High-quality fish oil for human consumption requires low levels of toxic elements. The aim of this study was to compare different oil extraction methods to identify the most efficient method for extracting fish oil of high quality with the least contamination. The methods used in this study were Soxhlet extraction, enzymatic extraction, wet reduction, and supercritical fluid extraction. The results showed that toxic elements in fish oil could be reduced using supercritical CO2 at a modest temperature (60°C) and pressure (35 MPa) with little reduction in the oil yield. There were significant reductions in mercury (85 to 100%), cadmium (97 to 100%), and lead (100%) content of the fish oil extracted using the supercritical fluid extraction method. The fish oil extracted using conventional methods contained toxic elements at levels much higher than the accepted limits of 0.1 μg/g.
The present study was conducted to investigate the prevalence and antibiotic resistance among Campylobacter jejuni in ulam at farms and retail outlets located in Kuala Terengganu, Malaysia. A total of 526 samples (ulam, soil, and fertilizer) were investigated for the presence of C. jejuni and the gene for cytolethal distending toxin (cdt) by using a multiplex PCR method. Antibiotic susceptibility to 10 types of antibiotics was determined using the disk diffusion method for 33 C. jejuni isolates. The average prevalence of contaminated samples from farms, wet markets, and supermarkets was 35.29, 52.66, and 69.88%, respectively. The cdt gene was not detected in 24 of the 33 C. jejuni isolates, but 9 isolates harbored cdtC. Antibiotic resistance in C. jejuni isolates was highest to penicillin G (96.97% of isolates) followed by vancomycin (87.88%), ampicillin (75.76%), erythromycin (60.61%), tetracycline (9.09%), amikacin (6.06%), and norfloxacin (3.03%); none of the isolates were resistant to ciprofloxacin, enrofloxacin, and gentamicin. In this study, C. jejuni was present in ulam, and some isolates were highly resistant to some antibiotics but not to quinolones. Thus, appropriate attention and measures are required to prevent C. jejuni contamination on farms and at retail outlets.
Sprouts have gained popularity worldwide due to their nutritional values and health benefits. The fact that their consumption has been associated with numerous outbreaks of foodborne illness threatens the $250 million market that this industry has established in the United States. Therefore, sprout manufacturers have utilized the U.S. Food and Drug Administration recommended application of 20,000 ppm of calcium hypochlorite solution to seeds before germination as a preventative method. Concentrations of up to 200 ppm of chlorine wash are also commonly used on sprouts. However, chlorine-based treatment achieves on average only 1- to 3-log reductions in bacteria and is associated with negative health and environmental issues. The search for alternative strategies has been widespread, involving chemical, biological, physical, and hurdle processes that can achieve up to 7-log reductions in bacteria in some cases. The compilation here of the current scientific data related to these techniques is used to compare their efficacy for ensuring the microbial safety of sprouts and their practicality for commercial producers. Of specific importance for alternative seed and sprout treatments is maintaining the industry-accepted germination rate of 95% and the sensorial attributes of the final product. This review provides an evaluation of suggested decontamination technologies for seeds and sprouts before, during, and after germination and concludes that thermal inactivation of seeds and irradiation of sprouts are the most practical stand-alone microbial safety interventions for sprout production.
Vibrio vulnificus is a highly invasive human pathogen that exists naturally in estuarine environment and coastal waters. In this study, we used different PCR assays to detect V. vulnificus in 260 seafood and 80 seawater samples. V. vulnificus was present in about 34 (13%) of the 260 seafood samples and 18 (23%) of the 80 seawater samples. Repetitive extragenic palindromic PCR (REP-PCR) and enterobacterial repetitive intergenic consensus PCR (ERIC-PCR) were applied to subtype the V. vulnificus isolates. Twenty-five REP profiles and 45 ERIC profiles were observed, and the isolates were categorized into 9 and 10 distinct clusters at the similarity of 80%, by REP-PCR and ERIC-PCR, respectively. ERIC-PCR is more discriminative than REP-PCR in subtyping V. vulnificus, demonstrating high genetic diversity among the isolates.
