Antimicrobial Resistance Patterns, Extended-Spectrum Beta-Lactamase Production, and Associated Risk Factors of Klebsiella Species among UTI-Suspected Patients at Bahir Dar City, Northwest Ethiopia

Introduction Klebsiella species cause pneumonia, UTI, and septicemia in human beings. Beta-lactam drugs are used extensively to treat patients infected with Klebsiella, but most of the Klebsiella species are resistant to third- and fourth-generation cephalosporins and monobactams to which data are scarce in the study area. Objective To determine the prevalence, antimicrobial resistance, ESBL production, and associated risk factors of Klebsiella species among UTI-suspected patients in Bahir Dar City, Northwest Ethiopia. Methods A multi-institution-based prospective cross-sectional study was conducted from January to May 2019. Midstream urines were collected from 385 patients and inoculated onto CLED and MacConkey agars. Identification of growth was done by a battery of biochemical tests. Antimicrobial resistance and ESBL production patterns were determined by using the disc-diffusion method on Mueller–Hinton agar. Quality of data was maintained by following SOPs and using Klebsiella pneumoniae (ACTT700603). Logistic regression statistical analysis was done using the SPSS, version 25, statistical package. A pvalue ≤ 0.05 was considered statistically significant. Results The median age of the study participants was 32 years. Majority of them were female, urban residents, and unable to read and/or write. The total Klebsiella species detected were 38 (9.9%). Of which, 25 (65.8%) were Klebsiella pneumoniae, followed by 6 (15.8%) Klebsiella ozaenae. 20 (80%), 19 (76%), and 19 (76%) Klebsiella pneumoniae showed resistance to amoxicillin/clavulanic acid, ampicillin, and cotrimoxazole, respectively. All K. oxytoca were resistant to ampicillin, and all K. rhinoscleromatis were resistant to cefoxitin and cefotaxime. Klebsiella species that showed resistance to ≥3 antimicrobials were 26 (68%). ESBL-producing Klebsiella species were 10 (26.3%). Patients who had history of antibiotic use, hospitalization, and tight dressing habit had more risk of getting UTI (p < 0.05) than their counterparts. Conclusions Overall UTI prevalence in our study was lower than that of previous Ethiopian studies. High MDR and ESBL-producing Klebsiella species were detected. Patients' history of antibiotic use, hospitalization, and tight dressing habit were risk factors for UTI. It calls up for improving prevention/control systems of Klebsiella species.


Introduction
Urinary tract infection (UTI) caused by bacteria is one of the major reasons for visiting health institutions, morbidity, and comorbidities in patients with underlying conditions worldwide [1][2][3].
ough different pathogens cause UTI, various studies described that the causative agents of UTI are predominantly associated with Gram-negative bacteria such as E. coli, Pseudomonas species, Klebsiella spp., Proteus spp., and Gram-positive bacteria such as coagulase-negative staph (CoNS) and S. aureus [2,[4][5][6]. e prevalence of Klebsiella species isolated from urine specimens in Nepal revealed that K. pneumoniae accounted for 145 (2.3%) [5]; whereas the prevalence of Klebsiella

Study Design, Period, and Area.
A multicenter institution-based cross-sectional study was conducted from January to March 2019 at Felege-Hiwot Comprehensive Specialized Hospital, Gammby Private General Hospital, and Bahir Dar Health Center, Bahir Dar City, Amhara National Regional State. Bahir Dar City, the capital of the Amhara National Regional State, is located 576 km northwest from the capital city of the country, Addis Ababa. Two comprehensive specialized referral hospitals, eight health centers, three private hospitals, and seven private specialized clinics provide health services to the community. Among these, one comprehensive specialized hospital, one health center, and one general private hospital were selected for this study. Felege-Hiwot Referral Hospital is a tertiary care referral hospital serving over 7 million people from the surroundings with 500 beds, 9 operating tables, and over 600 employers. Gammby General Hospital has 150 beds and 2 operating tables and provides service for around 2 million people. Both hospitals provide obstetrics, pediatrics, internal medicine, ophthalmology, orthopedics, and surgery services, whereas the Bahir Dar Health Center provides service for outpatients of acute cases and delivery services for around 3 million people.

