further educate artisan cheesemakers so that
they continue to produce their appreciated typical cheeses, yet with improved safety
standards. Close monitoring of hygiene on sheep and goat dairy farms and eradication
an
of disease are general on-farm prevention measures for quality milk. However, if such
milk is for processing into raw milk cheeses, even more strict preventive measures are
M
needed during processing, such as regular sterilisation of dairy equipment, process
monitoring and stringent hygiene of operators.
ed

Acknowledgments
Dr. Gonzales-Barron wishes to acknowledge the financial support provided by the
pt

Portuguese Foundation for Science and Technology (FCT) through the award of a five-
year Investigator Fellowship (IF) in the mode of Development Grants (IF/00570).
ce
Ac

Page 8 of 21
References

[1] EUROSTAT (2016) Agriculture, forestry and f ishery statistics. Theme: Agriculture
and fisheries Collection: Statistical books. Luxembourg: EUROSTAT.
[2] Boyazoglu J, Morand-Fehr P: Mediterranean dairy sheep and goat products and

t
their quality, Small Rumin Res 2001, 40(1): 1 – 11.

ip
[3] Ombarak R, Elbagory A: Bacteriological quality and occurrence of some

cr
microbial pathogens in goat's and ewe's milk in Egypt, Int Food. 2017, 24(2):
847-851.

us
[4] Verraes C, Vlaemynck G, Weyenberg SV, Zutter LD, Daube G, Sindic M,
Uyttendaele M, Herman L: A review of the microbiological hazards of dairy
products made from raw milk, Int Dairy J. 2015, 50: 32 – 44.
an
[5] Claeys WL, Cardoen S, Daube G, Block JD, Dewettinck K, Dierick K, Zutter LD,
Huyghebaert A, Imberechts H, Thiange P, Vandenplas Y, Herman L: Raw or
M
heated cow milk consumption: Review of risks and benefits, Food Control 2013,
31(1):251 – 262.
[6] EFSA(2016) The European Union summary report on trends and sources of
ed

zoonoses, zoonotic agents and food-borne outbreaks in 2015, EFSA Journal

2016, 14(12): 4634, 231. doi:10.2903/j.efsa.2016.4634.
pt

[7] Abou-Eleinin A-AM, Ryser ET, Donnelly CW: Incidence and seasonal variation
of Listeria species in bulk tank goat's milk, J Food Prot. 2000, 63(9): 1208-1213.
ce

[8] Akineden Ö, Hassan AA, Schneider E, Usleber E: Enterotoxigenic properties of

Staphylococcus aureus isolated from goats' milk cheese, Int J Food Microbiol.
Ac

2008, 124(2): 211-216.

[9] Albarracin C, Poutou P, Carrascal C: Listeria spp., y L. monocytogenes en leche
cruda de cabra, Rev.MVZ Córdoba 2008,13(2):1326-1332.
[10] Araya V, Gallo L, Quesada C, Chaves C, Arias M: Bacteriological evaluation
of goat milk and cheese distributed in the Metropolitan Area of San José, Costa
Rica, Arch Latinoam Nutr. 2008, 58(2):182-186.
[11] Soto-Beltran M, Gerba CP, Porto-Fett A, Luchansky JB, Chaidez C: Prevalence
and characterization of Listeria monocytogenes, Salmonella and Shiga toxin-

Page 9 of 21
 producing Escherichia coli isolated from small Mexican retail markets of queso
fresco, Int J Environ Health Res 2016, 25(2): 140-148.
[12] Bogdanovičová K, Šová M-K, Babák V, Kalhotka L, Koláčková I, Karpíšková
R.: Microbiological quality of raw milk in the Czech Republic, Czech J. Food.
2016, 34(3):189-196.
[13] Bogdanovičová K, Skočková A, Šťástková Z, Koláčková I, Karpíšková R: The

t
bacteriological quality of goat and ovine milk, Potravinarstvo 2015, 9(1): 72-76.*

ip
*The results of this study confirm the presence of pathogenic bacteria in goat and ovine

cr
raw milk. The study confirms that unpasteurised milk may be contaminated with
different types of microorganisms and can be an important source of foodborne

us
illnesses.
[14] Brito JRF, Santos EMP, Arcuri EF, Lange CC, Brito MAVP, Souza GN,
Luchansky JB: Retail survey of Brazilian milk and Minas Frescal cheese and a
an
contaminated dairy plant to establish prevalence, relatedness, and sources of
Listeria monocytogenes isolates, Appl Environ Microbiol 2008, 74(15): 4954–
M
4961.*
*This study shows the importance of finding specific niches wherein L. monocytogenes
can survive in the food-processing environment to more specifically direct the cleaning,
ed

disinfection, and renovation/repair efforts to these sites.