A total of 43 Salmonella enterica isolates belonging to different serovars (Salmonella Albany, Salmonella Agona, Salmonella Corvallis, Salmonella Stanley, Salmonella Typhimurium, Salmonella Mikawasima, and Salmonella Bovismorbificans) were isolated from catfish (Clarias gariepinus) and tilapia (Tilapia mossambica) obtained from nine wet markets and eight ponds in Penang, Malaysia. Thirteen, 19, and 11 isolates were isolated from 9 of 32 catfish, 14 of 32 tilapia, and 11 of 44 water samples, respectively. Fish reared in ponds were fed chicken offal, spoiled eggs, and commercial fish feed. The genetic relatedness of these Salmonella isolates was determined by random amplified polymorphic DNA PCR (RAPD-PCR) using primer OPC2, repetitive extragenic palindromic PCR (REP-PCR), and pulsed-field gel electrophoresis (PFGE). Composite analysis of the RAPD-PCR, REP-PCR, and PFGE results showed that the Salmonella serovars could be differentiated into six clusters and 15 singletons. RAPD-PCR differentiated the Salmonella isolates into 11 clusters and 10 singletons, while REP-PCR differentiated them into 4 clusters and 1 singleton. PFGE differentiated the Salmonella isolates into seven clusters and seven singletons. The close genetic relationship of Salmonella isolates from catfish or tilapia obtained from different ponds, irrespective of the type of feed given, may be caused by several factors, such as the quality of the water, density of fish, and size of ponds.
A total of 106 beef samples which consisted of local (n = 59) and imported (n = 47) beef and 180 milk samples from cows (n = 86) and goats (n = 94) were collected from Selangor, Malaysia. Overall, 30.2% (32 of 106) of beef samples were found positive for Arcobacter species. Imported beef was significantly more contaminated (46.80%) than local beef (16.9%). Arcobacter butzleri was the species isolated most frequently from imported (81.8%) and local (60%) beef, followed by Arcobacter cryaerophilus in local (33.3%) and imported (18.2%) beef samples. Only one local beef sample (10%) yielded Arcobacter skirrowii. Arcobacter species were detected from cow's milk (5.8%), with A. butzleri as the dominant species (60%), followed by A. cryaerophilus (40%), whereas none of the goat's milk samples were found positive for Arcobacter. This is the first report of the detection of Arcobacter in milk and beef in Malaysia.
The use of simple crude water extracts of common herbs to reduce bacterial attachment may be a cost-effective way to control bacterial foodborne pathogens, particularly in developing countries. The ability of water extracts of three common Malaysian herbs (Andrographis paniculata, Eurycoma longifolia, and Garcinia atroviridis) to modulate hydrophobicity and attachment to surfaces of five food-related bacterial strains (Bacillus cereus ATCC 14576, Escherichia coli ATCC 25922, Pseudomonas aeruginosa ATCC 10145, Salmonella Enteritidis ATCC 13076, Staphylococcus aureus ATCC 25923) were determined. The bacterial attachment to hydrocarbon assay was used to determine bacterial hydrophobicity. Staining and direct microscopic counts were used to determine attachment of bacteria to glass and stainless steel. Plating on selective media was used to determine attachment of bacteria to shrimp. All extracts were capable of either significantly ( P < 0.05) increasing or decreasing bacterial surface hydrophobicity, depending on the herb extract and bacteria combination. Bacterial attachment to all surfaces was either significantly (P < 0.05) increased or decreased, depending on the herb extract and bacteria combination. Overall, hydrophobicity did not show a significant correlation (P > 0.05) to bacterial attachment. For specific combinations of bacteria, surface material, and plant extract, significant correlations (R > 0.80) between hydrophobicity and attachment were observed. The highest of these was observed for S. aureus attachment to stainless steel and glass after treatment with the E. longifolia extract (R = 0.99, P < 0.01). The crude water herb extracts in this study were shown to have the potential to modulate specific bacterial and surface interactions and may, with further work, be useful for the simple and practical control of foodborne pathogens.
