Nontyphoidal Salmonella and Their Antimicrobial Susceptibility among Diarrheic Patients Attending Private Hospitals in Addis Ababa, Ethiopia

Nontyphoidal Salmonella (NTS) is one of the major causes of bacterial foodborne infection. It is mainly manifested by self-limiting gastroenteritis in healthy individuals but can also cause severe complications including blood stream infection and mortality. The emergence of multidrug-resistant strains of Salmonella is becoming a global public health concern. This study is aimed at estimating the prevalence of Salmonella, identifying serotypes involved, and investigating antimicrobial susceptibility of the isolates among diarrheic patients attending private hospitals in Addis Ababa. We collected a total of 298 stool samples from diarrheic patients attending five private hospitals in Addis Ababa and isolated Salmonella according to standard microbiological techniques; the isolates were serotyped using slide agglutination and microplate agglutination techniques. Antimicrobial susceptibility test of the isolates was carried out using Kirby-Bauer disc diffusion assay according to Clinical Laboratory Standards Institute guidelines. Fourteen stool samples (4.7%) were positive for Salmonella, and Salmonella Kiambu was the most dominant serovar (n = 7, 50%) followed by S. Saintpaul (n = 4, 28.6%) and S. Haifa (n = 2, 14.3%). Three (21.4%) of the isolates were resistant to sulfisoxazole and tetracycline each and 2 (14.3%) to ampicillin. Resistance to two antimicrobials was detected only in 2 (14.3%) of the isolates, and none of the isolates were resistant to more than two antimicrobials. In conclusion, the current study showed low prevalence of NTS in diarrheic patients attending private hospitals in Addis Ababa. Although multidrug resistance to several antimicrobials was not detected in the isolates, prudent use of antimicrobials is recommended to guaranty the long-term use of the available antimicrobials.


Introduction
Infection with NTS causes significant morbidity and mortality in humans and animals globally. Infection with NTS causes symptoms that can range from mild self-limiting gastroenteritis to severe invasive septicemia and can occasionally be fatal. Most of the infections caused by nontyphoidal Salmonella serovars are characterized by self-limiting acute gastroenteritis and diarrhea [1][2][3]. In addition to illness and death inflicted by NTS serovars, it was estimated to cause 4 million disability adjusted life years (DALYS) accounting for 22.2% of DALYS due to diarrhea causing agents that leads to huge economic loss [4]. In human, NTS infection usually causes foodborne outbreaks [3]. The common sources of infections are various animal products such as poultry, dairy, and pork [5]. Salmonella can also be transmitted from person to person, domestic animals like dogs, cats, and rodents, and consumption of contaminated products like sprouts, tomatoes, fruits, peanuts, and spinaches [5][6][7].
For decades, antimicrobials have reduced morbidity and mortality due to various bacterial infections worldwide. However, easy access to antimicrobials without prescription and extensive misuse both in public health and animal health sectors has led to emergence and spread of resistant strains of pathogenic bacteria [8,9]. The emergence and rapid spread of antimicrobial resistance (AMR) in Salmonella has become a worldwide concern. In addition to resistance to first-line antimicrobials like third-generation cephalosporins and fluoroquinolones, plasmid borne resistance to colistin (the last resort antimicrobial for treatment of multidrug gram negative bacterial infections) mediated by different variants of mcr gene has been reported in Salmonella isolates from different countries [10]. An increasing number of treatment failures are often linked to multidrug resistant (MDR) Salmonella strains [11]. MDR Salmonella isolates have been associated with high risk of invasive infection, an increased risk of death, and prolonged illness as compared to infections caused by susceptible strains [12].
In Ethiopia, there are studies that indicated occurrence of Salmonella and high rate of antimicrobial resistance in both public health and veterinary sectors. For example, a study that investigated diarrheic patients attending health care facility in Gondar reported that 1.2% (4/372) of the stool samples were positive for Salmonella [13]. The isolates detected in this study were resistant to tetracycline (100%), amoxicillin (100%), and ampicillin (75%). Other study conducted in Nekemte referral hospital reported Salmonella in 2.1% (9/422) of the diarrheic study participants, and all of the isolates were resistant to amoxicillin [14]. In a study conducted in Southern Ethiopia, Salmonella was detected in 1.0% (2/204) of diarrheic patients and both isolates were resistant to ampicillin and gentamicin [15].
Most of the previous studies in Ethiopia were conducted on patients attending public health facilities, and little data is available on prevalence and antimicrobial susceptibility of Salmonella isolates from patients attending private health facilities. Due to difference in income, educational status and exposure to different predisposing factors, rate of Salmonella infection and antimicrobial susceptibility status of isolates might be different in people attending private health facilities compared to those attending government-owned health facilities. Estimating the prevalence of Salmonella among diarrheic patients and establishing antimicrobial susceptibility status of isolates combined with investigation of susceptibility of isolates to antimicrobials are of paramount importance to advice clinicians on appropriate management of Salmonella infection. Results from such study will also strengthen epidemiological knowledge of the disease that could help policy makers while planning interventions for the at-risk populations.
The objectives of this study were therefore (a) to estimate the prevalence of Salmonella among diarrheic patients attending private hospitals in Addis Ababa and to identify serotypes involved, (b) to investigate factors that can potentially be associated with Salmonella positivity, and (c) to assess antimicrobial susceptibility of Salmonella isolates.

