Salmonella Typhi and Salmonella Paratyphi prevalence, antimicrobial susceptibility profile and factors associated with enteric fever infection in Bahir Dar, Ethiopia

Enteric fever (EF) is caused by Salmonella enterica serovars Typhi (S. Typhi) and Paratyphi (S. Paratyphi) causing significant health problems in developing countries including Ethiopia. Thus present study aimed to determine prevalence and antimicrobial resistance profile of S. Typhi and S. Paratyphi among EF suspected patients at Felege-Hiwot comprehensive specialized hospital, Bahir Dar, Ethiopia. Hospital based cross-sectional study was conducted from March-to-May 2020. Totally, 150 patients were included conveniently. Data were collected using questionnaires by face-to-face interview. Concurrently, venous blood and stool specimens were collected and processed following standard bacteriological technique. Antimicrobial susceptibility test (AST) was performed by disc diffusion method. Logistic regression was performed to identify factors associated with EF infection. The study indicated 5.3% EF prevalence where S. Typhi accounted 75%. S. Typhi and S. Paratyphi isolates were 100% sensitive to cephalosporins but at least 83.3% showed resistance against chloramphenicol and tetracycline. At least 66.7% of isolates were multidrug resistance (MDR). Using well water for drinking (AOR = 6.22, CI 1.4–27.5) and previous EF history (AOR = 10.74, CI 2.01–55.9) were significantly associated with EF infection. Thus high bacterial prevalence and MDR isolates was observed. Therefore, health professionals should consider AST and use antibiotics with cautions for EF patient management.

Clinical specimen collection. After the interview, venous blood and stool specimens were collected aseptically from each study participant. Ten ml and 3 ml blood specimens were collected aseptically from adults and children respectively using culture bottles with tryptone soya broth (Oxoid; Hampshire UK). Besides, fresh stool specimens were collected using sterile screw capped containers. Both specimens were transported to medical microbiology research laboratory of Bahir Dar University College of Medicine and Health Sciences within two hours of collection.
Bacterial isolation and identification. Each blood culture bottle was incubated at 35 °C and observed daily for 7 consecutive days for microbial growth evidenced by presence of hemolysis, gas formation or media color change. But blood culture broth with no visual evidence of bacterial growth after 7 days of incubation was sub-cultured before it was considered as negative. Culture bottles which showed growth were opened aseptically and small amount of broth was taken using sterile wire loop and sub cultured into blood agar plate and MacConkey agar (BIOMARK Laboratories, India). The liquid stool specimen was processed directly but fecal suspension was prepared from formed stool specimen using normal saline. Four drops of fecal suspension was added into bottles containing selenite F broth (Oxoid; Hampshire UK) and incubated at 35 °C for 18 h and then sub-cultured into xylose-lysine-deoxycholate agar (XLD) (BIOMARK Laboratories, India), and MacConkey agar (BIOMARK Laboratories, India) further incubated at 35 °C for 24 h. Identification of Salmonella genus were done based on colony morphology, Gram staining and biochemical test following standard bacteriological methods. Further serovar identification of S. Typhi and S. Paratyphi were performed by automated microbiological technique using VITEK 2 system (BioMérieux diagnostics, France) 22 .
Antibiotic susceptibility testing. Antibiotic  Quality control. Before data collection, the questionnaire was pre-tested and every questionnaire was checked for completeness after collection. All culture media was prepared following manufacturer's instruction. A sample of culture media plates prepared from each batch was incubated at 37 °C for 24 h to cheek for sterility. Before inoculation, the culture media was visually inspected for any microbial growth or deterioration. Moreover, McFarland standard was used to standardize inoculums density of bacterial suspension for the AST. Furthermore, Salmonella Typhomurium ATCC 14028 and Escherichia coli ATCC 25922 standard strains were used as quality control 23 . Furthermore all methods section including sample collection, bacterial isolation and AST were performed in accordance with CLSI and WHO guidelines 22 Ethics approval and consent to participate. Ethical clearance was obtained from the Institutional Review Board (IRB) of Bahir Dar University, College of Medicine and Health Sciences protocol number 0013/2020. Moreover before data collection, written informed consent was obtained from each study participant with age greater than or equal to 18 years. Furthermore, from study participants with age less than 18 years, a written informed assent was obtained from their parents or legal guardian. In addition, all the information obtained from the study participants was registered by code to maintain confidentiality and culture positive results were communicated with responsible physician for proper patient management.

