Urinary bacterial profile and antibiotic susceptibility pattern among pregnant women in Rahima Moosa Mother and Child Hospital, Johannesburg

Background Urinary tract infection (UTI) in pregnancy is associated with significant morbidity for both the mother and the foetus. The aim of this study was to determine the prevalence of UTI, urinary bacterial susceptibility, and resistance patterns among pregnant women with a possible UTI at Rahima Moosa Mother and Child Hospital (RMMCH) in Johannesburg. Methods In this retrospective study, we analysed mid-stream urine culture and antibiotic susceptibility data from both inpatients and outpatients of pregnant women who attended RMMCH from January 2017 to December 2017. Data were collected from patients’ files and then matched with urine microscopy, sensitivity and culture (MC&S) results from the National Health Laboratory Services (NHLS) data. Results Urine microscopy, cultures and sensitivities were performed on 1984 specimens belonging to pregnant women who presented with symptoms and/or signs of a UTI. A total of 333 patients (16.8%) had positive bacterial cultures. Escherichia coli (E. coli) was the commonest bacterial isolate (49.9%). Other microorganisms isolated included Klebsiella species (14.4%), Enterococcus faecalis (12.9%) and coagulase-negative staphylococci (CoNS); (8.9%). Approximately 98% of organisms were sensitive to cephalexin. Cefuroxime (95.2%), ceftriaxone/cefotaxime (94.4%) and nitrofurantoin (81.9%) demonstrated antimicrobial effectiveness as indicated. Most isolates were resistant to ampicillin/amoxicillin (84.4%), Trimethoprim/Sulfamethoxazole (55.6%) and amoxicillin-clavulanic acid (50.2%). Conclusion E. coli was the commonest pathogen causing UTIs in pregnancy with Enterococcus faecalis increasing in prevalence. The choice of antimicrobial therapy in pregnancy should be determined according to sensitivity and resistance and foeto-maternal safety.


Introduction
Several changes occur during pregnancy that predispose pregnant women to urinary tract infections (UTIs). Physiological, physical, mechanical and hormonal changes result in increased urinary stasis. Altered urine composition with elevated glucose levels coupled with a short urethra (3 cm -4 cm in women) increases the predisposition to UTIs in pregnant women. 1 The prevalence of UTI in pregnancy ranges between 2% and 10% globally. 2,6,7,8 indigent women having a five-fold increased risk of acquiring UTIs. 9 Other risk factors include: increasing maternal age, high parity, reduced immune function, poor perineal hygiene, a history of recurrent UTI, diabetes mellitus, neurogenic urinary retention, anatomic or functional urinary tract abnormalities, and increased frequency of sexual activity. 2,10,11 Studies in developing countries show that UTIs are usually present at the first antenatal visit and less than 1% of women develop bacteriuria after a negative screen in early pregnancy. 2 A UTI in pregnancy contributes to significant maternal and perinatal morbidity and mortality. Maternal complications include overt pyelonephritis in 25% -40% of previously asymptomatic women as the pregnancy progresses, and in 1% -2% in those with symptomatic infections. 2,12 Other maternal complications include: anaemia, miscarriages, preterm labour, hypertension, pre-eclampsia, puerperal sepsis, chronic pyelonephritis and occasionally, renal failure. 2,8,13 Urinary tract infections are also associated with foetal growth restriction, prematurity, low birthweight and foetal death. 13,14 The causative organisms arise from the normal vaginal, perineal and faecal flora. 3,13 These include: Escherichia coli (E. coli), Staphylococcus aureus, Enterococcus faecalis, Proteus mirabilis, Klebsiella species, and Streptococcus species, among others. 2 There are numerous reports of resistance to antimicrobials by urinary tract pathogens. 14,15 Antimicrobial resistance in these organisms occurs because of broad-spectrum antibiotic abuse in humans and in animal feeds. 13,14 Antibiotic resistance is frequently observed in nosocomial settings. However, it is also becoming apparent in communityacquired UTIs, with an increasing incidence of Gram-positive cocci, for example, Staphylococci sp. and Gram-negative organisms such as Klebsiella sp. becoming more prevalent. 2,16,17 Urinary tract infections may present as acute infections and the administration of antibiotics may be necessary while awaiting microscopy, sensitivity and culture (MC&S) results to prevent and/or reduce maternal and foetal morbidity and mortality especially in low-resourced countries.
The aim of this study was to determine the antibiotic sensitivity pattern among pregnant women with symptomatic UTIs and to describe the pathogenicity and antibiotic susceptibility among the causative bacterial organisms. 18 Knowledge of the local bacterial and susceptibility patterns can guide the judicious use of empiric therapy. 19

Setting and study design
The study was conducted at the Rahima Moosa Mother and Child Hospital (RMMCH), which is a regional hospital. The hospital serves a population of approximately 200 000 women and children in three regions in the Gauteng province. 20 An average of 1700 women attend the antenatal clinic on a monthly basis, including high-and low-risk women.
In this retrospective study, we analysed mid-stream urine culture and antibiotic susceptibility data from both inpatients and outpatients of pregnant women who attended RMMCH from January 2017 to December 2017. The National Health Laboratory Services (NHLS) was approached for a list of all samples sent in the specified time period. All available files from those that were culture positive were retrieved and reviewed. The NHLS data uses a pre-defined procedure for culturing, bacterial identification, and susceptibility testing.

