Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Experimental cross-contamination of chicken salad with Salmonella enterica serovars Typhimurium and London during food preparation in Cambodian households

  • Rortana Chea ,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Validation, Visualization, Writing – original draft, Writing – review & editing

    rortana.chea@slu.se (RC); s.dang@cgiar.org (SDX)

    Affiliations National Animal Health and Production Research Institute, General Directorate of Animal Health and Production, Phnom Penh, Cambodia, International Livestock Research Institute, Nairobi, Kenya, Department of Biomedical Sciences and Veterinary Public Health, Swedish University of Agricultural Sciences, Uppsala, Sweden

  • Hung Nguyen-Viet,

    Roles Conceptualization, Funding acquisition, Methodology, Project administration, Supervision, Writing – review & editing

    Affiliation International Livestock Research Institute, Nairobi, Kenya

  • Sothyra Tum,

    Roles Conceptualization, Funding acquisition, Investigation, Supervision, Writing – review & editing

    Affiliation National Animal Health and Production Research Institute, General Directorate of Animal Health and Production, Phnom Penh, Cambodia

  • Fred Unger,

    Roles Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Writing – review & editing

    Affiliation International Livestock Research Institute, Nairobi, Kenya

  • Johanna Lindahl,

    Roles Conceptualization, Data curation, Formal analysis, Methodology, Supervision, Writing – review & editing

    Affiliations International Livestock Research Institute, Nairobi, Kenya, Department of Clinical Sciences, Swedish University of Agricultural Sciences, Uppsala, Sweden, Department of Medical Biochemistry and Microbiology, Uppsala University, Uppsala, Sweden

  • Delia Grace,

    Roles Conceptualization, Funding acquisition, Methodology, Project administration, Supervision, Writing – review & editing

    Affiliations International Livestock Research Institute, Nairobi, Kenya, Natural Resources Institute, University of Greenwich, Kent, United Kingdom

  • Chhay Ty,

    Roles Conceptualization, Investigation, Methodology

    Affiliation Livestock Development for Community Livelihood Organization, Phnom Penh, Cambodia

  • Sok Koam,

    Roles Investigation, Methodology

    Affiliation National Animal Health and Production Research Institute, General Directorate of Animal Health and Production, Phnom Penh, Cambodia

  • Vor Sina,

    Roles Investigation

    Affiliation Livestock Development for Community Livelihood Organization, Phnom Penh, Cambodia

  • Huy Sokchea,

    Roles Conceptualization

    Affiliation Livestock Development for Community Livelihood Organization, Phnom Penh, Cambodia

  • Son Pov,

    Roles Investigation

    Affiliation Livestock Development for Community Livelihood Organization, Phnom Penh, Cambodia

  • Theng Heng,

    Roles Investigation

    Affiliation National Animal Health and Production Research Institute, General Directorate of Animal Health and Production, Phnom Penh, Cambodia

  • Or Phirum,

    Roles Investigation

    Affiliation National Animal Health and Production Research Institute, General Directorate of Animal Health and Production, Phnom Penh, Cambodia

  • Sinh Dang-Xuan

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing – review & editing

    rortana.chea@slu.se (RC); s.dang@cgiar.org (SDX)

    Affiliations International Livestock Research Institute, Nairobi, Kenya, Center for Public Health and Ecosystem Research, Hanoi University of Public Health, Hanoi, Vietnam

Abstract

Non-typhoidal Salmonellae are common foodborne pathogens that can cause gastroenteritis and other illnesses in people. This is the first study to assess the transfer of Salmonella enterica from raw chicken carcasses to ready-to-eat chicken salad in Cambodia. Twelve focus group discussions in four Cambodian provinces collected information on typical household ways of preparing salad. The results informed four laboratory experiments that mimicked household practices, using chicken carcasses inoculated with Salmonella. We developed four scenarios encompassing the range of practices, varying by order of washing (chicken or vegetables first) and change of chopping utensils (same utensils or different). Even though raw carcasses were washed twice, Salmonella was isolated from 32 out of 36 chicken samples (88.9%, 95% CI: 73.0–96.4) and two out of 18 vegetable samples (11.1%, 95% CI: 1.9–36.1). Salmonella was detected on cutting boards (66.7%), knives (50.0%) and hands (22.2%) after one wash; cross-contamination was significantly higher on cutting boards than on knives or hands (p-value < 0.05). The ready-to-eat chicken salad was contaminated in scenario 1 (wash vegetables first, use same utensils), 2 (wash vegetables first, use different utensils) and 3 (wash chicken first, use same utensils) but not 4 (wash chicken first, use different utensils) (77.8%, 11.1%, 22.2% and 0%, respectively). There was significantly higher Salmonella cross-contamination in scenario 1 (wash vegetables first, use same utensils) than in the other three scenarios. These results show how different hygiene practices influence the risk of pathogens contaminating chicken salad. This information could decrease the risk of foodborne disease in Cambodia and provides inputs to a quantitative risk assessment model.

Introduction

Food safety is a major concern worldwide [1]. Animal-source food (ASF) provides essential nutrients but is a common source of pathogens. The World Health Organization estimated that, in 2010, more than 600 million illnesses were caused by 31 common foodborne hazards: most were diarrhoeal and caused by zoonoses [2]. Foodborne disease (FBD) affects humans of all age groups, but children under five years are among the most vulnerable. According to the Foodborne Disease Burden Epidemiology Reference Group (FERG), one in 10 children worldwide suffer from FBD annually [3]. FBD also decreases human capital, entails prevention and treatment costs, and hinders trade [4]; economic losses are estimated at more than USD 100 billion a year across developing countries [1]. In Cambodia, over 5,000 people fell sick from FBD in 371 outbreaks in 2019 [5], but this is a huge underestimate as there is inadequate surveillance of FBD [6, 7].

Salmonella spp. is one of the most important causes of FBD and is often associated with ASF consumption [8]. Non-typhoidal Salmonella (NTS) was estimated to cause 59,000 deaths out of 420,000 annual deaths globally from foodborne hazards in 2010 [3].

