Profile of uropathogens in pregnancy over 5 years from a large tertiary center in South India

INTRODUCTION

Urinary tract infection (UTI) is usually seen as asymptomatic bacteriuria and is present in 8–10% of antenatal women.^[1] About 1.5–3% of postnatal women have UTI depending on the mode of delivery, with UTI being more common after cesarean delivery.^[2] It is known from studies that 30% of antenatal women with asymptomatic bacteriuria develop pyelonephritis. Moreover, the association of UTI and preterm birth^[3] and the consequent economic strain makes it an important entity for screening in pregnancy. Small studies from the west have described the profile of microorganisms in UTI complicating pregnancy^[4-8] and there are similar small studies from India.^[9,10] Antibacterial resistance has always been an area of concern with extended spectrum beta-lactamase (ESBL)-producing strains increasing in number, especially in India.^[11-14] Thus, the aim of our study was to assess the prevalence of significant UTI in pregnancy and to study the microbiologic profile and antibiotic sensitivity trend of these infections over 5 years in a large tertiary center.

MATERIAL AND METHODS

This retrospective study was done over 5 years, between 2013 and 2017 after the Institutional Review Board Clearance (Retro 11336 dated April 18, 2018, and IRB.No.11758 dated December 19, 2018). Our tertiary center in South India had more than 100,000 antenatal outpatients (OP) and around 13,000 booked deliveries per annum. During this period, spontaneous preterm labor was seen in 8% of antenatal women. The information on urine culture during this period was obtained from the microbiology laboratory data records and the hospital electronic medical record database.

The policy in the antenatal clinic was to screen for UTI using the urine dipstick method. Urine cultures were sent only if the screening test was positive, if the woman had significant urinary symptoms, if there was history of previous preterm delivery or a bad obstetric outcome, and if there was threatened preterm labor in the current pregnancy. Postnatally, urine culture was sent for woman with fever, urinary symptoms such as dysuria, and significant retention of urine. About 20% of antenatal and postnatal women had urine cultures done. A colony count of more than 10⁵ colony-forming units (CFU)/mL was considered as significant bacteriuria and 10³–10⁵/mL was taken as probably significant culture. Less than 10³ CFU/mL was taken as a negative culture.

Recurrence of UTI was defined as two or more occurrences within 6 months.^[15]

Urine specimens received were processed within 1 h by the semi-quantitative culture method.^[16] Gram stain was performed as an initial assessment of the uncentrifuged urine specimen to grade the presence of inflammatory reactive cells, epithelial cells, and bacteria and to determine the quality of the specimen.^[17-19] Those found to be smear negative were processed by the automated culture method (HB and L Alifax, Italy) with a limit of detection of 350 CFU/ml. The routine semi-quantitative culture was performed with 10 µl of the specimen on the sterile and quality passed 7% sheep blood agar and MacConkey media, prepared in the department of clinical microbiology.^[16,18] The cultures were incubated aerobically at 37°C for up to 48 h.

Culture interpretation was done using the following criteria:

1. Amount of inflammatory response as seen on Gram-stained smears of the urine specimen.

2. The concomitant symptoms at the time of culture

3. The growth in culture.

Culture growth up to two organisms with >100,000 CFU/mL was interpreted as significant bacteriuria.^{[15,17,18,20]} Colony counts of 1000–100,000 CFU/ml up to two organisms were interpreted as probably significant bacteriuria. The presence of more than two pathogens in counts more than 1000 CFU/mL and normal urethral flora was considered as a mixture of organisms and contaminants, respectively, while colony counts of <1000 CFU/mL of a single organism were reported as insignificant bacteriuria.

The Gram-negative bacilli were identified using the oxidase test and the preliminary screening media consisting of the mannitol motility media, triple sugar iron agar, peptone water, Simmons citrate, and Christensen’s urea media and characterized further to derive a species identification. Enterococcus species was identified using bile esculin agar.

Susceptibility testing was performed by the Kirby–Bauer disk diffusion method on Mueller-Hinton agar, Difco, India, in accordance with the Clinical Laboratories Standard Institute guidelines relevant for each year.^[21] Standard American Type Culture Collection control strains were used as quality control strains for the drugs tested. The drugs included in the panels for testing were based on the bioavailability in the urinary tract, the prevalence of resistance, and the spectrum of action. Enterobacterales were tested against, first-line agents cefpodoxime (10 µg), nitrofurantoin (300 µg), cotrimoxazole (1.25/23.75 µg), ciprofloxacin (5 µg), gentamicin(10 µg), and amoxicillin-clavulanate (20/10 µg) and if resistant to the first-line against, they were test against amikacin (30 µg), piperacillin-tazobactam (100/10 µg), and meropenem (10 µg). Cefpodoxime resistance was a marker of possible ESBL-producing bacteria.