[15] Cavicchioli V, Scatamburlo T, Yamazi A, Pieri F, Nero L: Occurrence of
pt

Salmonella, Listeria monocytogenes, and enterotoxigenic Staphylococcus in goat

milk from small and medium-sized farms located in Minas Gerais State, Brazil,
ce

J. Dairy Sci. 2015, 98(12): 8386-8390.

[16] Cortés C, de la Fuente R, Contreras A, Sánchez A, Corrales JC, Martínez S,
Ac

Orden JA: A survey of Salmonella spp. and Campylobacter spp. in dairy goat
faeces and bulk tank milk in the Murcia region of Spain, Ir Vet J. 2006, 59(7):
391-393.
[17] Cupáková Š, Pospíšilová M, Karpíšková R, Janštová B, Vorlová L:
Microbiological quality and safety of goat's milk from one farm, Acta Univ.
Agric. et Silvic. Mendel. Brun 2012, 60(6): 33-38.
[18] D'Amico D, Groves E, Donnelly C: Low incidence of foodborne pathogens of
concern in raw milk utilized for farmstead cheese production,

Page 10 of 21
 J Food Prot 2008, 71(8):1580-1589.
[19] D'Amico D, Donnelly C: Microbiological quality of raw milk used for small-
scale artisan cheese production in Vermont: Effect of farm characteristics and
practices, J. Dairy Sci. 2010, 93(1): 134-147.
[20] Durmaz H, Avci M, AYGÜN O: The presence of Listeria species in corn
silage and raw milk produced in southeast region of Turkey, Kafas Univ Vet

t
Fak. 2015, 21(1).

ip
[21] Eglezos S, Huang B, Dykes G, Fegan N, Bell K, Stuttard E: A survey of the

cr
microbiological quality of frozen unpasteurised goats' milk in Queensland,
Australia, Aust J Dairy Technol. 2008, 63(3):79-81.

us
[22] Farzan R, Rahimi E, Momtaz H: Virulence properties of Shiga Toxin-
producing Escherichia Coli isolated from Iranian raw milk and dairy products,
Slov Vet Zb. 2012, 49(4):159-166.
[23]
an
Foschino R, Invernizzi A, Barucco R, Stradiotto K: Microbial composition,
including the incidence of pathogens, of goat milk from the Bergamo region of
M
Italy during a lactation year, J Dairy Res. 2002, 69(2): 213-225.*
* This work assessed the incidence of pathogens (Escherichia coli O157, Listeria
monocytogenes, Salmonella spp., Staphylococcus aureus and their toxins) in raw goat
ed

milk intended for cheese-making,

[24] Guerra M, McLauchlin J, Bernardo F: Listeria in ready-to-eat and
pt

unprocessed foods produced in Portugal, Food Microbiol.2001, 18(4):423 - 429.

[25] Jakobsen RA, Heggebø R, Sunde EB, Skjervheim M: Staphylococcus aureus
ce

and Listeria monocytogenes in Norwegian raw milk cheese production. Food

Microbiol 2011, 28(3): 492 - 496.
Ac

[26] Jamali H, Radmehr B, Thong KL: Prevalence, characterisation, and

antimicrobial resistance of Listeria species and Listeria monocytogenes isolates
from raw milk in farm bulk tanks, Food Control. 2013, 34(1):121 - 125.
[27] Little C, De Louvois J: Health risks associated with unpasteurized goats' and
ewes' milk on retail sale in England and Wales. A PHLS Dairy Products
Working Group Study, Epidemiol Infect 1999, 122(3):403-408.*
* On the basis of the results from this small study, unpasteurised milk from cows, sheep
or goats may bear an unacceptable risk to public health.

