Introduction

Human immunodeficiency virus/Acquired immune deficiency syndrome (HIV/AIDS) infection which damages mucous membrane of intestine, baring the host to antimicrobial resistant infections [1]. Antimicrobial resistance including ESBL production is of significant impact globally mainly in the developing countries [2]. Moreover; the burden of ESBL among healthy people is increasing worldwide with an estimated of 5% rise per year [3]. More than two folds increased prevalence of ESBL producing Enterobacteriaceae in French children between 2010 and 2015 has grown from 4.8 to 10.2% [4]. In addition, a review by Lewis et al. in sub Saharan Africa reported a pooled estimate of ESBL carriage as 18%, indicating the highest burden in the region than reports from US and Europe (3.4 to 7.3%) [5].

ESBL enzymes confer resistance to penicillin, cephalosporin and monobactam [6]. Even some reports indicate the presence of carbapenem resistance [7]. Bacteria produce β-lactamases, a potent family of enzymes that breakdown β-lactam ring of antibiotics making it ineffective and become a major risk in an increasing burden of resistant infections [6]. Multidrug resistance (MDR) was defined as resistance to a minimum of one drug in three classes of antimicrobial agents [8].

East and South Africa region is the largest HIV burden area of the world. Ethiopia, as part of this region contributed to an estimated of 722,248 HIV carriers 2017 [9] and 62,000 children living with HIV/AIDS were reported in 2016 [10]. The immune status of HIV patients is weakened and are at an increased risk of infections, hospitalization and antibiotic consumption than HIV free individuals [11]. Health care associated infection, mechanical ventilation, use of invasive medical devices, prolonged hospital stay and antibiotic use are among factors contributing to increased burden of ESBL producing Enterobacteriaceae [6, 7].

Intestinal carriage of ESBL producing Enterobacteriaceae can cause infections with increased hospital stay, loss of treatment options and associated healthcare costs even leads to death [12]. Hence, the problem needs an emphasis, investigation of ESBL carriage is considered as one of the targets that help to fight antimicrobial resistant infections particularly in resource constrained settings [13]. Data on the carriage rate of ESBL producing Enterobacteriaceae in the general population is very limited in Ethiopia. In spite of their high risk to multiple infections and resulting exposure several antibiotics, the problem among children living with the HIV is overlooked in the country. Hence, this study was aimed to investigate the intestinal carriage rate of ESBL among HIV infected children at the University of Gondar comprehensive Specialized Hospital.

Methods

Study area and population

Institutional cross-sectional study was conducted at the University of Gondar comprehensive Specialized Referral Hospital (UoGCSH) antiretroviral therapy (ART) clinic from January to April 2020. The hospital is found in Gondar town, Amhara region Ethiopia. The town is found at 740 km in the northwest of the capital city, Addis Ababa. The University of Gondar Specialized Referral Hospital is the teaching hospital which provides teaching activities to medical and health science students and the oldest academic institution in Ethiopia. It provides medical, surgical, psychiatric, and many other services to more than 7 million people of the Gondar province and the neighboring regions. The hospital has both inpatient with more than 512 beds and outpatient services. It also provides HIV/AIDS intervention activities like free diagnosis, treatment and monitoring in its ART clinic.

Population

All HIV-infected children who were attending at UoGSRH ART clinic were the source population while, all HIV infected children who visit UOGSRH ART clinic during the study period were our study population. All HIV infected children who are under 15 years old were included in the study while children without legal guardian or unaccompanied children were excluded from the study.

Socio-demographic and clinical data collection

A pretested questionnaire was employed to collect socio-demographic data of the study participants including age, gender, residence, educational status, family size, family income, family occupation and food habit by face to face interview. In addition, clinical data of the study participants such as, history of invasive medications, World Health Organization (WHO) disease stage of HIV, opportunistic infections, presence of fever and diarrhea, HAART experience and type of HAART, duration of HAART and recent history of antibiotic use were collected by reviewing the medical record of HIV infected children. Anthropometric measurements including weight and height were measured by digital scale. The data collection was investigator administered on site to all participants and/or their guardians at the ART follow up clinic of UoGCSH.