Chicken is popular in Cambodia; it is affordable, widely available and provides protein and micronutrients essential for growth and health [9, 10]. In 2020, total meat consumption was 301,000 tons per year, of which poultry meat was 62,000 tons [11]. Poultry meat consumption is expected to increase by 5.5% annually due to increased demand driven by population growth, urbanization and increasing incomes [12]. Cambodians prefer local, backyard chicken, but the modern industrial sector, which uses exotic chicken, has higher productivity and is growing rapidly [13]. High and rising consumption of chicken meat is a health concern because it is a common source of Salmonella and other pathogens [14, 15]. Salmonella can persist in the chicken intestinal tract without causing clinical signs and is not detected by meat inspection. Therefore, Salmonella can contaminate carcasses at the slaughterhouse, especially if facilities and hygiene are poor: for example, if floors are dirty, or if the same tank filled with dirty water is used for all washing steps. In addition, transport and sale under humid tropical conditions may further facilitate Salmonella contamination and growth [1619]. A recent nationwide survey in Cambodian markets reported that 42.6% of chicken meat samples were contaminated with Salmonella, with an average most probable number (MPN) of 10.6 MPN/g [20]; previous studies on retailed chicken meat samples also showed high Salmonella prevalences which ranged from 20–60% [2124].

Chicken is widely used in Cambodian cuisine, including traditional salad, consisting of boiled chicken mixed with raw vegetables, herbs, spices and banana flowers [25]. Chicken salad is commonly prepared in households and restaurants and for wedding banquets. Salad is prone to contamination because it includes raw ingredients, is served cold, and its preparation requires equipment and several handling steps [16]. A common cause of bacterial cross-contamination of ready-to-eat (RTE) foods is poor hygiene practices allowing pathogens from raw meat to contaminate hands, knives and cutting boards and hence be transferred to other food [17, 25, 26]. Unfortunately, food hygiene often receives insufficient attention since Cambodian consumers are more concerned about chemicals than microbes [27]. Yet, considering the high Salmonella prevalence in chicken meat, consumers could be at risk from chicken salad. To quantify the risk of foodborne illness from eating chicken salad, information on microbial load and cross-contamination in this typical dish is needed. However, the levels and mechanisms of contamination of Cambodian salad with Salmonella have not been investigated. Therefore, this study aimed to investigate how Salmonella from raw chicken could cross-contaminate RTE chicken salad, given the usual practices in Cambodian households. These experimental results can model exposure assessment steps in quantitative microbial risk assessment of Salmonella in chicken salad and recommend reducing cross-contamination while preparing food.

Materials and methods

Household survey on hygiene practices when cooking chicken salad

In 2020, 12 focus group discussions (FGD) with consumers were held in four provinces (Siem Reap, Preah Sihanouk, Battambang and Phnom Penh) in Cambodia. In each province, three areas representing different geographical and social contexts, covering rural, peri-urban and urban areas, were purposely selected based on information from local authorities. In each area, individuals who were mainly in charge of purchasing and preparing food for their families, from six to eight households, were invited by the communal authorities to participate in the FGD. A total of 93 participants discussed food safety practices during purchasing, storing and washing of chicken carcasses and vegetables, use and cleaning of kitchen utensils, and hand washing when preparing the traditional chicken salad (‘ngam sach man sroyong chek’). After gaining written consent from the participants, the FGD was facilitated by a senior researcher while another researcher took notes and recorded the discussion. The FGD was conducted in the Khmer language and lasted approximately 1.5 to 2 hours. Transcripts were then translated into English for quantitative analysis.

Cooking chicken salad scenarios

Cambodian chicken salad is a mix of boiled chicken meat with sliced banana flowers, cucumber, tomato, lemon juice, basil, fresh chili, herbs, and spices (hereafter, these plant ingredients are referred to as ‘vegetables’). Fresh vegetables are washed and cut into small pieces. A whole chicken carcass bought from the market is washed, cut into pieces, boiled, deboned, then torn and sliced into shreds. Boiling lasts around 20 minutes, and the time for total preparation is one to two hours.

Four experimental scenarios were designed to imitate the process of preparing the chicken salad in the household as reported in FGDs (that is, washing vegetables either before or after washing the chicken carcass and using the same knife and cutting board (utensils) for chicken and vegetables or different utensils). All four experimental scenarios were carried out on the same day and repeated nine times per scenario. The main steps of cooking chicken salad in each scenario are described in Table 1 and Fig 1.

thumbnail
Fig 1. The scenario diagram of the preparation and practice steps of cooking chicken salad, including the sampling points.

Red dots indicate the sampling types and stages collected during the experiment of each scenario: 1washing with clean, Salmonella-free water; 2washing with clean, Salmonella-free water, dishwashing detergent and clean dishcloth.

https://doi.org/10.1371/journal.pone.0270425.g001

thumbnail
Table 1. Chicken salad preparation steps in each experimental scenario and the number of samples collected and analyzed.

https://doi.org/10.1371/journal.pone.0270425.t001

Scenario 1.

Wash vegetables first and use the same utensils (WVF-SU). The process was washing (twice) and chopping vegetables; washing and cutting the raw chicken carcass; washing the cutting board, knife and hands (once with dish detergent); boiling the chicken and using the same (washed) knife, cutting board and hands to debone, tear and slice the boiled chicken and, finally, mixing the salad.

Scenario 2.

Wash vegetables first and use different utensils (WVF-DU). The process was washing (twice) and cutting vegetables; washing and cutting raw chicken carcass; washing the cutting board, knife and hands (once with dish detergent); boiling the chicken; using a separate knife and cutting board, and washing hands to debone, tear and slice boiled chicken and mixing the salad.

Scenario 3.

Wash chicken first and use the same utensils (WCF-SU). The process was washing (twice) and cutting raw chicken carcass; washing and chopping vegetables; washing the cutting board, knife and hands (once with dish detergent); boiling the chicken; using the same (washed) knife, cutting board and hands to debone, tear and slice boiled chicken, and mixing the salad.

Scenario 4.

Wash chicken first and use different utensils (WCF-DU). The process was washing (twice) and cutting the raw chicken carcass; washing and chopping vegetables; washing the cutting board, knife and hands (once with dish detergent); boiling the chicken; using a separate knife and cutting board, and washing hands to debone, tear and slice boiled chicken, and mixing the salad.

Chicken carcass preparation and inoculation

Preparation of the chicken carcass.