Page 11 of 21
[28] Marozzi S, De Santis P, Lovari S, Condoleo R, Bilei S, Marciano R, Mezher Z:
Prevalence and molecular characterisation of shiga toxin-producing
Escherichia Coli in raw milk cheeses from Lazio region, Italy, Ital J Food Sci.
2015, 5(1):4566.
[29] Morales-Pablo R, de la Cruz A, Leyva-Ruelas G, Ybarra-Moncada
M: Bacteriological quality goat raw milk produced in Miravalles, Puebla

t
[Calidad bacteriológica de leche cruda de cabra producida en Miravalles,

ip
Puebla], Rev Mex Ing Quim. 2012, 11(1): 45-54.

cr
[30] Muehlherr J, Zweifel C, Corti S, Blanco J, Stephan R: Microbiological quality
of raw goat's and ewe's bulk-tank milk in Switzerland, J. Dairy Sci. 2003,

us
86(12): 3849-3856.
[31] Perez G, Belda F, Cardell E, Zarate V: Microbiological quality and
occurrence of Salmonella and Listeria monocytogenes in fresh Tenerife goat's
an
milk cheese, Milchwissenschaft 1998, 53(6): 324-327.
[32] Rey J, Sánchez S, Blanco J, Hermoso De Mendoza J, Hermoso De Mendoza M,
M
García A, Gil C, Tejero N, Rubio R, Alonso J: Prevalence, serotypes and
virulence genes of Shiga toxin-producing Escherichia coli isolated from ovine
and caprine milk and other dairy products in Spain, Int J Food Microbiol. 2006,
ed

107(2): 212-217.
[33] Rosengren Å, Fabricius A, Guss B, Sylvén S, Lindqvist R: Occurrence of
pt

foodborne pathogens and characterization of Staphylococcus aureus in cheese

produced on farm-dairies, Int J Food Microbiol. 2010, 144(2):263-269.
ce

[34] Solomakos N, Govaris A, Angelidis A, Pournaras S, Burriel A, Kritas S,

Papageorgiou D: Occurrence, virulence genes and antibiotic resistance of
Ac

Escherichia coli O157 isolated from raw bovine, caprine and ovine milk in
Greece, Food Microbiol. 2010, 26(8): 865-871.
[35] Sosnowski M, Rola J, Osek J: Alkaline phosphatase activity and
microbiological quality of heat-treated goat milk and cheeses. Small Rumin Res.
2016, 136: 132-136.

[36] Spanu V, Scarano C, Virdis S, Melito S, Spanu C, De Santis E: Population

structure of Staphylococcus aureus isolated from bulk tank goat's milk,
Foodborne Pathogens and Disease. 2013, 10(4): 310-315.

Page 12 of 21
[37] Stephan R, Schumacher S, Corti S, Krause G, Danuser J, Beutin L: Prevalence
and characteristics of shiga toxin-producing Escherichia coli in Swiss raw milk
cheeses collected at producer level, J. Dairy Sci. 2008, 91(7): 2561-2565.
[38] Vyletělova M, Hanuš O, Karpíšková R, Štástková Z: Occurrence and
antimicrobial sensitivity in Staphylococci isolated from goat, sheep and cow's
milk, Acta Univ. Agric. et Silvic. Mendel. Brun. 2011, 59(3): 209-213.

t
[39] Xing X, Zhang Y, Wu Q, Wang X, Ge W, Wu C: Prevalence and

ip
characterization of Staphylococcus aureus isolated from goat milk powder

cr
processing plants, Food Control 2016, 59: 644 – 650.
[40] Zweifel C, Giezendanner N, Corti S, Krause G, Beutin L, Danuser J, Stephan R:

us
Characteristics of Shiga Toxin-producing Escherichia coli Isolated from Swiss
raw milk cheese within a 3-year monitoring program, Journal of Food
Protection 2010, 73(1):88-91.
[41]
an
Caro I, García-Armesto M: Occurrence of shiga toxin-producing Escherichia
coli in a Spanish raw ewe's milk cheese, Int J Food Microbiol. 2007, 116(3):410-
M
413.
[42] Giammanco G, Pepe A, Aleo A, D'Agostino V, Milone S, Mammina C:
ed

Microbiological quality of Pecorino Siciliano "primosale" cheese on retail sale

in the street markets of Palermo, Italy, New Microbiologica. 2011, 34(2): 179-
185.
pt

[43] Viechtbauer W: Conducting meta-analyses in R with the metaphor Package.