Materials and Methods
2.1. Study Area, Context of the Study, and Study Design. The current study was conducted in Addis Ababa, the capital city of Ethiopia. The city lies at an elevation of 2,355 meters above sea level (m.a.s.l) and located at 9°1 ′ 48 ″ North and 38°44 ′ 24 ″ East, and it has a subtropical highland climate, with average annual temperature of 16.3°C and 1089 mm annual rainfall. Administratively, Addis Ababa was divided into 10 subcities, namely, Addis Ketema, Akaki-Kality, Arada, Bole, Gulele, Kirkos, Kolfe-Keranyo, Lideta, Nifas Silk-Lafto, and Yeka, when this study was conducted. According to World Population Review, population of Addis Ababa was estimated to be 4,591,983 in 2019 with average population density of 5,165 individuals per square kilometer [16].
Residents of Addis Ababa receive health services both from health institutions owned by the government and by private sectors. According to Addis Ababa City Administration Food, Medicine and Health Care Administration and Control Authority, there were more than 22 higher private hospitals in Addis Ababa in 2018 (personal communication). Although seven private hospitals were randomly selected, only five hospitals located in 4 of the 10 subcities in Addis Ababa agreed to participate in the study. Hence, diarrheic patients were recruited from these five private hospitals, namely, Bethezatha located in Kirkos subcity, Ethio Tebib in Addis Ketema subcity, iCMC and Migbaresenay Hospitals located in Yeka subcity, and Teklehaimanot Hospital located in Arada subcity. A cross-sectional study design was used to estimate prevalence of Salmonella in diarrheic patients by examining human stool samples and to determine antimicrobial susceptibility of the isolates.

Stool Sample Collection and Interviewing of Study
Participants. Stool sample collection and patient interviews were conducted from September 2018 to June 2019. Two hundred and ninety-eight stool samples were collected from diarrheic patients attending five private hospitals during the study period. Screw-caped clean plastic containers were used for the collection of stool samples. The samples were transported to Microbiology Laboratory of Aklilu Lemma Institute of Pathobiology in an icebox within 3-4 hours of collection. Various patient-related data including age, sex, history of recent antimicrobial usage, marital status, occupation, educational status, hand washing habit, and habit of consumption of raw vegetables and raw meat were collected directly by interviewing the study participants using structured questionnaire to investigate possible factors associated with occurrence of Salmonella and antimicrobial resistance.

Isolation, Identification, and Serotyping of Salmonella.
Salmonella isolation was conducted according to Global Foodborne Infections Network, laboratory protocol [17]. Briefly, 1 g of stool sample was suspended in 9 ml of buffered peptone water (Himedia, India) and incubated for 24 hours at 37°C. Then, 100 μl of the suspension was transferred to 10 ml of Rappaport Vassiliadis enrichment broth (RVB), (Oxoid, UK) and incubated for 24 h at 42°C. Suspension of 1 ml of each sample was also transferred to 9 ml of Muller-Kauffmann-Tetrathionate broth (Himedia, India) and Selanite-F broth (Becton-Dickinson, USA) and incubated for 24 h at 37°C. Samples from these three enrichment broths were streaked on to Xylose-Lysine Deoxycholate Agar (Himedia, India), and the plates were incubated at 37°C for 24-48 hours.
Biochemical tests were conducted for presumptive Salmonella colonies using Urea, Triple Sugar Iron Agar, Lysine 2 BioMed Research International Iron Agar, and Citrate as described elsewhere. Typical Salmonella colonies were confirmed using genus specific PCR as described previously [18].  (Table 3).

Serovar Distribution and Antimicrobial Susceptibility of Salmonella Isolated from Diarrheic Patients. Four different
Salmonella serovars were detected in the current study. Salmonella Kiambu (n = 7) was the most frequently isolated followed by S. Saintpaul (n = 4), S. Haifa (n = 2), and S. Enteritidis (n = 1). The findings of antimicrobial susceptibility test showed that 3 isolates (21.4%) were resistant to sulfisoxazole, 3 isolates (21.4%) were resistant to tetracycline, and 2 isolates (14.3%) were resistant to ampicillin. Three out of 7 S. Kiambu isolates (42.9%) were resistant to sulfisoxazole. Resistance to tetracycline was recorded in the three serovars except S. Enteritidis. All Salmonella isolates were susceptible to amikacin, cephalothin, amoxicillin+clavulanic acid, chloramphenicol, ceftriaxone, ciprofloxacin, gentamicin, streptomycin, and sulfamethoxazole+trimethoprim (Table 4). Coresistance to ampicillin and tetracycline was detected only in 14.3% (2/14) of the isolates, one from S. Kiambu and the other from S. Saintpaul. Majority of the isolates were susceptible to most of the antimicrobials investigated, and MDR to over two antimicrobials was not detected.