Seafood samples obtained in seafood markets and supermarkets at 11 sites selected from four states in Malaysia were examined for the presence of nine potentially pathogenic species from the genus Vibrio between July 1998 and June 1999. We examined 768 sample sets that included shrimp, squid, crab, cockles, and mussels. We extensively examined shrimp samples from Selangor State to determine seasonal variation of Vibrio populations. Eight potentially pathogenic Vibrio species were detected, with overall incidence in the samples at 4.6% for V. cholerae, 4.7% for V. parahaemolyticus, 6.0% for V. vulnificus, 11% for V. alginolyticus, 9.9% for V. metschnikovii, 1.3% for V. mimicus, 13% for V. damsela, 7.6% for V. fluvialis, and 52% for a combined population of all of the above. As many as eight Vibrio species were detected in shrimp and only four in squid and peel mussels. The overall percent incidence of any of the eight vibrios was highest (82%) in cockles (Anadara granosa) among the seafoods examined and was highest (100%) in Kuching, Sarawak State, and lowest (25%) in Penang, Pulau Penang State, among the sampling sites. Of 97 strains of V. cholerae isolated, one strain belonged to the O1 serotype and 14 to the O139 serotype. The results indicate that the various seafood markets in Malaysia are contaminated with potentially pathogenic Vibrio species regardless of the season and suggest that there is a need for adequate consumer protection measures.
Twenty-two strains of vancomycin-resistant Enterococcus faecalis were isolated from 9 (6%) of 150 samples of frozen beef and beef products imported to Malaysia. The isolates were obtained from eight samples of beef and one sample of minced beef patty. No E. faecalis was isolated from frankfurters. Twelve of the 22 isolates (54.5%) were beta-hemolytic, and all isolates harbored the vanA gene. All vancomycin-resistant isolates were also resistant to streptomycin, erythromycin, kanamycin, bacitracin, ceftazimide, gentamycin, tetracycline, nalidixic acid, and teicoplanin; 95.4% were resistant to trimethoprimsulfamethoxazole; 68.8% were resistant to chloramphenicol; and 41% were resistant to ampicillin and penicillin. Small plasmids ranging in size from 1.5 to 5.8 kb were detected in 8 (36.4%) of 22 strains. The 22 isolates were classified into 20 random amplified polymorphic DNA types. Isolates were divided into two groups, each containing subclusters, that may reflect their clonal lineages. It is concluded that several clones of vancomycin-resistant E. faecalis are represented in the isolates obtained from beef imported to Malaysia.
While there is good epidemiological evidence for foods as vehicles for norovirus transmission, the precise means of spread and its control remain unknown. The feline calicivirus was used as a surrogate for noroviruses to study infectious virus transfer between hands and selected types of foods and environmental surfaces. Assessment of the potential of selected topicals in interrupting such virus transfer was also made. Ten microliters of inoculum of feline calicivirus deposited onto each fingerpad of adult subjects was allowed to air dry and the contaminated area on individual fingerpads was pressed (10 s at a pressure of 0.2 to 0.4 kg/cm2) onto 1-cm-diameter disks of ham, lettuce, or brushed stainless steel. The virus remaining on the donor and that transferred to the recipient surfaces was eluted and plaque assayed. Virus transfer to clean hands from experimentally contaminated disks of ham, lettuce, and stainless steel was also tested. Nearly 46 +/- 20.3, 18 +/- 5.7, and 13 +/- 3.6% of infectious virus was transferred from contaminated fingerpads to ham, lettuce, and metal disks, respectively. In contrast, approximately 6 +/- 1.8, 14 +/- 3.5, and 7 +/- 1.9% virus transfer occurred, respectively, from ham, lettuce, and metal disks to hands. One-way analysis of variance test showed that pretreatment (washing) of the fingerpads either with water or with both topical agent and water significantly (P < 0.05) reduced virus transfer to < or = 0.9%, as compared with < or = 2.3 and < or = 3.4% transfer following treatments with either 75% (vol/vol) ethanol or a commercial hand gel containing 62% ethanol, respectively. Despite wide variations in virus transfer among the targeted items used, intervention agents tested reduced virus transfer significantly (P < 0.05) when compared with that without such treatments (71 +/- 8.9%). These findings should help in a better assessment of the potential for cross-contamination of foods during handling and also assist in developing more effective approaches to foodborne spread of norovirus infections.
