3.1 Socio-demographic characteristics of apparently healthy humans.
A total of 400 apparently healthy humans were sampled. The mean age was 33.13 years with an age range of < 1 year to > 70 years. Majority of samples were obtained among age group 20-27 years (27.8%) and 30-39 years (20.0%) and most were female (58.8%) against male counterpart (41.72%) Table1.
3.2 Characteristics of animals sampled
Table 2 shows the characteristics of animals sampled. A total of 300 animals were sampled, cattle (n=200) and pig (n=100). Animal age ranges between 5 months to 1½ years with a median age of 6 months. Majority of individual animals were more than 1 years (63.0%). More female animals (73.3% were sampled than male counterpart (26.7%).
3.3 Prevalence of E. coli and ESBL-EC from human, cattle, pig, beef and soil sources.
An overall E. coli isolation rate of 9.3% (140/1500) was detected from all specimens. E. coli isolation rate among the various samples were humans 17.5% (70/400), cattle 20.0% (40/200) and pigs 20.0% (20/100). However, no E. coli was recovered from pork, lettuce and spring onions Table 3. Overall, the prevalence of ESBL-EC isolates from all sampling sources was 21.5% (30/140).
3.4 Distribution of E. coli cultured from human stool samples by study location.
The distribution of E. coli among human stool samples in Accra and Tema is shown in Table 4. While Agbobloshie had the least isolation rate of E. coli. Generally, there was no statistical difference in the isolation rate of E. coli among the different study locations in Accra and Tema (p > 0.05).
3.5 Distribution of E. coli cultured from food sample, soil and fecal samples by study locations
The distribution of E. coli among food sample, soil and fecal sample is shown in Table 5. Generally, there was no statiscal difference between isolation rate of E. coli in Accra and Tema (p>0.05). However, soil samples in Tema were more contaminated with E. coli than Accra (p<0.05).
3.6 Resistant Patterns of E. coli isolates
Resistance patterns of E. coli isolates from diverse sources is shown in Table 6. E. coli exhibited high levels resistance to ampicillin (84.4%), ceftazidime (61.4%), cefotaxime (71.4%), and cefuroxime (96.4%). However, moderate levels of resistance to trimethoprim (31.2%) and tetracycline (25.7%) were observed. For ciprofloxacin, isolates recovered from humans exhibited a higher resistance levels compared to those from cattle (p ≤ 0.05). Similarly for tetracycline, isolates from cattle exhibited a higher resistance compared to those from humans (p ≤ 0.05). For ampicillin and nalidixic acid, isolates recovered from pigs showed a higher resistance levels compared to those from humans (p ≤ 0.05) Table 6.
3.6 Multi-drug Resistant (MDR) Patterns of E. coli isolates recovered from diverse sources.
Overall, 35.7% (n=50) of all E. coli isolates were MDR, but resistance varied among the various sample types; human (28.5%; n=20), cattle (55.0%; n=22), pig (35.0%; n=7) and soil (12.5%; n=1) Table 7. Also, 21.4% (n=30) of E. coli isolates were resistance to 1 antibiotic agent, 28.6% (n=40) to 2 antibiotics agents and 35.7% (n=50) to 3 or more antibiotic agents. Several MDR phenotypes were observed, and more than half of the MDR isolates, 56% (n=28) were co-resistant to ampicillin, cefuroxime, trimethoprim and tetracycline and usually in combination with other antibiotic agents like chloramphenicol. About one-quarter of MDR isolates, 24% (n=12) were resistant to the nalidixic acid and ciprofloxacin Table 7. Table 7 is at the end of the article.
3.4 Resistance Patterns of ESBL-EC from diverse sources
Overall, ESBL-EC isolates exhibited high levels resistance of 100%, 100%, 75.8%, 78.0%, and 58.2% to ampicillin, cefuroxime, ceftazidime, cefotaxime and tetracycline. However, low resistance of 17.2% and 0.0% to ceftriaxone and meropenem. ESBL-EC isolates from human, cattle, pig and soil-sourced exhibited resistant prevalence of 100% to ampicillin, 42.9%, 55.6%, 100% and 33.3% resistance to tetracycline (Figure 2). Among human sourced isolates, there was no statistical significant difference between resistant pattern of ESBL-EC isolates and non-ESBL-EC isolates, with the exception of ampicillin, ciprofloxacin, gentamicin and nitrofurantoin (p ≤ 0.05) Table 8. Among cattle sourced isolates, with the exception of an ESBL-EC isolate that exhibited a higher resistance of (20.0% vrs 0%,) to ciprofloxacin as compared to non-ESBL-EC isolate (p ≤ 0.05) Table 8. Overall, there was no statistical significant difference between antibiotic resistance of ESBL-EC isolates vrs non-ESBL-EC isolates Table 8. Among pig-sourced, there was no statistical significant difference in resistant pattern of ESBL-EC isolates as compared to non-ESBL-EC isolates. However, ESBL-EC isolates exhibited a higher resistance of (66.7% vrs 0%) and (100% vrs 5.9%) to ciprofloxacin and nalidixic acid (p ≤ 0.05) Table 9. Overall, MDR prevalence of ESBL-EC isolates was 66.7% (n=20) and among sourced samples, human 71.4% (n=10), cattle 66.7% (n=6), pig 100% (n=3) and soil 33.3% (n=1).
3.6 Prevalence of AMR genes in ESBL-EC isolates
A total of 16 different AMR genes were from major antibiotic class were observed in the ESBL-EC isolates Table 10. Across all ESBL-EC isolates, the most common AMR genes was blaTEM-1B (32.0%; n=8), (tetA) (48.0%; n=12), (sul2) (36.0%; n=9), (aph(3'')-Id) (24.0%; n=6) and (dfrA14) (16.0%; n=4). Also identified were plasmid-mediated quinolone resistant genes (qnrS1) (12.0%; n=3) and quinolone resistant determining region at the gyrA (4.0%; n=1) and blaCTX-M-15 (4.0%; n=1) from soil sourced. Majority of AMR genes were observed from human and soil ESBL-EC isolates Table 10.
3.7 Prevalence of plasmid replicon types in ESBL-EC isolates from diverse sources
A total of 21 different plasmid replicon types were identified from in silico analysis of ESBL-EC isolates Table 11. The most prevalent IncF plasmid replicon types were IncFIB(Apoo1918) (40.0%; n=10), followed by IncFII(pCoo) (36.0%; n=9). Overall, diverse plasmid replicon types were observed in human ESBL-EC isolates, followed by soil ESBL-EC isolates. IncFIB(Apoo1918) (40.0%; n=5) was most dominant among cattle ESBL-EC isolates, followed by IncFII(pCoo) (40.0%; n=10) Table 11.
3.8 Distribution of sequence types in ESBL-EC isolates from diverse sources
A total of 17 different sequence types were identified from in silico analysis of ESBL-EC isolates from diverse sources Figure 3. The most prevalent sequence types were ST10 (12.0%; n=3), ST206 (12.0%; n=3), ST9312 (12.0%; n=3) and ST4151 (12.0%; n=3). However, 4.0% (n=1) was observed for each of the following; ST73, ST835, ST159, ST1237, ST6311, ST6237, ST5557, ST2061, ST8535, ST960, ST1684, ST999 and ST697, respectively (Figure 3). Overall, ST10 was observed in both cattle and pig ESBL-EC isolates, while ST206 was found in both human and cattle ESBL-EC isolates. Furthermore, ST73 was observed in cattle.