The non-fermenting Gram-negative bacilli were tested against ceftazidime (30 µg), levofloxacin (5 µg), amikacin (30 µg), piperacillin-tazobactam (100/10 µg), and meropenem (10 µg).

Enterococcus species were tested against first-line agents such as ampicillin (10 mg), nitrofurantoin (300 mg), high-level gentamicin (120 mg) plus vancomycin (30 mg), teicoplanin (30 mg), and linezolid (30 mg). If these species were resistant to the first line agents or for Staphylococcus species the susceptibility to cefoxitin(30 μg) and nitrofurantoin(300 μg) were also determined.

Data were summarized using mean standard deviation for continuous variables and frequency along with percentage for categorical variables.

RESULTS

A total of 20,203 urine specimens for culture were sent from the OP and inpatients of the obstetrics department from 2013 to 2017 [Table 1]. The OP specimens were mostly antenatal and the inpatient (IP) specimens were mostly postnatal. Among the OP specimens, 9.49% (1021/10755) had significant organisms on culture while 13.58% (1283/9448) were significant among IP specimens [Figure 1]. Among the total urine cultures sent, 9448 were postnatal samples and 10,755 antenatal samples. Of the antenatal samples, the approximate incidence of significant UTI was 1.57% (1021/64,933) and 2% (1283/64,933) in postnatal samples.

Close

Figure 1:

Study flow diagram.

Export to PPT

Among 10,755 antenatal urine cultures, 1027 were probably significant, 50 were improperly collected or mixture of organism, and 9627 had no growth. Among the 9448 postnatal samples, 879 had probably significant growth, 2214 had improperly collected or mixture of organism, and 5072 samples had no growth. Among the postnatal samples that were significant, 35% had delivered normally and 65% by operative delivery. The demographic features of women with positive cultures compared to baseline characteristics of the birth cohort are recorded in [Table 2]. The number of patients presenting with significant UTI rose steadily between 4 and 40 weeks among the antenatal patients with its frequency, being highest during the third trimester. Postnatally, the majority of significant bacteriuria cases were seen at 1 week postpartum with 90.9% occurring in hospital [Figure 2a and b].

Close

Figure 2:

Time period for the development of UTI in the (a) antenatal period and (b) postnatal period.

Export to PPT

Recurrent significant UTI was seen in 25% of women in the antenatal period and 10% of women in the postnatal period.

Uropathogens

The spectrum of microorganisms causing significant bacteriuria is described in [Figure 3]. Escherichia coli accounted for 66 and 60% of infections among antenatal and postnatal women, respectively. Enterococcus species accounted for 13% in both while Klebsiella species was 4.9 and 7.3%, respectively.

Close

Figure 3:

Uropathogens isolated in urine cultures among obstetric patients.

Export to PPT

E. coli sensitivity [Table 3] suggests that overall 68.3 and 59.2% of isolates in antenatal and postnatal period were cefpodoxime susceptible. Nearly 75% of ante- and postnatal isolates were susceptible to amoxicillin-clavulanate. Susceptibility to oral agent nitrofurantoin in E. coli isolated among antenatal and postnatal cultures was seen in 90.2% and 92.5%, respectively. Among Enterococci species, susceptibility to ampicillin was seen in 84.4 and 97.1% of isolates in ante- and postnatal samples while susceptibility to high-level gentamicin was present in 64.4 and 77.4%, respectively. This susceptibility pattern remained the same in 5-year period with no significant change.

Table 3: Cumulative antimicrobial susceptibility pattern of pathogens causing significant bacteriuria.

