This study aimed at detecting the presence of antibiotic-resistant Gram-negatives in samples of meals delivered at the University General Hospital of Palermo, Italy. Antibiotic resistant Gram negatives were isolated in July—September 2007 ffrom cold dishes and food contact surfaces and utensils. Bacterial strains were submitted to susceptibility test and subtyped by random amplification of polymorphic DNA (RAPD). Forty-six of 55 (83.6%) food samples and 14 of 17 (82.3%) environmental swabs were culture positive for Gram negative bacilli resistant to at least one group of antibacterial drugs. A total of 134 antibiotic resistant strains, 51 fermenters and 83 non-fermenters, were recovered. Fermenters and non-fermenters showed frequencies as high as 97.8% of resistance to two or more groups of antibiotics and non fermenters were 28.9% resistant to more than three groups. Molecular typing detected 34 different profiles among the fermenters and 68 among the non-fermenters. Antibiotic resistance was very common among both fermenters and non-fermenters. However, the wide heterogeneity of RAPD patterns seems to support a prominent role of cross-contamination rather than a clonal expansion of a few resistant isolates. A contribution of commensal Gram negatives colonizing foods to a common bacterial resistance pool should not been overlooked.
Resistance to various antibacterial drugs is rapidly emerging and posing a major challenge to Public Health. A comprehensive understanding of the most important dissemination routes of antimicrobial resistant bacterial (ARB) strains and resistance encoding genetic sequences is crucial to effectively control and minimize the problem [
Food appears to be an effective source for the acquisition by humans of drug resistant bacteria and drug resistance genes, but the extension and the actual consequences of this exposure are still insufficiently investigated [
Such transfers will likely be most successful when the host is simultaneously submitted to a selective pressure by an antimicrobial substance to which the involved organisms are resistant [
Hospital food service systems are considered one of the most critical segment of hospitality industry, where the client base is often a vulnerable group, and a strict and systematic monitoring of foodborne hazards has to be consistently applied [
This study aimed at detecting the presence of multidrug resistant Gram-negatives (MDR-GN) in food samples delivered through a plated service at the University general hospital “Azienda Ospedaliero-Universitaria Policlinico” (AOUP), Palermo, Italy. Food contact surfaces of equipment and utensils at the Hospital caterer food premise, where food was prepared, portioned and placed into the personalized trays were also sampled during an inspective visit. Patterns of antibacterial drug susceptibility and genetic heterogeneity of the MDR-GN were assessed.
This investigation was conducted at the University general hospital “Azienda Ospedaliero-Universitaria Policlinico” (AOUP), Palermo, Italy, in the period July–September 2007. At this hospital, the food service was contracted out to an external caterer who was employing a traditional cook and serve production scheme and a plated meal distribution system. In particular, food was being ordered according to the patient’s choice up to 24 hours in advance. At the caterer plant, meals were prepared, assembled, and then plated using a conveyor belt with food handlers standing either side and serving appropriate portions into plates. Hot dishes were then placed into trays that were being in turn stacked into preheated cabinets, with cold dishes being placed into separate compartments, before transport and delivery to the hospital wards.
For the purpose of the study, cold dishes only that had not been submitted to thermal treatment were selected. Samples of 50 g approximately were daily collected at receipt in the hospital wards from lunch meals, by taking them from a preordered tray similar to that of a patient. Samples were immediately placed in a refrigerated container and kept at
The inspection and the environmental sampling were carried out on September 2007. For food contact surfaces and utensils of the caterer premise, a swab sampling technique was used. The tip of each sterile cotton swab was moistened with sterile saline, pH 7.0, and then rolled repeatedly over each 10 cmq surface area. After the sampling, the swabs were placed aseptically into 10 mL of peptone water and transferred to the laboratory at chilled temperature. After the swabs were delivered to the laboratory, each tube containing the swab was vortexed 10 seconds to assure mixture of the sample and then incubated for 24 hours at
One MacConkey agar plate was inoculated with 0.2 mL of the 24-hour enrichment culture to obtain a continuous lawn after overnight incubation in ambient air at
After incubation, plates were examined, and all colonies of different morphology growing into each antibiotic inhibition halo were Gram stained and subcultured for purity. All isolates were submitted to a biochemical screening by testing for oxidase and catalase activity and glucose fermentation and classified as Gram negative fermenters or non-fermenters. A complete biochemical identification by the system API 20E or API 20NE (bioMérieux, Marcy-l’Etoile, France) was deserved to some clustered isolates and to two extended-spectrum
Susceptibility of each isolate to a panel of nine antimicrobial substances was assessed by disk diffusion on Mueller-Hinton agar plates, according to the Clinical and Laboratory Standards Institute (CLSI) guidelines [
ESBL production was detected by decreased susceptibility or resistance to third-generation cephalosporins and the synergy between disks containing cefotaxime, ceftazidime, cefepime, and aztreonam and a disk containing amoxicillin-clavulanic acid [
For the purpose of the study, a Gram negative organism resistant to at least two different groups of antimicrobial agents (amoxicillin-clavulanic acid, cephalosporins, aminoglycosides, quinolones, tetracycline) was defined as MDR-GN.