Urine sample collection, primary inoculation, and analysis
The samples were collected during routine clinical care, at the discretion of the treating clinician, and it is not possible to comment on the correctness of the technique or time taken to reach the laboratory. The reason for sending a sample was not always reflected in the available notes.
Urine samples were cultured on 5% blood agar and MacConkey agar using calibrated loops in a semi-quantitative assessment and incubated in aerobic conditions at 35 °C -37 °C for 18 h -24 h. 2 Isolates were identified and confirmed using standard methods including Gram staining; colony morphology on media; growth on selective media; lactose and mannitol fermentation; hydrogen sulphide production; catalase, oxidase, coagulase, and indole tests; citrate utilisation; and urease testing. Urine infection cultures were considered positive with bacterial counts ≥ 10 5 /mL. 2,21 All patients with positive urine cultures were treated. 2

Antimicrobial susceptibility testing
For reliable detection, laboratories may use conventional, quantitative susceptibility testing methods or specially developed, single concentration agar screening tests for some resistant species. 2 Antimicrobial susceptibility testing for UTIs in the laboratory is performed using two groups of antimicrobial discs for cascade reporting, a strategy recommended by the Clinical and Laboratory Standards Institute. In this strategy, the reporting of antimicrobial susceptibility test results for the second group of agents (e.g. broader-spectrum, more costly) may only be reported if an organism is resistant to primary agents within a particular drug class. If a pathogen shows resistance to all of these, the laboratory will move to the second stage for testing broad-spectrum antibiotics. 2 Identification and sensitivity testing were done if the culture was pure and growth was significant (≥ 10 5 CFU/mL). If the culture growth involved a mix of two pathogens and no isolate was dominant, or more than two types of colonies were grown, then it was reported as a mixed growth (contaminants) and clinical correlation was needed to make a determination. In such cases, no sensitivity testing was carried out. 2,22

Statistical analysis
All the data collected were managed using Research Electronic Data Capture (REDCap) electronic data capture tools hosted at a University in Johannesburg. The REDCap is a secure, web-based application designed to support data capture for research studies. 23,24 The biostatisticians at the University of the Witwatersrand (WITS) in Johannesburg assisted in the analysis phase of the study. The data were analysed using Stata® version 13.0.15. 25 Descriptive data was expressed using means with ranges and medians with standard deviations (s.d.).

Ethical considerations
Ethical clearance was obtained from the Human Research Ethics Committee (Medical) of the University of the Witwatersrand, reference number: M181070.

Results
A total of 1984 urine samples from pregnant women with suspected UTI were selected for isolation and identification of bacteria and antimicrobial susceptibility testing in both inpatients and outpatients at our centre, 333 urine samples (16.8%) were culture positive for UTI, while 1599 urine samples (80.6%) were culture negative and 52 urine samples (2.6%) were identified as contaminants and contaminated urine were determined by more than five epithelial cells on microscopic evaluation, specific gravidity > 1.035 and urine culture with polymicrobial growth -that is, urine culture that contain more than one organisms ( Figure 1).
In all, 324 women (97.3%) were booked for antenatal care. The ages of the study cohort ranged between 16 and 44 years, with the majority (58.9%) in the 20 and 29-year age group. There were 52 (15.6%) human immunodeficiency virus (HIV) infected women. The highest frequency of UTIs observed (n = 229, 68.8%) were from specimens taken in the third trimester. Gestational ages ranged between 26 weeks and 40 weeks and birthweights from 619 g to 4495 g. Of the 333 patients, 29.1% (97/333) were inpatients and 70.9% (236/333) were outpatients ( Table 1).
All the Klebsiella were sensitive to amikacin, ertapenem and imipenem, while nearly half of Klebsiella isolates were sensitive to different cephalosporins whereas only 20.9% were sensitive to nitrofurantoin. Enterobacter species showed 100% sensitivity to amikacin, ciprofloxacin, co-trimoxazole, ertapenem and imipenem. These cultures were most resistant to co-amoxiclav and cefuroxime. The three Proteus species were sensitive to most tested antibiotics except ceftriaxone, with a 75% reported resistance.
Acinetobacter species was cultured twice and showed resistance to ampicillin/amoxicillin, co-amoxiclav, cefuroxime, nalidixic acid and nitrofurantoin. Both were sensitive to the carbapenems.

