Prevalence of bacterial uropathogens and their antimicrobial susceptibility patterns among pregnant women in Eastern Ethiopia: hospital-based cross-sectional study

The present study was undertaken at antenatal care clinics of HFSUH which is affiliated with Haramaya University College of Medicine and Health Sciences. It is found in the Harari Regional State, Harar, Eastern Ethiopia which is found 525 km from the capital city, Addis Ababa.

A hospital-based cross-sectional study design was conducted from November 2017 to January 2018. All pregnant women who had a follow-up and attended the antenatal care clinic of HFSUH, and willing to participate in the study during the study period were consecutively recruited. However, pregnant women who received antibiotics within 15 days before ANC follow-up and who were treated for another infection were excluded from the study.

A single population formula was used to calculate the sample size using the following parameter, a 14% prevalence rate of UTI among pregnant women attending the antenatal clinic of Dil Chora Referral Hospital, Dire Dewa Ethiopia [21]. Tolerable margin of error (d) = 5%, Z score for 95% confidence interval, with 10% non-response rate.

where z = Z score for 95% confidence interval, a = 1.96, P = prevalence, d = margin of error (5%).m; n = \(\frac{{{\text{z}}2{\text{ p }}\left( {1 - {\text{p}}} \right)}}{{{\text{d}}2}}\) = \(\frac{{1.962{*}0.14{ }\left( {1 - 0.14} \right)}}{{0.05{*}0.05}}\) = 185; Non-response rate = 185 ~ 10% = 18.5 ~ 19; Total sample size was = 185 + 19 = 204.

All pregnant women attending the antenatal clinics of HFSUH and that fulfilled the inclusion criteria till the required sample size attained were selected using a convenient sampling technique.

A semi-structured questionnaire (see Additional file 1: File 1) that was adopted from Derese et al. [21], and modified based on the study objective, was used for the collection of clinical/socio-demographic data. It demonstrates an identification number of the patient, age, pregnancy (gestational age) and marital status, urinary signs and symptoms, predisposing factors, and other related information. Culture and susceptibility results were also included in the questionnaire to record their culture and susceptibility test results.

After face to face interview about sociodemographic and clinical data related to UTI was completed with each pregnant woman, they were sent to the hospital laboratory with their request form. Then, all study participants requested to bring ∼10–20 mL clean-catch midstream urine specimens after urine sample collection techniques were explained by trained medical laboratory professionals. Accordingly, all study participants have thoroughly washed their hands with water and drying. And they told to separate their labial by one hand for cleaning around the urinary opening area with water in the backward direction and dry thoroughly. Then, by separating the labia, the first 20–30 ml urine was voided in the toilet, and ∼10–20 mL clean- catch midstream urine samples were collected into a sterile universal utensil. Finally, they closed the cap of the urine bottle immediately considering not to touch either the edge of the bottle or the inner side of the bottle cap. The bottle was labeled with the unique client’s identification (ID) number, date, and collection time by medical laboratory professionals.

We could not conduct urine sample processing immediately within 30 min of its collection since sample collection and urine processing for culture were not conducted at the same site. Hence, to maintain the quality of the samples collected, we stored them in the refrigerator of the central medical laboratory of HFSUH until we sent all together within two hours of its collection to the microbiology department of Harari Health Research and Regional Laboratory which was around 150 m away from the study area. The samples were stored in the cold box during transportation to the microbiology department. Then, once each sample reach the microbiology department, it was processed immediately for culture. Four samples were excluded (i.e.1 contaminated with stool and 3 not refrigerated properly) while the rest 200 samples were processed for culture after proper mixing. For more, please see Fig. 1.

Once each urine sample was delivered to the microbiology laboratory; it screened for the presence of bacterial agents according to the standard procedures for the diagnosis of bacterial UTI and isolation/identification of bacteria from urine [28]. Hence, the well-mixed, as well as non-centrifuged specimens, were inoculated using a standard wire loop that can deliver 1µL of urine specimen onto Cystine Lactose Electrolyte-Deficient (CLED) and MacConkey agar (with and without crystal violet, Titan, Biotech LTD, India) through a surface streak procedure. CLED was used because it gives consistent results and allows the growth of both Gram-negative and Gram-positive bacterial pathogens.

The plates were put in an incubator at 37 °C overnight (18–24 h) under aerobic conditions. After 24 h of incubation, the plates were examined for the presence or absence of the bacterial growth macroscopically (unaided eyes). Then those plates with positive cultures were examined macroscopically to evaluate the colony appearance and its size. Subculturing of not well-isolated colonies was performed using a sterile wire loop to ensure pure cultures. Then, the confirmed colonies were counted from CLED and multiplied by 1000 to determine the number of bacteria per milliliter (CFU/ml) of the original urine specimen. Plates with significant bacteria, pure growth, and colonies ≥ 10⁵ CFU/ml, were further subjected to identification and susceptibility testing.

