Barone Fistula Catheter Trial: 7-Day versus 14-Day Bladder Catheterization After Simple Fistula Repair
The Barone fistula catheter trial was an open-label, eight-hospital randomized noninferiority trial in 524 women whose repair of a simple genital fistula was still closed on a dye test 7 days after surgery, comparing bladder catheter removal on day 7 with removal on day 14. Dye-test-defined breakdown from 8 days after catheter removal to 3 months after surgery occurred in 10/250 (4.0%) of the 7-day group and 8/251 (3.2%) of the 14-day group (risk difference 0.8 percentage points, 95% CI -2.8 to 4.5), inside the 10-point noninferiority margin.[1]
- Design and population
- Multicenter open-label noninferiority randomized trial across eight African hospitals. Women at eight hospitals in eight African countries with a surgeon-judged simple genital fistula that was closed on dye test at the end of surgery and still closed 7 days after repair.
- Arms
- 7-day bladder catheterization versus 14-day bladder catheterization
- Report and follow-up
- 2015 · 3 months after surgery (primary window from 8 days after catheter removal)
| Endpoint | Results | Effect | Finding |
|---|---|---|---|
| Dye-test-defined repair breakdown from 8 days after catheter removal to 3 months after surgery, per-protocol; risk difference in percentage points (7-day minus 14-day); noninferiority margin 10 points | 10/250 (4.0%) vs 8/251 (3.2%) | difference 0.8; CI -2.8–4.5 | Met |
Harms: Urinary retention during hospital stay: 31/261 (11.9%) vs 25/262 (9.5%); Infection possibly related to treatment: 8/261 (3.1%) vs 16/262 (6.1%); Febrile episode (above 38 C) possibly related to treatment: 5/261 (1.9%) vs 10/262 (3.8%)
Design
Hospitals in the Democratic Republic of the Congo (90 participants), Ethiopia (23), Guinea (109), Kenya (37), Niger (50), Nigeria (76), Sierra Leone (96) and Uganda (43) enrolled women from March 7, 2012 to May 6, 2013. Of 1007 women assessed, 524 (52.0%) were randomized; 381 were excluded after surgery because the fistula was not simple. The operating surgeon decided at the end of surgery whether the fistula was simple, using personal criteria and no fixed definition. When 17 surgeons classified fistula drawings as simple or not simple, inter-rater agreement was moderate to good (kappa 0.55). Radiation-induced fistulas, fistulas associated with cancer or lymphogranuloma venereum, and pregnancy were exclusions; multiple fistula became an exclusion on July 31, 2012.[1]
Eligibility was screened at three points: before surgery, at the end of surgery (fistula closed on dye test) and at day 7 (still closed on dye test). Randomization took place at day 7 by a research assistant at each site, 1:1 in blocks of 4 to 6 stratified by country, from a sequence generated centrally by WHO and concealed in sealed opaque envelopes. The 7-day group had the catheter removed on the day of randomization and the 14-day group 7 days later. Participants stayed at the facility for 7 days after removal and had a dye test on day 7 after removal (postoperative day 14 versus day 21). A positive test ended participation and counted as a breakdown; women with a negative test returned at 3 months for an interview, examination and dye test.[1]
The primary outcome was breakdown on dye test at any time from 8 days after catheter removal to 3 months after surgery, assessed by the site's fistula surgeon, who was not masked. The noninferiority margin of 10 percentage points was set by expert clinical judgment. The sample-size calculation assumed 13% breakdown in the 14-day group, one-sided alpha 0.025 and power 80%, giving 177 per arm and a planned total of 507 after allowance for loss to follow-up. The analysis was done in the per-protocol population. A data and safety monitoring committee reviewed one interim analysis after a third of participants had returned for the 3-month visit.[1]
Results
The primary outcome was established in 250 women in the 7-day group and 251 in the 14-day group. Breakdown occurred in 10/250 (4.0%) and 8/251 (3.2%): risk difference 0.8 percentage points (95% CI -2.8 to 4.5), relative risk 1.3 (95% CI 0.5 to 3.1). The upper confidence limit of 4.5 points lay below the 10-point margin, so 7-day catheterization met the noninferiority criterion. In the worst of nine sensitivity scenarios for the 11 women lost to follow-up in each group, breakdown was 13% versus 8% (difference 1.9 points, 95% CI -2.1 to 5.2), still below the margin. A multilevel logistic regression found no significant site effect.[1]
Breakdown from catheter removal to 3 months, which includes the dye test 7 days after removal, occurred in 23/261 (8.8%) versus 20/262 (7.6%) (difference 1.2 points, 95% CI -3.9 to 6.3). Of these, 13/261 (5.0%) versus 12/262 (4.6%) were detected at the day-7 test after removal. Seven primary events (4/250 and 3/251) were identified before the 3-month visit in women who returned with urine leakage. At 3 months, residual incontinence was recorded in 9/250 (3.6%) versus 6/251 (2.4%) (difference 1.2 points, 95% CI -2.2 to 4.6); the diagnosis was stress incontinence in 11 women, overflow in 2 and urge in 2, by the surgeon's history and examination without urodynamics.[1]
Surgery and treatment delivery
Mean age was 31.9 years in the 7-day group and 30.6 years in the 14-day group, and leakage began after delivery in 243/261 (93.1%) and 252/262 (96.2%). Previous repair attempts were reported in 57/261 (21.8%) and 58/262 (22.1%). The median largest fistula diameter was 1 cm in both groups, and moderate or severe scarring was recorded in 36/261 (13.8%) and 33/262 (12.6%). Juxtaurethral fistulas were recorded in 69/261 (26.4%) and 58/262 (22.1%), juxtacervical in 67/261 (25.7%) and 84/262 (32.1%) and midvaginal in 70/261 (26.8%) and 65/262 (24.8%); more than one location was possible. Repairs were vaginal in 244/261 (93.5%) and 245/262 (93.5%) and done under spinal anesthesia in 254/261 (97.3%) and 257/262 (98.1%). Care before randomization was not standardized: empirical postoperative antibiotics were given to 136/261 (52.1%) and 145/263 (55.1%), and 25 women (10%) in each group had bladder training.[1]