J Stat Softw 2010, 36 (3):1-48.**
ce

** An educational and easy-to-follow brief manual on how to conduct meta-analysis.

[44] Xavier C, Gonzales-Barron U, Paula V, Estevinho L, Cadavez V: Meta-
Ac

analysis of the incidence of foodborne pathogens in Portuguese meats and their

products, Food Res Int 2014, 55:311-323.**
** A very detailed explanation of the meta-analysis methods applied to combine
prevalence data is presented in this work.
[45] Klinger I, Rosenthal I: Public health and the safety of milk and milk products
from sheep and goats, Rev Sci Tech Off Int Epiz 1997, 482-488.*
* The authors discuss the impact of low levels of sheep and goat milk production per
head, milking system, and raising conditions of herds on the sanitary quality of milk.

Page 13 of 21
[46] Schoder D, Melzner D, Schmalwieser A, Zangana A, Winter P, Wagner M:
Important vectors for Listeria monocytogenes transmission at farm dairies
manufacturing fresh sheep and goat cheese from raw milk, J Food Prot
2011, 74(6), 919-924.
[47] Osman KM, Zolnikov TR, Samir A, Orabi A: Prevalence, pathogenic
capability, virulence genes, biofilm formation, and antibiotic resistance of

t
Listeria in goat and sheep milk confirms need of hygienic milking

ip
conditions, Pathog Glob Health, 2014, 108(1):21-29.

cr
[48] Baylis CL: Raw milk and raw milk cheeses as vehicles for infection by
Verocytotoxin producing Escherichia coli., Int J Dairy Technol. 2009, 62.3 : 293-

us
307.*
* The authors revise outbreaks of VTEC O157:H7 infections attributed to the
consumption of raw milk and associated dairy products. E. coli O157 strains can indeed
survive the cheese-making process.
an
M
[49] Fagundes H, Oliveira-Fernandes CA: Infecções intramamárias causadas por
Staphylococcus aureus e suas implicações em saúde pública, Ciencia
Rural 2004, 34.4.
ed

[50] EFSA (2015) Scientific Opinion on the public health risks related to the
consumption of raw drinking milk, EFSA Journal 2013, 13(1): 3940, 95.
pt

doi:10.2903/j.efsa.2015.3940
ce
Ac

Page 14 of 21
FIGURE CAPTIONS

Fig. 1: Galbraith plots of the meta-analyses of the incidence of the four pathogens in
sheep raw milk (top; overall prevalence 0.0745; 95% CI: 0.0389 – 0.1379) and goat raw
milk (bottom; overall prevalence 0.0663; 95% CI: 0.0415 – 0.1043)

t
Fig. 2: Galbraith plots of the meta-analyses of the incidence of pathogens in sheep milk

ip
cheeses (top; overall prevalence 0.0294; 95% CI: 0.0132 – 0.0645) and goat milk
cheeses (bottom; overall prevalence 0.0983; 95% CI: 0.0582 – 0.1613) sampled at
retail.

cr
us
TABLE CAPTIONS
Table 1: Meta-analyses of the incidence of pathogens in raw sheep and goat milk
an
sampled from bulk tanks, showing pooled prevalence, 95% confidence intervals,
number of observations (n), between-study variability (t2), residuals variance (σ2) and
study size effect (test for publication bias)
M
Table 2. Meta-analyses of the incidence of pathogens in sheep and goat cheeses,
showing pooled prevalence, 95% confidence intervals, number of observations (n),
between-study variability (t2), residuals variance (σ2) and study size effect (test for
ed

publication bias)

Table 3. Meta-analyses of the incidence of S. aureus and STEC in goat and sheep
pt

cheeses by type, showing pooled prevalence, 95% confidence intervals, number of

observations (n), between-study variability (t2) and residuals variance (σ2)
ce
Ac

Page 15 of 21
Figure 1

t
ip
yi 2 0.82
zi = 0 0.55
vi + τ2 0.25

cr
-2 0.08
0.02
0.01

us
0.00

0.0 0.2 0.4 0.6 0.8

xi = 1 vi + τ2
an
M
ed

2 0.92
yi
pt

0.72
zi = 0
vi + τ2 0.36
-2 0.11
ce

0.03
0.01
0.00
Ac

0.0 0.2 0.4 0.6

xi = 1 vi + τ2

Page 16 of 21
Figure 2

t
ip
yi 2
zi = 0
vi + τ 2 -2

cr
0.17
0.10
0.06
0.03
0.02
0.01

us
0.01

0.0 0.5 1.0 1.5

xi = 1
an vi + τ2
M
ed

0.99
0.95
2
pt

yi 0.80
zi = 0 0.45
vi + τ 2
0.14
ce

-2
0.03
0.01
Ac

0.0 0.2 0.4 0.6

xi = 1 vi + τ2

Page 17 of 21
Table 1. Meta-analyses of the incidence of pathogens in raw sheep and goat milk
sampled from bulk tanks, showing pooled prevalence, 95% confidence intervals,
number of observations (n), between-study variability (t2), residuals variance (σ2) and
study size effect (test for publication bias)

