LACORS/HPA Co-ordinated Food Liaison Group Studies:
The Microbiological Examination of External Packaging of
Raw Meat and Offal from Retail Premises
F Burgess1, CL Little1‡, G Allen2, K Williamson2, RT Mitchell1 and the Food,
Water and Environmental Surveillance Network†.
1, Environmental Surveillance Unit, Health Protection Agency Communicable Disease
Surveillance Centre, 61 Colindale Avenue, London, NW9 5EQ.
2, Food and Environmental Microbiology Services North West, Preston Microbiology Services,
Royal Preston Hospital, Sharoe Green Lane North, Preston, PR2 9HG
On behalf of the Local Authorities Co-ordinators of Regulatory Services and the Health
Protection Agency.
Summary
During September and October 2002 3662 pre-packaged raw meat/offal samples were
collected in a LACORS/PHLS Co-ordinated Food Liaison Group Study with the aim of
identifying the extent and nature of microbiological contamination on external surfaces of
the packaging which could potentially cross-contaminate ready-to-eat foods during and
after purchase.
Salmonella was detected from two (<1%) samples of external
packaging (both from raw chicken) and Campylobacter from 41 (1.1%) samples of
external packaging. The external packaging of game fowl exhibited the highest
contamination from Campylobacter (3.6%), followed by raw chicken (3.0%), lamb (1.6%),
turkey (0.8%), pork (0.2%), and beef (0.1%), with C. jejuni and C. coli accounting for
59% (24/41) and 24% (10/41) of the contaminating Campylobacters, respectively.
Escherichia coli (an indicator of faecal contamination) was isolated from the external
packaging on 4% of the raw meat samples, and at levels in the range of 40 to 10 5
cfu/swab.
The external packaging of raw meats is a vehicle for potential cross-contamination of
Campylobacter, Salmonella, and E. coli in retail premises and consumers’ kitchens. The
external surface of heat sealed packaging was less frequently contaminated with
Campylobacter and E. coli compared to other types (overwrapping, bag and tie tape) of
packaging (p<0.0001 to p=0.01). In addition, external packaging of raw meats was
contaminated less frequently with Campylobacter and E. coli when packaging was intact,
packaging and display areas were visually clean, the display temperature was below
8°C, and when hazard analysis was in place.
‡, Author for correspondence
†, FWES Network comprises Laboratories listed in Annex I
1
Introduction
Pre-packed raw meat is meat packed in advance of sale. Plastic bags, rigid trays with a
covering of a gas-permeable film (overwrap), vacuum packing, or modified atmosphere
packing may be used to pre-pack raw meat1. Packaging plays a key role in the efficient
storage, distribution, retail display of meat, and self-service to the consumer2. High
standards of hygiene and temperature control are essential prerequisites for the quality
and safety of pre-packed meats. The use of basic food hygiene procedures in the
handling and preparation of meat products is vital if gastrointestinal illness is to be
avoided. As the potential hazards involved in the handling of pre-packed raw meat
would not be obvious to the general consumer, it is suggested that appropriate
instruction should be given on the label3.
The Advisory Committee on the
Microbiological Safety of Food (ACMSF) has noted that on-pack hygiene advice (with the
exception of cooking instructions) is infrequently provided on raw meat and poultry
products4, despite the existence of an industry code of practice5.
In terms of foodborne disease burden in England and Wales during 1992 to 2000 the
most important pathogens include Campylobacter and Salmonella6. These pathogens
are known to colonise the intestines of farm animals and may contaminate meat of
cattle, sheep, pigs, and poultry at the time of slaughter7. Raw meat and poultry and their
products can therefore inevitably contain pathogenic microorganisms7-9.
Both
Salmonella (6%) and Campylobacter (ranging from 3% to 6%) have also been isolated
from the external wrapping of raw chicken purchased from retail stores10-12. There is
however no published information on the prevalence of Salmonella and Campylobacter
on the external packaging of raw red meats. The potential for cross-contamination
between raw meats and individuals or surfaces in the kitchen environment and the
subsequent risk for ready-to-eat food becoming contaminated has been well
documented13,14. The infectious dose of Campylobacter infection in humans is thought
to be very low, and information from some outbreaks also suggest that a low infectious
dose of Salmonella may cause disease. It can therefore be difficult for consumers to
control the spread of these pathogens, and as a consequence cross-contamination is a
major concern15.
It has been highlighted that in order to work towards a 20% reduction in foodborne
disease by April 2006 in line with the Food Standards Agency’s target, more emphasis
needs to be placed on Campylobacter. Although the commonest cause of acute
bacterial gastroenteritis, the epidemiology of Campylobacter infection remains poorly
understood. While the widespread contamination of poultry and poultry products with
Campylobacter is well documented8-10, the possible spread of infectious intestinal
disease via handling raw meat packaging is less clear. The aim of the study was to
identify the extent of Campylobacter, Salmonella, and Escherichia coli (a faecal indicator
organism) contamination of external surfaces of raw meat and offal packaging. The
external packaging of various meat and products from a range of retail premises were
examined in order to determine the potential for such packaging to act as a source of
microbiological cross-contamination to ready-to-eat foods, either before or during and
after purchase.
2
Materials and Methods
Sample Collection
The external packaging of pre-packed raw meats collected from retail premises were
examined by Public Health Laboratory Service (PHLS; now the Health Protection
Agency) and non-PHLS laboratories in the UK between 1 September and 31 October
2002 according to a standardised protocol. Pre-packed fresh meats included in the
study were red meats, poultry or offal. Frozen or canned meat products were specifically
excluded from the study. Pre-packed meats were collected from a range of premises
including supermarkets, butchers and greengrocers by staff from local Environmental
Health Departments and were transported to the laboratory in accordance with the Food
Safety Act 1990, Code of Practice No 716.
