Multidrug-resistant bacterial isolates from patients suspected of nosocomial infections at the University of Gondar Comprehensive Specialized Hospital, Northwest Ethiopia

Objectives As the hospital environment favors the circulation of drug resistant bacteria, continuous surveillance of antibiotic resistant patterns is an important approach for a better patient management. This study is therefore, aimed to assess multidrug resistant bacterial isolates from patients suspected of nosocomial infections at the University of Gondar Comprehensive Specialized Hospital, Gondar, Ethiopia. Results Of the 260 patients, 173 (66.5%) of them were culture positive. Among culture positive patients a total of 216 bacterial isolates were recovered, of which the most common species were S. aureus 77 (35.6%), followed by E. coli 33 (15.3%) and Klebsiella spp 29 (13.4%). Of the S. aureus isolates, 67.5% were cefoxitin (methicillin) resistant. Citrobacter spp (100%), Klebsiella spp (79.3%) and E. coli (75.3%) were the leading MDR Gram-negative isolates. The overall MDR resistant rate was 152 (70.4%).


Introduction
The impact of nosocomial infections (NIs) is an important public health concern due to the presence of increasing numbers of immunocompromised hospitalized patients, emergence of new strains of antibiotic resistant pathogens [1,2]. It is largely known that bacteria are the leading cause of NIs [3]. The most common types of NIs that could occur in a hospital set up are: surgical site infections, blood stream infections, urinary tract infections [4] respiratory infections, gastroenteritis, pneumonia and meningitis and other soft tissue infections [5].
Currently, antibiotics remain the leading therapy for treating bacterial infections. However, the irrational use of antibiotics is no longer viewed as benign and certain strains of multidrug resistant bacteria have emerged by selection pressure; as a result, bacteria that have been once sensitive, re-emerged as resistant to different antibiotics and create limited therapeutic options, increased risks of treatment failure and poor patient management. As the incidence of antibiotic resistance rises, so do costs associate with its consequences. The worldwide emergence of multidrug resistant (MDR) among Gram-negative and Gram-positive bacteria has resulted in a great threat to efforts against bacterial pathogens [6]. It is clear that if there is no effective timely response, the challenge of antibiotic resistance is becoming alarming and will be a great challenge in the years to come. The aim of this study is therefore, to assess the rate of multidrug resistant bacterial isolates from patients suspected of nosocomial infections at the University of Gondar Comprehensive Specialized Hospital, Northwest Ethiopia. Feleke et al. BMC Res Notes (2018) 11:602 Main text

Study area, study period and study population
The study was conducted at the University of Gondar Comprehensive Specialized Hospital, Northwest of Ethiopia from February 1 to May 31, 2016. All inpatients in the hospital wards that were tentatively diagnosed by the responsible physicians who were suspected of nosocomial infection were study populations. Nosocomial infection is an infection occurring after 48 h of hospital admission or 3 days after discharge or infection occurs within 30 days after the operation [7].

Sample size and sampling technique
The sample size was determined by using a single population proportion formula using the prevalence rate of 16.4% [8] a total of 260 samples were included in the study. The study participants were enrolled consecutively using convenience sampling technique.

Data collection and laboratory methods
Data were collected using structured questionnaire consisting of the patient's demographic information. Clinical samples such as blood (2 mL for children and 10 mL for adults), pus, wound discharge, morning midstream urine (3 mL), stool, eye discharge and body fluids were aseptically collected using sterile containers and transported promptly to the Microbiology laboratory with appropriate transport media. Identification of bacteria to species level was done using colony characteristics, gram reaction and different biochemical tests following standard procedure [9] and antimicrobial susceptibility testing of each isolates was checked using Muller-Hinton agar (Oxoid, England). Different antibiotic discs were used (Oxoid, England) following Kirby Bauer disc-diffusion method and interpreted following CLSI guidelines [10].

Data analysis and interpretation
Data were entered, and analyzed using SPSS version 23 software. Results were presented through graphs and tables. Statistical significance association was measured by using Chi square test and p value < 0.05 was considered as statistically significant.

