Twice-Daily Nitrofurantoin Administration Following Short-term Transurethral Catheterization After Pelvic Reconstructive Surgery: A Randomized Clinical Trial

Abstract

Objective 

The objective of this study was to evaluate if twice-daily nitrofurantoin for 5 days after discontinuation of transurethral catheterization decreases the rate of urinary tract infection (UTI) in women with postoperative urinary retention (POUR) after pelvic reconstructive surgery.

Methods 

This was a double-blind, placebo-controlled, randomized trial conducted across 2 clinical sites between October 2017 and April 2019. Women with acute POUR after pelvic reconstructive surgery were included and randomized to nitrofurantoin (100 mg) or placebo twice-daily for 5 days. The primary outcome was clinically suspected UTI (defined as dysuria, frequency, and irritation in the absence of vaginal discharge) and/or culture-proven UTI (defined as greater than 10⁵ colony forming units of a single organism) within 30 days of surgery. Secondary outcomes included evaluation of adverse events related to study medication and medication adherence.

Results 

Data from 164 participants were eligible for intention-to-treat analysis (nitrofurantoin, n = 82; placebo, n = 82). There were no significant demographic or intraoperative differences except for body mass index and race. Median duration of catheterization was 3 days (interquartile range, 2–5 days, P = 0.12). Fifteen women in the nitrofurantoin group and 14 women in the placebo group experienced UTI within 30 days (18.3% vs 17.1%; P = 0.84; odds ratio, 1.09; 95% confidence interval, 0.49–2.43). There were no study medication allergies; however, nausea was the most common intolerance. Most women in each group completed the study drug treatment (91.5% vs 86.4%, P = 0.30).

Conclusions 

Nitrofurantoin prophylaxis after transurethral catheter removal did not reduce the risk of UTI in women with acute POUR after pelvic reconstructive surgery.

Urinary tract infections (UTIs) are common after pelvic organ prolapse (POP) and urinary incontinence (UI) surgery, with a risk ranging from 7% to 40%.^1–6 Risk factors for UTIs in women undergoing these surgical procedures include history of recurrent UTI, longer operative times, and increased duration of postoperative catheterization.⁴

Acute postoperative urinary retention (POUR) is also common after POP and UI surgery, with incidences ranging from 2.5% to 50%.^4,6 Often managed with short-term catheterization, women with POUR have an increased risk of UTI.^1,2,6–8 There is a 15%–20% risk of developing a UTI with short-term catheterization, even if the catheter remains in place for 24 hours, and increases 5%–10% each day the catheter is in situ.^8,9 The Stress Incontinence Surgical Treatment Efficacy (SISTEr) trial, a multicenter, prospective, randomized trial comparing Burch colposuspension and autologous fascial slings, demonstrated an increased rate of UTIs in the fascial sling group.^2,10,11 Although not the direct cause of UTIs, this group did have POUR and prolonged catheterization.^2,10,11 Complications of catheter-associated UTIs increase the morbidity for patients undergoing these surgical procedures, especially in the elderly.^2,7 As the lifetime risk of surgery for POP and UI increases to 20%, it is imperative to reduce the risk of postoperative UTI and its complications.⁶

Numerous studies have evaluated prophylactic antibiotic regimens to decrease the rate of postoperative catheter-associated UTIs.^{1,7,8,12–14} Nitrofurantoin has been used in some studies because of its low bacterial resistance and high rate of tolerability.⁶ Although some studies demonstrate nitrofurantoin use during catheterization reduces postoperative UTI, contradictory findings have also been reported.^1,3,6,13,15 The American Urological Association (AUA) Best Practice Statement recommends 24-hour empiric administration of fluoroquinolones or trimethoprim-sulfamethoxazole (TMP-SMX) at the time of catheter removal in patients with risk factors for urologic complications based on evidence supporting lower UTI rates with treatment after catheter removal.^14,16–19 However, it is unclear if this recommendation extends to the urogynecology population. Consequently, there is variability in the practice of antibiotic prophylaxis at the time of catheter discontinuation. In addition, the Infectious Disease Society of America recommends against routine use of fluoroquinolones because of growing concerns regarding antimicrobial resistance and potential adverse events.²⁰ Although nitrofurantoin has been studied during catheterization, there are insufficient data evaluating nitrofurantoin prophylaxis after catheter removal. The objective of our study was to identify if twice-daily administration of nitrofurantoin after transurethral catheter discontinuation decreases the risks of postoperative UTI in women undergoing pelvic reconstructive surgery.

