High prevalence of extended-spectrum beta-lactamase-producing Gram-negative pathogens from patients attending Felege Hiwot Comprehensive Specialized Hospital, Bahir Dar, Amhara region

Background Infections caused by extended-spectrum beta-lactamases (ESBL) producing Gram-negative bacteria has emerge as a global threat in clinical practices. The treat is more serious in developing countries due to inappropriate use, poor adherence, use of counterfeit, sub-standard antibiotics and poor infection control practices. Data on ESBL producing Gram-negative bacteria are limited in developing countries including Ethiopia. The aim of this study was therefore, to describe the burden of ESBL producing Gram negative pathogens isolated from patients attending at Felege Hiwot Comprehensive Specialized Hospital, Bahir Dar, Amhara region. Materials and methods A total of 532 clinical samples of blood, urine, stool, wound, abscess, ear discharge, nasal discharge, cervical discharge and body fluid specimens were aseptically collected and bacteriologically processed. Identification of the bacterial species was performed using an automated system (Vitek-2 Compact 27530, USA) and antibiotic susceptibility test was determined by disk diffusion method and selection of antibiotics were in accordance with CLSI guidelines. The MDR pattern of the Gram-negative pathogens was assessed using phenotypic methods of ESBL and carbapenemase production following standard procedure. Result A total of 532 samples were processed and 263 pathogens were isolated. Of these, 185 (70.3%) were Gram-negative and 78 (29.7%) Gram-positive. Of the Gram-negative bacteria the high proportion of the isolates were identified from blood 146/185 (78.9%) and 29/185 (15.7%) were from urine cultures. The most common isolate in all clinical samples was Klebsiella pneumoniae 97/185 (52.4%) followed by Escherichia coli 23/185 (12.4%), Acinetobacter baumannii 15/185 (17.6%) and Enterobacter aerogenes 12/185(6.5%). Of the total Gram negatives, the prevalence of MDR was 148/185 (80.0%). Of the MDR isolates the prevalence of ESBL producers were, 127/148 (85.8%) and 24/148 (16.2%) were carbapenemase producers. Conclusion and recommendation Prevalence of MDR and ESBL producing Gram-negative pathogens in this hospital is alarmingly high. Therefore, continuous monitoring of the problem with effective infection prevention and careful selection of empirical therapy are warranted in the study area.


Introduction
Antibiotics play a critical role in reducing the burden of communicable diseases all over the world. Among the many classes of antibiotics, the β-lactam antibiotics are the most commonly used because of their broad-spectrum activities and better safety profiles [1]. However, emergence of antibiotic resistance threatens the effectiveness of successful treatment of infections and is a public health issue with national and global dimensions. The problem is more alarming in developing countries, where there is high burden of infectious diseases, lack of surveillance networks, laboratory capacity, and appropriate diagnostics [2,3]. Recent advances in drug resistance revealed that emergence of extended spectrum β-lactamase producing Enterobacteriaceae are increasing from time to time. Gram-negative bacilli like Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, and Acinetobacter spp. are important pathogens and may cause blood stream, abdominal and urinary tract infections. A few studies in Africa reported high case-fatality rate associated with blood stream infection, mainly caused by Gram-negative bacteria, and being 2-fold higher than that of malaria (43.5% versus 20.2%) [4,5].
The overall pooled estimate of ESBL proportion in East African hospitals reported to be 42% [6]. A recent report from Addis Ababa, Ethiopia also confirm that there is a high prevalence of ESBL producing Enterobacteriaceae (78.7%) reported from blood stream and urinary tract infections [7]. However, there is little information about ESBL in developing countries like Ethiopia. Therefore, it is crucial to continuously monitor and evaluate the threat posed by ESBL in the community and aware the policymakers and health care professionals to prioritize the public health problems in the region. The aim of the present study was therefore, to evaluate the burden of ESBL at Felege Hiwot Comprehensive Specialized Hospital, Bahir Dar, Amhara region.

Study area, design and period
The study was conducted at Felege Hiwot Comprehensive Specialized Hospital, Bahir Dar, Amhara region. Bahir Dar is a capital city of Amhara region, in the Northwest of Ethiopia. The city is located 565 kms far from Addis Ababa, capital city of Ethiopia. It has a total projected population of 649,426 (city health office 2018 report). Felege Hiwot Comprehensive Specialized Hospital is a tertiary referral hospital with around 423 beds and provide specialized and referral health care services for more than 7 million people who are living in the city and its surrounding regions. The study was an institution based cross-sectional study, conducted from December 2017 to April 2018.