A cut point of ≥ 10⁵ CFU/ml was used to define UTI. The bacterial isolates were identified using standard biochemical tests [29, 30]. Accordingly, Gram-negative bacteria were identified by performing a series of biochemical tests, namely oxidase, indole, Simmons citrate agar(citrate utilization), lysine decarboxylation, lactose fermentation, gas, and Hydrogen Sulfide (H₂S) production as well as motility tests. Gram-positive bacteria were also identified based on their catalase test and coagulase tests. After identification of the specific bacteria was completed, the antimicrobial susceptibility test of isolated pathogens was followed ensuring that appropriate and adequate antibiotics were provided.

Antibiotic susceptibility testing was done by using Kirby Bauer (disk diffusion) method on Muller Hinton agar according to the standard procedures [29, 30]. The suspension of the bacterium was prepared by picking a pure colony with a sterile wire loop, suspended and emulsified into a test tube containing 5 ml of nutrient saline, and then mixed gently until the uniform suspension was formed. Standard inoculums were adjusted to 0.5 McFarland which yield a uniform suspension containing 10⁵–10⁶ cells/mL.

Using sterile applicator cotton-swab, a sample of the standardized inoculum was taken and streaked back and forth on the entire surface of the dried Mueller–Hinton agar plate (Biomark, Laboratory, India). The streaking procedures were repeated and the plates were turned at a 60° angle between each streaking to ensure even distribution and then the inoculums were allowed to dry for 5–15 min with the lid in place. Then, using sterile forceps, the selected antibiotics disks were applied to the plates. The antimicrobial agents used for susceptibility testing were: augmentin, (AMC, 30 µg), ampicillin (AMP, 10 µg), chloramphenicol (C, 30 µg), ciprofloxacin (CIP, 5 µg), norfloxacin (NOR, 10 µg), trimethoprim + sulphamethazole (SXT, 25 µg, 1.25/23.75 µg), amikacin(AK, 10 µg), gentamicin (GEN, 10 µg), ceftriaxone (CRO, 30 µg), ceftazidime (CAZ, 30 µg), nalixidic acid (NA, 30 µg), and nitrofurantoin (F, 300 µg). All antimicrobials used for the study were Oxoid Ltd. Bashing store Hampshire, UK products.

The discs were placed at least 24 mm away from each other and 15 mm from the edge to avoid the overlapping of the zone of inhibition and pressed down to ensure complete contact with the agar surface. The plates were inverted upside down and incubated aerobically at 37 °C for 18–24 h. The diameter of the zone of inhibition around each disc was measured to the nearest wholly millimeter (mm) from the back of the plate by using reflected light and a digital caliper. Then, the bacterial isolates were classified as susceptible (S), intermediate (I), or resistance (R) by comparing them against the zone of inhibition diameter of interpretative standards as indicated in the Clinical and Laboratory Standards Institute (CLSI) guideline [29, 31] (For more see Additional file 2: File 2). Bacterial isolates resistant to three or more antimicrobials belonging to different structural classes were classified as MDR [32]. The overall procedure layout is depicted in Fig. 1.

The semi-structured questionnaire was first prepared in English language and translated to local Amharic language, back-translated to English to ascertain its consistency, and pretested on 5% of the sample size at HFSUH. Before the actual work commenced, practical refreshment training was given to data collectors (i.e. medical laboratory technicians) on the aim of the study and the methodology part regarding urine sample collection and processing for bacterial isolation, identification, and AST for half-day by investigators.

Participants were oriented on how to collect self-midstream urine samples by trained data collectors. The specimens were stored in the refrigerator of HFSUH Central Medical laboratory with controlled temperature, and transported to the Harari Health Research Center and regional laboratory within 2 h of its collection in the cold box, and immediately processed by the Medical Microbiologist. Each culture media was checked daily to observe crack contamination, and decolorization formed during the culture process. Standard operating procedures and the manufacturer’s instruction manual were strictly followed for isolation and identification, and the drug susceptibility testing of the isolated pathogen. Moreover, the reference strains such as Escherichia coli (American Type Culture Collection (ATCC) 25922), Staphylococcus aureus (ATCC25923), and Pseudomonas aeruginosa (ATTC 27853) were used as a control to check the culture performance and drug disks. All the standard strains were obtained from the Harari Health Research Center and regional laboratory.

The collected data were checked for completeness, coded, entered, and cleaned using Epi-data version 3.1, and was exported and analyzed using SPSS version 25 software. Descriptive statistics such as frequency, percentage, and cross-tabulation were used to present the findings in the form of a graphs and tables.