Sampling Technique of Institutions and Sample Size
Determination. A simple random sampling method was employed to select two hospitals and one health center. UTI patients who attended one month before the study period the selected health institutions were found to be 395, and the estimated number of patients during the data collection period, from January to March 2019, was 1185. erefore, 384 study participants were allocated to the three health institutions proportionally based on their patient flow in such a way that 277 were from Felege-Hiwot Comprehensive Specialized Hospital, 83 from Gammby Private General Hospital, and 24 from Bahir Dar Health Center.

Source Population.
All patients who were attending their medical treatments at the three selected health institutions at Bahir Dar City during the study period were the source population.

Study Population.
Suspected patients for UTI and who were attending their medical treatments at the three selected health institutions at Bahir Dar City during the study period were the study population.

Inclusion
Criteria. All UTI-suspected patients who attended the selected health institutions during the study period and who did not take any antibiotics with in the previous two weeks and were voluntary to participate on the study were included in the study.

Sociodemographic Data and Specimen Collection.
After informed consent/assent was obtained, a predesigned structured questionnaire was administered to collect the sociodemographic variables (age, sex, marital status, education level, occupation, residence) and clinical risk factors (history of catheterization, history of UTI, history of previous hospitalization, and history of antibiotic use). To ensure the reliability of information, all the questionnaires were administered by native Amharic speakers and were checked for completeness and consistency every day. e patients or patients' guardians were given a sterilized container to collect midstream urine and were transported to microbiology laboratory for examination.

Culture and Identification.
Each urine sample was inoculated onto cystine-lactose-electrolyte-deficient (CLED) agar and incubated at 37°C for 24 hours. e counted colonies yielding bacterial growth of 10 5 CFU/mL for single midstream urine was considered significant for positive bacteriuria. Pure colony was transferred to MacConkey and incubated at 37°C for further 24 hours. Isolates were preliminarily screened by their colony morphology, pigment production (pink), flat or mucoid colonies, and Gram staining reaction. Further identifications of isolates were made by doing a battery of biochemical tests such as triple sugar iron agar, indole, Simon's citrate, and urea, and motility tests were performed by picking lactose fermenter or pink colonies from MacConkey agar for identification of isolates to the species level by using the Clinical Laboratory Standards Institute (CLSI) guideline 2018 [24].

Extended-Spectrum β-Lactamase Detection.
e screening for extended-spectrum beta-lactamase (ESBL) was done by using ceftazidime (≤22 mm) and cefotaxime (≤27 mm). More than one antimicrobial discs were used for screening to improve the sensitivity of ESBL detection as recommended by CLSI guideline 2018 [24]. e organisms that showed zone of inhibition lower than the minimum for any of the above antimicrobial discs were considered as ESBL positive. e phenotypic confirmations were done by testing the isolated bacterial species against ceftazidime (30 μg) and ceftazidime-clavulanic acid combination (30/10 μg/discs) (Oxoid Ltd., UK). In this test, an overnight culture suspension of the bacterial isolate was adjusted to 0.5 McFarland's standard. Lawn cultures were done on the Mueller-Hinton agar (MHA) plate. e ceftazidime and ceftazidime-clavulanic acid discs were placed at 20 mm apart on the agar surface. After incubating at 37°C for 24 hours, a ≥5 mm increase in the zone diameter in comparison to ceftazidime were taken as indicative for ESBL positive/producer.

Quality Control.
e questionnaire was prepared in English and translated into the local language which the study participants use (Amharic), and then it was translated back to English for analysis and report. It was pretested before the commencement of the actual study to make sure that whether the questionnaire was appropriate and understandable to the study participants. Microbiology standard operating procedures were strictly followed throughout the laboratory processes. Escherichia coli ATCC25922 as an ESBL-negative and Klebsiella pneumoniae ATCC700603 as ESBL-positive reference strains were used.

Data
Analysis. Data were entered and coded in Epi-Data and exported and analyzed using SPSS software (version 25). e characteristics of the study patients were summarized using frequencies and mean and standard deviation. Binary logistic regression was done to determine the association of independent and dependent variables. Adjusted odds ratios were computed using multivariable logistic regression for variables with p value < 0.2 in the crude odds ratio. Findings of the study were presented in tables and figures. A p value ≤ 0.05 was considered statistically significant.

