Antibiotics Resistance pattern of gram negative bacilli
The level of antimicrobial resistance of identified gram negative isolates was ranging from 0%-75.4%. The present study showed that there was higher resistance to ampicillin (75.4%), followed by amoxicillin with clavulanic acid (64.0%), sulfamethoxazole-trimethoprim (55.6%), aztreonam and cefuroxime (48.8%), cefotaxime (47.0%), cefepime (45.6%), ceftriaxone (44.9%), ceftazidime (44.1%). There were also significant level of resistance to ciprofloxacin (40.2%) and gentamycin (21.3). Comparable result were reported in Ethiopia such as in Gondar: ampicillin (84.6%) and sulfamethoxazole-trimethoprim (79.5%) and gentamicin (35.9%) [16], Debre Markos: ampicillin (70.4%), amoxicillin with clavulanic acid (58.8%), sulfamethoxazole-trimethoprim (53.1%) [21].However, the resistance level was lower than a study conducted in Tanzania: ampicillin (100%), amoxicillin with clavulanic acid (98.7%), sulfamethoxazole-trimethoprim (95.2%), ceftazidime (74.0%) [22], Southeast Iran: sulfamethoxazole-trimethoprim (93.8%) and amoxicillin with clavulanic acid (91.4%) [23] and in Sierra Leone: sulfamethoxazole-trimethoprim (91.4%), gentamycin (72.9%) [24]. The possible reason for this difference might be due to indiscriminate use of antibiotics, patient condition and majority of bacteria in these countries were β-lactamases producing GNB.
Multi drug resistance pattern of gram negative bacilli
In the present study, the overall magnitude of MDR among all GNB isolate was 73.7%. There were also similar findings from studies conducted in Gondar (68.0%) [25], Dessie (74.6%) [26], Debre Markos: (72.2%) [21] and Nepal (64.0%) [27]. However, our result was lower than studies done in Sierra Leone (85.7 %) [24], Gondar (87.4%) [16], Bahir Dar (93.1%) [28], Nepal (96.8%) [29].The difference in magnitude of MDR isolates might be due to patient condition, definition for MDR and empirical treatment trend. In addition, our result was higher when compared to a previous study done in Jimma (59.3%) [30], Nepal (54.2%) [27], another study in Nepal by Lamichhane et al [31] reported (33.14%). However, the increased proportion of MDR seen in this study was considered as alarming because only a few treatment options remain for infections. Therefore, implementing strong infection control strategies is required to reduce MDR burden.
The present study showed that Pseudomonas spp.(100%), Enterobacter spp. (90.0%) and Citrobacter spp (90.0%) were found to be the principal MDR isolates which agreed with a study done in Jimma: Citrobacter spp. 100% [30], in Nepal: Enterobacter spp. 71.4% [31]. Different studies showed different pathogens as a major MDR isolates, in Gondar: K. pneumoniae (95.6%) and E. coli (92.9%) [16], Sierra Leone: K. pneumoniae (73.3%) and E. coli (61.5%) [23], Nepal: K. pneumoniae 100% & E. coli 95.5% [29] were found to be the predominant MDR isolates. These pathogens are the most commonly found in both hospital and community acquired infections. In addition, these bacteria are resistance to multiple groups of antimicrobial agents, this makes treatment difficult [27].
Magnitude of ESBLs producing gram negative bacilli
The overall magnitude of ESBLs producing GNB in the present study was 38.8%, which was in agreement with study reported in Harrer 33.3% [32], Nepal 34.5% [27], Spain 42.8% [33], India 44.0% [34]. ESBLs producing organisms are the major cause of treatment failure, reduced rate of clinical and microbiological responses, longer hospital stay and increased cost of hospital [35].
However, the magnitude of ESBLs production among GNB in our study was lower than studies done in Addis Ababa 57.7% [36], Bahir Dar 57.6% [28], North West Nigeria 58.0% [37], Southwestern Uganda 89% [38] and Southeast Iran 53.8% [23]. This wide variation might be due to differences in study population, type of specimen, sample size, the extent of antibiotic use and mainly the methodology used.