After randomization, care was standardized: daily perineal and vaginal care, walking and free fluids, and no prophylactic antibiotics, anthelmintics or bladder training. Postvoid residual volume was measured by urethral catheter 1, 3 and 7 days after catheter removal. Retention, defined as a residual above 50% of the voided volume, was managed with intermittent catheterization at least three times daily until the residual was below 50% of the voided volume on two consecutive occasions.[1]
Adverse events
Urinary retention during the hospital stay occurred in 31/261 (11.9%) versus 25/262 (9.5%) (difference 2.3 points, 95% CI -3.3 to 8.0). After catheter removal there were 100 retention episodes (54 in the 7-day group and 46 in the 14-day group) in 56 women; 48/54 (88.9%) and 42/46 (91.3%) of the episodes resolved with intermittent catheterization. Infection possibly related to treatment occurred in 8/261 (3.1%) versus 16/262 (6.1%), comprising 10 urinary tract infections (6 versus 4), 4 vaginal infections (all in the 14-day group), 5 wound infections (2 versus 3) and 5 other infections (all in the 14-day group), all mild or moderate and treated with antibiotics. Febrile episodes occurred in 5/261 (1.9%) versus 10/262 (3.8%), catheter blockage in 5/261 (1.9%) versus 11/262 (4.2%), and extended hospital stay for a possibly treatment-related reason in 5/261 (1.9%) versus 1/262 (0.4%) (relative risk 5.0, 95% CI 0.6 to 42.7). None of these secondary comparisons differed significantly.[1]
Intraoperative complications, recorded before randomization, were hemorrhage in 1 woman in the 7-day group and 5 in the 14-day group and one right ureteric transection in the 14-day group.[1]
Limitations
The trial was open-label; masking participants, staff or the outcome-assessing surgeon was not practical, and few qualified surgeons were available at each site. Twelve women with multiple fistulas (3 in the 7-day group, 9 in the 14-day group) were randomized before that exclusion was added, and because the unit of analysis was the fistula, including them could have increased apparent precision. Follow-up ended at 3 months, residual incontinence rested on clinical impression, and 11 women in each group were lost to follow-up. Breakdown in the 14-day group (3.2%) was well below the 13% assumed for the sample-size calculation. Funder staff took part in the design and in the interpretation of the analysis.[1]
In conversation
An earlier randomized noninferiority trial compared 10-day with 14-day catheterization after fistula repair and measured breakdown at catheter removal; more than three quarters of breakdowns in the 10-day group occurred before the catheter was removed, so that trial could say little about the role of catheter duration. This trial defined its primary outcome to begin 8 days after removal, so that every participant analysed was dye-test negative at the start of follow-up.[1]
Guidelines and approvals
- WHO (2018). For women after surgical repair of a simple obstetric urinary fistula, short-duration bladder catheterization (7 to 10 days) is recommended as an alternative to longer duration. The panel defined a simple fistula as a mid-anterior vaginal wall fistula with minimal scarring and a diameter of 3 cm or less .[2]
Report history
| Published | Report | Cutoff | Headline results | Citation |
|---|---|---|---|---|
| 2015-04-22 | Randomized noninferiority report · primary | — | Dye-test-defined repair breakdown from 8 days after catheter removal to 3 months after surgery, per-protocol; risk difference in percentage points (7-day minus 14-day); noninferiority margin 10 points: 10/250 (4.0%) vs 8/251 (3.2%); difference 0.8; CI -2.8–4.5 Met Dye-test-defined breakdown from catheter removal to 3 months after surgery; risk difference in percentage points (7-day minus 14-day): 23/261 (8.8%) vs 20/262 (7.6%); difference 1.2; CI -3.9–6.3 Dye-test-defined breakdown at the dye test 7 days after catheter removal; risk difference in percentage points (7-day minus 14-day): 13/261 (5.0%) vs 12/262 (4.6%); difference 0.4; CI -3.6–4.4 Residual incontinence after fistula closure at the 3-month visit; risk difference in percentage points (7-day minus 14-day): 9/250 (3.6%) vs 6/251 (2.4%); difference 1.2; CI -2.2–4.6 Primary outcome with missing observations imputed as a four-fold breakdown proportion in the 11 lost to follow-up in the 7-day group and a 99% reduction in the 11 lost in the 14-day group; risk difference in percentage points: 13% vs 8%; difference 1.9; CI -2.1–5.2 | Barone MA, Widmer M, Arrowsmith S, et al. Breakdown of simple female genital fistula repair after 7 day versus 14 day postoperative bladder catheterisation: a randomised, controlled, open-label, non-inferiority trial. *Lancet.* 2015;386(9988):56-62. doi:10.1016/S0140-6736(14)62337-0 DOI |
References
1. Barone MA, Widmer M, Arrowsmith S, et al. "Breakdown of simple female genital fistula repair after 7 day versus 14 day postoperative bladder catheterisation: a randomised, controlled, open-label, non-inferiority trial." Lancet. 2015;386(9988):56-62. doi:10.1016/S0140-6736(14)62337-0
2. Widmer M, Tunçalp Ö, Torloni MR, et al. "Improving care for women with obstetric fistula: new WHO recommendation on duration of bladder catheterisation after the surgical repair of a simple obstetric urinary fistula." BJOG. 2018;125(12):1502-1503. doi:10.1111/1471-0528.15276