Four whole chicken carcasses (approximately 1.2 ± 0.2 kg each) were purchased from a shop that had tested negative for Salmonella in a recent Cambodian market survey [20]. The butcher was supervised to ensure hygienic processing, including thorough washing and disinfection of hands and equipment (knives and buckets). After cleaning, each chicken carcass was washed thoroughly twice with clean and Salmonella-free bottled drinking water (Vital, Phnom Penh, Cambodia) to minimize bacterial contamination. Immediately after that, the washed carcasses were placed into separate sampling bags. The packed carcasses were placed in a cool box and transported immediately to the laboratory to start the experiment within 1 hour.

Preparation of Salmonella inoculum.

According to recent studies in Cambodia and Vietnam, Salmonella Typhimurium and Salmonella London were commonly found in retailed ASF, including chicken meat and pork [24, 28]. This study utilized two Salmonella strains (S. Typhimurium and S. London) isolated from Vietnam [16] to prepare an inoculation culture. Both strains were cultured separately in a 100 mL glass bottle (Schott Duran, Mainz, Germany) consisting of 50 mL Buffered Peptone Water (BPW; Merck, Darmstadt, Germany) for 12–14 h at 37°C with gentle agitation. Next, the Salmonella load in each strain culture was enumerated using a spread plate count method on Xylose-Lysine Deoxycholate Agar (XLD; Hi-Media, Mumbai, India). Based on the determined concentration in each initial cultured strain, a final suspension of 104 Salmonella CFU/mL medium was made of appropriate mixed-volume proportions of the two strains. The final Salmonella suspension was then used to inoculate the chicken carcass.

Inoculation of chicken carcass.

Immediately after preparing the Salmonella medium (concentration of 104 CFU/mL), each chicken carcass was weighed (in grams) and inoculated with a corresponding volume (in microlitres) of Salmonella medium; for instance, a 1,200-gram carcass was inoculated with 1,200 μL Salmonella medium. Salmonella medium was dropped over the entire chicken carcass surface using 10–200 μL tips and pipette (Corning, NY, US), which created a 10 CFU Salmonella per one gram of chicken. After the inoculation, carcasses were kept at room temperature for 30 min for stable absorption, following the methodology of previous experiments [16, 29].

Washing of vegetables, chicken carcasses, hands and equipment, and preparation of vegetables and chicken carcasses.

Vegetables comprised banana flowers, lemon, fresh chilli, cucumber, tomato and basil bought from a shop in the early morning. Before using, the researchers washed them twice with clean water containing 1% sodium chloride and immersed them in saline water for 30 min to minimize contamination and then cut them into small slices using a washed cutting board, knife and hands.

The inoculated chicken carcass was washed twice in a basin using clean, Salmonella-free water (approximately 5 L per chicken) with clean, bare hands. The washed carcass was put on the cutting board and cut into parts (neck-head, two wings, two drumsticks, two thighs and two breasts). The chicken parts were then boiled in a pot with 2–2.5 L of water for 20 min, after which a spoon was used to remove and place them on a sterile plate to cool down for approximately 40–45 min. The boiled chicken was then deboned, sliced and torn into small pieces using clean bare hands, knives and a cutting board. The cutting board, knife and hands used to cut the fresh carcass were washed once in a different basin with clean water using dishwashing detergent (Sunlight, Unilever, Vietnam) and dishcloth (Sunlight, Unilever, Vietnam) for about 3–4 min as described in an earlier study [16]. The cutting board and knife were kept at ambient temperature to dry at least 15 min before the next step. Vegetables were prepared and processed according to four pre-defined scenarios (Table 1).

Sampling.

In each scenario, just before chopping for boiling, the chicken carcass was sampled using a 5 x 5 cm stainless steel frame and a pre-moistened gauze to swab the breast surface. In scenarios 3 and 4, after washing and cutting, but before mixing with meat, vegetables were sampled by taking approximately 50 g of the mixed vegetables with sterile forceps. In scenario 1 (WVF-SU) and 3 (WCF-SU), just before slicing the boiled chicken, a set of three surface swabs and pre-moistened gauze samples were collected from the cutting board (25 cm2 in the centre), knife (both sides of the blade, 25 cm2 each) and two hands (palms, fingers, interdigital folds). In scenarios 2 (WCF-SU) and 4 (WCF-DU), the surface of two hands (palms, fingers and interdigital folds) was swabbed just before slicing boiled chicken. After finishing the last step (mixing salad) in each scenario, samples of RTE chicken salad were collected by taking approximately 50 g each of chicken meat and vegetables. The number of samples taken in each step by scenarios and Salmonella analyses is presented in Table 1.

Microbiological test.

Following the ISO procedure, all samples (n = 162, Table 1) underwent Salmonella isolation [30]. A sample of 25 g chicken salad or vegetables or swab was homogenized with appropriate BPW volume and incubated for 18±2 h at 37°C as a pre-enrichment step. Selective enrichment step was done by adding 1 mL of suspension into 9 mL Muller Kauffmann Tetrathionate (MKTT; Merck, Germany) incubated for 24 ± 3 h at 37°C and 0.1 mL into 10 mL Rappaport-Vassiliadis Soya (RVS; Merck, Germany) incubated for 24 ± 3 h at 41.5°C. Selective plating was performed by streaking one loopful (approximately 10 μL) each of MKTT and RVS onto XLD agar. MacConkey (Merck, Darmstadt, Germany) was the second selective plating agar. Two presumptive Salmonella colonies per plate were selected to test biochemically (lactose, indole, lysine and hydrogen sulfide) for Salmonella confirmation.

A 3-tube MPN method was used to quantify Salmonella, as described previously [31]. In brief, the sample was diluted first at 10−1 by adding 25 g of chicken salad in 225 mL BPW. Three tubes containing 10 mL of this 10−1 dilution were prepared, after which 1 mL of the 10−1 dilution was added to three tubes containing 9 mL BPW to make the second series of 10−2; then 1 mL of 10−2 was added to the last three tubes containing 9 mL of BPW to make the third series of 10−3. The three-tube set of three consecutive dilutions was incubated at 37°C for 18 ± 2 h. The steps to detect Salmonella in each tube were followed according to the isolation procedures mentioned above. The presence of Salmonella in the three tube-set was used to calculate the MPN index according to the method described earlier [32, 33].