Laboratory procedures

Fresh stool specimen was collected using a codded clean leak proof plastic cup and transported to Medical microbiology laboratory for culture within 2 h of collection. The collected stool specimen was inoculated on MacConkey agar medium and then incubated at 37℃ for 24 h for selective growth of Gram-negative bacteria and lactose fermentation characteristics. Biochemical tests were performed for species identification. Once the species of Enterobacteriaceae were identified, antimicrobial susceptibility test (AST) was performed using Kirby-Bauer Disk Diffusion susceptibility test method on Mueller Hinton agar (MHA). Antimicrobial agents such as ampicillin, amoxicillin-clavulanate, Trimethoprim-Sulfamethoxazole, chloramphenicol, cefixime, cefoxitin, ceftazidime, cefotaxime, tetracycline and ciprofloxacin were used for AST. Discs were selected based on their availability in the local treatment following CLSI-2019. Isolates were also screened for ESBL enzyme production by applying Cefotaxime 30 µg and Ceftazidime 30 µg discs. Phenotypic confirmation of ESBL production was performed using combination disc diffusion method. After screening, simultaneous application of Cefotaxime 30 µg and Ceftazidime 30 µg with their respective combination with clavulanate 10 µg (Cefotaxime–clavulanate 30–10 µg, Ceftazidime–clavulanate 30–10 µg) was used for confirmation following the guideline of Clinical Laboratory Standards Institute (CLSI-2019). Susceptibility was done by preparing suspension of pure isolates comparable to 0.5Mcfurland standard. The difference in their zone of inhibition were measured after inoculation on MHA and incubation aerobically at 37˚C for 16-18hrs. A change in diameter zone of 5 mm and above for either cefotaxime or ceftazidime or both from their combined form was reported as ESBL producing isolate. Klebsiella pneumoniae ATCC 700,603 (positive control) and Escherichia coli strain ATCC 25,922 (negative control) were used for quality control (CLSI-2019) [14].

Data analysis

Data were entered to epi-data version 4.1 and exported to statistical packages for social sciences (SPSS) version-20 for analysis. The results were presented in frequency and percentage through table and text. Univariate and multivariate logistic regression were used to assess the association between the independent variables and the occurrence of ESBL producing Enterobacteriaceae. A variable with P-value of < 0.05 was considered as statistically significant.

Results

Sociodemographic variables of the study population

A total of 161 study participants were recruited in the study with male to female ratio of 1:1.09. Majority of participants were above 10 years (77%), urban residents (87%) and in the WHO HIV stage I (96.3%). Moreover; 46.6% of participants had stunted growth, 34.8% had viral load of > 1000 copies/ml and 90.1% had been taking HAART for at least a year (Table 1).

Table 1 The Sociodemographic and clinical characteristics of HIV infected children at the University of Gondar Comprehensive Specialized Hospital, 2020
Full size table

Distribution and antimicrobial resistance profile of Enterobacteriaceae

Among 161 study participants who brought stool specimen, there were 186 Enterobacteriaceae isolated from stool culture. E. coli was the most common isolate 59.7% followed by K. pneumoniae 16.13%. All of isolates were highly resistant to Amoxicillin (95.1%) and Ampicillin (85%). Specifically, more than 96% of E. coli and K. pneumoniae were resistant to Amoxicillin while, 85.6% E. coli and 90% K. pneumoniae were resistant to Ampicillin. Similarly, 47.7% of E. coli and 46.7% of K. pneumoniae were resistant tetracycline. On the other hand, 90%, 92.8%, 97.3% of E. coli were sensitive to CAZ, CHL and CXT respectively. High proportion of K. pneumoniae were resistant to CXM (90%), CAZ (93.3%) and CHL (93.3%) (Table 2).

Table 2 Antimicrobial Susceptibility pattern of Enterobacteriaceae isolates among HIV infected children at the University of Gondar Comprehensive Specialized Hospital, 2020
Full size table

Distribution of MDR and ESBL producing Enterobacteriaceae

The proportion of MDR from the total Enterobacteriaceae isolates was 71/186 (38.7%) in this study. Of the MDR isolates, 46/71 (64.8%) were E. coli followed by 8/71 (11.3%) K. pneumonia. Moreover; thirteen of the MDR isolates (twelve E. coli and one K. pneumonia) were ESBL producers. Of the total (161) participants 19.9% were carriers of ESBL producing isolates. There was no double or multiple carriage of ESBL isolates. All ESBL producing isolates were E. coli (16.2%) and K. pneumonia (3.7%) (Table 3).

Table 3 Frequency of Enterobacteriaceae and distribution of MDR and ESBL producing isolates among HIV infected children at the University of Gondar Comprehensive Specialized Hospital, 2020
Full size table

Factors associated with ESBL production

Among the factors analyzed in bivariate logistic regression, family education and history of antibiotic use had P-value of less than 0.2 and were fitted to multivariate analysis. But in multivariate analysis, only history of antibiotic use had statistically significant association with ESBL carriage (AOR, 3.23 95% CI 1.054–9.88) and P-value of 0.04 (Table 4).