MATERIALS AND METHODS

This was a randomized, double-blind, placebo-controlled trial of women with acute POUR after POP and/or UI surgery managed with transurethral catheterization across 2 teaching hospital sites within Atrium Health Women’s Care — Urogynecology and Pelvic Surgery Division between October 2017 and April 2019. Funding for this study was provided by the Atrium Health Department of Obstetrics and Gynecology Education and Research Fund. Institutional review board approval was obtained. The study was registered at ClinicalTrials.gov (ID: NCT03287089) and reported using Consolidated Standards of Reporting Trials guidelines.²¹

Consecutive women were prescreened for eligibility and introduced to the study before surgery. Women with a positive UTI screen (defined as treatment for UTI or an affirmative response to dysuria, urinary frequency, or bladder irritation in the absence of vaginal discharge) within 30 days before surgery were ineligible. Additional exclusionary criteria are outlined in Supplemental Digital Content 1 (Supplemental Digital Content 1, Supplementary Table, https://links.lww.com/FPMRS/A190). Written informed consent was collected either preoperatively at time of study introduction or at their postoperative void trial appointment before initiation of the in-office void trial. All women who provided consent preoperatively and who passed a postoperative void trial at the time of hospital discharge were not randomized.

All women received perioperative antibiotic prophylaxis within 1 hour of skin incision with intravenous cefazolin or an appropriate alternative if allergic. All surgical procedures were performed by female pelvic medicine and reconstructive surgery board-certified surgeons and female pelvic medicine and reconstructive surgery fellows. A sterile transurethral catheter was then placed. Before hospital discharge, all women underwent a retrograde fill void trial, regardless of procedure type. Per our practice guidelines, the bladder was backfilled with 300 mL of sterile water followed by removal of the transurethral catheter. The patient voided immediately, and a postvoid residual volume was measured with a portable bladder ultrasound. For our study, we defined POUR as a postvoid residual volume greater than one third of the total urine volume (which is equal to the total voided amount plus postvoid residual amount). A transurethral catheter was placed before hospital discharge. Postoperative in-office void trials were scheduled within 1–7 days after hospital discharge and followed the same criterion as the initial voiding trial.

Randomization was performed by computer-generated permuted blocks of 4 and 6 and stratified by site. All participants, clinicians, and study team members were masked to study allocation except the Atrium Health Investigational Drug Service, which prepared and distributed the study drug. Randomized participants were allocated 1:1 to 100 mg of nitrofurantoin (monohydrate/macrocrystals) or an identical appearing placebo capsule that was taken twice-daily for 5 days after catheter discontinuation. The first dose of the study medication was dispensed in the office by study personnel and recorded on the medication adherence log. Participants were instructed to record the date and time of each dose of study medication administration. Participants were seen by research staff and returned unused study drug, study drug containers, and medication adherence logs at or before their 6-week postoperative visit. Each participant’s medical chart was queried during their enrollment to identify incidences of UTIs, adverse drug events, or additional antibiotic use not previously reported. Participants were queried via telephone calls and in-person visits to confirm all outcomes. A data and safety monitoring board reviewed the progress and safety throughout the study. Study data were managed using Research Electronic Data Capture, which is a secure, web-based software platform designed to support data capture for research studies.^22,23

Participants were instructed to contact our office with concerns of UTI for evaluation. If a participant had clinical symptoms of a UTI, a clean-catch urine specimen was sent for culture and empiric treatment was started by the treating clinician. Each clinician was educated before study implementation on treatment protocol. Empiric treatment was started with one of the following antibiotics at the discretion of the treating provider: cephalexin (500 mg every 12 hours for 7–14 days) or fosfomycin (3 g once). Ciprofloxacin (250 mg every 12 hours for 3 days or 250–500 mg every 12 hours for 7–14 days) was used based on personal treatment history, contraindications, or allergies to other empiric antibiotic options. We did not offer TMP-SMX as empiric treatment because it displayed only a 75% sensitivity for treating Escherichia coli in our region based on our institution’s antibiotic susceptibility report. Once the culture identified the uropathogen, the appropriate antibiotic regimen was selected if the empiric therapy was shown to inadequately treat the infection. Participants discontinued empiric antibiotic therapy if the urine culture returned negative. Participants who were treated by outside providers or treated based on only clinical symptoms were deemed treatment failures and were included in the study analysis as such.