Data analysis and modelling.

Data were entered into Excel spreadsheets (Microsoft, 2016) and analyzed descriptively (proportion, mean, standard deviation). A chi-squared test or Fisher exact test was used to evaluate the Salmonella cross-contamination proportions among sample types and scenarios. R version 3.3.2 (R Core Team, 2020) was used to compute testing and bootstrapping [34]. To describe the distributions of Salmonella concentration in the chicken salad in each scenario, both non-parametric and parametric bootstrapping techniques were used. Salmonella was quantified as MPN/g and thus followed log-normal distribution. A Bayesian statistic was used to assess variability and uncertainty during the simulation of Salmonella load and reduction rates. The parameters and functions used to carry out the bootstrapping, and simulated sample data distributions followed the steps described earlier [16]. The function fitdist() in the fitdistrplus package in R was used to estimate the mean and standard deviation of Salmonella CFU/g [34]. For the presentation of distributions, kernel density was calculated in density() function based on the simulated sample data and plotted using R. We model the reduction of Salmonella CFU/g by using the equation: Reduction rate = (10—Salmonella CFU/g in RTE chicken salad)/10 x 100, where 10 was an inoculated CFU/g in raw chicken carcass from the beginning of the test. The distribution of the reduction rate was calculated using iterations and presented in a histogram. The experiments showed four MPN values (110, 110, 15 and 15 MPN/g in scenario 1) which were higher than the inoculation level. The simulation performed without and with these four values was named Scenario 1 and the worst-case Scenario 1, respectively. The simulation was not carried out for Scenario 4 since no Salmonella positive salad samples were found in this scenario.

Ethical statement.

The experiments were conducted at the National Animal Health and Production Research Institute (Phnom Penh, Cambodia). Participants invited to the focus group discussions were asked for consent before starting (S1 File). All information on the participants was used among the research team only and not shared with any third party. Written consent was obtained from researchers participating in the experiment, including instructions on safety procedures (S2 File). The chicken salad was sterilized and hygienically discarded after finishing the experiment. Ethical approvals of this study were under the Safe Food, Fair Food Cambodia project and granted by the National Ethical Committee of Cambodia (S3 File), No. 300NECHR, dated 26th December 2017, and the International Livestock Research Institute Institutional Research Ethics Committee (S4 File), No. ILRI-RC010 18/IBC/010/CR, dated 5th July 2018.

Results

Hygiene practices when preparing and cooking chicken salad in Cambodian households

Most (86%, 80/93) households reported that they first washed chicken carcasses two to three times with water before washing and preparing vegetables; only 14% (13/93) washed and prepared vegetables before washing chicken carcasses. All participants washed knives and cutting boards at least once, with soap or dishwashing detergent, immediately after cutting fresh chicken carcasses. However, almost all (97%, 90/93) used the same knife and cutting board to prepare raw vegetables and chicken carcasses, as well as to prepare raw and boiled chicken, while the use of separate knives and cutting boards between raw and cooked chicken was less common (3.2%, 3/93, Table 2).

thumbnail
Table 2. Food safety practices for preparing chicken salad in Cambodian households.

https://doi.org/10.1371/journal.pone.0270425.t002

Salmonella contamination from the raw chicken after washing to vegetables/herbs, hands, cutting board and knife during chicken salad preparation

After washing the chicken carcasses twice, Salmonella was isolated from 32 out of 36 samples (88.9%, 95%CI: 73.0–96.4). Two out of 18 vegetable samples were cross-contaminated with Salmonella (11.1%, 95%CI: 1.9–36.1, Table 3). Eight out of 36 hand swabs (palms, fingers and interdigital folds of two hands) (22.2%, 95%CI: 10.7–39.6), which had been washed once after handling contaminated chicken carcasses, were positive for Salmonella, which was significantly lower compared to 66.7% of washed cutting boards being contaminated (12/18, χ2 = 8.35, df = 1, p-value = 0.004) and lower compared to contamination on washed knives (50.0%, 9/18, χ2 = 3.10, df = 1, p-value = 0.07, Table 3). Salmonella from chicken carcasses was most often transferred to the cutting boards, followed by knives and hands, even though hands and equipment were washed once with clean water and dishwashing detergent. From the simulated data, Salmonella cross-contamination to cutting boards was significantly higher than to knives and hands and was higher to knives than to hands (p-value < 0.001); average contamination on hands, knives and cutting boards was 23.7% (95%CI: 5.1–28.3), 50.1% (95%CI: 42.5–57.8) and 65.2% (95%CI: 58.3–72.6), respectively (Table 3, Fig 2).

thumbnail
Fig 2. Simulated probability distribution of Salmonella cross-contamination from raw chicken to hands (a), cutting board (b) and knife (c) after washing once after washing and cutting the fresh chicken carcass.

https://doi.org/10.1371/journal.pone.0270425.g002

thumbnail
Table 3. Salmonella contamination from the raw chicken after washing twice and vegetables, hands, knives and cutting boards during chicken salad preparation.

https://doi.org/10.1371/journal.pone.0270425.t003

Cross-contamination of Salmonella to ready-to-eat chicken salad

In Scenario 1 (WVF-SU), seven out of nine (77.8%) chicken salads were positive with Salmonella, while the number of salad samples positive with Salmonella in Scenarios 2 (WFV-DU) and 3 (WCF-SU) was one out of nine (11.1%) and two out of nine (22.2%), respectively. Scenario 4 (WCF-DU) showed no positive Salmonella in salad in all nine experiments. Scenarios 1 and 3 used the same knife, cutting board, and hands for handling salad, while in Scenarios 2 and 4, a separate knife and cutting board were used. There was a significantly higher cross-contamination rate in Scenario 1 compared to the other three scenarios (Fisher exact test, p-value = 0.05). The average Salmonella contamination in the salad was highest in Scenario 1 (37.3 MPN/g) and 0.36 MPN/g in both Scenarios 2 and 3. The overall proportion of Salmonella contamination in all four scenarios was 27.8% (95%CI: 14.8–45.4, Table 4).

thumbnail
Table 4. Contamination of Salmonella in ready-to-eat chicken salad during preparation and handling in four different experiment scenarios.

https://doi.org/10.1371/journal.pone.0270425.t004

Based on simulated data (5000 iterations), the average Salmonella contamination in salad was higher in Scenario 1 (8.58 CFU/g) (including 4 values that exceeded the initial value of 10 CFU/g) with a mean of 77.78 CFU/g. In contrast, the average Salmonella contamination in Scenarios 2 and 3 was 0.8 and 0.78 CFU/g, respectively (Table 5 and Fig 3). (Salmonella was absent in Scenario 4).

thumbnail
Fig 3. Salmonella concentration (CFU/g) on contaminated RTE chicken salad in Scenario 1* (a), Scenario 1** (b), Scenario 2 (c) and Scenario 3 (d) based on the experiment values simulated 5000 times.