Discussion
In the current study, prevalence of Salmonella among diarrheic patients was 4.7% which is lower than what was reported in a previous study (7.2%) in Addis Ababa among patients attending government-owned health centers but higher than 1.1% reported from a study in Gondar, North Ethiopia [13] and diarrheic children in Ambo town (1.2%) [22]. Prevalence of Salmonella in the current study is in close agreement with previous reports in Addis Ababa (4.5%) [23] and 5.4% prevalence among HIV positive and 4.5% among HIV-negative diarrheic patients in Dessie town, North Ethiopia [24]. Meta-analysis of works done in Ethiopia showed 5.7% prevalence of Salmonella in diarrheic adult patients [25]. The difference in isolation rate of Salmonella might be explained by difference in socioeconomic status of the diarrheic patients which could be expressed through difference in life style like hand washing and other hygienic practices. Since health services in private hospitals are expensive in Addis Ababa as compared to government hospitals, better ability to pay might be main driving force to attend private hospitals. Most previous studies in Ethiopia recruited study participants from outpatient clinics of government-owned health centers [24][25][26][27]. It is believed that only patients with good income can afford to visit private hospitals. In addition, season of sample collection, geographic location, and habit of consuming raw vegetables and raw animal products and illness due to other enteric pathogens could affect the rate of detection of Salmonella from diarrheic patients.
In the current study, none of the patient-related factors were shown to have statistically significant association with detection of Salmonella from stool of diarrheic patients unlike previous studies where consumption of raw vegetables and stool being watery [26], and poor hand washing practices and consumption of raw/uncooked foods were associated with detection of Salmonella in diarrheic patients [28]. The reason why we did not observe such association in this study could be due to small sample size and better practices of cleaning raw vegetables before consumption among study participants.
The most prevalent Salmonella serovar in the current study was S. Kiambu followed by S. Saintpaul. In the previous studies conducted in Addis Ababa, S. Typhimurium was the dominant serotype followed by S. Virchow [26] whereas these serovars were not isolated in the current study. Salmonella Concord and S. Typhimurium were reported to be the dominant serovars in diarrheic patients in Addis Ababa in other previous studies [27,29]. Previous studies in other sub-Saharan African countries also showed dominance of S. Typhimurium [30,31]. Different serovars of Salmonella are known to be commonly detected in a given area at different times depending on the type of serovars of Salmonella circulating in food animals and food sources in a study area. Epidemiology of NTS is shown to be characterized by temporal variation in dominance of certain Salmonella serovars at a time followed by replacement with other serovars [32].
The present study revealed that irrational use of antimicrobials such as failure to finish the whole course of antimicrobials, taking antimicrobials without prescription, and sharing antimicrobials with relatives and or friends were the commonly observed behavior of the study participants. This finding agrees with previous studies conducted in Ethiopia [33] and elsewhere [34]. Such irrational use of antimicrobials, particularly purchase of antimicrobials without prescription, low level of finishing prescribed antimicrobials, and low level of awareness about antimicrobial resistance contributes to emergence and spread of antimicrobial resistance [35]. Hence, special attention should be paid to increase awareness of the community on rational use of antimicrobials.  Relatively low proportions of the Salmonella isolates were resistant to sulfisoxazole, tetracycline and ampicillin in this study. Previous study in Ethiopia also showed similar low rate of resistance to tetracycline (13.4%) and ampicillin (14.3%) whereas high rate of resistance to sulfisoxazole (38.9%) was reported [26]. The finding of the current study agrees with previous studies from Greece [36], Libya [37], and England [38]. The current finding is however very low compared to previous reports in Ethiopia [39,40] where high rate of resistance was recorded in isolates from diarrheic patients and food handlers, respectively. Similarly, unlike previous studies conducted among diarrheic patients in Addis Ababa that reported 3% resistance to ceftriaxone (third-generation cephalosporin) and 4.5% to ciprofloxacin [26], resistance to these antimicrobials was not detected among Salmonella isolates tested in the current study. This previous study reported MDR to 5 or more antimicrobials in 25.4% of the isolates. This could be either due to low sample size or relatively better understanding of the study population in the current study on rational use of antimicrobials and may not be exposed to frequent empirical antimicrobial therapy. Moreover, strains of Salmonella isolated from patients in the current study might have originated from reservoir hosts or food sources where there is minimal exposure to antimicrobials. Trend analysis of antimicrobial resistance among S. Typhimurium isolates from animal meat and humans in China for 20 years showed fluctuation over time [41]. The limitation of this study is low number of samples investigated and the number of samples collected was not equally distributed across each hospital due to low number of diarrheic patients coming to some of the hospitals during the study period.

Conclusion
The current study showed low prevalence of Salmonella in diarrheic patients attending private hospitals in Addis Ababa. Although MDR to several antimicrobials was not detected in the isolates, the observed resistance to some of antimicrobials suggests the need for prudent use of antimi-crobials to protect the long-term use of the available antimicrobials.

Data Availability
All data generated or analyzed during this study are included in this published article.

Conflicts of Interest
The authors declare that they have no competing interests.