Recently, flies were recognized as potential vectors for AMR in hospital and non-hospital environments (7, 8). In the study center, the proportion of ESBL-producing GNB isolated from clinical samples was very high (Figure. 1). We conducted this study in order to compare the colonization of flies with ESBL-producing GNB at various locations inside and one location with high density of flies outside of the hospital compound. We found a high proportion of ESBL-producing bacteria among isolates from flies collected inside the hospital compared to the near absence of ESBL genes in bacterial isolates from flies collected 1.5 km away from the hospital. The ESBL proportion was highest at the NICU and at the orthopedic ward, and only slightly lower at the hospital waste disposal area (Table 1). Our findings could partly be explained by exposition of bacteria to different antibiotics in the environment of the hospital or more likely by the accumulation of resistant bacteria in and on flies in the patients’ environment. Similar to our findings, a study conducted in Iran shows that bacterial isolates from houseflies in a hospital compound had a significantly higher frequencies of antimicrobial resistance against various antibiotics, than bacteria isolated from houseflies in non-hospital environment (7). A study conducted in Berlin, Germany showed that the prevalence of ESBL in flies trapped from two different residential areas differed (0% vs. 18%) (9). According to this study, the distribution of ESBL-producing bacteria among flies in certain geographical locations is not uniform.
In our study, the proportion of ESBL-producing GNB was very high among common pathogenic bacteria like Escherichia coli, Klebsiella spp., Enterobacter spp., Citrobacter spp., and Raoultella spp. compared with opportunistic bacteria. Kluyvera spp. are opportunistic bacteria with the highest rate of ESBL-production (Table 2). Half of the bacteria isolated from the butchery in Asella town were not commonly pathogenic bacteria. This different distribution of bacteria colonizing the isolated flies might also influence the proportion of ESBL isolates from different sites based on pathogenicity of the bacteria and their exposure to cephalosporin antibiotics (10).
In this study, the most frequently detected resistance genes in confirmed ESBL-producing GNB colonized flies were CTX-M-1-like gene, followed by TEM-like gene and SHV-like gene, respectively. The frequency and characterization of ESBL genes of clinical samples and flies isolates showed similarities (Table 3) (11) and also similar findings reported by Boulesteix G. et al. in 2005 from Dakar, Senegal (12). This suggests that flies may acquire the bacteria from the hospital environment. Similar findings were reported by Fotedar R. et al. in 1992 (13); however, to clearly identify the source of the ESBL-producing bacteria on flies needs further investigation. On the basis of our results, no statement can be made on the question of whether flies can be considered as vectors for MDR bacteria. In this context, however, it is interesting to note that Rahuma N. et al (2005) reported earlier that flies may be potential vectors for the transmission of MDR bacteria from hospitals to surrounding communities.(14).flies colonized with ESBL-producing bacteria found in the hospital compound can possibly spread AMR to surrounding residential areas or restaurants, thereby endangering public health (15). As described in (Table 2), not only well-known pathogenic bacteria but also opportunistic bacteria can carry clinically relevant resistance genes and enhance the spread of AMR in the community.
Published investigations from Ethiopia demonstrate that MDR GNB commonly express the blaCTX-M-1 gene encoded in ESBLs and the blaNDM-1 gene in carbapenemase-producing bacteria (5, 16-19). For molecular detection of ESBL, blaCTX-M-1 ESBL gene can be used as target gene by either conventional PCR or a loop-mediated isothermal amplification (LAMP) technique, which is rapid, effective and affordable to detect the presence of blaCTX-M-1 ESBL gene in RLS like Ethiopia (20).
In this study, the susceptibility to non-β-lactam antibiotics such as ciprofloxacin, gentamicin, and trimethoprim-sulfamethoxazole was significantly lower in ESBL-producing bacteria compared to ESBL-negative bacteria (p<0.001). This finding is an indicator for the limited options of appropriate antibiotic therapy regimen for ESBL-producing bacterial infection management. Poudel et al. (2019) also reported high rates of resistance of E. coli and K. pneumoniae isolated from flies against tetracycline and ampicillin due to emergence of ESBL-producing strains (21). This might probably be caused by plasmid-mediated mobile resistance genes such as quinolone-resistance (qnr) genes, aminoglycoside acetyltransferase (aac), dfr (trimethoprim resistance) and sul (sulfamethoxazole resistance) genes, being more frequent in ESBL-producing bacteria compared to ESBL-negative bacteria (22-25). However, the identification of other resistance genes than the described ESBL genes was not part of this investigation.
In tropical regions where poor hospital hygiene is common, hand hygiene and patient isolation or implementation of antimicrobial stewardship programs may not be sufficient to control the expansion of AMR. Our findings can be considered as indicators for a possible dissemination of antimicrobial resistance inside and outside of hospital compounds and to the nearest environments by flies. Therefore, to tackle the expansion of ESBL-producing bacteria, fly-control measures in critical areas of the hospitals might be essential (7, 15). In order to inhibit further expansions of ESBL-carrying bacteria from hospitals to residential areas, environmental and health professionals and municipality administration should work together and strengthen a one health approach. Future AMR prevention and control protocols may consider screening of flies for AMR and eradication-measures to control the population density of flies at health care facilities in tropical regions (21). As distribution of ESBL genes in clinical samples and flies caught in the hospital show comparable results, flies might be used as an indicator organism for ESBL-prevalence in hospital facilities.
Our study has certain limitations. Fly species identification was not performed and we recognize that the different species Musca sorbens and Musca domestica have a different life style and thus may be involved in bacterial transmission to different degrees. Nevertheless, an identification of the different species in this study is unlikely to have an impact on the results in general. ESBL-colonization in the community was reported as low, but flies were sampled from a single butchery only. Even though the colonization with ESBL to be common in flies at the hospital, the source of the ESBL was not addressed. The study design also fails to point out whether the external organs like legs and mouth or the gut of the flies are more involved in carrying ESBL-producing bacteria, a factor with possible impact for the transmission of the bacteria. The role of flies in transmission of nosocomial infections and the source of ESBL-producing bacteria in the hospital needs further investigation.
A further limitation is the use of non-selective media for the screening process in flies and the fact that we tested phenotypically different isolates obtained from one fly (in some case we found more than one bacteria from a single fly). This procedure was chosen in order to harmonize the study protocol with diagnostics in the clinical setting. In consequence, our data reflect the proportion of ESBL among GNB isolates from flies, but not the prevalence of ESBL-carriage among flies, which might have been higher, if a selective screening approach was used. However, even with the non-selective culture technique used, ESBL-producing bacteria were detected in a major proportion of flies.