The objective of this study was to determine the level of preservatives and microbiological loads in various brands of commercially available chili bo (paste). Fifteen different brands of chili bo obtained from the local market and hypermarkets were analyzed for pH, moisture and benzoic acid content, microbiological loads (aerobic, anaerobic, aerobic spores, and fungi), and thermophilic microorganisms. Results showed that both moisture content and pH vary among samples. The concentrations of benzoic acid detected in chili bo were found to be in the range of 537 to 5,435 mg/kg. Nine of fifteen brands were found to exceed the maximum level permitted by the Malaysian Food Law in accordance with the Codex Alimentarius (1,000 mg/kg for benzoic acid). An apparent correlation between benzoic acid concentration and microbiological loads present in the chili bo was observed. The microbiological loads were found to be relatively low in the end products containing high amounts of benzoic acid. The heat-resistant (70 to 80 degrees C) microorganisms present in chili bo were identified as Ochrobacterum tritici, Stenotrophomonas rhizophila, Microbacterium maritypicum, Roseomonas spp., CDC group II-E subgroup A, Flavimonas oryzihabitans, and Pseudomonas aeruginosa, with M. maritypicum being the most frequently found (in 9 of 15 samples) microorganism. Most of these identified microorganisms were not known to cause foodborne illnesses.
This study examined whether the survival of Vibrio cholerae O1 on contaminated cooked rice was influenced by the type of rice. Vibrios survived unchanged on clumps of glutinous white rice (wet, grains adhered) held at room temperature for 24 h. On nonglutinous white rice (slightly moist, grains separate), 30% viable vibrios remained at 24 h. On nonglutinous brown rice (moist, separate, covered with a mucus-like substance), the number of vibrios increased 2.7-fold at 24 h. Survival rates of vibrios on the surfaces of a row of five cooked rice grains after 2 h of exposure at room temperature were 86, 29, 12, and 4% for glutinous rice, white rice, and the endosperm and pericarp of brown rice, respectively. (Each boiled brown rice grain surface was partly pericarp and partly endosperm, which became exposed by a rupture of the pericarp.) Covering each inoculated grain with a similar cooked rice grain surface increased the corresponding figures to 93, 99, 60, and 94%. Scanning electron microscopy revealed that each type of cooked grain surface possessed a distinct microtopography. For example, the surfaces of glutinous rice grains consisted of separated overlapping strips with many holes, while the pericarps of brown rice were flat interspersed with small pits. In conclusion, each type of boiled rice produced a distinct survival pattern of V. cholerae O1 caused by both the distinct gross features and the fine surface characteristics of the rice. The significance of this finding is that the type of rice consumed can be a factor in cholera transmission by contaminated rice.