The primary outcome was clinically suspected UTI (defined as dysuria, frequency, and bladder irritation in the absence of vaginal discharge) and/or culture-proven UTI (defined as greater than 10⁵ colony forming units per milliliter of a single uropathogenic organism)^16,17 within 30 days of surgery. Secondary outcomes included evaluation of adverse events related to study medication and study medication adherence. Frequency of adverse events and medication intolerances related to study medication administration were recorded at postoperative follow-up visits. Medication adherence was calculated as the number of study pills taken divided by the total number of study pills dispensed (10). A closer ratio to 1 (range, 0–1) indicates better medication adherence; conversely, a ratio closer to 0 indicates poor medication adherence.

Calculations of sample size were based on comparisons of UTI rates in the nitrofurantoin and placebo arms. According to published results from randomized trials, the UTI rate in the placebo arm was 32%, and approximately a 14%–15% reduction was observed in the treatment arm.^1,13 Using this information, we designed our study to mimic these findings and assumed that a UTI reduction rate of 18% in the nitrofurantoin arm could be observed in our clinical setting. Based on this assumption, a total sample size of 164 patients was projected to achieve 80% power with a 2-sided α error of 0.05. We increased the total sample size to 180 to allow for 10% attrition. However, because of primary investigator timeline and lack of resources to overrecruit, recruitment ceased once the sample size of 164 was reached. The power was calculated by using the PASS 15 (2017, NCSS, LLC, Kaysville, UT). All analyses were conducted using an intention-to-treat analysis, which included all women randomized in the study. A per-protocol analysis was also performed and only included patients who completed the study.

Categorical data were compared using χ² or Fisher exact test where appropriate. Continuous variables were compared using Student t test or Wilcoxon rank sum where appropriate and are presented as mean ± SD or median (interquartile range), respectively. We estimated risks for UTI with odds ratio (OR) and corresponding 95% confidence intervals (CIs) calculated using logistic regression with covariates that were determined a priori. A P value of <0.05 defined statistical significance. All analyses were conducted using SAS Enterprise Guide version 7.1 (SAS Institute, Cary, NC). No interim analysis was performed.

RESULTS

A total of 932 women were screened for eligibility between October 2017 and April 2019 across 2 clinical sites, and 190 women consented (Fig. 1). Of those, 26 women who provided preoperative consent passed postoperative voiding trials and, therefore, were not eligible for randomization. There were 82 participants randomized to each arm in the intention-to-treat analysis, including 2 participants who were lost to follow-up, 4 participants who were withdrawn (2 per patient request, 1 by primary investigator because of recurrent urinary retention with subsequent transurethral catheterization, and 1 with postoperative wound dehiscence of a posterior colporrhaphy incision). We continued to collect data on these 6 participants, and they were labeled as a negative UTI in the intention-to-treat analysis. Consequently, a total of 158 participants were included in the per-protocol analysis, 82 in the nitrofurantoin group, and 76 in the placebo group.

FIGURE 1: Consolidated standards of reporting trials flow diagram.

There were no significant demographic or intraoperative differences between groups except for body mass index and race (Table 1). Median duration of postoperative transurethral catheterization was 3 days (Table 1). Of the total study population, 61 women (37%) underwent prolapse surgery, 24 (15%) incontinence surgery, 74 (45%) combined prolapse and incontinence surgery, 3 (2%) vaginal mesh excision, and 2 (1%) for combined vaginal mesh excision and prolapse or incontinence surgery.