Results
Prevalence of enteric fever. Among 150 study participants, 81 (54%) and 95 (63.3%) were females and urban residents respectively. The age range of study participants were 8-80 with a mean of 34.1 and median of 32.5 age in years. The overall prevalence of culture confirmed enteric fever was 5.3%. Enteric fever was more prevalent on age group of 1-10 years (11.1%), females (6.2%) and rural residents (7.3%) than the other age groups, males and urban residents respectively. Moreover, enteric fever was more prevalent on participants who cannot read and write (7.4%) than participants who are educated (Table 1).
In the present study, the prevalence of TF (4%) was higher than PTF (1.3%). From the total 8 culture confirmed EF patients, 75% was caused by S. Typhi and 25% by S. Paratyphi A with no co-infection. Besides, 75% and 25% of Salmonella species were isolated from blood and stool specimens respectively (Fig. 1).
Antimicrobial resistant profile of S. Typhi and S. Paratyphi isolates. S. Typhi revealed the highest resistance rate for tetracycline and chloramphenicol with 83.3% for each. Similarly, all S. Paratyphi A isolates were resistant for tetracycline, chloramphenicol and amoxicillin-clavulanate. On the other hand, all S. Typhi and S. Paratyphi A isolated were susceptible for cefotaxime, ceftazidime and cefoxitin (Table 2). Furthermore, 4 (66.7%) and all isolates of S. Typhi and S. Paratyphi A were multidrug resistance (MDR) respectively.

Multivariable analysis on risk factors of EF infection.
Based on multivariable analysis, using well water for drinking and a previous history of EF were significantly associated explanatory factors for enteric fever infection. Patients who have previous history of EF infection and used well water for drinking had 16.4 and 14.9 times more chance of developing EF than those who didn't have a history of EF infection and used pipe water for drinking respectively. Though it was not significant (P = 0.365), a considerable difference in EF infection was observed among participants who didn't have toilet (9.5%) than those having toilet (4.7%). Despite higher EF infection compared with their counter parts, consuming raw meat (P = 0.402), raw vegetable (P = 0.510), street food (P = 0.573) and drinking raw milk (P = 0.569) was not significantly associated with EF infection (Table 3).

Discussion
Enteric fever is a widespread public health problem in low and middle income countries including Ethiopia 21,24 , where 88% urban and 92% rural residents don't treat drinking water and only 6% of households use improved toilet facilities 25 . In the present study, the culture confirmed prevalence of EF was 5.3%. This was comparable with previous studies in Ethiopia 26 and Bangladesh 16 which reported 4.1% and 5% respectively. Similarly the prevalence of TF (4%) in the present study was comparable with previous studies in Shashemene, Ethiopia 27 and abroad in Egypt 28 . But the prevalence of EF in the present study was higher than a study done in Ethiopia 18  www.nature.com/scientificreports/     www.nature.com/scientificreports/ time and even in consecutive years at the same geographical location 21 and seasonal variability 34 . Furthermore sensitivity differences in the laboratory detection methods used might also contribute for the difference observed. In contrast to our findings, previous studies in Jigjiga, Ethiopia 19 , Nepal 14,35 , India 17 , Nigeria 36 and Indonesia 37 reported a higher EF prevalence among febrile patients ranging from 11 to 14.1%. This might be due to the varied incidence of TF in different study areas and periods 38 . Moreover, the geographical heterogeneous nature of TF burden 21,32 might also contribute for the difference observed. Furthermore, higher prevalence of S. Typhi than S. Paratyphi was documented in the present study. Previous studies have reported comparably higher TF than PTF prevalence in Ethiopia 19,26 , Indonesia 37 , Bangladesh 16 , India 12 , Fiji 13 , Nepal 14 , India 17,30 , and Indonesia 37 . On the other hand, few studies in Ethiopia and abroad reported higher prevalence of S. Paratyphi than S. Typhi 18,36 . In a meta-analysis done on EF 24 , S. Typhi accounted 76.1% which might indicate its higher share in causing EF than S. Paratyphi.