Survival of rotavirus in fresh fruit juices of papaya (Caraca papaya L.), honeydew melon (Cucumis melo L.), and pineapple (Ananas comosus [L.] Merr.) was studied. Clarified juices were prepared from pulps of ripe fruits and sterilized by ultrafiltration. One milliliter of juice from each fruit was inoculated with 20 microl of 1 x 10(6) PFU of SA11 rotavirus and sampled immediately (0-h exposure) and 1 and 3 h later at 28 degrees C. Mean viral titers in juices of papaya (pH 5.1) and honeydew melon (pH 6.3) at 1 and 3 h were not significantly different from titers at 0-h exposure. Mean viral titers in juices from pineapples with ripening color indices of 3 (pH 3.6) and 6 (pH 3.7) at 1-h exposure (color index 3: 4.0 +/- 1.7 x 10(4); color index 6: 2.3 +/- 0.3 x 10(5)) and 3-h exposure (color index 3: 1.1 +/- 0.4 x 10(4); color index 6:1.3 +/- 0.6 x 10(5)) were significantly lower than titers at 0-h exposure (color index 3: 5.7 +/- 2.9 x 10(5); color index 6: 7.4 +/- 1.3 x 10(5)). Virus titers in pineapple juices of color index 3 were significantly lower than titers of the virus in juices of index 6. In cell culture medium (pH 7.4), SA11 titer remained stable over 3 h at 28 degrees C. However, at pH 3.6, the virus titer was reduced to a level not significantly different from that of the virus in pineapple juice of color index 6 (pH 3.7). In conclusion, papaya and honeydew melon juices, in contrast to pineapple juice, have the potential to transmit rotavirus. Inactivation of SA11 virus in pineapple juice can be possibly attributed to low pH and constituent(s) in the juice.
Campylobacter jejuni was found to occur at high prevalence in the raw salad vegetables examined. Previous reports describe cross-contamination involving meat; here we investigated the occurrence of cross-contamination and decontamination events in the domestic kitchen via C. jejuni-contaminated vegetables during salad preparation. This is the first report concerning quantitative cross-contamination and decontamination involving naturally contaminated produce. The study was designed to simulate the real preparation of salad in a household kitchen, starting with washing the vegetables in tap water, then cutting the vegetables on a cutting board, followed by slicing cucumber and blanching (heating in hot water) the vegetables in 85 degrees C water. Vegetables naturally contaminated with C. jejuni were used throughout the simulation to attain realistic quantitative data. The mean of the percent transfer rates for C. jejuni from vegetable to wash water was 30.1 to 38.2%; from wash water to cucumber, it was 26.3 to 47.2%; from vegetables to cutting board, it was 1.6 to 10.3%; and from cutting board to cucumber, it was 22.6 to 73.3%. The data suggest the wash water and plastic cutting board as potential risk factors in C. jejuni transmission to consumers. Washing of the vegetables with tap water caused a 0.4-log reduction of C. jejuni attached to the vegetables (most probable number/gram), while rapid blanching reduced the number of C. jejuni organisms to an undetectable level.
Oxalic acid was evaluated as a treatment for reducing populations of naturally occurring microorganisms on raw chicken. Raw chicken breasts were dipped in solutions of oxalic acid (0, 0.5, 1.0, 1.5, and 2.0%, wt/vol) for 10, 20, and 30 min, individually packed in oxygen-permeable polyethylene bags, and stored at 4 degrees C. Total plate counts of aerobic bacteria and populations of Pseudomonas spp. and Enterobacteriaceae on breasts were determined before treatment and after storage for 1, 3, 7, 10, and 14 days. The pH and Hunter L, a, and b values of the breast surface were measured. Total plate counts were ca. 1.5 and 4.0 log CFU/g higher on untreated chicken breasts after storage for 7 and 14 days, respectively, than on breasts treated with 0.5% oxalic acid, regardless of dip time. Differences in counts on chicken breasts treated with water and 1.0 to 2.0% of oxalic acid were greater. Populations of Pseudomonas spp. on chicken breasts treated with 0.5 to 2.0% oxalic acid and stored at 4 degrees C for 1 day were less than 2 log CFU/g (detection limit), compared with 5.14 log CFU/g on untreated breasts. Pseudomonas grew on chicken breasts treated with 0.5% oxalic acid to reach counts not exceeding 3.88 log CFU/g after storage for 14 days. Counts on untreated chicken exceeded 8.83 log CFU/g at 14 days. Treatment with oxalic acid caused similar reductions in Enterobacteriaceae counts. Kocuria rhizophila was the predominant bacterium isolated from treated chicken. Other common bacteria included Escherichia coli and Empedobacter brevis. Treatment with oxalic acid caused a slight darkening in color (decreased Hunter L value), retention of redness (increased Hunter a value), and increase in yellowness (increased Hunter b value). Oxalic acid has potential for use as a sanitizer to reduce populations of spoilage microorganisms naturally occurring on raw chicken, thereby extending chicken shelf life.