TABLE 1 - Baseline Characteristics

Characteristic	Nitrofurantoin (n = 82)	Placebo (n = 82)	P
Age, y	61.6 ± 11.7	61.1 ± 12.4	0.77
Race
Caucasian	79 (96.3)	72 (87.8)	0.04*
Black	1 (1.2)	8 (9.8)	0.03*
Other	2 (2.4)	2 (2.4)	0.99
BMI, kg/m2	27.1 ± 4.7	28.6 ± 5.0	<0.05*
Parity	2 [1, 3]	2 [2, 3]	0.36
POP-Q stage	2 [2, 3]	2 [2, 3]	0.26
Menopausal status			0.88
Premenopausal	18 (22.0)	17 (21.0)
Postmenopausal	64 (78.0)	64 (79.0)
Postmenopausal hormone use			0.99
No hormonal therapy	58 (70.7)	58 (70.7)
Vaginal estrogen therapy	20 (24.4)	19 (23.2)
Systemic hormonal therapy	4 (4.9)	5 (6.1)
Current smoking	6 (7.3)	6 (7.3)	0.99
History of recurrent UTI	1 (1.2)	1 (1.2)	0.99
Diabetes mellitus	9 (11.0)	10 (12.2)	0.81
Preoperative PVR volume, mL	57.9 ± 92.6	45.3 ± 55.7	0.38
Creatinine clearance, mL/min	84.1 ± 30.1	94.2 ± 34.9	0.06
Procedure
Midurethral sling surgery	50 (61.0)	47 (57.3)	0.63
Burch colposuspension	1 (1.2)	2 (2.4)	0.99
Anterior repair	15 (18.3)	20 (24.4)	0.34
Posterior repair	53 (64.6)	59 (72.0)	0.31
Hysterectomy	32 (39.0)	35 (42.7)	0.63
Vaginectomy/colpocleisis	7 (8.6)	2 (2.4)	0.10
Apical suspension
Uterosacral ligament suspension	18 (22.0)	24 (29.3)	0.28
Sacrospinous ligament fixation	0 (0.0)	1 (1.2)	0.99
Iliococcygeus vault suspension	1 (1.2)	1 (1.2)	0.99
Sacral colpopexy	30 (36.6)	29 (35.4)	0.87
Vaginal mesh excision	4 (4.9)	1 (1.2)	0.37
Operative time, min	134.0 ± 80.8	142.9 ± 86.9	0.50
Estimated blood loss, mL	45.5 ± 35.5	60.9 ± 61.6	0.05
Length of hospital stay, d	1 [0, 1]	1 [0, 1]	0.70
Duration of catheterization, d	3 [2, 5]	3 [2, 5]	0.12

Some columns do not equal 100% because of missing data.

Data are presented as mean ± SD, n (%), or median (interquartile range [25th percentile, 75th percentile]).

*Represents statistical significance.

Abbreviations: BMI, body mass index; POP-Q, Pelvic Organ Prolapse Quantification; PVR, postvoid residual volume.

Of the 164 patients in the intention-to-treat primary analysis, 15 women in the nitrofurantoin group and 14 women in the placebo group had UTIs within 30 days of surgery (18.3% vs 17.1%; P = 0.84; OR [95% CI], 1.09 [0.49–2.43]). Based on these numbers, the number needed to treat or the number of women who would receive nitrofurantoin prophylaxis to prevent 1 woman from experiencing a postoperative UTI at 2 weeks and 30 days was 13 and 83, respectively. A per-protocol analysis was also performed with no significant differences (Table 2).

TABLE 2 - Primary Outcome: Comparison of UTI Between Groups

	Nitrofurantoin	Placebo	OR (95% CI)	P
At 2 wk
Intention-to-treat	5/82 (6.5)	11/82 (14.3)	0.42 (0.14–1.26)	0.11
Per-protocol	5/82 (6.5)	11/76 (14.5)	0.39 (0.13–1.17)	0.08
At 30 d
Intention-to-treat	15/82 (18.3)	14/82 (17.1)	1.09 (0.49–2.43)	0.84
Per-protocol	15/82 (18.3)	14/76 (18.4)	0.99 (0.44–2.22)	0.98

Data are presented as n/group total (%).