Among 204 samples collected, only 200 samples were processed for culture and susceptibility test which gives a 98.04% response rate. The majority, 72(36%), of the study participants, were aged between 23 and 27 years. More than 89% of the study participants lived in an urban area. Moreover, 61(30.5%) had educational level of secondary cycle (9–12 grade), 185 (92.5%) married, and 61(30.5%) were civil servant. Their obstetric data showed that 79% had gravidity of 1–3 times. Among the study participants, 27% had a history of UTI, 6.5% underwent obstetric and gynecologic surgery, and 4% reported prior use of an indwelling catheter (Table 1).

Among the 200 urine samples investigated, 31 of them grew bacteria in culture with significant bacterial growth, which gives an overall prevalence of 15.5%. A higher rate of UTI, 45.2%, was reported among age groups 23–27 years and followed by 32.5% among the age group of 28–32 years. Among the study participants with significant bacteriuria, 27/31(87%) lived in an urban area. Those study participants, who had a history of catheterization, prior Obstetric and Gynecologic Surgery and history of UTI showed that 4/8(50%), 5/13(38.5%), and 19/54(35.2%) prevalence of UTI respectively. Out of the 200 midstream urine samples processed, 13/62(20.9%) and 18/138(13%) had shown significant bacteriuria among symptomatic and asymptomatic participants respectively. Unfortunately, among 31 study participants with positive urine culture (i.e. significant bacteriuria), 18(58.1%) of them did not complain of any symptom of UTI on presentation (see Table 2).

In our study, out of the total midstream urine sample culture test, a total of (n = 31) bacteria were isolated with five different species. Most of the isolated bacteria were Gram-negative organisms 28(90.3%) while only 3(9.7%) were Gram-positive Staphylococcus aureus. Escherichia coli was the most frequently isolated bacteria 14(45.2%) that followed by Proteus spp.7(22.6%), Klebsiella pneumoniae 5(16.1%), Staphylococcus aureus 3(9.7%), and Pseudomonas aeruginosa 2(6.4%) (see Fig. 2).

Among the identified bacterial uropathogens, 30 (96.8%) of them were susceptible to amikacin, 28(90.3%) were susceptible to nitrofurantoin, 26(83.9%) were susceptible to gentamicin, and 22(71%) showed susceptibility to chloramphenicol. However, high rates of resistance were observed to ampicillin 18(58.1%), amoxicillin-clavulanate 15(51.6%), cotrimoxazole 15(51.6%), and nalixidic acid 13(42.0%). For more see Table 3 and Fig. 3.

Concerning specific isolate, all isolated Escherichia coli (100%) were susceptible to amikacin. Besides, 13(92.3%) and 12(85.7%) of Escherichia coli were also susceptible to gentamicin and nitrofurantoin respectively. However, 7(50%) of isolated Escherichia coli showed resistance to ampicillin, while 6(42.9%) of them developed resistance to each amoxicillin-clavulanate and co-trimoxazole.

All Proteus spp. 7(100%) isolated were susceptible to amikacin and nitrofurantoin while 5(71.4%) of them were susceptible to gentamicin. On the contrary, 5(71.4%) of the Proteus spp. were resistant to each cotrimoxazole and nalixidic acid. In addition, all isolates of Klebsiella pneumoniae also showed susceptibility to amikacin, gentamicin, and nitrofurantoin though still majority of them were resistant to ampicillin 4(80%) and amoxicillin-clavulanate 3(60%). The two (100%) Pseudomonas aeruginosa identified were also resistant to each ampicillin and amoxicillin-clavulanate. However, all of them showed susceptibility to ceftriaxone.

Similar to the majorities of identified Gram-negative bacteria, the only Gram-positive Staphylococcus aureus, 3(100%), were also susceptible to kanamicin, gentamicin, and nitrofurantoin. Two (66.6%) of the Staphylococcus aureus were showed resistance to each co-trimoxazole and ceftazidime (see Table 3).

Among the total isolates (n = 31), multi-drug resistance was observed in 16 (51.6%) of all bacterial uropathogens. Conversely, 2(100%) of Pseudomonas aeruginosa, 5(71.4%) of Proteus spp, and 3(60%) of Klebsiella pneumoniae were developed multi-drug resistance. In general, MDR features are presented in Fig. 4.

Among the isolated bacteria, 6(19.4%) of them were resistant to three structurally different antimicrobials while 5(16.1%), 4(12.9%), and 1(3.2%) of them were resistant to four, five, and six structurally different antimicrobials respectively. Only 8 (25.8%) of the bacterial isolates were susceptible to all antibiotics (see Table 4).