An unusual cause of ‘incontinence’ after hysterectomy


L. Cardozo, Department of Urogynaecology, 6th Floor, Ruskin Wing, King's College Hospital, London SE5 9RS, UK.


A 54-year-old woman was referred to the urogynaecology unit complaining of continuous vaginal leakage 6 weeks after a total abdominal hysterectomy and bilateral salpingo-oophorectomy for abdominal discomfort secondary to uterine fibroids. She had no preoperative urinary symptoms. Having previously had two Caesarean sections the bladder was adherent to the cervix and a small bladder injury was sustained at the time of surgery, although this was identified and closed in two layers. After surgery she made an uneventful recovery and an indwelling urethral catheter was left in situ for 10 days, after which she was discharged home.

On review at 6 weeks she remained well and was asymptomatic. However, shortly after this she started to leak fluid from the vagina. This leakage could occur day or night and was sufficient to require the use of an incontinence pad. A vesico-vaginal fistula was suspected and she was referred for a urological opinion. Both IVU and a cystogram were unremarkable, and cystoscopy, urethral dilatation and cystodistension were performed, although no fistula was identified. Subsequently she was treated empirically with tolterodine 2 mg twice daily, which did little to improve her symptoms. At presentation to the urogynaecology unit she complained of continuously losing fluid from the vagina and having to wear incontinence pads daily. There was no association with coughing, sneezing or exercise, and she gave no history of symptoms of urgency or frequency. She had previously had two child children, both by Caesarean section, but no other abdominal or pelvic operations. Generally she was well, taking no medication, and drank ≈ 70 units of alcohol a week. She was not taking hormone-replacement therapy. On examination the abdomen was not tender and the suprapubic scar was well healed. Vaginal examination revealed no evidence of a fistula but the vagina was noted to be moist. The vaginal vault and anterior vaginal wall were well supported, although there was a moderate rectocele. Urine analysis (midstream) and liver function tests were all normal. Cystometry showed a stable bladder with no evidence of detrusor overactivity, and the bladder morphology was normal on videocysto-urethrography. Specifically there was no evidence of a vesico-vaginal fistula or demonstrable sphincter incompetence. As her symptoms were not explained by laboratory testing an ambulatory urodynamic study was conducted, which also showed a stable bladder with no evidence of leakage. Repeat IVU showed normal upper urinary tracts with no evidence of a ureterovaginal fistula, although there was a suspicion of a vesico-vaginal fistula on a lateral view at the bladder base. However, Bonney's triple swab test with 300 mL of methylene blue instilled in the bladder was normal. In view of these findings she was admitted for examination under anaesthesia, cystoscopy and posterior vaginal wall repair.

On examination she had a small sinus tract at the vaginal vault which admitted a probe and was draining straw-coloured fluid. However, simultaneous cystoscopy revealed a normal-looking bladder and the probe could not be seen. The bladder was then filled with methylene blue but the fluid discharge remained clear and colourless. A diagnosis of peritoneal fistula was suspected and this was confirmed on diagnostic laparoscopy, when the probe could clearly be seen in the pouch of Douglas. As she complained of symptomatic urogenital prolapse and the need to digitate to defecate, a posterior repair was made and the vaginal incision extended to the vaginal vault to excise the fistula tract. The excess peritoneum was excised and then closed with two 2/0-polydioxanone purse-string sutures. A posterior repair was than made with a continuous 2/0 polyglycolic suture and an interlocking fascial suture. She made an uncomplicated recovery and was discharged home 4 days after surgery. Histology revealed marked chronic inflammation, although no definite fistula tract was identified. On review in clinic at 6 weeks she remained well and was asymptomatic; a vaginal examination was unremarkable.


De novo urinary incontinence after hysterectomy is a rare but serious complication requiring urgent investigation and diagnosis. Simple causes, e.g. lower UTI, should first be excluded before a full urogynaecologcal investigation to determine the cause and plan further management. Vaginal discharge after hysterectomy is very common and initially probably simply reflects the healing process, although if protracted, may indicate local infection or a vaginal vault haematoma. Other less common causes of vaginal discharge after pelvic surgery include lymphatic fistulae [1] and Fallopian tube drainage (hydrops tubae profluens) [2].

Urinary incontinence may also present after hysterectomy and either represent unmasking of previously occult pathology or incontinence arising ‘de-novo’. In such cases it is mandatory to exclude the possibility of iatrogenic fistula, although these remain rare, the incidence of vesicovaginal fistula being 0.1–0.2%[3] and that of ureterovaginal fistula lower still.

Peritoneal-vaginal fistulae after hysterectomy are very unusual; this was first described in 1989, in a patient with pneumoperitoneum after sexual intercourse [4]. More recently a series of five patients was reported [5], all developing a peritoneal-vaginal fistula at 2.5–15 years after hysterectomy (one vaginal and four abdominal). Interestingly, three of these patients had had incontinence surgery before the diagnosis of a peritoneal fistula, despite a full evaluation with urodynamics, voiding cysto-urethrography and cystoscopy. The fistula was repaired surgically in all cases by excising the vaginal vault, although the fistula was only clearly identified in three patients and confirmed histologically in one.

In the present case both urodynamic and radiological investigations were unremarkable, and the fistula was finally diagnosed at examination under anaesthesia, when the sinus tract was clearly identified entering the peritoneal cavity, by using the laparoscope. Whilst time-consuming, thorough investigation of the lower urinary tract will avoid unnecessary incontinence procedures which can be associated with significant morbidity.

The cause of peritoneal-vaginal fistulae is unclear. It is common practice to leave the peritoneum, and often the vaginal vault, open after hysterectomy, allowing the peritoneum to close spontaneously; this does not cause a high complication rate. Local infection may have initially played a role in delayed healing, although it is unclear why a fistula should develop. About 60 mL of peritoneal fluid is produced daily and this is sufficient to cause considerable symptoms of ‘incontinence’ corresponding to a pad loss of ≈ 60 g/day. In cases of ascites the risk of developing a peritoneal-vaginal fistula may be higher, although in this particular case the liver function tests were normal (despite a high alcohol intake) and inspection of the abdominal cavity unremarkable.

The present case illustrates the importance of a full evaluation and investigation of women complaining of LUTS after previous pelvic surgery, and that ‘leakage’ should not be treated unless incontinence is identified. Accurate diagnosis will avoid unnecessary intervention and reduce morbidity. Whilst the cause of peritoneal-vaginal fistulae remains uncertain it should be considered in all women who present with ‘incontinence’ after hysterectomy