Information on the packaged meat samples and retail premises was obtained by
observation and enquiry and recorded on a standard proforma. Food hygiene
inspections of premises are carried out by environmental health officers to assess
hygiene and compliance with public health protection aspects of food law17. Some food
premises and businesses pose a greater risk to the consumer than others, which is
reflected by the frequency of inspection. Premises rated Inspection Rating Category A
pose the greatest risk and are visited at least once every six months while premises
rated Inspection Rating Category F pose the least risk and are visited at least once every
five years. Environmental health officers also consider the number of customers likely to
be put at risk if there is a failure in food hygiene and safety procedures in a particular
premise, and award a consumer at risk score accordingly. Scores range from 0 (very
few customers at risk) to 15 (a substantial number of customers at risk). Confidence in
management and food safety management systems are also assessed and scored
accordingly. Confidence in management scores range from 0 (highly confident) to 30 (no
confidence). Additional information collected on packaged meats included type of meat,
packaging used, display area arrangement, and vendor handling of the meats.
Sample Examination
Entire external packaging surfaces were sampled in the laboratory by the surface swab
technique using two 5 x 5 cm2 sterile filmated (gauze) unfolded swabs (moistened from a
20ml Buffered Peptone Water (BPW)). One gauze swab was placed into 90 ml prewarmed (room temperature or 37°C) Bolton broth for detection of Campylobacter, and
the other placed back into the same 20 ml BPW for detection of Salmonella and
enumeration of Escherichia coli. The Bolton broth and BPW were recapped and mixed
carefully to release the microorganisms. Within 20 minutes of placing the gauze into the
BPW the enumeration procedure for E. coli was carried out as described below, and
remainder incubated for detection of Salmonella. Swabs were supplied by Johnson &
Johnson Ltd (P55140, P55405) and NHS Supplies (ENL031, ENL011).
Enumeration of E. coli and detection of Salmonella spp. was carried out in accordance
with HPA Standard Microbiological Methods18,19. Campylobacter spp. were detected by
enrichment in Bolton Selective Enrichment Broth with incubation at 37°C for 4 hours,
followed by further incubation at 41.5°C and subculture to Campylobacter selective agar
(CCDA) after 44  2 h. Inoculated plates were incubated at 41.5°C for 48 h, and colonies
identified as described in HPA Standard Microbiological Method F2120. Isolates of
Campylobacter spp. and Salmonella spp. were sent to the Laboratory of Enteric
3
Pathogens (LEP) at the Central Public Health Laboratory (CPHL) for confirmation and
typing.
Statistical analysis
Descriptive and statistical analysis of the data was undertaken using Microsoft Excel
version 7 and Epi Info version 6.04d. Relative proportions were compared using the
Fishers’ exact test.
Results
The external packaging from 3662 pre-packed raw meat and offal samples collected
from 2304 retail premises was examined in 38 laboratories (27 PHLS and 11 non-PHLS)
in England, Wales, Scotland and Northern Ireland. Samples were submitted by 362
Local Authorities, involving 52 Local Authority Food Liaison Groups (Annex 1). A further
5 samples did not fit the criteria described in the study protocol and were not included in
the analysis.
Microbiological Results
External packaging sampled from 3662 raw meats included packaging from beef (31%),
chicken (25%), pork (24%), lamb (15%), turkey (4%) and other meats (1%; game fowl,
venison) (Table 1). The types of packaged meat products collected included whole
poultry (191; 5.2%), joints (201; 5.5%), steaks (555; 15.1%), chops (428; 11.7%),
portions (873; 23.8%), cubed diced meat (256; 7%), mince (587; 16%), and offal (506;
13.8 %) that consisted of liver (364; 9.9%), kidney (89; 2.4%) and heart (42; 1.1%).
E. coli was isolated from external packaging on 4% (139/3662) of raw meat products,
and at levels in the range of 40 to 105 cfu/swab (Table 1). Salmonella spp. (S.
Goldcoast, S. unnamed) and Campylobacter spp. were detected from the external
packaging of two (0.05%) and 41 (1.1%) raw meat products, respectively (Table 1).
Table 1: Microbiological results of raw meat/offal external packaging (n=3662)
External packaging of
raw meat
Type
Number
Beef
1142 (31.2)
Lamb
563 (15.4)
Pork
877 (23.9)
Chicken
895 (24.9)
Turkey
129 (3.5)
Game fowl
28 (0.8)
Venison
9
(0.2)
Other
17 (0.5)
(rabbit,
mixed
meats)
Not
1
(<0.1)
recorded
Total
3662
Salmonella
/swab (%)
0
0
0
2 (0.2)
0
0
0
0
Campylobacter /swab
(%)
1 (0.1)
9 (1.6)
2 (0.2)
27 (3.0)
1 (0.8)
1 (3.6)
0
0
0
0
2 (0.1)
41 (1.1)
Escherichia coli cfu/swab(%)
<40†
40 102-<103
2
<10
1125 (98.5)
10 (0.9)
5 (0.4)
543 (96.4)
8 (1.4)
7 (1.2)
842 (96.0)
16 (1.8)
11 (1.3)
843 (94.2)
29 (3.2)
19 (2.1)
122 (94.6)
4 (3.1)
2 (1.6)
24 (85.7)
1(3.6)
1 (3.6)
7 (77.7)
1(11.1)
1 (11.1)
13 (76.5)
0 (0)
2 (11.8)
1(100)
352 (96.2)
103-<104
2 (0.2)
4 (0.7)
6 (0.7)
3 (0.3)
0 (0)
2 (7.1)
1 (5.9)
104<105
0 (0)
1 (0.2)
1 (0.1)
0 (0)
1 (0.8)
0 (0)
1 (5.9)
-
-
-
-
69 (1.9)
48 (1.3)
18 (0.5)
4 (0.1)
†, lower limit of detection 40 cfu/swab
4
The external packaging of game fowl and raw chicken exhibited the highest
contamination from Campylobacter (3.6% and 3%, respectively), followed by lamb
(1.6%), turkey (0.8%), pork (0.2%), and beef (0.1%) (Table 1). This finding was
significant when comparing the external packaging of chicken to pork and beef products
(p<0.0001). In the case of game fowl, only a small number of samples were collected
relative to other meat types. Salmonella was only detected from the external packaging
of raw chicken (0.2%; Table 1). The external packaging of other meat (rabbit, mixed
meats) showed the highest contamination from E. coli (23%), followed by venison (22%),
game fowl (14%), chicken (6%), turkey (5%), lamb (4%), pork (4%), and beef (1%)
(Table 1).