To the contrary, our finding was higher than studies reported in Adama 25.0% [39], Nigeria 15.8% [40], Nepal 26.8 % [29] and Italy (6.3%) [41]. This indicated that ESBL-producing organisms are increasing from time to time.
In the current study, the predominant ESBLs producing GNB was K. pneumoniae (56.1%) than E. coli (43.8%). This finding was supported by previous studies done in Addis Ababa, K. pneumoniae (78.6%) and E. coli (52.2%) [36], Bahir Dar: K. pneumoniae (69.8%) and E. coli (58.2%) [28], North West Nigeria: K. pneumoniae (62.9%) and E. coli (54.2%) [37], Southwestern Uganda: K. pneumoniae (52%) and E. coli (44%) [38], Nairobi: K. pneumoniae (78.8%) and E. coli (60.7%) [42], Uganda: K. pneumoniae (72.7 %) and E. coli (58.1% [43]. On the other side, our finding was in contrary to study conducted in Sri Lanka: E. coli (86.8%) and K. pneumoniae (13.1%) [44], India: E. coli (50.14%) and K. pneumoniae (48.3%) [45] in which E. coli was the predominant ESBLs producer than K. pneumoniae.
Magnitude of AmpC producing gram negative bacilli
AmpC β-lactamases producing GNB have been responsible for several nosocomial outbreaks and high rate of treatment failure [46].
In the present study, from the total specimens, 2.4% (8/338) AmpC producing GNB were isolated. This finding was in line with the study done in Iran 1.5% [47], Greek 2.6% [48], India (8%) [49]. However, it was lower than the finding in Nigeria (15.2%) [46], Spain (14.2%) [33], India (37%) [34]. Differences in finding between these countries might be related to detection methods, study participants, geographic area and AmpC genes prevalence difference.
In present study, K. pneumoniae 7.3% (3/41) followed by E. coli 2.2% (5/224) were the principal AmpC producing pathogen which was in line with study done in Turkey: K. pneumoniae 3.6% [50] and in Spain: K. pneumoniae 5.8% were common AmpC producer [33]. This might be due to the fact that plasmid mediated AmpC β-lactamases are seen in Enterobacteriaceae and these genes are easily transferable horizontally [13].
The current study also demonstrated co-existence of ESBL and AmpC enzymes in five isolates 3.6% (5/139). This finding was concordant with study conducted in Nigeria (6.04%) [40], South India (4.4%) [2], India (9.9%) [9] Simultaneous production of ESBL and AmpC enzymes in a bacterium causes false negative confirmatory test for ESBL production. In other word, existence of plasmid mediated AmpC beta-lactamase enzyme can mask the presence of ESBL [51]. Therefore, simultaneous detection of these enzymes helps to prevent missing of ESBL.
Antibiotics susceptibility pattern of ESBLs and AmpC producing gram negative bacilli
In present study, the highest rate of susceptibility of ESBLs producing isolates were found toward amikacin (100%) followed by imipenem (99.2%), meropenem (97.7%), ertapenem (98.5%), piperacillin/tazobactam (87.0%) and cefoxitine (84.0%). This finding was in harmony with the findings of other studies conducted in Nepal: imipenem (100%), piperacillin/tazobactam (93.3 %), and amikacin (91.8%) [27], Sri Lanka: meropenem (95%), imipenem (73.7%) and amikacin (60.6%) [44], India: imipenem (100%), piperacillin/tazobactam (89.3%), meropenem (87.5%), and amikacin (83.9%) [45], Ghana: 100% sensitive to meropenem [52], and study in India [34] showed 98.0% isolates were sensitive to imipenem.