Discussion
Pregnant women are at increased risk of developing UTI, mainly because of a shift in the position of the urinary tract and hormonal changes that occur throughout pregnancy, thus making it easier for bacteria to reach the kidney and leading to both symptomatic and asymptomatic bacteriuria. 2,26,27,28  E. coli was the most common gram-negative bacteria isolated in this study, which is in line with the findings of previous     studies, such as that by Tandan et al. and several other studies. 1,12,13,15 This finding suggests that most organisms causing UTI are from the lower gastrointestinal tract which acts as a reservoir for organisms such as E. coli. 1,9 Klebsiella species was the next most common gram-negative organism isolated in this study, accounting for 14.4% of positive cultures, and is similar to the finding in a study conducted in KwaZulu-Natal (20%). 1,29 Enterococcus faecalis (12.9%) was the most frequent grampositive organism detected and had been noted as a significant bacterial isolate from women with UTI in pregnancy in other studies. 1,30 Empiric therapy should be commenced as soon as urine samples are taken and modified once culture results become available to prevent serious morbidity. 1,31 Hence, the recommendation of nitrofurantoin as a first line drug for the treatment of UTI in pregnancy. 7,12 The University's obstetric protocol recommends the use of nitrofurantoin at a dose of 100 mg orally, 6 hourly for 5 days or cefuroxime at a dose of 250 mg orally stat (this is a non-Essential Medicines List (EML) item in SA) 32 for uncomplicated UTI and twice daily for 5 days as empirical therapy for complicated UTI. 33,34,35 The use of nitrofurantoin in pregnancy is supported by the most recent American College of Obstetricians and Gynaecologists (ACOG) Committee opinion which concluded that in the second and third trimester it was a suitable choice. In the first trimester it can be used if there are no other suitable alternatives. 36 The use of nitrofurantoin in pregnancy shows no increased risk for cardiovascular malformations, oral cleft, or craniosynostosis. 2,37 Chloramphenicol, tetracyclines, and cotrimoxazole should be avoided in pregnancy. 2,34 However, during early pregnancy chloramphenicol treatment presents little, if any, teratogenic risk to the foetus in humans. 2,38 The exposure in utero to clindamycin, doxycycline, quinolones, and macrolides are linked to organ-specific malformations whereas exposure to amoxicillin, cephalosporins, and nitrofurantoin are not associated with major congenital malformations. 2,38 Beta-lactams, vancomycin, nitrofurantoin, and clindamycin are generally considered safe and effective in pregnancy whereas fluoroquinolones and tetracyclines should generally be avoided in pregnancy. 2,39,40 There was poor documentation of repeat cultures, as well as the use of prophylactic urinary antimicrobials to prevent recurrent infections, particularly in those patients with acute  pyelonephritis. Only 50% of cases of pyelonephritis had repeat cultures either during admission or at follow up. There was a 0.45% incidence of pyelonephritis in this study, which is similar to reported ranges from 0.5% to 2.0% in the literature. 29 Urinary tract infections recur in approximately 4% -5% of pregnancies in patients with structural abnormalities of renal system and a single, postcoital dose or daily suppression with cephalexin or nitrofurantoin is an effective preventive therapy. 41 Some studies demonstrated a relationship between UTI in pregnant women and the risk of poor perinatal outcomes. 42 However, authors like Chen et al. concluded that there were no increased risks of adverse pregnancy outcomes in women, and neonates born to women with UTIs. 43

Limitations
The limitations of this study were the small sample size, missing data in the hospital file and microscopy results, symptoms at time of testing especially as symptoms of UTI and pregnancy are similar, the inherent shortcomings of a retrospective study and that it was confined to one hospital. Approximately 14% of the files were not retrieved, as some were lost. This study did not differentiate between nosocomial and community acquired infections.
Future prospective studies evaluating the impact of UTI in pregnant women in our setting using a larger sample size is recommended to address this limitation. And an improved patient record system will be beneficial to future studies.

Conclusion
E. coli, Klebsiella, Enterococcus faecalis and coagulase-negative staphylococcus were the most common microorganisms identified in this study. 2 Empirical therapy with oral nitrofurantoin and cefuroxime or parenteral ceftriaxone are appropriate. Most of these antibiotics are relatively safe to be used in pregnancy and breastfeeding. 2 The choice of antimicrobial therapy in pregnancy should be determined according to the sensitivity and resistance for foeto-maternal safety. 2