The prevalence of Campylobacter present on external packaging also varied according
to the type of raw meat product (Table 2). Of the chicken packaging samples that were
contaminated with Campylobacter, chicken liver (8%) and whole chicken (6.5%) were
more frequently contaminated than chicken portions (2.3%). This finding was significant
when comparing whole chicken to chicken portions (p= 0.0112). Lamb offal external
packaging displayed higher contamination rates with Campylobacter than lamb chops
(Table 2), however, these differences were not statistically significant. Regarding
different Campylobacter spp., C. coli was predominantly isolated from the external
packaging of raw poultry products (chicken, guinea fowl, turkey), whereas C. jejuni was
predominantly isolated from the external packaging of raw chicken, lamb or pork
products (Table 2).
Table 2. Prevalence of Campylobacter spp. present on external packaging of raw
meat products
External packaging
Meat type
Product
Beef
Steak
No.
samples
330
Campylobacter/swab
(%)
1
(0.3%)
Campylobacter isolates†
C. jejuni
C. coli
0
1
Chicken
Whole
Portions
Offal (Liver)
170
614
24
11
14
2
(6.5%)
(2.3%)
(8.0%)
8
4
1
1
5
1
Game Fowl
Whole
9
1
(11.1 %)
0
1
Lamb
Chops
Offal (Liver)
Offal (Heart)
150
173
32
1
6
2
(0.7%)
(3.5 %)
(6.3 %)
1
6
2
0
0
0
Pork
Diced/Cubed
Offal (Liver)
46
133
1
1
(2.2 %)
(0.8 %)
1
1
0
0
Turkey
Portions
68
1
(1.47%)
0
1
†, 7 campylobacter isolates were not further characterised
Campylobacter isolate types
Of the 36 Campylobacter isolates that were further characterised, over two-thirds were
C. jejuni (69%; 25) and 31% (11) were C. coli (Table 3). A breakdown of HS-serotypes
and phage types among C. jejuni and C. coli isolates is provided in Table 3. There were
5
a total of seven and six HS-serotypes of C. jejuni and C. coli, respectively. A wider
range of phage types of C. jejuni (14 phage types) was identified than of C. coli (2 phage
types). Both of the Salmonella positive samples and 39% (16) of the Campylobacter
samples were also contaminated with E. coli.
Table 3. Sero- and phage types of Campylobacter isolated from raw meat/offal
external packaging
Campylobacter
spp.
C. jejuni
Serotype (number of
isolates)
HS1 (2)
HS9 (2)
HS13 (3)
HS18 (3)
HS50 (5)
HS60 (1)
UT† (9)
Phage type (number of isolates)
C. coli
HS25
HS34
HS49
HS51
HS56
UT
44 (2),
44 (1),
2 (2), 44 (1)
2 (1)
44 (2)
2 (1), 44 (1)
(2)
(1)
(3)
(1)
(2)
(2)
1 (1), 2 (1)
39 (1), UT (1)
1 (1), 33 (1), 39 (1)
8 (1), 14 (1), 19 (1)
5 (1), 6 (2), 33 (1), 34 (1)
65 (1)
1 (2), 2 (1), 33 (3), 36 (1), 44 (1), UT (1)
†, UT; Untypeable
The microbial drug resistance of C. jejuni and C. coli isolates is outlined in Table 4.
Sixty-one percent (14/23) of C. jejuni isolates and 18% (2/11) C. coli isolates were
resistant to one antimicrobial drug (trimethoprim). Nine (39%) of C. jejuni and nine (82%)
of C. coli isolates were resistant to two or more antimicrobial drugs. Multiple resistance
(to four or more antimicrobial drugs, including resistance to ciprofloxacin) was seen in
27% (3/11) of C. coli isolates but in none of the C. jejuni isolates.
Table 4. Microbial drug resistance of Campylobacter isolated from raw meat
product external packaging
Antimicrobial agent
Tm
ATm
TTm
ATTm
TTmNx
ATTmNxCp
AKTTmNxCpNe
C. jejuni
(n=23)
14 (60.9%)
4 (17.4%)
2 (8.7%)
3 (13.0%)
0
0
0
C. coli
(n=11)
2 (18.2%)
1 (9.1%)
4 (36.3%)
0
1 (9.1%)
1 (9.1%)
2 (18.2%)
Total
(n=34)
16 (47.1%)
5 (14.7%)
6 (17.7%)
3 (8.8%)
1 (2.9%)
1 (2.9%)
2 (5.9%)
A, ampillicin; Cp, ciprofloxacin; K, kanamycin; Ne, neomycin; Nx, naladixic acid; T, tetracycline; Tm, trimethoprim
Packaging of raw meat products
Most raw meat product samples were packed in a polystyrene tray with an over wrap
(40%), in a plastic tray and heat sealed lid (23%) or in a polystyrene tray and heat sealed
lid (16%) (Table 5). More samples with Campylobacter and Salmonella detected were
6
on the external surface of polystyrene tray and overwrap packaging (2.1%) when
compared to all other packaging types (0%-1.3%; Table 5). This finding was significant
when comparing polystyrene tray and overwrap packaging with plastic tray and heat
sealed plastic lid packaging (p<0.003). Significantly fewer samples packaged with a heat
sealed lid and either a plastic or polystyrene tray had E. coli present (1%) compared to
their counterparts with an overwrap (4%-7%; Table 5) (p<0.0001).
Over half of the raw meat products were packaged in normal atmosphere (57%), 27%
were modified atmosphere packed (MAP), and 5% were vacuum packed. For the
remaining 11% of samples this information was not recorded. More samples with
Campylobacter and Salmonella detected were on the external packaging surface of
samples packed in normal atmosphere (1.7%) when compared to MAP (0.1%) and
vacuum packed products (0.6%) (Table 5). This finding was significant when comparing
normal atmosphere packed samples with those that were MAP (p<0.0001). Similarly,
significantly more samples with E. coli present were on the external packaging surface of
samples packed in normal atmosphere (5%) compared to MAP (1%; p<0.0001) and
vacuum packed products (3%; p=0.01) (Table 5). MAP (87%) and vacuum packed
(45%) samples were more likely to be heat sealed compared to samples packed in a
normal atmosphere (17%).