* Scenario 1 had only values below the initial concentration (10 CFU/g); ** worst-case Scenario 1 included four MPN/g values (two 110 CFU/g and two 15 CFU/g), which exceeded the initial concentration (10 CFU/g).

https://doi.org/10.1371/journal.pone.0270425.g003

thumbnail
Table 5. Salmonella (CFU/g) concentration was simulated in each scenario in the ready-to-eat chicken salad, based on the experiment values with 5000 iterations.

https://doi.org/10.1371/journal.pone.0270425.t005

The reduction rate of Salmonella contamination from raw chicken to ready-to-eat chicken salad

Most Salmonella was transmitted from raw chicken to salad in Scenario 1 (WVF-SU). In this scenario, four out of seven positive samples had higher levels of Salmonella after processing. On the other hand, in Scenarios 2 (WVF-DU) and 3 (WCF-SU) Salmonella was reduced by 92%. The proportion of the simulated values that exceeded initial CFU/g in Scenarios 1, 2 and 3 were 53.8, 18.6, 3.8 and 3.6, respectively (Table 6 and Fig 4).

thumbnail
Fig 4. The reduction rate of Salmonella concentration (CFU/g) in the contaminated RTE chicken salad in Scenario 1* (a), worst-case Scenario 1** (b), Scenario 2 (c) and Scenario 3 (d) based on the experiment values to bootstrap 5000 times.

* Scenario 1 had only values below the initial concentration (10 CFU/g); ** Worst-case Scenario 1 included four MPN/g values (two 110 CFU/g and two 15 CFU/g), which exceeded the initial concentration (10 CFU/g).

https://doi.org/10.1371/journal.pone.0270425.g004

thumbnail
Table 6. Simulated reduction rate (percentage) of Salmonella concentration (CFU/g) in the RTE chicken salad in each scenario based on the experiment values simulated 5000 times.

https://doi.org/10.1371/journal.pone.0270425.t006

Discussion

This study examined how food handling might affect the risk of cross-contamination by Salmonella in households under different preparation scenarios. The results indicate interventions to reduce risk. In addition, findings can be used to model exposure assessment steps in conducting a quantitative microbial risk assessment of Salmonella in chicken salad. In Cambodia, several foodborne outbreaks have been associated with contamination during food preparation [5, 35]. The FGDs of the 93 households described the common practices that were used to develop scenarios for the experiment. Chicken and pork salad are typically consumed in households, restaurants or ceremonies in Cambodia [3638], Southeast Asia and Middle Eastern countries [39]. FBD cases have been associated with these [35, 36]. Salmonella in raw chicken carcasses sold in the market is a source of contamination, especially when using the same hands, knife, or cutting board without adequate washing [40]. Several factors could contribute to the current high prevalence of Salmonella in retailed chicken and pork in Cambodian markets (exceeding 40%) [20]. These include reused or unsafely used water for cleaning (for example, washing intestines in the same basin as carcasses), absence of appropriate storage facilities for food, low frequency of cleaning and disinfection of the shop, etc. [6, 14, 41]. Despite washing the chicken carcasses twice in water, this study found that 88.9% (95% CI: 73.0–96.4) of experimental carcasses still harboured Salmonella, similar to a finding in an experiment of washing contaminated pork in Vietnam, but the washing steps significantly reduced the number of Salmonella bacteria by up to 92.2% (Table 5) [16, 42, 43]. In this study, handling and preparing raw meat (including washing and cutting) increased Salmonella contamination of hands, cutting boards, knives and vegetables. Washing of food products before preparation is often observed in many low- and middle -income countries [39] and was also a common practice in the households interviewed in the present study.

In high-income countries, washing raw meat before cooking is usually not recommended because the meat is often contaminated with bacteria, and washing raw meat can spread pathogens around the kitchen. On the other hand, in developing countries, it is often believed that washing meat before cooking removes dirt, slime and blood and makes it safer. In Cambodia, only a few supermarkets or minimarts in urban areas provide packed and cooled or chilled meat and the level of hazards present in these is not well established. Indeed, studies have found that contamination levels are not always lower in modern retail shops in developing countries [20, 43]. Nevertheless, separating equipment used for raw and RTE foods is strongly recommended in all contexts. Cleaning and disinfection of hands, knives and cutting board surfaces right after contact with raw meat/chicken can minimize contamination with bacteria [43].

The current study detected Salmonella cross-contamination from chicken meat to RTE Cambodian chicken salad, with 27.8% of all salads being contaminated. This is the first study to quantify cross-contamination of bacteria during simulated home preparation of ASF products in Cambodia. Studies in other countries on cross-contamination from pork [16] and chicken [29] found a similar trend of Salmonella cross-contamination to RTE food. Salmonella from ASF can contaminate hands, equipment and containers and cross-contaminate RTE chicken salad during preparation, consequently causing foodborne illness in consumers. Several reports show that FBDs, especially in developing countries, are often underreported, and there is a lack of food safety surveillance and traceability systems [1, 3, 57, 44]. This implies there are more foodborne illness cases, including salmonellosis, in Cambodia than the number officially reported by the health authority (5000 cases of FBD in 2019) [5, 35].