To determine the occurrence of Salmonella and Shigella in infant formula from Southeast Asia, 74 packages of dehydrated powdered infant follow-on formula (recommended age, > 4 months) from five different manufacturers, four from Indonesia and one from Malaysia, were analyzed. None of the 25-g test portions yielded Salmonella or Shigella. However, further identification of colonies growing on selective media used for Salmonella and Shigella detection revealed the frequent occurrence of several other Enterobacteriaceae species. A total of 35 samples (47%) were positive for Enterobacteriaceae. Ten samples (13.5%) from two Indonesian manufacturers yielded Enterobacter sakazakii. Other Enterobacteriaceae isolated included Pantoea spp. (n = 12), Escherichia hermanii (n = 10), Enterobacter cloacae (n = 8), Klebsiella pneumoniae subsp. pneumoniae (n = 3), Citrobacter spp. (n = 2), Serratia spp. (n = 2), and Escherichia coli (n = 2). To our knowledge, this is the first report to describe the contamination of dehydrated powdered infant formula from Indonesia with E. sakazakii and several other Enterobacteriaceae that could be opportunistic pathogens. Improper preparation and conservation of these products could result in a health risk for infants in Indonesia.
Experiments were performed to determine the thermal resistance of hepatitis A virus (HAV) in three types of dairy products containing increased amounts of fat content (skim milk, homogenized milk; 3.5% MFG, and table cream; 18% MFG). HAV-inoculated dairy products were introduced into custom-made U-shaped microcapillary tubes that in turn were simultaneously immersed in a waterbath, using custom-made floating boats and a carrying platform. Following exposure to the desired time and temperature combinations, the contents of each of the tubes was retrieved and was tested by plaque assay to determine the reduction in virus titer. Our data indicated that < 0.5 min at 85 degrees C was sufficient to cause a 5-log reduction in HAV titer in all three dairy products, whereas at 80 degrees C, < or = 0.68 min (for skim and homogenized milk), and 1.24 min (for cream) were needed to cause a similar log reduction. Using a nonlinear two-phase negative exponential model (two-compartment model) to analyze the data, it was found that at temperatures of 65, 67, 69, 71, and 75 degrees C, significantly (P < 0.05) higher exposure times were needed to achieve a 1-log reduction in virus titer in cream, as compared to skim and homogenized milk. For example, at 71 degrees C, a significantly (P < 0.05) higher exposure time of 0.52 min (for cream) was needed as compared to < or = 0.18 min (for skim and homogenized milk) to achieve a 1-log reduction in virus titer. A similar trend of inactivation was observed at 73 and 75 degrees C where significantly (P < 0.05) higher exposure times of 0.29 to 0.36 min for cream were needed to cause a 1-log reduction in HAV in cream, as compared to < or = 0.17 min for skim and homogenized milk. This study has provided information on the heat resistance of HAV in skim milk, homogenized milk, and table cream and demonstrated that an increase in fat content appears to play a protective role and contributes to the heat stability of HAV.