Table 5: Raw Meat Packaging in relation to the presence of Campylobacter,
Salmonella and E. coli
Number of
Samples
(n=3662)
(%)
Samples with
Campylobacter or
Salmonella (%)
Samples with E. coli
(40 – 105 cfu/swab)
(%)
Packaging Type
Bag & Tie Tape
Plastic Tray / Heat Sealed Plastic Lid
Plastic Tray / Over Wrap
Polystyrene Tray / Heat Sealed Plastic Lid
Polystyrene Tray / Over Wrap
Other (bags, cling & heat sealed film & vac)
Not Recorded
313
857
129
598
1486
227
52
(8.55)
(23.40)
(3.52)
(16.33)
(40.58)
(6.20)
(1.42)
2
4
1
32
3
-
(0.64)
(0.47)
(0.78)
(2.15)
(1.32)
-
14
9
9
4
87
13
3
(4%)
(1%)
(7%)
(1%)
(6%)
(6%)
(6%)
Packing Atmosphere
MAP
Normal
Vacuum-packed
Not Recorded
972
2076
343
271
(26.54)
(56.69)
(4.88)
(7.32)
1
36
2
3
(0.10)
(1.73)
(0.58)
(1.11)
9
105
10
15
(1%)
(5%)
(3%)
(6%)
Packaging Intact
Yes
No
Not Recorded
3527
82
53
(96.31)
(2.24)
(1.45)
39
2
1
(1.12)
(2.44)
(1.89)
104
33
2
(3%)
(41%)
(4%)
Packaging appearance visually clean
Yes
No
Not recorded
3006
137
519
(82.08)
(3.74)
(14.17)
27
5
10
(0.90)
(3.65)
(1.93)
98
22
21
(3%)
(16%)
(4%)
Packaging details
Over three-quarters of raw meat product samples (78%) did not have safe handling
instructions on the packaging, 15% did, and for 7% this information was not recorded.
The external packaging was intact for the majority (96%) of samples (Table 5). The
proportion of samples with Campylobacter or Salmonella detected on external packaging
7
was higher when the packaging was not intact (2.4%) when compared to samples with
intact packaging (1.1%; Table 5), although this finding was not significant. However,
significantly more samples that did not have intact packaging had E. coli present (41%)
compared to those that had intact packaging (3%) (p=0.0001) (Table 5).
The external packaging was visually clean for most samples (82%) (Table 5).
Significantly a higher proportion of samples with Campylobacter or Salmonella present
were from packaging that was not visually clean (3.65%) compared to when it was
visually clean (0.9%) (p=0.0114). The proportion of samples with E. coli present was
also higher when the packaging was not clean (16%) compared to when it was clean
(3%) (p<0.0001) (Table 5).
Display of packaged raw meat products
Most (84%) samples were in display areas that did not have plastic bags available for
the customer to place the packaged raw meat in, 10% were, and for 6% this information
was not recorded. The display area was visually clean for most (90%) of samples (Table
6). More samples with Campylobacter or Salmonella present on the external packaging
surface were from display areas that were not visually clean (2.14%) when compared to
those that were clean (1.09%). Similarly, more samples with E. coli present were from
display areas that were not visually clean (6%) compared to those that were clean (4%).
However, these findings were not statistically significant.
Table 6. Display area of packaged raw meats in relation to presence of
Campylobacter, Salmonella and E. coli
Number of
Samples
(n=3662)
(%)
Samples with
Campylobacter or
Salmonella (%)
Samples with E. coli
(40 – 105 cfu/swab)
(%)
Display area visually clean
Yes
No
Not recorded
3294
234
134
(89.95)
(6.39)
(3.66)
36
5
1
(1.09)
(2.14)
(0.75)
119
14
6
(4%)
(6%)
(4%)
Display area & presence of:
Blood
Meat juices
Water
Other#
Not recorded
267
228
26
38
3103
(7.29)
(6.23)
(0.71)
(1.04)
(84.74)
4
4
1
2
31
(1.50)
(1.75)
(3.85)
(5.26)
(1.00)
17
17
1
1
103
(6%)
(7%)
(4%)
(3%)
(3%)
Meat Juices Under Sample
Yes
No
Not Recorded
354
3026
282
(9.67)
(82.63)
(7.70)
9
32
1
(2.54)
(1.06)
(0.35)
28
103
8
(8%)
(3%)
(3%)
Meat Sample in Contact with Another
Yes
No
Not Recorded
2143
1163
356
(58.52)
(31.76)
(9.72)
23
18
1
(1.07)
(1.55)
(0.28)
86
42
11
(4%)
(4%)
(3%)
Display Temperature
Equal/below 8° C
Above 8° C
Not Recorded
3377
111
174
(92.21)
(3.03)
(4.75)
37
2
3
(1.10)
(1.80)
(1.72)
122
11
8
(4%)
(10%)
(5%)
Packaged raw meat display area
details
#, Other contaminants included ice, meat residues, crumbs, dead fly, dust, dirt, or sticky substances
8
Where details were provided on contaminants that were present in the display area of
packaged raw meats (15%), 7% contained blood, 6% meat juices, 1% water, and 1%
contained other contaminants (Table 6). Presence of Campylobacter, Salmonella or E.
coli on the external packaging of raw meat ranged from 1 to 7% depending on the type
of contaminant present in the display area, although this was not statistically significant.
Most (83%) sampled packaged raw meats did not have meat juices present under the
samples in the display area (Table 6). A significantly higher proportion of samples with
Campylobacter or Salmonella present were found when meat juices were present under
the sample (2.5%) compared to when they were not (1%) (p=0.0315) (Table 6). The
proportion of samples with E. coli present was also significantly higher when meat juices
were present under the sample (8%) compared to when they were not (3%) (p=0.0002).
Over half (59%) of the sampled packaged raw meats were in contact with another
packaged raw meat product in the display area (Table 6). There was no significant
correlation between the presence of Campylobacter, Salmonella or E. coli on the
external packaging of raw meats and contact or no contact with another packaged raw
meat product.
The majority (92%) of packaged raw meat was displayed at or below 8°C (Table 6).
There was no significant correlation between the proportion of samples with
Campylobacter or Salmonella present on external packaging and the display
temperature (Table 6). However, twice as many samples with E. coli present were from
samples displayed above 8°C (10%) compared to those displayed at or below 8°C (4%)
(p=0.0030) (Table 6).