Table 3 illustrates treatment of UTI based on positive urine culture alone, clinical symptoms alone, or both. Within the first 2 weeks of surgery, 5 women in the nitrofurantoin group and 11 women in the placebo group reported clinical symptoms (6.5% vs 14.3%; P = 0.11; OR [95% CI], 0.42 [0.14–1.26]). There were a total of 12 participants (6 in each group) who had UTI symptoms with a negative urine culture at 30 days. Of all the positive urine cultures, 5 were Escherichia coli, 5 were Enterococcus faecalis, 2 were Enterobacter cloacae, 1 was Pseudomonas aeruginosa, and 1 was Serratia marcescens.

TABLE 3 - UTI Diagnosis

Total UTI (n = 29)	Nitrofurantoin	Placebo	P
At 2 wk
Culture alone	0 (0)	3 (10)	0.25
Symptoms alone	2 (7)	4 (14)	0.68
Culture + symptoms	0 (0)	3 (10)	0.25
At 30 d
Culture alone	3 (10)	0 (0)	0.99
Symptoms alone	2 (7)	1 (3)	0.99
Culture + symptoms	6 (21)	0 (0)	0.50

Totals do not equal 100% because of 5 participants being treated for UTI by outside providers.

Data are presented as n/group total (%).

A logistic regression model was constructed to evaluate the effect of prophylactic nitrofurantoin while controlling for confounding factors. When adjusting for age, menopausal status, duration of catheterization, and overnight hospital stay, nitrofurantoin did not reduce the risk of UTI compared with placebo at 30 days (OR, 1.04; 95% CI, 0.46–2.36).

There were no study medication allergies (defined as hives, rashes, swelling, or anaphylaxis). However, there were a total of 9 women with intolerances to the study medication. The most common medication intolerance was nausea (Table 4). All women with intolerances, with the exception of one woman in the placebo group with a headache, discontinued study medication after 1–5 doses.

TABLE 4 - Study Medication Adverse Events at 30 Days

	Nitrofurantoin (n = 82)	Placebo (n = 82)	P
Allergy*	0	0	n/a
Intolerance			0.44
Nausea	5 (6.1)	2 (2.4)	0.44
Vomiting	0	0
Diarrhea	0	0
Other:			0.50
Headache	0	1 (1.2)
Joint pain	0	1 (1.2)

*Includes hives, rash, swelling, and anaphylaxis.

Data are presented as n (%).

Abbreviation: n/a, not applicable.

Study medication adherence was also assessed. Most women in each group completed the study drug treatment (91.5% vs 86.4%, P = 0.30). There were no significant differences in study medication adherence between groups (0.95 ± 0.18 vs 0.96 ± 0.15, P = 0.68).

DISCUSSION

Our study findings demonstrate that women with acute POUR after POP or UI surgery managed with short-term transurethral catheterization did not have a reduced risk of postoperative UTI when treated with 100 mg of nitrofurantoin or placebo twice daily for 5 days upon transurethral catheter discontinuation. This finding was consistent in both the intention-to-treat and per-protocol analyses.

We found that the UTI rate in the nitrofurantoin group was higher at 30 days compared with placebo, which was also demonstrated in other studies.^3,6 This, however, was not the case at 2 weeks, where the UTI rate was 6.5% in the nitrofurantoin group versus 14.3% in the placebo group. This difference at 2 weeks may suggest that the nitrofurantoin may initially decrease bacteriuria, but this benefit decreases with time. Further investigation on the time-dependent effect on the urinary microbiome needs to be undertaken to provide further evidence.

Prior published prospective studies have shown precedent for the use of antibiotic prophylaxis at the time of transurethral catheter removal to decrease UTI in catheterized women.^7,8 The AUA recommends prophylactic antibiotic treatment at the time of transurethral catheter removal because of the risk of bacteremia from bacteriuria.¹⁶ Fluoroquinolones and TMP-SMX decrease the risk of postoperative UTIs after short-term catheterization.^18,19 However, the use of these antibiotics is not always ideal. There are growing resistance patterns to TMP-SMX, and in 2008, the U.S. Food and Drug Administration recommended against the prophylactic use of fluoroquinolones.^18,24