Results
A total of 260 hospitalized patients suspected of having nosocomial infections were assessed by culture. The 152/260 (58.5%) of the study subjects were females. The age of the patients ranged from 1 to 85 years with the median age of 27.61 ± 20.3 years. The clinical samples collected were mainly wound discharge, pus, urine, blood, stool, eye discharge and body fluids (Table 1).

Discussion
The present study showed that 66.5% of the inpatients had bacterial infections. This result is in line with a study reported from Dessie, Northeast Ethiopia (70.5%) [11]. However, our result is lower than studies from Jimma Hospital (87.3%) and Addis Ababa Hospitals (84.1%) [12,13]. These variations may be due to differences in clinical samples, methodological variations, and study settings. Of the 173 culture positive patients, 131 (75.7%) had single infections, 41 (23.7%) double infections and 1 (0.6%) triple bacterial infections. These findings were consistent with a report in Brazil, 23.3% with polymirobial infections [14]. However, slightly lower results of polymicrobial growth were reported by Sikka et al. [15], 14.3% and Singh et al. [16], 11.3%. The difference in polymicrobial infection might be due to the fact that, most of the patients were with surgical site infections, the patient-related factors, poor equipment quality, financial resources as well as the competence of the surgical team [17].
In the present study, high rate of Gram negative (53.2%) bacteria were the causative agents of nosocomial infections than Gram-positives (46.8%). Similarly, Gram-negative bacteria as the leading cause of nosocomial infection were reported from Addis Ababa Hospitals 76 (73%) [13] and Jimma Hospital 53% [12]. Furthermore, Gramnegative organisms constituted more than three quarters (78.3%) of all causative organisms in hospital associated infections in Ibadan, Nigeria than Gram positives [18]. The higher proportion of the Gram-negative bacteria might be ascribed that most study participants had under gone abdominal surgery and Gram negatives are predominantly reported tobe involved in intra-abdominal procedures [19].
S. aureus 77 (35.6%) and E. coli 33 (15.3%) were the most common isolates reported in this study. Similar results reported from Jimma, revealed that the predominant organisms isolated from wound infection were S. aureus 47 (32.4%) and E. coli 29 (20%) [12]. The most common clinical samples that revealed high frequency of isolation were wound/pus 136 (63%), followed by urine 40 (18.5%) and blood 29 (13.4%). These findings are in line with other studies done in Gabon and Ethiopia [17,20]. On the contrary, the present study is different from other studies elsewhere [21][22][23]. The variations in this report may be due to host, microbial and environmental factors [24].
Among Gram-positive bacteria, S. aureus and CoNS were the most common isolates that causes nosocomial infections. This is in line with reports from India revealed that S. aureus and CoNS were commonly isolated Gram positive bacteria [25]. Of the total isolates of nosocomial infections S. aureus was the most common isolate, 35.6%. This was consistent from studies in Jimma, Nigeria and India [12,26,27]. S. aureus is a common pathogen due to its ubiquitous nature both as a normal skin flora and a pathogen in human host [28,29]. Furthermore, MRSA isolation rate in this study was 67.5%. This was higher than 55% MRSA reported by Abera et al. [30] in Ethiopia. The high rate of prevalence in the present study may be due to the fact that like many other African countries MRSA prevalence is increasing from time to time [31].
Among Gram-negative isolates, E. coli (15.3%), Klebsiella spp (13.4%), Citrobacter spp (7.0%) were commonly isolated pathogens. This is consistent with culture results of different specimens in Iran which revealed that E. coli and Klebsiella pneumoniae were the most common isolates [32].

Limitation of the study
Addressing MRSA and MDR Gram negative bacterial pathogens from culture of 260 clinical samples may be timely. However, complete characterization of resistance mechanisms and clonal relatedness was not made and considered as a limitation of this study.
Abbreviations MDR: multidrug resistance (acquired resistance to three or more different classes of antimicrobial agents); CoNS: coagulase negative Staphylococci; MRSA: methicillin resistant S. aureus; NIs: nosocomial infections.

Authors' contributions
TF was conceived the study, designed data collection, conducted data analysis and interpretation. FM, MT, SE, ME, MD and WA interpreted the results and reviewed the initial and final drafts of the manuscript. All authors read and approved the final manuscript.