Table 4 Factors associated with ESBL Production among HIV infected children at the University of Gondar Comprehensive Specialized Hospital, 2020
Full size table

Discussion

The overall carriage rate of ESBL producing Enterobacteriaceae among children living with HIV was 19.9% (95% CI 14.4–26.1). The result was comparable to reports among healthy children (16%) [15] and (22%) [16] in Cameroon, 21% Madagascar [17], (24.8%) Lebanon [18] and 16.8% Sweden [19]. On the other hand, the result was lower compared to reported results 31% in Niger [20], 34.3% Tanzania [21], 59% Central Africa Republic [22], 32.6% Guinea Bissau [23] and 49.6% Lebanon [24]. The fecal carriage of ESBL producing Enterobacteriaceae in this study was a little higher than 13.7% in Zimbabwe [25] and higher than reports among HIV negative children from different geographic regions: (10%) in Kenya [26], (5.0%) Ghana [27], (3.5%) United States of America [28], (4.6%) France [29], (4.7%) South Africa [12]. Geographic variation, variation in the method of ESBL detection or measures taken for the prevention of antimicrobial resistance might be responsible for the difference.

The rate of antimicrobial resistance in this study were highest against Amoxicillin (95.1%), Ampicillin (85%), Amoxicillin-clavulanic acid (46.8%), Tetracycline (46.8%) and Trimethoprim-sulfamethoxazole (43.5%). However; lowest resistance to ceftazidime (91.9), cefoxitin (91.9%) and chloramphenicol (94.1%) was observed. The result concords to a report from Arba Minch, Ethiopia [30]. This high proportion of resistance indicates isolates could probably adapt themselves to the commonly prescribed antibiotics. Trimethoprim-sulphamethoxazole is a prophylactically prescribed to HIV infected children in Ethiopia as per WHO recommendations to resource limited settings [31]. Comparable resistance to trimethoprim-sulphamethoxazole was reported in Nepal 48.9% [32], Nigeria 54% [33]. But resistance rate found in this study was lower than 91.3% in Madagascar [17]. The difference might be due to difference in population.

The overall MDR rate of Enterobacteriaceae isolates among HIV infected children were 71/186 (38.2%). The two most common MDR isolates were E. coli 46/71 (64.8%) and K. pneumonia 8/71 (11.3%). In addition, 13/71 (18.3%) MDR isolates (twelve E. coli and one K. pneumonia) were ESBL producers. This result was lower than 68.3% MDR in Addis Ababa, Ethiopia [34]. This may be due to the difference in the study population and this ESBL producing Enterobacteriaceae may transfer their resistant trait to the naïve enteric commensals.

Among the total 161 HIV infected children included in this study, 32(19.9%) were carriers of phenotypically confirmed ESBL producing Enterobacteriaceae. The result was higher than 5.3% prevalence in Uganda [35] but lower than 28.46% in Nepal [32]. In this study there was no double or multiple carriage of ESBL isolates and all of the ESBL producing isolates were E. coli 26 (16.2%) and K. pneumonia 6 (3.7%). This high resistance to multiple antimicrobial agents in addition to ESBL production is a bottle neck in the treatment of infectious diseases and pushes to the utilization of last resort drugs resulting in loss of effective treatment option [36]. Over use, frequent and intermittent use together with ease of access to the antibiotics without prescription from private pharmacies could be the possible rationale for the increased resistance.

The history of antibiotic use has shown statistically significant association. The ESBL carriage rate among children with drug use was more than 3 times compared to their counter parts (AOR 3.2, 95% CI 1.05–9.9). History of antibiotic use is also reported as a risk factor by several studies too [15, 21]. A study also has reported high family income as an independent factor associated with increased risk of ESBL carriage. Because low family income limits the rate of exposure to antibiotics which intern reduces the risk of antibiotic resistance and ESBL carriage [22]. On the other hand, low family income was reported as significant factor associated with increased risk of ESBL carriage [21]. The result of this study in contrast showed that ESBL producing isolates were distributed irrespective of the income level and no significant association was found. Family size appeared to be associated with ESBL carriage in bivariate analysis but in multivariate analysis it was not significantly associated. Residence was reported as a factor by a study done among Hospitalized patients at Arba Minch, Ethiopia [30]. But it didn’t show statistically significant association with ESBL carriage in this study. In addition, in this study, age was not significantly associated with ESBL carriage which is in agreement with a report in Madagascar [17]. This study included only HIV infected children who had visited the hospital during the study period. Advanced techniques of ESBL detection were not used in this study.

Conclusion

Fecal carriage of MDR and ESBL producing Enterobacteriaceae among HIV infected children was considerable. The History of antibiotic use was the independent factor associated with the carriage of ESBL producing Enterobacteriaceae. Regular screening of HIV patients for the carriage of ESBL producing isolates need to be strengthened. In addition, large scale antibiotic resistance survey including healthy community could be important.