Ethical approval and consent to participate
Ethical approval was obtained from the institutional review board of the University of Gondar with reference number O/VIP/RCS/05/478/2015. Informed written consent was obtained from each study participants. Children less than 18 years who are not able to give consent were also asked an assent and/or written consent taken from their parents or guardians.

Source and study populations
The source population was all patients who were attending the Felege Hiwot Comprehensive Specialized Hospital seeking treatment during the study period. The study populations were all patients suspected for having bloodstream, UTI, wound and others infections at the Felege Hiwot Comprehensive Specialized Hospital during the study period.

Sample size determination
A total of 532 study subjects were recruited in this study who were suspected for bacterial infections. Using convenient sampling technique, blood, urine, stool, wound, abscess, ear discharge, nasal discharge, cervical discharge and body fluid specimens were aseptically collected from study participants for cultures, from December 2017 to April 2018.

Sample processing and isolation of pathogens from clinical samples
Blood culture. Blood samples of 10 ml from adults, 5ml from pediatrics age group and 2ml from neonates were aseptically collected in duplicate (2 bottles for each patient at different time). The blood samples were inoculated in tryptic soya broth and incubated immediately aerobically at 35-37˚C for 18hrs. After 18 hour of incubation, Gram stain was done and subcultured on to 5% Chocolate agar and MacConkey agar. The chocolate agar plates were incubated in a carbon dioxide atmosphere for up to 48 hours, and the MacConkey agar plate aerobically overnight. Those blood culture bottles which do not show growth, were continuously monitored for potential growth of pathogens until 7 days and if no growth after 7days, the blood culture were reported as negative [8].
Urine samples. Mid-stream urine specimens were collected using wide sterile mouth cupped and inoculated onto blood agar, Cysteine Lactose Electrolyte Deficient agar (CLED) and incubated overnight at 37˚C using calibrated loop (measures 1.3mm diameter, delivering 1 μL). The samples with significant bacteriuria (= 10 5 CFU/mL) were further processed.
Other samples Bacteria from other clinical samples such as stool, wound and other body fluids were processed following standard procedures [8].
All isolates of Gram-negative bacteria were subjected and identified using an automated system (Vitek-2 Compact 27530, USA) and antibiotic susceptibility were performed by disk diffusion methods and selection of antibiotics were in accordance with CLSI guidelines.

Antibiotic susceptibility testing
Antibiotic susceptibility testing was performed to all isolates by using the disk diffusion method against; ampicillin (AMP, 10μg), amoxicillin-clavulanic acid (AMC, 20/10μg), cotrimoxazole (SXT, 25μg), tetracycline (TET, 30μg), ciprofloxacin (CIP,5μg), chloramphenicol (CHL, 30μg), gentamycin (GEN, 10μg), cefepime (FEP, 30μg), cefixime (CFM 5μg), ceftriaxone (CRO, 30μg), cefoxitin (FOX, 30μg), and ceftazidime (CAZ, 30 μg) (all from Abtekbio. Ltd UK) were selected in accordance with CLSI guidelines. Bacterial inocula were prepared by suspending the freshly grown bacteria in 3-5 ml normal saline and turbidity was adjusted to 0.5 McFarland standards. Sterile cotton swab was dipped and rotated several times, and was pressed against the wall of the test tube. It was then swabbed over the entire surface of the agar. Antimicrobial impregnated paper disks were placed on the plate and incubated aerobically at 37˚C for 24 hours. The zone of inhibition was measured by calibrated ruler and interpreted as sensitive, intermediate or resistant by using standard chart [9].
Screening for ESBL. This was done as part of the routine susceptibility testing, according to criteria recommended by the CLSI. The disk diffusion technique was used for screening ESBL using ceftazidime (30 μg) and cefotaxime (30 μg). If the result showed an inhibition zone of < 22 mm for ceftazidime and < 27 mm for cefotaxime indicated that the strain probably produces ESBL [9].
Modified Hodge test (MHT). Carbapenemase producing isolates of Gram-negative bacteria were phenotypically investigated by MHT. A 0.5 McFarland suspension of carbapenem susceptible strains of E. coli ATCC25922 was used as a lawn over Mueller Hinton agar plates and meropenem (10 μg) disc were placed at the center of the plates. Isolates to be tested along with a positive control using K. pneumoniae ATCC1705. Inoculated plates were kept for 15 minutes at room temperature and incubated at 37˚C for 24 hours. Observing a clover leaf like shape was considered as positive for carbapenemase production.