Scenarios 1 (WVF-SU) and 3 (WCF-SU), using the same utensils for chopping chicken and vegetables and assumed to be the scenarios with the poorest hygiene practices, had the highest proportion of Salmonella contamination of salad (77.8% and 22.2%, respectively) and the highest quantity of Salmonella (37.3 MPN/g, Scenario 1). About 90% of the surveyed households practiced these sub-optimal procedures for preparing chicken salad in their homes. Other studies have also found that unwashed knives, cutting boards and hands increased the risk of cross-contamination in frequency and CFU/g [16, 41, 45]. For example, a Cambodian study of a foodborne outbreak revealed that the unhygienic practices could have led to cross-contaminated food [36]. Cross-contamination of bacteria to RTE food was also reported in studies in China, Indonesia, Malaysia, Thailand and Vietnam [17, 46], where daily food preparation practices were examined. In a similar experiment comparing Salmonella contamination when using the same or different utensils to prepare boiled pork in Vietnam, the Salmonella prevalence varied from 22.2–77.8%, and average Salmonella concentrations were from 0.12 to 5.79 CFU/g in cooked pork [16].

This study shows that washing did not eliminate Salmonella from the chicken. In addition, washing vegetables after washing the raw chicken carcasses resulted in the transfer of Salmonella to the vegetables via hands or equipment, or both. However, we did not expect that levels of Salmonella in the chicken salad in scenario 1 (vegetables washed first) would be worse than in scenario 3 (chicken washed first). In addition, in Scenarios 3 and 4 of our experiment, two out of eighteen washed and prepared vegetables were positive with Salmonella (Table 3) would explain the less involvement of hands and sink in transferring Salmonella to vegetables during washing contaminated chicken first. Practices of washing chicken carcass before washing vegetables were reported in 86% of interviewed households (Table 2); however, actual of using the sink, basket, hands, or contact with rinsed water in washing vegetables can be different by households thus, the level of cross-contamination to vegetables could be higher than in our experiment. Washing meat could decrease bacterial contamination; nonetheless, the varied practices also result in different levels of bacteria remaining in food [42]. Cutting boards have been reported to contribute the most to the cross-contamination of bacteria from raw ASF to other food [28]. A study in China using Campylobacter spp. as an indicator of in-home cooking procedures also found that cutting boards were an important source of cross-contamination [47].

Scenarios 2 (WVF-DU) and 4 (WCF-DU) were more hygienic as a separate knife and cutting board were used to prepare raw and cooked chicken. A lower proportion of Salmonella-contaminated chicken salads was observed in scenarios 2 and 4 (11.1% and 0.0%, respectively) than in scenarios 1 (WVF-SU) and 3 (WCF-SU). However, there was only one Salmonella-positive vegetable sample in scenario 3 and one in scenario 4 (WCF-DU) when vegetables were washed and chopped after the chicken. It is generally agreed that washing vegetables before handling and preparing raw meat would significantly reduce the risk of cross-contamination [42, 43]. Our study did not find any evidence of this, with scenario 1 having the highest contamination despite washing vegetables first and scenario 4, with chicken, washed first, resulting in no contamination of the chicken salad.

Information on Salmonella prevalence, load and reduction rate in each hygiene practice scenario, will be helpful in exposure assessment; a step often inadequately addressed in risk assessment [48]. This is the first study investigating cross-contamination by Salmonella in Cambodian households when preparing salad. The findings show different ways that bacteria can contaminate RTE food and may be generalized to other types of salad prepared by similar procedures and used to assess the risk of cross-contamination in other types of raw meat and seafood. Furthermore, the results can be used to design and disseminate more targeted, evidence-based food safety practice messages, such as the need to use a separate cutting board and knife for raw and RTE food and adequately clean and disinfect hands and equipment surfaces after contact with raw meat.

This study had some limitations. The experiment used raw chicken and ‘vegetables’ (cucumber, tomato, basil, fresh chili and banana flower) purchased from hygienic slaughterhouses and shops; however, these might not always have been Salmonella-free during all nine experimental days. In addition, during the experiment, the variation in contact time, pressure and moisture and the nature (surface) of meat, vegetables and equipment between replications might have affected Salmonella (cross) contamination [4951]. Future studies should also assess the risk of washing instead of not washing chicken carcasses before cooking.

Conclusions

Our finding that the median Salmonella load in the chicken salad was between 0.36 and 12.41 CFU/g raises health concerns. We described the usual practices of preparing chicken salad in Cambodia and examined how these could lead to Salmonella being transferred from chicken carcasses to salad, identifying risky practices including the use of the same cutting board and knife for meat and vegetables and inadequate handwashing. Different salad preparation practices result in very different contamination levels, with washing vegetables before the chicken and using the same utensils for chopping chicken and vegetables resulting in higher levels of salad contamination. Risk communication messages should focus on the need for separate kitchen utensils and frequent and adequate washing and disinfecting of food contact surfaces (cutting board, knife, hands). The finding that washing chicken carcass before the vegetables resulted in less contamination was not expected and requires further investigation. Data on Salmonella levels under different preparation scenarios will be used to support quantitative microbial risk assessments through eating salad.

Supporting information

S1 File. Consent form for participant in research: Household.

https://doi.org/10.1371/journal.pone.0270425.s001

(PDF)

S2 File. Consent form for participant in research: Laboratory research team.

https://doi.org/10.1371/journal.pone.0270425.s002

(PDF)

S3 File. Ethical approvals of this study granted by the National Ethical Committee of Cambodia.

https://doi.org/10.1371/journal.pone.0270425.s003

(PDF)

S4 File. Biosafety check by International Livestock Research Institute Institutional Research Ethics Committee.

https://doi.org/10.1371/journal.pone.0270425.s004

(PDF)

Acknowledgments

The authors thank the research team at the Laboratory of Bacteriology and Mycology at the National Animal Health and Production Research Institute, General Directorate of Animal Health and Production, Ministry of Agriculture Forestry and Fisheries of Cambodia, Livestock Development for Community Livelihood Organization Cambodia, as well as participants from the provincial animal health offices in Siem Reap, Preah Sihanouk, Battambang and Phnom Penh provinces for supporting this study. We thank Tezira Lore and Ainsley Smith (International Livestock Research Institute) for editing the paper. We also thank Sofia Boqvist and Kistina Obsjer from the Swedish University of Agriculture Sciences for her kind guidance. The authors also sincerely thank field and laboratory work facilitators, including Khuoch Tet and Roeurn Sophai.