Vendor hand contact
Most vendors (84%) touched the external packaging of raw meat by hand (Table 7), and
a quarter (26%) of the same vendors also touched unwrapped ready-to-eat food by
hand, 69% did not, and for 5% this information was not recorded. There was no
significant correlation between the proportion of samples with Campylobacter,
Salmonella or E. coli present on external packaging of raw meats and whether the
vendor touched the raw meat packaging by hand or not.
Regarding vendor/checkout handling, facilities were within easy reach for washing hands
at the point of sale for a quarter of the samples (26%), and these were used about half
the time after direct handling of raw meat product packaging (47%) (Table 7). Where
handwashing facilities were not available within easy reach at the point of sale (70%),
most samples (84%) were handled by vendors that did not take other precautions to
avoid direct handling/cross contamination from raw meat packaging (Table 7). There
was no significant correlation between the proportion of samples with Campylobacter or
Salmonella present on external packaging of raw meats and the handwashing facility
information collected (Table 7). However, twice as many samples with E. coli present
were from samples that had been handled by the vendor where hand-washing facilities
were available within easy reach at the point of sale (7%) compared to where they were
not (3%) (p<0.0001) (Table 7).
9
Table 7. Vendor hand contact in relation to the presence of Campylobacter,
Salmonella and E. coli
Number of
Samples
(n=3662)
(%)
Samples with
Campylobacter or
Salmonella (%)
Samples with E.
coli (40 – 105
cfu/swab) (%)
Vendor touched raw meat packaging by
hand
Yes
No
Not Recorded
3091
456
115
(84.41)
(12.45)
(3.14)
37
4
1
(1.20)
(0.88)
(0.87)
117
17
5
(4%)
(4%)
(4%)
Handwashing facilities within easy reach at
point of sale
Yes
No
Not Recorded
936
2571
155
(25.56)
(70.21)
(4.23)
12
27
3
(1.28)
(1.05)
(1.94)
60
73
6
(7%)
(3%)
(4%)
Handwashing facilities used (n=936)
Yes
No
Not Recorded
436
431
69
(46.58)
(46.05)
(7.37)
5
6
1
(1.15)
(1.39)
(1.45)
25
30
5
(6%)
(7%)
(7%)
Handwashing facilities not used & other
precautions taken* (n=2571)
Yes
346
(13.46)
3
(0.87)
No
2153
(83.74)
24
(1.11)
Not Recorded
72
(2.80)
*, Use of sanitisers, sanitiser wipes, gloves, or the provision of separate wrapping at the checkout
8
64
1
(2%)
(3%)
(1%)
Premises details
Among the 2304 retail premises visited to collect samples, 45% were licensed as
butcher shops, 52% were not and for the remainder (3%), this was not recorded. Over
half of the retail premises (56%) handled or wrapped raw meat on the premises, 41% did
not and for the remainder (3%), this was not recorded. There was no significant
correlation between the proportion of raw meat samples with Campylobacter or
Salmonella present on the external packaging from premises licensed as butcher shops
(1.5%; 24/1621) or not (0.9%; 17/1945), or if raw meat was handled or wrapped on the
premises (1.2%; 25/2089) or not (1%; 15/1466). Similarly, there was no significant
correlation between the proportion of samples with E. coli present and whether the
premise was licensed as a butcher shop (5%; 79/1621) or not (3%; 56/1944) or whether
raw meat was handled or wrapped on the premises (5%; 94/2088) or not (3%; 42/1466).
Nearly all of the premises visited were permanent premises (99%; 2276), the remainder
comprised of temporary market stalls or mobile vendors (1%; 28). Taking account of the
main activity of the premises concerned, over two-thirds (68%) were supermarkets and
approximately a quarter (23%) were high street butchers (Table 8). Significantly more
samples with Campylobacter or Salmonella present on the external packaging were from
other premises (3.47%) when compared to supermarkets (0.84%; p=0.0007) and high
street butchers (1.37%; p=0.0355) (Table 8). Although represented in comparatively low
numbers, the proportion of samples from delicatessens with E. coli present on the
external packaging (18%) was higher when compared to all other premises types (2-8%)
(Table 8). This finding was significant when comparing samples from delicatessens with
those from supermarkets (p=0.011).
10
Table 8. Premises details in relation to the presence of Campylobacter,
Salmonella, and E. coli present on external packaging of raw meats
Premises details
No.
premises
(n=2304)
(%)
Samples with
Campylobacter or
Salmonella (n=3662)
(%)
Samples with E.