Milk Pathogen Pooled n 95% CI Heterogeneity

prevalence* analysis

t
Sheep L. monocytogenes 0.0356b 9 [0.0153 – 0.0811]

ip
Salmonella 0.0137a 4 [0.0027 – 0.0662]
c
S. aureus 0.3939 6 [0.2275 – 0.5890] τ2=0.870**

cr
STEC 0.0483b 8 [0.0218 – 0.1039] σ2=1.294
Publication bias <.0001

us
a
Goat L. monocytogenes 0.0292 18 [0.0149 – 0.0566]
Salmonella 0.0236a 10 [0.0087 – 0.0626]
S. aureus
STEC
0.3515c
0.0433b
an
19
13
[0.2318 – 0.4932]
[0.0215 – 0.0855]
τ2=1.142**
σ2=1.732
Publication bias 0.762
M
(*) Different superscript letters indicate significant differences in mean pooled prevalence (p<0.10)
(**) Residual heterogeneity is significant (p<0.05)
ed
pt
ce
Ac

Page 18 of 21
Table 2. Meta-analyses of the incidence of pathogens in sheep and goat cheeses,
showing pooled prevalence, 95% confidence intervals, number of observations (n),
between-study variability (t2), residuals variance (σ2) and study size effect (test for
publication bias)
Cheese Pathogen Pooled n 95% CI Heterogeneity
prevalence* analysis
Sheep L. monocytogenes 0.0361a 4 [0.0070 – 0.1644] τ2=0.000

t
ip
a
STEC 0.0276 5 [0.0109 – 0.0679] σ2=1.333
Publication bias 0.108

cr
Goat L. monocytogenes 0.1284b 7 [0.0449 – 0.3160]
Salmonella 0.0591a 6 [0.0146 – 0.2100]

us
b
S. aureus 0.1604 21 [0.0792 – 0.2980] τ2=1.120**
STEC 0.0431a 13 [0.0178 – 0.1004] σ2=3.621
Publication bias 0.116
an
(*) Different superscript letters indicate significant differences in mean pooled prevalence (p<0.10)
(**) Residual heterogeneity is significant (p<0.05)
M
ed
pt
ce
Ac

Page 19 of 21
Table 3. Meta-analyses of the incidence of S. aureus and STEC in goat and sheep
cheeses by type, showing pooled prevalence, 95% confidence intervals, number of
observations (n), between-study variability (t2) and residuals variance (σ2)

Pathogen Cheese type Pooled n 95% CI Heterogeneity

prevalence* analysis
S. aureus Raw 0.3869b 17 [0.0928 – 0.7955] 2
τ =5.513**
a
Pasteurised 0.0420 5 [0.0063 – 0.2326] σ2=6.7235

t
S. aureus Hard/raw milk 0.3457a 6 [0.0627 – 0.8066] τ2=5.515**

ip
Soft/raw milk 0.2568a 11 [0.0501 – 0.6935] σ2=8.377
0.1000b τ2=0.597**

cr
STEC Raw 12 [0.0325 – 0.2689]
Pasteurised 0.0470a 6 [0.0200 – 0.1068] σ2=2.376

us
(*) Different superscript letters indicate significant differences in mean pooled prevalence (p<0.10)
(**) Residual heterogeneity is significant (p<0.05)

an
M
ed
pt
ce
Ac

Page 20 of 21
Highlights

S. aureus is the main pathogen detected in sheep/goat raw milk (22 – 58%)
Salmonella spp. has the lowest pooled incidence in raw milk (1.4 – 2.4%)
L. monocytogenes is more frequently detected in dairy (3.6 – 12%) of goat origin

t
ip
In goat milk cheeses, S. aureus has the highest pooled incidence (16%)
E. coli O57 has been detected in sheep/goat pasteurised milk cheeses (4.7%)

cr
us
an
M
ed
pt
ce
Ac

Page 21 of 21