Ethical Consideration.
is study was ethically approved by the Ethical Review Committee of the School of Biomedical and Laboratory Sciences, University of Gondar. Moreover, legal permission was obtained from the Felege-Hiwot Comprehensive Specialized Hospital, Gammby Private General Hospital, and Bahir Dar Health Center of higher management prior to data collection. Written consent and assent were taken from each participant after they understood the purpose of the study. Data were kept in full confidentiality and were not disclosed to an unauthorized person. Positive cases were referred to the attending physician for their better management.

Sociodemographic Characteristics.
A total of 385 clinically UTI-suspected patients were included in the study; of these, 205 (53.2%) were females, and majority of the study International Journal of Microbiology participants (37.9%) were in the age groups of 16-30 years with median age of 32 years. Moreover, 60.8% of the participants were from urban areas and 24.4% were illiterate (Table 1).
K. ozaenae (4, 10.5%) and K. rhinoscleromatis (3,7.9%) were also identified in the current study, which were not detected in the Iraq and Nigerian studies stated above. e present study showed that majority (22,57.9%) of the total isolates causing UTI were from female participants, which is in line with studies that demonstrated 81.2% of culture positivity of urine specimen were from female patients in Ethiopia [19]; 29 (88%) of Klebsiella species isolated were from the female participants in Iraq [30], and 70 (60.35%) of UTI were recorded in females in Saudi Arabia [32]. However, a report from Yemen indicated that the majority (55.9%) of culture positivity of UTI was obtained from the male study participants [8]. ese discrepancies in the rates of isolation of Klebsiella species causing UTI in different studies might be resulted from variations in the sample size and type of the populations studied, laboratory methods practiced, subjectivity on definition of significant bacteriuria during culture reading, the length of the study period, and specimen's quality. In addition, the possible reason of gender variability of UTI prevalence might be related with the history of urinary tract operation, no use of antibiotics in the preceding time periods and infection outside the urinary tract [33]; and the other cause might be due to the difference in the target setting/population [34], where those male patients might have different predisposing factors such as age (prevalence may range from 20% to 50% in old age), enlarged prostate, prostatism, debilitation, and subsequent instrumentation of the urinary tract. On the other hand, the high prevalence of UTI in females is unambiguously due to anatomical factors such as the shorter urethra and the closeness of the urethral opening to the anus that shortens the distance that bacteria must travel to reach the bladder and sexual intercourse, bad personal hygiene, and incontinence.
Klebsiella species identified in the current study showed high levels (66.7% to 100%) of resistance to ampicillin, amoxicillin-clavulanate, and cotrimoxazole. On the other hand, meropenem, chloramphenicol, ciprofloxacin, ceftazidime, cefoxitin, cefotaxime, gentamicin, and nitrofurantoin were found to be the drug of choice for Klebsiella species associated UTI therapy since the isolates showed better sensitivity (0% to 28% resistance) to these antimicrobials tested. A similar result was reported in Ethiopia where Klebsiella species isolated from UTI-suspected patients demonstrated 66.7% resistance to cotrimoxazole and 100% to ampicillin. However, isolates demonstrated a lower resistance rate (5.6% to 22.2%) to cefoxitin, amoxicillinclavulanate, ciprofloxacin, and gentamicin [19]. Isolates in  the present study showed better sensitivity to nitrofurantoin but higher resistance to amoxicillin-clavulanate, which is different from Bitew et al report where isolates showed better sensitivity to amoxicillin-clavulanate but higher resistance to nitrofurantoin [19]. ese variations might be because of the differences in the levels of consciousness of the inhabitants on how to use antimicrobials, and unavailability of bacteriology laboratory facilities, habit of irrational use of antimicrobials (use of broad-spectrum antimicrobials empirically and from over the counter), and unavailability of nitrofurantoin were more encountered in the setting of the present study (730 km far from Addis Ababa) compared with the capital of the country where Bitew et al.'s work was done. In line with the current results, the studies in the northwestern part of Ethiopia revealed that K. pneumoniae showed a high level of susceptibility to meropenem 5/5 (100%), and ceftazidime and