References

  1. 1. Jaffee S, Henson S, Unnevehr L, Grace D, Cassou E. The safe food imperative: Accelerating progress in low-and middle-income countries. Washington DC, USA: World Bank Publications; 2018 https://openknowledge.worldbank.org/handle/10986/30568.
  2. 2. Grace D. Food safety in low and middle-income countries. Int J Environ Res Public Health. 2015;12(9):10490–507. pmid:26343693
  3. 3. Havelaar AH, Kirk MD, Torgerson PR, Gibb HJ, Hald T, Lake RJ, et al. World Health Organization global estimates and regional comparisons of the burden of foodborne disease in 2010. Plos Med. 2015;12(12). pmid:26633896
  4. 4. World Health Organization. The burden of foodborne diseases in the South-East Asia Region. World Health Organization. Regional Office for South-East Asia.; 2016. Report No.: 9290225033.
  5. 5. Ministry of Health Cambodia. Report of the achievement of the health sector in 2019 and the plan for 2020. Ministry of Health, Cambodia; 2020.
  6. 6. Nguyen-Viet H. Food safety control: Improving food safety in markets in Cambodia. Presentation at a webinar on ’The enabling environment for animal-source food market systems: Lessons from the field’. 2020.
  7. 7. Kristina R and Sophal C. Evidence on foodborne diseases in Cambodia. safe food, fair food for Cambodia project annual project meeting 2018; Phnom Penh, Cambodia. 2018. https://cgspace.cgiar.org/bitstream/handle/10568/115418/Report.pdf?sequence=1&isAllowed=y
  8. 8. Crump JA, Wain J. Salmonella. In: Quah SR, editor. International Encyclopedia of Public Health (Second Edition). Oxford, UK: Academic Press; 2017. p. 425–33.
  9. 9. Darapheak C, Takano T, Kizuki M, Nakamura K, Seino K. Consumption of animal source foods and dietary diversity reduce stunting in children in Cambodia. Int Arch Med. 2013;6(1):1–11.
  10. 10. Sary S, Shiwei X, Wen Y, Darith S, Chorn S, editors. Household food consumption in rural, Cambodia is almost ideal demand system analysis. Journal of Physics: Conference Series; 2019: IOP Publishing.
  11. 11. General Directorate of Animal Health and Production. Annual report of animal health and production in Cambodia for 2020 and action plan for 2021. Phnom Penh, Cambodia: The General Directorate of Animal Health and Production, Phnom Penh, Cambodia, 2021.
  12. 12. Ministry of Agriculture Forestry and Fisheries. A strategic planning framework for livestock development: 2016–2025: A future direction for livestock development in Cambodia. Phnom Penh, Cambodia: Ministry of Agriculture Forestry and Fisheries; Phnom Penh, Cambodia 2015.
  13. 13. People in Need (PIN). A value chain analysis of chicken production by Cambodian smallholders: based on an assessment conducted in Pursat and Kampong Chnnang Provinces, Cambodia. Cambodia; 2015.
  14. 14. Nair DV, Johny AK. Salmonella in poultry meat production. Food Safety in Poultry Meat Production: Springer; 2019. p. 1–24.
  15. 15. Eng S-K, Pusparajah P, Ab Mutalib N-S, Ser H-L, Chan K-G, Lee L-H. Salmonella: a review on pathogenesis, epidemiology and antibiotic resistance. Front Life Sci. 2015;8(3):284–93.
  16. 16. Dang-Xuan S, Nguyen-Viet H, Pham-Duc P, Grace D, Unger F, Nguyen-Hai N, et al. Simulating cross-contamination of cooked pork with Salmonella enterica from raw pork through home kitchen preparation in Vietnam. Int J Environ Res Public Health. 2018;15(10):2324. pmid:30360454
  17. 17. Carrasco E, Morales-Rueda A, García-Gimeno RM. Cross-contamination and recontamination by Salmonella in foods: a review. Food Res Inter. 2012;45(2):545–56.
  18. 18. Xiao X, Wang W, Zhang J, Liao M, Rainwater C, Yang H, et al. A quantitative risk assessment model of Salmonella contamination for the yellow-feathered broiler chicken supply chain in China. Food Control. 2021;121.
  19. 19. Dang-Xuan S, Nguyen-Viet H, Unger F, Pham-Duc P, Grace D, Tran-Thi N, et al. Quantitative risk assessment of human salmonellosis in the smallholder pig value chains in urban of Vietnam. Int J Public Health 2017;62(1):93–102. pmid:27837223
  20. 20. Rortana C, Nguyen-Viet H, Tum S, Unger F, Boqvist S, Dang-Xuan S, et al. Prevalence of Salmonella spp. and Staphylococcus aureus in chicken meat and pork from Cambodian markets. Pathogens. 2021;10(5):556. pmid:34064354
  21. 21. Trongjit S, Angkititrakul S, Tuttle RE, Poungseree J, Padungtod P, Chuanchuen R. Prevalence and antimicrobial resistance in Salmonella enterica isolated from broiler chickens, pigs and meat products in Thailand–Cambodia border provinces. Microbiol. Immunol. 2017;61(1):23–33. pmid:28042666
  22. 22. Nadimpalli M, Fabre L, Yith V, Sem N, Gouali M, Delarocque-Astagneau E, et al. CTX-M-55-type ESBL-producing Salmonella enterica are emerging among retail meats in Phnom Penh, Cambodia. J Antimicrob Chemother. 2019;74(2):342–8. pmid:30376113
  23. 23. Lay KS, Vuthy Y, Song P, Phol K, Sarthou JL. Prevalence, numbers and antimicrobial susceptibilities of Salmonella serovars and Campylobacter spp. in retail poultry in Phnom Penh, Cambodia. J Vet Med Sci. 2011;73(3):325–9. pmid:21060246
  24. 24. van Cuyck H, Farbos-Granger A, Leroy P, Yith V, Guillard B, Sarthou JL, et al. MLVA polymorphism of Salmonella enterica subspecies isolated from humans, animals, and food in Cambodia. BMC research notes. 2011;4(1):1–8. pmid:21861934
  25. 25. Saorath C. Chiken & banana flower salad/#20(chay saorath). 2019. https://www.youtube.com/watch?v=JTgxctodp7U
