As far as we know the present investigation is the first prevalence report of the isolation of E. coli strains from raw and cocked hospital foods as well as molecular characterization and antimicrobial resistance properties of their STEC strains in the world. We found that 6.72% of hospital food samples were contaminated with E. coli strains and the prevalence of STEC strains were 41.02% which was considerable high. Gram-negative bacteria like E. coli are responsible for 30-70% of hospitalization in developed countries and E. coli was the most common etiologic gram-negative organism in the hospital environment [22, 23].
There were some probable reasons for the high prevalence of E. coli and also STEC strains in the hospital food samples. At first. high-volume food production and long process of catering caused several problems such as lack of adequate precision and accuracy in the preparation and washing of raw materials and even their well cooking, lack of enough time and even temperature for cooking of raw materials, cooling of foods during processing, lack of reaching of sufficient heat to the center of meat and other food materials, lack of enough time to withdraw meat from the frozen state and finally cooking of meat and its products more than the daily requirement and then their storage at improper temperature and conditions. These mentioned circumstances maybe lead to survival and even growth of pathogenic microorganisms in food stuffs. At second, using from unsanitary and also contaminated equipment and dishes for production of foods in hospitals. At third, presence of nurses and servants, which maybe the sources of dangerous pathogenic agents, in the distribution of food to patients. Prevalence of E. coli in raw and cooked fish samples were 1.42% and 2.72%, respectively. Presence of powerful competitor microflora in raw fish and also their destruction and addition of E. coli strains derived from contaminated staffs of the hospital’s kitchen are two main reasons for above findings. Because the large numbers of E. coli isolates recovered from raw meat, proper preparation of the raw meat can eliminate the distribution of bacteria. Unfortunately, basic principles of meat inspections were not observed in Iranian slaughterhouses. Therefore, close contact of animal carcasses with each other and even slaughterhouse floor, blood, content of the digestive tract and wool and skin of slaughtered animal caused transmission and distribution of pathogenic agents like E. coli to meat of slaughtered animals. Besides, the role of possible colonizers such as meat inspectors, butchers and miscellaneous people which mainly have come into the slaughterhouse for buying of meat and finally animals like rats, cats and birds which have been entered from outside the slaughterhouse as a sources of pathogenic E. coli should not be overlooked. Survival of STEC strains of raw food samples even after cooking procedure and occurrence of cross contamination after cooking procedure are two important routes of hospital foods contamination.
We found marked seasonality in distribution of E. coli strains in hospital food samples. High incidence of E. coli strains in summer season (64.10%) could be related to the low levels of individual hygiene in this season. The higher prevalence of STEC strains may be related to higher growth of them in hot seasons of the year. Of studies that have been conducted in this field [24–26], all have shown a seasonal distribution for E. coli with a higher prevalence of strains in warmer months of the year [24–26].
The results of the current study showed that resistant and virulence STEC strains had the high prevalence in the hospital food samples. Limit studies have been conducted in this field. Ifeadike et al., [27] revealed that the prevalence of E. coli strains in food handlers of Nigerian hospitals were 1.8% which was lower than our findings in hospital foods. They concluded that poor and faulty food-handling practices have been identified as the leading cause of the majority of food-borne disease in hospitals. Ha et al., [7] in a research which was conducted in order to study the contamination rate of raw foods in schools, factories, and hospitals of Vietnam reported that the prevalence of E. coli in raw poultry, meat, fish and vegetable samples were 45, 21.3, 6.6 and 18.5%, respectively which all were higher than our results.
STEC strains of our investigation harbored resistance against multiple antibiotics. The results of disk diffusion method obtained from STEC strains of our investigation were also confirmed by the presence of specific antibiotic resistance genes encode resistance against determined antibiotics. Stewardson et al., [5] reported that 92% of all food samples which were collected from hospitals in Switzerland were positive for Extended-Spectrum-Beta-lactamase-producing-Enterobacteriaceae (ESBL-PE). ESBL-producing E. coli was the most commonly detected (44.77%). They showed 6.45% of eligible food handlers were positive for ESBL-PE and prevalence of E. coli strains among ESBL-PE strains were 83.33%. ESBL-PE strains were generally not multidrug resistant, with 100, 90, 87, 79, and 98% of strains susceptible to meropenem, gentamicin, ciprofloxacin, cotrimoxazole, and fosfomycin, respectively. Antibiotic resistance-based finding of Stewardson et al., [5] study was in contrast with our results which showed the high prevalence of resistance and also antibiotic resistance genes. This part of our study was similar with those of India [28] (high prevalence of resistance against erythromycin, cephalothin, amikacin, kanamycin and gentamicin antibiotics), South Africa [29] (high presence of CITM, blaSHV and tetA antibiotic resistance genes), Korea [30] (high prevalence of resistance against ampicillin, tetracycline, streptomycin and amikacin antibiotics) and Mexico [31] (high presence of tet, blaSHV, qnr and aac (3)-IV antibiotic resistance genes and also high prevalence of resistance against ampicillin, trimethoprim-sulfamethoxazole, chloramphenicol and cephalotine antibiotics). Momtaz et al., [20] reported that aac (3)-IV (68.03%), sul1 (82.78%), blaSHV (56.55%), aadA1 (60.65%) and tetA (51.63%) and also resistance against tetracycline (86.88%), penicillin (100%), gentamycin (62.29%) and streptomycin (54.91%) were the most commonly reported antibiotic resistance-based finding of STEC strains of diarrheic patients which was similar to our results.