	Antenatal	Sensitivity (%)	Postnatal	Sensitivity (%)
3A Escherichia coli (n=1384)
Nitrofurantoin	547	90.2	679	92.5
Cotrimoxazole	386	64.5	447	60.3
Ciprofloxacin	369	62.1	48	16
Cefpodoxime	398	68.3	441	59.2
Amoxicillin-clavulanate	645	79.3	556	73.9
Piperacillin-tazobactam	28	77.7	97	84.2
Cefoperazone-sulbactam	42	89.3	103	91.1
Gentamicin	509	83	603	80.4
Netilmicin	36	97.2	109	94.7
Amikacin	47	95.9	110	94.8
Imipenem	35	97.2	112	97.3
Meropenem	37	97.3	112	97.3
3B Enterococcus species (n=330)
Ampicillin	98	84.4	172	97.1
Nitrofurantoin	95	84	168	95.4
High-level gentamicin	76	64.4	137	77.4
Teicoplanin	5	100	4	100
Linezolid	5	100	4	100
Vancomycin	5	100	2	100
3C Klebsiella species (n=318)
Nitrofurantoin	133	6	149	5
Cotrimoxazole	139	84.2	135	97
Ciprofloxacin	139	92.8	164	80
Cefpodoxime	139	84.2	165	66.7
Amoxicillin-clavulanate	139	84.2	165	69
Piperacillin-tazobactam	10	50	26	65.4
Cefoperazone-sulbactam	13	80	31	64.5
Gentamicin	60	92.8	165	81.2
Netilmicin	8	70	26	69.2
Amikacin	13	80	31	71
Imipenem	8	70	26	80.8
Meropenem	8	70	26	80.8

DISCUSSION

This data of UTI from 64,933 births can give us important information that will be useful to tailor screening and treatment of UTI in pregnancy in this geographic region. Only 11.4% (2304/20,203) of all cultures that were sent during this period in pregnancy were positive. Approximately 9% of antenatal cultures were positive in a setting where cultures were not done universally but as the second step following a primary screening protocol. Postnatal cultures that were sent only for symptomatic women were positive in 13.7% of samples. This was common in the 1^st–2^nd week following delivery as seen in other studies.^[2,22] Overall incidence of 1–2% that was seen in this study^[1,2,4] was similar to other studies.

This study has also been able to find recurrence rates of infection both antenatally and postnatally. UTI with both instrumental delivery and cesarean delivery was more common than with normal delivery, and this could be mainly due the increased catheterization associated with operative deliveries as seen in another study.^[23] Optimization of collected urine samples has been a challenge especially when collected from postnatal women. This study has shown properly collected samples in 90% of antenatal women and 70% postnatal women.

E. coli known to be the most common uropathogen universally in UTI for both postnatal and antenatal women was seen in 60% of the samples. Enterococcus (seen in 14%) and Klebsiella species like most other studies were the second most common. Klebsiella infections being more common among postnatal women as compared to antenatal women (7.3 vs. 4.9%) could reflect the presence of a hospital-acquired infection in postnatal women. The spectrum of uropathogens was largely similar in both antenatal and postnatal samples.

The increase of the prevalence of multidrug-resistant organisms is always a challenge and is also specific to each geographical region. It is the regional susceptibility data that guide the appropriate use of antibiotic treatment. Nitrofurantoin with its excellent sensitivity remained the best choice for asymptomatic or lower UTI. However, it is known that this drug is not useful when women with UTI have systemic signs and symptoms. The sensitivity data suggest that nearly 30% of antenatal and 40% of postnatal E. coli isolates could be potential ESBL producers. This study showed 80–90% sensitivity with use of gentamicin for Gram-negative bacilli and could still be used as first-line management in acute pyelonephritis in pregnancy before cultures sensitivity reports are available. Good susceptibility of uropathogens to amoxicillin made it a convenient drug to use in UTI with Gram-positive enterococcal infection.

The strength of our study is that the large amount of data collated over 5 years. The limitations of our study are that it is a retrospective study from a large tertiary center where even though most of the women revisited the hospital with their complications, there may be considerable under reporting, and therefore, the prevalence of UTI may not be accurate. The data were collated from laboratory and hospital records of women having for antenatal and postnatal care.

Antimicrobial stewardship is an important aspect of healthcare and a heightened awareness of susceptibility patterns of common uropathogens to commonly used antibiotic is paramount for rational drug use in pregnancy. This study has made us aware of the bacterial profile and its sensitivity to commonly used antibiotics in our setting.

CONCLUSION

Thus, the importance of microbiological evidence prior to administration of antimicrobials is evidenced by 79% negative cultures in this 5-year cohort. About 9.5% culture positivity among antenatal and 13.28% among postnatal patients may be expected with Asymptomatic bacterial carriage (ASB) 2.1% of the antenatal cases in this region and 40% of all significant culture growths. Escherichia coli accounted for 60-66% of significant bacteriuria followed by Enterococcus and Klebsiella species with 30-40% E.coli probable ESBL producers. Nitrofurantoin followed by amoxicillin-clavulanate were found to be the best oral antimicrobial options.