coli (40 – 105
cfu/swab (%)
1574
528
13
12
176
(68)
(23)
(1)
(1)
(8)
22/2607
10/728
10/288
(0.84)
(1.37)
(3.47)
57
58
3
1
20
(2)
(8)
(18)
(5)
(7)
1
(<0.1)
-
-
-
-
53
(2)
5/79
(6.33)
7
(9)
479
1196
(21)
(52)
13/706
19/1968
(1.84)
(0.97)
36
70
(5)
(4)
283
145
34
114
(12)
(6)
(1)
(5)
2/446
1/226
0/48
2/189
(0.45)
(0.44)
10
6
2
8
(2)
(3)
(4)
(4)
Consumers at Risk Score
0 (Very Few)
5 (Few)
10 (Intermediate)
15 (Substantial)
Not Recorded
14
1314
753
49
174
(<1)
(57)
(33)
(2)
(8)
3/16
19/1945
17/1336
0/87
3/278
(18.75)
(0.98)
(1.27)
3
76
47
2
13
(19)
(4)
(4)
(2)
(5)
Confidence in Management Score
0 (High)
5 (Moderate)
10 (Some)
20 (Little)
30 (No Confidence)
Not Recorded
176
1040
795
117
13
163
(8)
(45)
(35)
(5)
(<1)
(7)
1/277
20/1696
18/1224
1/189
0/16
2/260
(0.36)
(1.18)
(1.47)
(0.53)
(0.77)
4
62
53
10
1
9
(1)
(4)
(4)
(5)
(6)
(3)
1881
86
116
(81)
(4)
(5)
36/3038
0/123
2/165
(1.18)
(1.21)
107
6
7
(4)
(5)
(4)
107
114
(5)
(5)
3/150
1/186
(2.00)
(0.54)
12
7
(8)
(4)
2091
940
694
95
251
111
88
125
(91)
(45)
(33)
(5)
(12)
(5)
(4)
(5)
39/3313
8/1466
18/1131
3/172
9/383
1/161
1/141
1/165
(0.94)
(0.55)
(1.59)
(1.74)
(2.35)
(0.62)
(0.71)
(0.48)
125
49
43
9
24
5
9
(4)
(3)
(4)
(5)
(6)
(4)
(4)
Premises type
Supermarket
High street butchers
Delicatessen
Greengrocer
Other (Butcher’s concession,
convenience/general store, service
station, farm shop, market stall, cash &
carry)
Not recorded
Inspection Rating Category
Points
Minimum Frequency of
Category
Inspection
A
At least every 6
months
B
At least every year
C
At least every 18
months
D
At least every 2 years
E
At least every 3 years
F
At least every 5 years
Not recorded
Hazard Analysis system
In place & documented
In place & undocumented
In place; documentation status not
recorded
Not in place
Not recorded
Management food hygiene training
Received training & attended;
a basic 6 hour course
an intermediate course
an advanced course
another recognised course
training not specified
No training
Not recorded
(1.06)
(1.08)
11
Of the premises visited, over half had an inspection rating category of C (52%; inspected
at least every 18 months) and had consumer at risk scores of 5 (57%; few at risk), and
most had a confidence in management score of 5 (45%; moderate confidence) or 10
(34%; some confidence) (Table 8). Significantly more samples with Campylobacter or
Salmonella present on the external packaging of raw meats were from premises rated
category A (6.33%) compared to premises in all other categories (p=0.001-0.027) (Table
8). Likewise more samples with E. coli present on external packaging were from
premises rated category A (9%) compared to premises in all other categories (p=0.00750.0450) (Table 8).
A higher proportion of samples with Campylobacter or Salmonella present on the
external packaging of raw meats were from premises with a consumer of risk score of 0
(very few) (19%) when compared to premises in all other risk scores (few to substantial;
p=0.0005-0.0031) (Table 8). Likewise more samples with E. coli present on external
packaging were from premises with a consumer of risk score of 0 (19%) when compared
to premises in all other risk scores (p=0.0190-0.0258). There was no significant
correlation between numbers of samples that had Campylobacter or Salmonella or E.
coli present and confidence in management scores of premises (Table 8).
Most (81%) premises visited had a documented hazard analysis system in place and a
further 4% had an undocumented hazard analysis system in place (Table 8). The
proportion of samples that had Campylobacter or Salmonella present on the external
packaging was higher where premises had no hazard analysis system in place (2%)
compared to premises that had a hazard analysis system in place (1.11%), although this
finding was not statistically significant (Table 8). However, significantly twice as many
samples that had E. coli present on the external packaging were from premises that did
not have hazard analysis in place (8%) compared to those premises that did (4%)
(p=0.0128).
The majority of premises had managers that had received some form of food hygiene
training (91%) (Table 8). Of those managers with food hygiene training, almost half
(45%) had attended a basic 6-hour food hygiene course, a third (33%) were trained to an
intermediate, and 5% to an advanced level. A further 12% had attended another
recognised course (Meat and Livestock Commission, in-house training). There was no
significant correlation between numbers of samples that had Campylobacter, Salmonella
or E. coli present on the external packaging and the food hygiene training of managers
(Table 8).
Discussion
Although there have been numerous studies investigating the prevalence of
campylobacter and salmonella in animals, poultry and raw meats, there is limited
information on the occurrence of these organisms on the external packaging of raw
meats, and this has previously only covered that of raw chicken and offal. This is the
first such study to provide information on the prevalence of Campylobacter, Salmonella
and E. coli (an indicator of faecal contamination) on the external packaging of raw red
meats. Of the external packaging examined from raw red meat, offal, and poultry
products in this study, Salmonella were only detected from the external packaging of raw
chicken (0.2%) and at a lower prevalence compared to that previously found by
Jørgensen et al. (6%)10. In contrast Harrison et al.11 did not recover Salmonella from the
12
external packaging of raw chicken. Overall Campylobacter were detected from 1.1% of
external packaging, out of which the external packaging of game fowl exhibited the
highest contamination from Campylobacter (3.6%), followed by raw chicken (3.0%), lamb
(1.6%), turkey (0.8%), pork (0.2%), and beef (0.1%). The prevalence of Campylobacter
found on the external packaging of raw chicken concurs with that reported by Harrison et
al. (3%)11 but is lower than that previously found by both Bolton et al. (5.4%)12 and
Jørgensen et al. (6%)10. E. coli was isolated from the external packaging on 4% of the
samples in this study, and at levels in the range of 40 to 105 cfu/swab. In proportion, the
external packaging of game meats (rabbit (23%), venison (22%)) and game fowl (14%)
showed the highest contamination from E. coli compared to other red meats (3%) and
poultry (5.5%).
Of the campylobacters recovered from the external packaging of raw meats most were
C. jejuni, with C. coli accounting for the remainder. This is in accordance with that found
both in retail raw meats and human isolates in England and Wales8,21. In general, C. coli
isolates from the external packaging were more multi-resistant to antimicrobial drugs,
including quinolones such as ciprofloxacin, than C. jejuni. Resistance to ciprofloxacin
has also been shown to be higher in human C. coli isolates (26%) than in C. jejuni
(6%)21.
Previous studies have demonstrated the incidence of Campylobacter to be higher than
Salmonella in raw meats and poultry9-11,22,23. Moreover, raw chicken is more frequently
contaminated with campylobacter compared to that found in lamb, beef and pork
meat8,24. This concurs with that found on the external packaging of raw meats, as
previously described above. Harrison et al.11 and the Food Standards Agency9 have
also shown that whole chickens are more contaminated with Campylobacter than that of
portions. The external packaging of whole chicken in this study was also contaminated
more frequently with campylobacter (8%) than that from portions (2.3%). Offal has also
shown to be frequently contaminated with campylobacter (54% to 73%)8.