Quality control
The reliability of the findings was guaranteed by implementing quality control measures throughout the whole processes of the laboratory work. The reference strains used as control were E. coli (ATCC 25922), K. pneumoniae (ATCC1705).

Data analysis
Data were entered and analyzed using SPSS version 20. Simple frequency was applied to see the distribution of sociodemographic variables.

Results
Of the 532 study participants enrolled in the present study, 290 (54.5%) were male, 263 (49.4%) were children aged less than 6 years and only 8 (1.5%) were greater than 60 years. Three hundred and nine (58.9%) were rural dwellers and 459 (86.3%) were inpatients. As most of the study participants came from rural areas, 99 (18.6%) of the study participants were illiterate. Regarding to patient setting majority of the patients were admitted patients, 459 (86.3%). (Table 1).

Discussions
Half of the clinical samples processed in our study were culture positive of which the Gramnegative pathogens were in the ratio of 3.3:1 compared with the Gram-positive isolates. This is   similar to a study done in Iran that the ratio of Gram-negative to Gram-positive bacteria was 3.2:1 [10]. As pointed out by Abe et.al [11], the reason may be because Gram-negative bacteremia induces greater magnitude of inflammatory response than Gram-positive bacteremia as a result Creactive protein and IL-6 levels are significantly higher in Gram-negative bacteremia than in Gram-positive bacteremia. These observations suggest a distinct immuno-pathophysiologic behavior of sepsis in patients with Gram-negative bacteremia that may influence clinical outcomes. K. pneumoniae, E. coli, A. baumanii and E. aerogenes were the most common Gram-negative isolates in all clinical samples of the present study. Similarly, a study on septicemia revealed that E. coli (27/70, 38.6%), K. pneumoniae (24/70, 34.3%), A. baumanii (9/70, 12.8%) were the commonest Gram-negative isolates [12].
The overall prevalence carbapenemase production in the present study was 24 (16.2). The carbapenemase producing isolates reported in different hospital settings ranged from 2.3% to 67.7% in North Africa and from 9% to 60% in sub-Saharan Africa [19]. It was lower than the studies from China carbapenem-resistant Klebsiella pneumoniae (CRKp) to be 31 (65%) [20] and slightly higher in Addis Ababa, 12.1% [7].
Comparing with other isolates relatively high number of isolates of K. pneumoniae 16 (10.8%) were carbapenamse producers. This is lower than a study from Brazil where 50% of K. pneumoniae isolates were positive for carbapenemase production [15].
K. pneumoniae, E. aerogenes, E. coli, and E. cloacae were the commonest MDR isolates in our study. Majority of the isolates were resistant to more than 9 antibiotics tested. This pattern was also true for ESBL producing isolates and the overall MDR rate was 80.0%. This was similar to a study done at Mekele Hospital, where 82.9% multi drug resistant to the commonly used antibiotics [23]. However, it was higher than a study in Marikos, where 72.2% [24] and Tanzania, 61.4%, [18] but lower than a study from Jimma, 85% [25]. Many of the isolates in the present study becomes resistant to commonly used antibiotics. This high prevalence of MDR in the present study may be due to the reason that, majority of our study subjects were admitted patients (83.6%) where many antibiotics are circulating in the hospital which may serve as a selective pressure for increasing MDR isolates.

Conclusion and recommendations
The overall MDR isolates in the present study was 80%. Among phenotypically tested Gramnegative bacteria 85.8% were ESBL producers and 16.2% were carbapenemase producers. The Gram negatives showed high level of drug resistance towards the tested cephalosporin except cefoxitin. Although there is an increasing intermediate resistance in cefoxitin and ciprofloxacin, compared to other antibiotics tested; these antibiotics holds promise as an alternative choice of therapy for Gram-negative bacterial infections. Finding high rate of ESBL production in Gram-negative bacteria require strict infection control measures and careful selection of empirical therapy in the study area.