Single colonies growing on solid media were removed with a sterile plastic tip and resuspended in 100
RAPD was performed as previously described [
The primers used in this study were ERIC2 (
Amplified PCR products were separated using 2% agarose gels, visualized by UV transillumination and photographed. A 1 Kbp or a 100 bp DNA ladder (Promega) was used as molecular weight. DNA fingerprints were compared by visual inspection and considered unique when they differed by at least one band, irrespectively of band intensity.
Data were analyzed by the EpiInfo software (version 6.0, CDC, Atlanta, GA, US). Frequency analysis was performed with the chi-square test. Cross tabulation and chi-square or Fisher exact tests were performed to determine the relationship between resistance and some characteristics of isolates and food products processed for isolation of resistant Gram negatives. For some statistical analysis, susceptibility to antibiotics was categorized as a dichotomous variable by interpreting intermediate susceptibility as resistance. In all analyses, differences were considered statistically significant at
During the period July–September 2007, 55 food samples and 17 swabs from food contact surfaces and utensils were examined. Forty-six of 55 (83.6%) food samples and 14 of 17 (82.3%) environmental swabs proved to be culture positive for Gram negative bacilli resistant to at least one group of antibacterial drugs.
A total of 115 different isolates of Gram negative bacilli resistant to at least one antimicrobial group were identified from the following food products: 22 from 18 samples of soft cheese, three from three samples of sliced ham, 20 from 12 samples of mixed ham and cheese dishes, and 70 from 24 samples of vegetables.
The 19 resistant environmental isolates were, respectively, six from countertops, a centrifuge, a sink, and a conveyor belt in the washing area of fresh vegetables, nine from cutting boards, trays, and knives in the vegetables processing area, and four from a slicing machine and food contact surfaces in the cured meat working area.
The 134 Gram negative isolates were biochemically categorized as 51 fermenters and 83 non-fermenters. Prevalence of resistant Gram negative fermenters versus non-fermenters did not significantly differed in the four groups of food products examined (
Prevalence of the resistant, intermediate, or susceptible phenotype towards the antibacterial drugs tested among all isolates is illustrated in Figure
Frequency of resistance, intermediate susceptibility, and susceptibility to the antibacterial drugs tested in fermenters and nonfermenters.
Na | Cn | |||||||||||||||||
F | NF | F | NF | F | NF | F | NF | F | NF | F | NF | F | NF | F | NF | F | NF | |
S (%) | 29.4 | 7.2 | 96.1 | 78.3 | 94.1 | 20.0 | 94.1 | 8.4 | 0 | 15.7 | 74.5 | 98.8 | 96.1 | 91.6 | 100 | 88.0 | 37.3 | 67.5 |
I (%) | 17.7 | 4.8 | 0.0 | 6.0 | 2.0 | 50.6 | 0.0 | 41.0 | 13.7 | 2.4 | 13.7 | 0.0 | 0.0 | 0.0 | 0.0 | 1.2 | 3.9 | 22.9 |
R (%) | 52.9 | 88.0 | 3.9 | 15.7 | 3.9 | 25.3 | 5.9 | 50.6 | 86.3 | 81.9 | 11.8 | 1.2 | 3.9 | 8.4 | 0.0 | 10.8 | 58.8 | 9.6 |
Prevalence of resistance, intermediate susceptibility, and susceptibility towards the antibacterial drugs tested among 134 Gram negative bacterial strains isolated from foods, food contact surfaces, and utensils.
A total of 29 different resistance patterns, 12 among fermenters and 20 among non-fermenters, respectively, were identified. The patterns and their distribution are summarized in Table
Resistance patterns to antibacterial drugs tested in fermenters and nonfermenters.