Campylobacter contamination of the external packaging of offal products in this study
ranged from 0-8%, out of which packaging of chicken liver exhibited the highest
frequency (8%) of Campylobacter being present. In contrast, Bolton et al.12 found the
external packaging of ox liver to be most frequently contaminated with Campylobacter
(9%).
This study has demonstrated that the external packaging from samples that were heatsealed were much less likely to have Campylobacter or E. coli present compared to
those from other packaging types, such as over-wrapping. Similarly, Bolton et al.12
reported that the external packaging of raw chicken and turkey that were heat sealed
had no Campylobacter present compared to 6.4% of those packaged in over-wrapping
with polystyrene trays12. The integrity of packaging seals on food packages is of critical
importance. A major problem is seal leakage, which can give rise to environments that
favour the growth of micro-organisms. The seal integrity of heat sealed packaging is
more robust, i.e. intact, than over-wrapping or bags with tie tapes and therefore less
likely to allow leakage of any meat juices or moisture, and thereby contamination on to
the external packaging of raw meats and other surfaces.
The external packaging of raw meats collected from retail premises was contaminated
more frequently with Campylobacter and E. coli when packaging and display areas were
not visually clean, the display temperature was above 8°C, and when hazard analysis
was not in place. The contamination on the external packaging of raw meats with micro-
13
organisms could also therefore be a reflection of environmental contamination of the
packaged product. Improving the cleaning and sanitisation schedules used by retail
premises, in conjunction with implementing HACCP principles as the basis for the
retailers’ product safety management systems would reduce environmental crosscontamination of the packaged meats.
This study has provided further evidence that external packaging of raw meats is a
vehicle for potential cross-contamination of Campylobacter, Salmonella, and E. coli.
This and other studies10-12 have highlighted a risk factor that may have implications for
the transmission of Campylobacter and Salmonella. The presence of Campylobacter
and Salmonella on the external packaging of raw meats, poultry and offal raises concern
as consumers would not expect products to be contaminated on the outside. The
practice of providing separate plastic bags for consumers to place the packaged raw
meat in is not widespread, only 10% of samples were collected from display areas where
plastic bags were available. Provision of separate plastic bags in which to place
packaged raw meat may reduce the risk of cross-contamination. Within the home
contamination may occur after handling raw meat packaging followed by contact with
ready-to-eat foods, kitchen work surfaces and utensils. Studies have shown that the
packaging of raw meats is often not removed from the preparation area during food
preparation25-27. Seventy-eight percent of packed raw meat samples in this study did not
have any safe handling instructions on the packaging. The ACMSF has recommended
that food safety advice on raw meat products should include measures for effective
cooking, and avoidance of cross-contamination4. Targeted advice and promotion of
good hygiene practice to consumers is also included in the Food Standards Agency food
hygiene campaign28, an initiative launched as part of the Agency’s strategy to reduce
foodborne disease29.
Acknowledgements
The authors would like to thank all the staff in the Environmental Health Departments
throughout the UK who collected samples for this study, all the staff in the PHLS (now
the HPA) and non-PHLS laboratories who performed microbiological examination.
Thanks are also extended to LEP (CPHL) for typing isolates, to David Lock at LACORS
for coordinating the participation of Environmental Health Officers and advice from the
LACORS Food Examination Focus Group, to the Regional FWE Coordinators Forum,
and Mr Kevin Bertram for their advice in preparing the sampling protocol, and to Lilian
Hucklesby for entering the data.
References
1. Meat Livestock Commission (MLC). Shelf Life of Fresh Meat. Matching Product Life to
Requirements. Milton Keynes: MLC, 2000.
2. Cowan C. Trends in retail display of fresh meat in Europe. Farm & Food 1997; 7: 20-21.
3. The Food Labelling Regulations 1996. London: HMSO.
4. Advisory Committee on the Microbiological Safety of Food, Draft Second ACMSF Report on
Campylobacter. Available at: http://www.foodstandards.gov.uk/multimedia/pdfs/acm651.pdf
14
5. Institute of Grocery Distribution. Voluntary guidelines for the Provision of Food Safety Advice
on Product Labels.2000.IGD, London
6. Adak GK, Long SM, O’Brien SJ. Trends in indigenous foodborne disease and deaths,
England and Wales: 1992 to 2000. Gut 2002; 51: 832-41.
7. Department
for
Environment,
Food
and
Rural
Affairs.
Zoonoses
Report
United Kingdom 2000. Available at:
http://www.defra.gov.uk/animalh/diseases/zoonoses/zoonoses_reports/zoonoses_2000.pdf
8. Kramer, JM, Frost, JA, Bolton, FJ and Wareing, DRA. Campylobacter contamination of raw
meat and poultry at retail sale: Identification of multiple types and comparison with isolates
from human infection. J. Food Protection 2000; 63: 1654-1659.
9. Food Standards Agency (FSA). UK-wide Survey of Salmonella and Campylobacter
Contamination of Fresh and Frozen Chicken on Retail Sale.
Available at:
http://www.food.gov.uk/multimedia/pdfs/campsalmsurvey.pdf
10. Jorgensen, F., Bailey, R., Williams, S., Henderson, P, Wareing, DRA, Bolton, FJ, Frost, JA,
Ward, L. and Humphrey, TJ. Prevalence and numbers of Salmonella and Campylobacter spp.
On raw, whole chickens in relation to sampling methods. Int. J. Food Micro. 2002; 76: 151164.
11. Harrison, WA, Griffith, CJ, Tennant, D. and Peters, AC. Incidence of Campylobacter and
Salmonella isolated from retail chicken and associated packaging in South Wales. Letters in
Applied Microbiology 2001; 33: 450-454.
12. Bolton FJ, Williamson JK, Allen G, Wareing DRA, Frost JA. Prevalence of C. jejuni and C. coli
in meat products and packaging sold at retail: a potential public health problem. Abstracts
and Final Program of the 10th International Workshop on Campylobacter, Helicobacter and
Related Organisms, Baltimore, MD, 1999.
13. Dawkins HC, Bolton FJ, Hutchinson DN. A study of the spread of Campylobacter jejuni in four
large kitchens. J Hyg Camb 1984; 92: 357-364.
14. Humphrey TJ, Martin KW, Slader J, Durham K. Campylobacter spp. in the kitchen: spread
and persistence. J App. Microbiol 2001; 91: 115S-120S.