The antioxidant and microbial stabilities of galangal (Alpinia galanga) extract in raw minced beef were examined at 4 +/- 1 degree C. Raw minced beef containing galangal extracts (0 to 0.10%, wt/wt) were prepared. Lipid oxidation during refrigerated storage was assessed by monitoring malonaldehyde formation, using the thiobarbituric acid reactive substances method. In minced beef, added galangal extract improved oxidative stability. Galangal extract at higher concentrations of 0.05% and 0.10% (wt/wt) were also found to extend the shelf-life of minced beef. Addition of alpha-tocopherol (0.02%, wt/wt) to galangal extract (0.05%, wt/wt) were observed to increase the oxidative but not the microbial stability of minced beef during the storage of 7 days. Galangal extract may prove useful in inhibiting lipid oxidation and increasing microbial stability of minced meat.
Thermophilic Campylobacter and Salmonella enterica are major causes of gastrointestinal foodborne infection. Survival of these pathogens on food-associated surfaces is a risk contributing to their spread through the food system. This study examined the transfer of two strains each of C. jejuni , C. coli , Salmonella Enteritidis, and Salmonella Typhimurium from chicken meat to a knife or scissors used on either a plastic or wooden cutting board. Each strain of Campylobacter and Salmonella at ∼10(8) CFU mL(-1) was inoculated (5 mL) onto 25 g of chicken meat with skin and allowed to attach (for 10 min). The meat was then cut (20 times per implement) into 1-cm(2) pieces with either a knife or scissors on either a plastic or wooden cutting board. The numbers of pathogens transferred from meat onto cutting implements and cutting board surfaces were enumerated. The surfaces were subsequently either rinsed with water or rinsed with water and wiped with a kitchen towel to mimic commonly used superficial cleaning practices for these implements, and the numbers of pathogens were enumerated again. The bacterial numbers for both pathogens were determined on thin-layer agar. The attachment of the Salmonella strains to chicken meat (∼7.0 to 7.8 log CFU cm(-2)) was higher than the attachment of the Campylobacter strains (∼4.6 to 6.6 log CFU cm(-2)). All four Salmonella strains transferred in higher numbers (∼1.9 to 6.3 log CFU cm(-2)) to all surfaces than did the Campylobacter strains (∼1.1 to 3.9 log CFU cm(-2)). The transfer rates of both pathogens from the chicken meat to all the surfaces examined varied substantially between ∼0 and 21.1%. The highest rate of transfer (∼21.1%) observed was for C. coli 2875 when transferred from the chicken meat to the scissors. Most cleaning treatments reduced the numbers of both pathogens (∼0.3 to 4.1 log CFU cm(-2)) transferred to all the surfaces. Our study gives insights into the risks associated with the transfer of Campylobacter and Salmonella from poultry to the surfaces used in poultry preparation.
Asymptomatic Salmonella carriers who work as food handlers pose food safety and public health risks, particularly during food preparation, and this has serious implications for the disease burden in society. Therefore, we conducted a study to determine the number of Salmonella carriers in a migrant cohort in several food establishments in three major cities in Peninsular Malaysia. Sociodemographic data and stool samples were collected and analyzed using standard methods of detection and isolation. Antimicrobial susceptibility tests of the positive samples were also performed. A total of 317 migrant food handlers, originating from South and Southeast Asian countries, were recruited voluntarily. Nine (2.8%) stool samples were confirmed to be Salmonella positive. PCR serotyping and pulsed-field gel electrophoresis identified four serotypes as Typhimurium (n = 3), Corvallis (n = 2), Hadar (n = 1), Agona (n = 1) and two unknown serovars. Antimicrobial susceptibility tests revealed that all nine isolates were susceptible to amoxicillin-clavulanic acid, cefotaxime, ceftazidime, ceftriaxone, and gentamycin. However, seven isolates were found to be multidrug resistant to ampicillin, chloramphenicol, trimethoprim-sulfamethoxazole, sulfonamides, streptomycin, and tetracycline. This study highlights that carriers of nontyphoidal Salmonella exist among migrant food handlers, which poses a health risk to consumers through food contamination. Our results indicate a need for authorities to enhance food safety awareness in the migrant workers and to reevaluate current health screening methods to include preventive measure such as mandatory stool screening as part of the preemployment and routine health examinations.