The present study indicated that ESBL producers had significant levels of resistance to third generation cephalosporin, penicillin & sulfonamide. Similar finding observed in Adama, Uganda and Ghana [39, 43, 52]. Highest levels of resistance to ampicillin (99.2%), ceftazidime (98.5%), ceftriaxone (98.5%), amoxicillin with clavulanic acid (98.0%) and sulfamethoxazole-trimethoprim (81.0%) observed in this study was in agreement with the study done in Nepal: amoxicillin with clavulanic acid (100%), sulfamethoxazole-trimethoprim (59%) [27], India: ceftazidime (97%), ceftriaxone (76%) [53], another study in India done by Shashwati et al, amoxicillin with clavulanic acid (89.3%), sulfamethoxazole-trimethoprim (94.6%) [45].
Most of the AmpC producing GNB were resistant to the commonly used antibiotics as seen in different studies [51, 54, 40]. In our finding, among cefoxitin resistance isolates, 10% (8/80) were AmpC producers which was in agreement with the study done in Iran, 5.1% [47], Turkey, 8.7% [50] but lower than the finding in India, 37.0% [53]. The possible reasons for cefoxitin resistance in the absence of AmpC production might be due to resistance mechanism like loose of permeability of porins and other [55].
The present study showed that AmpC producers were 100% sensitive to imipenem and meropenem (100%) but highly resistance againest ampicillin (100%), amoxicillin with clavulanic acid (97.0%), ceftriaxone (89.0%), sulfamethoxazole-trimethoprim (75.0%), ceftazidime (68.0%), and cefepime (50.0%). This finding was in line with study conducted in Turkey showed AmpC producers were 100% sensitive to imipenem and meropenem, 92.0% to amikacin and 82.0% resistance against amoxicillin with clavulanic acid, 68.0% to ceftazidime and 49.0% to cefepime [50], Likewise, in Nigeria: AmpC producers were resistant to amoxicillin with clavulanic acid (77.9%), ceftazidime (75.0%) [46], India: resistance level to amoxicillin with clavulanic acid was (95.9%), to sulfamethoxazole-trimethoprim (82.9%), and to ceftazidime (87.1%) [53]. AmpC producers seem susceptible to cepholosporins in-vitro but when cepholosporins are used in vivo, they result in failure of treatment [56, 57]. Therefore, cepholosporins could not be useful in treating infections caused by AmpC producing bacteria.
Distribution of ESBL and AmpC beta-lactamase producing gram negative bacilli in different specimens
The present study showed that ESBL and/or AmpC producing GNB were predominantly found in urine 44.7% (109/244) followed by in pus 34.9% (22/63). This might be due to the larger number of urine samples were included in this study. In this study maximum ESBL producers 41.8% (102/244) were found in urine. Fairly similar finding was also reported in different countries such as central India (52.28%) [45], Northwest Nigeria (63.5%) [37], Uganda 64.9 % [43], Ghana 66.7 % [52], Sierra Leone 64.3 % [24], India 52.3% [45] and India by Tewari et al, 35% [53]. However, study done in Adama, showed that major source of ESBLs producing GNB (53.0%) were isolated from pus [39].The difference might be due to proportion of urine specimen difference, study participant difference and risk factors.
All AmpC producing pathogen in our study was isolated from urine 3.3% (8/244). This finding was in close agreement with study conducted in Turkey and Nigeria [50, 57] but it was not supported by Ogefere et al [46] in Nigeria where isolates from sputum (50.0%) were the predominant producers of AmpC β-lactamase. This indicated that the prevalence of AmpC β-lactamase may differs significantly among bacteria recovered from different clinical specimens [34].
Strength of the study
This is the first study done on the magnitude of ESBL and AmpC β-lactamase producing GNB in Ethiopia using automation for screening and CLSI recommended conventional methods for confirmation of these β-lactamase producing GNB. The study tried to show the antimicrobial susceptibility pattern of ESBL and AmpC β-lactamase producing GNB.
Limitation of the Study
- Magnitude of ESBLs and AmpC producing GNB from blood culture were not addressed due to the down time of the Bactech 9050 machine during the study period.
- Since specimens were collected from different hospitals and arrived to ICL, we are unable to realize possible risk factors and the outcome of the patients infected with these β-lactamase producing bacteria.