STEC strains of hospital food samples and especially cooked samples harbored the high levels of resistance against human-based antibiotics such as meropenem, levofloxacillin, mezlocillin, cefipime, polymyxin B, cefotaxime, ciprofloxacin, cotrimoxazole, ceftazidime and imipenem which can indirectly confirm their anthropogenic origin of these strains. The prevalence of antibiotic resistance genes and especially those that were encode resistance against human-based antibiotics were also high among STEC strains recovered from cooked hospital food samples which also can indirectly approved the transmission of anthropogenic STEC strains probably from staffs of hospital kitchens to foods after cooking process. Prevalence of resistance against chloramphenicol antibiotic in the STEC strains of meat and chicken samples of our investigation were 25 and 50%, respectively. There were no positive results in other raw and also all of the cooked food samples. The Iranian Food and Drug Administration (FDA) listed chloramphenicol as a forbidden antibiotic for treatment of diseases of animals and poultries. It is because of prescription of this antibiotic may cause dangerous effects on animals and even humans who use from their edible sources like meat and milk. The presence of high chloramphenicol resistance showed its irregular and unauthorized use in veterinary treatment and especially field of poultry in Iran. Veterinary practitioners of the field of poultry use from this antibiotic as a primary choice for treatment of diseases. Therefore, in a very short period of time, antibiotic resistance appears. Similar results have been reported by Momtaz and Jamshidi [32], Ranjbar et al., [33] and Colello et al., [34]. Momtaz and Jamshidi [32] reported that 73.17% of STEC strains recovered from poultry meat samples were resistant against chloramphenicol which was higher than our findings. Ranjbar et al., [33] found that the prevalence of resistance against chloramphenicol in the STEC strains recovered from various types of food samples were 21.95% which was considerable. Colello et al. [34] showed that majority of STEC strains in the Argentina harboured the gene that encodes resistance against chloramphenicol.
The most commonly detected serogroups in the STEC strains of hospital food samples of our study were O26 and O157. Presence of O157 serogroup showed that some of our strains had animal-based origin. O157 serogroup is predominant in foods with animal origin. STEC strains of raw foods with animal origins like those of meat and chicken were remained even after cooking of foods. Presence of this serogroup. High prevalence of O157 and O26 serogroups have been reported previously from Iran (Hemmatinezhad et al., [35]), Nigeria (Mamza et al., [36]), Canada (Neher et al., [37]) and Turkey (Avaz et al., [38]). Dehkordi et al., [11] reported similar profile for distribution of STEC O-serogroups. They showed that total prevalence of O157, O145, O128, O121, O113, O111, O103, O91, O45 and O26 serogroups in the STEC strains of food stuffs were 26, 6, 8, 4, 6, 6, 6, 4, 4 and 12%, respectively, which was similar to our findings.
Another part of our investigation focused on the distribution of stx1, stx2, eaeA and ehly virulence factors in the E. coli strains of hospital foods. High presence of these factors in EHEC and AEEC subtypes showed their high pathogenicity for human and especially patients. Simultaneous presence of stx1 and eaeA and stx2 and eaeA virulence factors in some strains of E. coli of hospital foods indicated the important public health problem facing Iranian hospitals and health centers which can attributed to occurrence of dangerous food poisoning and STEC-based food-borne diseases in hospitals. Simultaneous presence of stx1 and eaeA and stx2 and eaeA virulence factors have been reported previously [39–41]. In a study which was conducted by Momtaz et al., [42] the prevalence of EHEC and AEEC subtypes in E. coli strains of food samples were 15.06 and 49.31%, respectively. They showed that all of the EHEC strains were positive for all stx1, eaeA and ehly virulence genes, while the prevalence of these genes in AEEC subtypes were 77.77, 13.88 and 55.55%, respectively. Higher prevalence of AEEC subtypes was also reported by various investigations [9–11, 20, 32, 33]. Prevalence of AEEC subtypes in the studies of Momtaz et al., [9] and Dehkordi et al., [11] were 79.41 and 62%, respectively which were higher than EHEC subtype in both studies. Momtaz et al., [10] reported that the prevalence of AEEC subtypes in the E. coli strains recovered from the meat samples of beef, sheep, goat and camel species were 45.54, 44.87, 39.02 and 50.00%, respectively which were higher than those of EHEC. Momtaz et al., [20] reported that the prevalence of AEEC subtypes among the E. coli strains of Iranian diarrheic patients had a range of 47 to 75% which was entirely high.