15. House of Commons Agriculture Committee Fourth Report. Food Safety. Vol I. 22 April 1998.
London: HMSO
16. Food Safety Act 1990, Code of Practice No 7:- Sampling for Analysis and Examination
(Revised October 2000). London: FSA, 2000.
17. Food Safety Act 1990, Code of Practice No 9:- Food Hygiene Inspections. (Second Revision
October 2000). London: FSA, 2000.
18. Health Protection Agency (HPA). Standard Methods for Food Products. Direct Enumeration
of Escherichia coli. Standard Method: F20. London: HPA, 2003
19. Health Protection Agency (HPA). Standard Methods for Food Products. Detection of
Salmonella spp. Standard Method: F13. London: HPA, 2003
20. Health Protection Agency (HPA). Standard Methods for Food Products. Detection of
Campylobacter spp. Standard Method: F21. London: HPA, 2003
15
21. Health Protection Agency (HPA). The Campylobacter Sentinel Surveillance System – data
from the first two years of the study. CDR Weekly 13 (19), 9 May 2003. Available at:
http://www.hpa.org.uk/cdr/PDFfiles/2003/cdr1903.pdf
22. Little, CL, Gillespie, I, de Louvois, J and Mitchell, R (1999) Microbiological Investigation of
halal butchery products and butchers’ premises. Commun. Dis. Public Health 2, 114-8.
23. Zhao C, Ge B, de Villena J, Sudler R, Yeh E, Zhao S, White DG, Wagner D, Meng J.
Prevalence of Campylobacter spp., Escherichia coli, and Salmonella serovars in retail
chicken, turkey, pork, and beef from the Greater Washington, D.C., area. Appl. Env. Microbiol
2001; 67: 5431-5436.
24. Pezzotti G, Searfin A, Luzzi I, Mioni R, Milan M, Perin R. Occurrence and resistance to
antibiotics of Campylobacter jejuni and Campylobacter coli in animals and meat in
northeastern Italy. Int J Food Microbiol 2003; 82: 281-287.
25. Worsfold, D and Griffith, CJ (1997) An assessment of the standard of consumer food safety
behaviour. Journal of Food Protection 60, 399-406.
26. Jay, LS, Comar, D and Govenlock, LD (1999) A video study of Australian domestic foodhandling practices. Journal of Food Protection 62, 1285-1296.
27. Griffith, C, Davidson, C, Peters, A and Lewis, A (2000) The use of notational analysis to
assess cross-contamination during domestic food preparation. In Proceedings of the 87th
Annual Meeting of the International Association of Food Protection Atlanta, USA, 6-9 August
2000. IOWA, USA: International Association for Food Protection.
28. Food
Standards
Agency.
Hygiene.
Online.
http://www.food.gov.uk/hygcampaign/> (accessed 26 January 2004).
Available
HTTP:
29. Food Standards Agency. Microbiological Foodborne Disease Strategy, July 2001. Online.
Available
HTTP:
http://www.food.gov.uk/multimedia/pdfs/fdscg-strategy-revised.pdf>
(accessed 22 December 2003).
16
Annex 1: Participating Laboratories and Local Authority Food Liaison Groups
Table I: Participating PHLS (now HPA and HPA Collaborating) Laboratories and
number of samples
HPA Region
East
London
South East
West Midlands
North West
North East, Yorkshire & the Humber
South West
East Midlands
Laboratory Name
Chelmsford
Norwich
London FWEM1
Ashford
Brighton
Reading
WEMS2
Birmingham
Coventry
Hereford
Shrewsbury
Stoke
Chester
Preston
Carlisle
Hull
Leeds
Middlesbrough
Newcastle
Sheffield
Bristol
Exeter
Gloucester
Plymouth
Truro
Leicester
Lincoln
Total
Number of Samples
145
129
383
130
240
74
139
67
147
43
120
57
158
193
66
95
55
106
90
167
175
100
36
26
60
128
207
3336
1, London Food, Water & Environmental Microbiology Laboratory
2, Wessex Environmental Microbiological Service
17
Table II: Participating Other Laboratories and number of samples
Nation
England
Northern Ireland
Scotland
Wales
Total
Laboratory
Kings Lynn & West Norfolk
Derbyshire Royal Infirmary
Belfast City Hospital
Aberdeen City Council Public Analyst
Dumfries & Galoway Royal Infirmary
Dundee Scientific Services
Edinburgh A & S Services
Glasgow Scientific Services
Royal Alexandra, Paisley
Bangor
Cardiff
Number of Samples
11
8
100
13
2
6
22
34
4
40
86
326
18
Table III: Participating Food Safety Liaison Groups and number of samples
Local Authority Food Liaison Group
Berkshire
Buckinghamshire
Cambridgeshire
Cheshire
Cornwall
Cumbria
Derbyshire
Devon
Dorset
Durham
East Sussex
Essex
Gloucestershire
LFCG1 Greater London NE Sector
LFCG Greater London NW Sector
LFCG Greater London SE Sector
LFCG Greater London SW Sector
Greater Manchester
Hampshire & Isle Of Wight
Hereford & Worcester
Hertfordshire & Bedfordshire
Humberside
Kent
Lancashire
Leicestershire
Lincolnshire
Merseyside
North Yorkshire
Northamptonshire
Northern Ireland Food Group2
Northumberland
Norfolk
Nottinghamshire
Oxfordshire
Scottish Food Enforcement Liaison Committee3
Shropshire
Somerset
South West Yorkshire
Staffordshire
Suffolk
Surrey
Tees Valley
Tyne & Wear
Wales North Group
Wales South East Group
Wales South West Group
Warwickshire
West Midlands
West of England
West Sussex
Wiltshire
Total
Number of Samples
35
43
37
97
60
80
114
99
53
36
93
52
36
69
75
59
58
88
60
76
99
80
131
91
128
92
48
70
73
100
11
106
102
20
81
59
28
83
72
64
76
67
72
54
75
17
88
85
94
79
97
3662
1, London Food Co-ordinating Group
2, Northern Ireland Food Group comprises of the Eastern, Northern, Southern and Western Groups
3, SFELG comprises of Central Scotland, Fife & Tayside, Lothian & Scottish Borders, North Scotland, and West of
Scotland
19