nitrofurantoin 4/5 (80%) each [11]; and in the southern part of Ethiopia, 91% to 100% of the Klebsiella species isolated were susceptible to gentamicin, amikacin, nitrofurantoin, doxycycline, and ceftriaxone [35]; and a study in Yemen also revealed that isolates causing UTI were resistant to ampicillin (95.2%) and amoxicillin (97.6%)       [18]. Besides chloramphenicol, ceftazidime, and ciprofloxacin, results of the present study indicate that meropenem is one of the most effective antimicrobials against Klebsiella species since the isolates demonstrated susceptibility ranging from 66.7% to 100%. is was in agreement with another study in India where 65.7% of K. pneumoniae were meropenem susceptible [36]. A 2018 study conducted in Germany also revealed that the carbapenem nonsusceptibility in K. pneumoniae isolates is still low even though it is slowly growing and in the light of the strong increase of K. pneumoniae isolates over the last year; this poses a significant challenge to public health [37].
Twenty-six (68%) of the Klebsiella species isolated in the current study were resistant to ≥3 antimicrobial drugs, and thus, they exhibited MDR. In addition, 15 (39.5%) of Klebsiella pneumoniae exhibited MDR (resistant to ≥3 antimicrobials), which is nearly in line with a study done at Dessie, Ethiopia, where 63% of Klebsiella spices showed MDR [26]; however, it is lower than another study conducted in Northwest Ethiopia where the highest MDR was observed in Klebsiella specie 6/7 (85.7%) [11]; and in Equatorial Guinea, 91.7%, 33.3%, and one isolate of K. pneumoniae demonstrated MDR, extensive drug resistance (XDR), and pan-drug resistance (PDR), respectively [38]. Rates of MDR K. pneumoniae in the current study are higher than a report from Iran where 96/198 (48.5%) of the isolates exhibited MDR and 14/96 (14.6%) of these MDR isolates were completely (100%) resistant to all antibiotics investigated [39].
In the case of ESBL production patterns, results of the present study indicates 10/25 (40%) prevalence of ESBLproducing K. pneumoniae, which is slightly in agreement with a study conducted in India where 44.7% Klebsiella pneumoniae were ESBL producers by using MIC [40] but is different from previous findings in Ethiopia where (1/7, 14.3%) [11], (4, 23.5%) [28], and 76.4% of K. pneumoniae were ESBL producers [41]; in Sri Lanka, Klebsiella pneumoniae with 4 (25%) ESBL producers [42]; Bangladesh, 18.57% Klebsiella spp. ESBL [43]; Pakistan, 15.25% (9/59) in K. pneumoniae [44]; and Iraq 85.7% (66) of K. pneumoniae isolated from UTI patients were ESBL producers based on phenotyping ESBL detection by ESBL E-strip tests [45]. One of the possible reasons for the high rates of resistance to routinely prescribed antimicrobials observed in the current study might be associated with the occurrence of substantial numbers of ESBL producer isolates. An irrational and frequent use of these antimicrobials, easy availability, and the over-the-counter sale of the antimicrobials without a right prescription and inappropriate dosing schedule might also attribute to the occurrence of these high rates of resistance. ough statistically not associated, being women, old age, rural residency, lower educational status, farmer, and inpatient were with high prevalence of UTI compared with their counterparts. However, history of antibiotic use, previous hospitalization, and tight dressing habit were independent associated risk factors of UTI in the current study (p < 0.05). It is supported by a previous study reported in India where previous antibiotic exposure was identified as a risk factor of getting UTI and ESBL-producing organisms [46]. Similarly, a study in Israel showed that the length of any hospital stay is significantly associated with UTIs [47].

Conclusion and Recommendation
e prevalence of Klebsiella spp. in this study is comparatively higher than that of many other studies reported across the globe. Klebsiella pneumoniae, K. ozonae, K. rhinoscleromatis, and K. oxytoca were identified and of these, K. pneumoniae was the predominant isolate. More than two-thirds of isolates showed MDR. e rate of ESBLproducing K. pneumoniae is higher than earlier Ethiopian reports. Chloramphenicol, ceftazidime, ciprofloxacin, and meropenem remain the most effective antimicrobials for UTI. History of antibiotic use, earlier hospitalization, and tight dressing habit were risk factors associated with UTI (p < 0.05). us, at least minimizing empirical treatment of UTI by using urine culture and awaiting antimicrobial susceptibility reports are compulsory.

Data Availability
All the data supporting the conclusion of the study are included in the paper.

Conflicts of Interest
e authors declare that they have no conflicts of interest.