Fermenters | Non-fermenters | |||
Resistance patter | Number of isolates | Resistance patter | Number of isolates | |
Amc Na | 18 | Amc Cro Ctx Na | 19 | |
Amc Na Te | 9 | Amc Cro Ctx Na Te | 17 | |
Na Cip Te | 8 | Amc Ctx Na | 11 | |
Amc Te | 4 | Amc Caz Cro Ctx Na | 9 | |
Na Te | 3 | Amc Cro Ctx Cn Net Te | 4 | |
Amc Na Cip Te | 2 | Amc Na | 4 | |
Na Cn | 2 | Caz Cro Ctx | 3 | |
Amc | 1 | Amc Cro Ctx | 2 | |
Amc Caz Cro Ctx | 1 | Amc Cro Ctx Na | 2 | |
Amc Caz Cro Ctx Te | 1 | Amc Ctx Na Te | 2 | |
Cro Ctx Na Te | 1 | Amc Caz Cro Ctx Cn Net | 1 | |
Na Cip | 1 | Amc Caz Cro Ctx Na Net Te | 1 | |
Amc Caz Cro Ctx Net | 1 | |||
Amc Cro Ctx Te | 1 | |||
Amc Cro Na | 1 | |||
Amc Na Net | 1 | |||
Amc Na Te | 1 | |||
Caz Cn Net | 1 | |||
Caz Ctx | 1 | |||
Cro Ctx Na Te | 1 |
Figures
(a) Percent distribution among the 134 resistant Gram negative bacterial strains of resistance (including intermediate susceptibility) to at least two, at least three, and more than three antibacterial drugs. (b) Comparison between frequency of resistance to at least two, at least three, and more than three antibacterial drugs in Gram negative fermenters and nonfermenters.
Two fermentative isolates that proved to be ESBL producing by the modified double-disk synergy test were identified as
Molecular typing of the resistant Gram negative bacilli by RAPD with the ERIC2 primer detected 34 different profiles among the 51 fermenters and 68 among the 83 non-fermenters (Figures
ERIC-2 RAPD patterns of representative Gram negative fermenters (a) and non-fermenters (b). MW = molecular weight. (a) 100 bp; (b) 1 Kbp.
All clusters were confirmed by using the M13 primer. The two ESBL positive
Clustered isolates were significantly (
Clustering was not significantly associated to any food product (
Except for the two larger ones, the remaining clusters included isolates recovered from food samples in an interval of time ranging from 0, when their isolation was made from different food products sampled in the same date, to 10 days. The 10 isolates of
The role of food within the overall framework of human exposure to drug resistant bacteria has been until now insufficiently investigated from a Public Health perspective. Many studies have focused on the contribute of antimicrobial resistance to the severity of the hazard posed by foodborne pathogens and the specific measures to be adopted along the food chain to minimize the current trend of some pathogenic bacteria, such as
To contribute additional information about food-mediated exposure to Gram negative ARB, we investigated the occurrence of these organisms in food products processed in a catering premise and delivered to hospitalized patients. Indeed, a considerable portion of these subjects belongs to a population subgroup of consumers where colonization by resistant bacteria and selective pressure due to use of antibacterial drugs may interact within a supportive environment and generate more severe health risks [
A proportion as high as 83.6% of food samples tested positive for Gram negative resistant to one or more groups of antibiotics, with a great heterogeneity of resistance patterns and RAPD patterns among both fermenters and nonfermenters. Moreover, both Gram negative groups showed frequencies as high as 97.8% of resistance to two or more groups of antibiotics. Nonfermenters, in particular, were 28.9% resistant to more than three groups. Markedly lower resistance prevalences have been previously described in
Of particular concern appears the high prevalence of resistance to nalidixic acid and, to a less extent, cefotaxime. Literature suggests that resistance to nalidixic acid determined by the disk diffusion method may be a reliable indicator of decreased susceptibility to ciprofloxacin [
A further finding that deserves consideration is the large heterogeneity of RAPD patterns, that excludes clonal expansion as a possible reason of the high prevalence of antibiotic resistances in foods. Moreover, identification of some clusters of fermenter and non-fermenter isolates in intervals of time ranging between 0 and 50 days proves the persistence in the food processing environment of resistant organisms and, consequently, the potential effectiveness of Good Hygienic Practices in minimizing their diffusion.
Our study has some limits. Firstly, sampling has been carried out in a single food catering premise. Consequently, the results could have been heavily influenced by the hygienic conditions of the plant, and their generalizability could be questionable. Furthermore, the issue of location and horizontal transferability of resistance genetic determinants have not been addressed. Intrinsic resistances to some antibiotics, for example, presence of AmpC mediated
Based on our results, a contribution of commensal ARB Gram negatives colonizing processed foods consumed without further thermal treatment or processing to a common resistance pool should not been overlooked. This is a disturbing finding when considering the possible impact of a daily administration of resistant bacteria to a susceptible population, particularly those with defective immune systems or comorbidities and those receiving antibiotic treatment. Understanding the routes connecting the resistant bacteria and genetic resistance determinants to humans, including the role of food vehicles, is critical to define effective strategies to control this problem. The consistent and effective application of good food hygiene practices is a key issue in the prevention and control of food contamination with antimicrobial-resistant pathogenic and commensal bacteria.