Nitrofurantoin was chosen as the interventional medication for this study because of its efficacy in reducing UTIs in women during postoperative catheterization, tolerability, and low resistance rates.^1,6,13 Although most of nitrofurantoin’s antimicrobial activity is limited to the lower urinary tract, it has broad antimicrobial activity against both gram-positive and gram-negative organisms.¹⁸ We chose a twice-daily dosage to ensure maximum coverage for UTI based on significant UTI reduction rates with a twice-daily dosage.¹ In addition, E. coli, the most common uropathogen, displays low resistance to nitrofurantoin in our geographical region based on our institution’s antibiotic susceptibility report demonstrating a 98% sensitivity for treating E. coli. Our study demonstrated E. coli and E. faecalis as the primary uropathogens. We also limited the exposure to the antibiotic treatment to only 5 days, which falls within the recommended treatment courses for nitrofurantoin (range, 5–7 days).^25,26

Because of the aforementioned qualities, nitrofurantoin has been evaluated and shown to be successful as a prophylactic regimen during short-term catheterization. Rogers et al¹³ conducted a prospective, multicenter, randomized controlled trial that demonstrated once-daily nitrofurantoin compared with placebo given for the duration of suprapubic catheter placement decreased the risk of UTI in women after POP or UI surgery (46% vs 61%, P = 0.002). A randomized controlled trial by Jackson et al¹ showed similar findings with a 3-day course of twice-daily nitrofurantoin significantly reducing the incidence of UTI in the first 6 weeks postoperatively in patients who were discharged with short-term catheterization (17.6% nitrofurantoin vs 32% placebo, P = 0.04) after outpatient midurethral sling surgery.

Conversely, there is conflicting evidence that nitrofurantoin may not be suitable for antibiotic prophylaxis during short-term postoperative catheterization. Hampton and Erekson¹⁵ published a retrospective study that compared 30-day bacteriuria rates in women with POUR after midurethral sling surgery and found that women who received nitrofurantoin prophylaxis during transurethral catheterization did not have reduced bacteriuria compared with nontreated women. Recent larger, randomized controlled trials confirmed these results and found that prophylactic once-daily nitrofurantoin during the catheterization period did not decrease the risk of UTI after POP or UI surgery.^3,6,27 Despite designing this study to evaluate nitrofurantoin prophylaxis with regards to AUA recommendations on antibiotic administration after catheter removal, our findings echo these robust, negative outcome studies. Together, these studies further support the statement put forth by the Infectious Disease Society of America that prophylactic antimicrobials should not be routinely used in patients with catheterization.²⁰

Strengths of our study include the prospective randomized double-blind, placebo-controlled trial design, large sample size, and multiple clinical sites. Our definition for UTI was similar to prior studies, which helps to establish diagnosis guidelines and improves the generalizability of our results.^1,3,10 In addition, our study provided data on antibiotic prophylaxis after transurethral catheterization. In a recent survey among surgeons, transurethral catheterization was the preferred route of management for POUR after pelvic surgery; however, these findings cannot be applied to women managed with suprapubic or clean-intermittent catheterization.²⁸ Additional limitations include recall bias concerning medication use and patient-reported UTIs; these were minimized with the use of a medication diary and verification of UTI treatments by outside providers. Our sample size of 164 was inflated by 10% to account for potential attrition. However, overrecruitment was likely not necessary because only 6 patients were lost to follow-up or withdrawn and, therefore, may not have significantly affected our findings. Finally, our primary outcome defined UTI by a pragmatic approach, contrasting from the International Continence Society definition that requires microbiologic confirmation in the setting of lower urinary tract symptoms.²⁹

This study was designed to address the clinical practice variability surrounding antibiotic prophylaxis after catheter removal, using recommendations from the AUA, as well as the methodology of other urogynecologic studies regarding nitrofurantoin prophylaxis. Although nitrofurantoin has attributes that would suggest that it is a viable prophylactic antibiotic, previous studies evaluating nitrofurantoin prophylaxis during catheterization have mixed results in reducing the UTI rate.^3,6,15 Our study is one of the first to demonstrate that nitrofurantoin does not decrease risk of postoperative UTI when initiated at the time of transurethral catheter discontinuation and, therefore, does not support the AUA recommendations. Consequently, we recommend against nitrofurantoin prophylaxis after discontinuation of transurethral catheterization in the urogynecology population.