  26. 26. Cliver DO. Cutting boards in Salmonella cross-contamination. Journal of AOAC International. 2006;89(2):538–42. pmid:16640304
  27. 27. Schwan CL. Defining the food safety landscape in Cambodia: the ecology of Salmonella enterica in informal markets 2020. PhD thesis, Kansas State University. https://krex.k-state.edu/dspace/handle/2097/40880
  28. 28. Stéfani T, Dantas ST, Rossi BF, Bonsaglia EC, Castilho IG, Hernandes RT, et al. Cross-contamination and biofilm formation by Salmonella enterica serovar Enteritidis on various cutting boards. Foodborne Pathog Dis. 2018;15(2):81–5. pmid:29053370
  29. 29. Ravishankar S, Zhu L, Jaroni D. Assessing the cross-contamination and transfer rates of Salmonella enterica from chicken to lettuce under different food-handling scenarios. Food Microbio. 2010;27(6):791–4. pmid:20630321
  30. 30. International Standard Organization (ISO). ISO 6579–1:2017: Microbiology of the food chain—Horizontal method for detecting, enumeration and serotyping of Salmonella—Part 1: Detection of Salmonella spp., International Standardization Organization: Geneva, Switzerland, 2007.
  31. 31. Pavic A, Groves P, Bailey G, Cox J. A validated miniaturized MPN method, based on ISO 6579: 2002, for the enumeration of Salmonella from poultry matrices. J Applied Microbio. 2010;109(1):25–34. pmid:20059618
  32. 32. De Man EJoAM. MPN tables, corrected. Eur J Appl Microbiol Biotechnol. 1983;17(5):301–5.
  33. 33. Blodgett R. BAM Appendix 2: most probable number from serial dilutions, Bacteriological Analytical Manual. Food and Drug Administration, Silver Spring, MD https://www.fda.gov/food/foodscienceresearch/laboratorymethods/ucm109656.htm. 2010.
  34. 34. Team Core R. R: a language and environment for statistical computing. 2015;3:2.
  35. 35. Kimsean P, Sreng K, Has P, Ly S, Sim S, Chhay S, et al. An outbreak of gastrointestinal illness associated with Khmer Noodles: A multipronged investigative approach, Kandal Province, Cambodia, June 2014. OSIR Journal. 2017;9(4):1–6.
  36. 36. Vandy S, Leakhann S, Phalmony H, Denny J, Roces MC. Vibrio parahaemolyticus enteritis outbreak following a wedding banquet in a rural village–Kampong Speu, Cambodia, April 2012. Western Pacific surveillance and response journal: WPSAR. 2012;3(4):25. pmid:23908935
  37. 37. Recipe C. 2020. https://www.cambodiarecipe.com/recipe/cambodian-chicken-salad-recipe/.
  38. 38. Baker D. A Taste of Cambodian Cuisine 2009.
  39. 39. Habib I, Harb A, Hansson I, Vågsholm I, Osama W, Adnan S, et al. Challenges and Opportunities towards the Development of Risk Assessment at the Consumer Phase in Developing Countries—The Case of Campylobacter Cross-Contamination during Handling of Raw Chicken in Two Middle Eastern Countries. Pathogens. 2020;9(1):62. pmid:31963109
  40. 40. Kuan CH, Lim LWK, Ting TW, Rukayadi Y, Ahmad SH, Radzi CWJWM, et al. Simulation of decontamination and transmission of Escherichia coli O157:H7, Salmonella Enteritidis, and Listeria monocytogenes during handling of raw vegetables in domestic kitchens. Food Control. 2017;80:395–400.
  41. 41. Aizaabi SE, Khan MA. A study on foodborne bacterial cross-contamination during fresh chicken preparation. Arab Journal of Nutrition and Exercise (AJNE). 2017:128–38.
  42. 42. IFIS-USDA. Washing Food: Does it promote food safety? 2021. https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/washing-food-does-it-promote-food#:~:text=Significantly%20decrease%20your%20risk%20by,washing%20or%20rinsing%20the%20poultry.
  43. 43. Bernstein C, Cates SC, Lavallee A, Shumaker E, Blake C, Brophy J, et al. Food Safety Consumer Research Project: Meal Preparation Experiment on Raw Stuffed Chicken Breasts. 2020.
  44. 44. Paul Ebner J, Lyda Hok. Food Safety in Cambodia: Current Programs and Opportunities. Feed the Future Innovation Lab for Food Safety 2020.
  45. 45. Luber P, Brynestad S, Topsch D, Scherer K, Bartelt E. Quantification of Campylobacter species cross-contamination during handling of contaminated fresh chicken parts in kitchens. Appl. Environ. Microbiol. 2006;72(1):66–70. pmid:16391026
  46. 46. Dewanti-Hariyadi R, Gitapratiwi D. Foodborne Diseases: Prevalence of Foodborne Diseases in South East and Central Asia. 2014. World Health Organization. https://apps.who.int/iris/handle/10665/327655
  47. 47. Bai Y, Lin X-H, Zhu J-H, Cui S-H, Guo L-X, Yan S-F, et al. Quantification of cross-contamination of Campylobacter jejuni during food preparation in a model kitchen in China. J Food Protection. 2021;84(5):850–6. pmid:33232459
  48. 48. Langiano E, Ferrara M, Lanni L, Viscardi V, Abbatecola AM, De Vito E. Food safety at home: knowledge and practices of consumers. J Public Health. 2012;20(1):47–57. pmid:22347771
  49. 49. Mafu AA, Plumety C, Deschenes L, Goulet J. Adhesion of pathogenic bacteria to food contact surfaces: influence of pH of culture. Int J Microbiol. 2011;2011:972494. pmid:20981289
  50. 50. Van Asselt E, De Jong A, De Jonge R, Nauta M. Cross‐contamination in the kitchen: estimation of transfer rates for cutting boards, hands and knives. J Appl Microbiol. 2008;105(5):1392–401. pmid:18713282
  51. 51. Miranda RC, Schaffner DW. Longer contact times increase cross-contamination of Enterobacter aerogenes from surfaces to food. Appl Environ Microbiol. 2016;82(21):6490–6. pmid:27590818