Pregnancy following obstetric fistula repair, the management of delivery


A Browning, Medical Director, Barhirdar Hamlin Fistula Centre, PO Box 1739, Barhirdar, Ethiopia. Email


Obstetric fistula patients invariably have dreadful obstetric histories. The antecedent delivery usually ends in a stillbirth and even multiparous obstetric fistula patients often have previous histories of multiple stillbirths and neonatal deaths. There is very little published on subsequent pregnancies following obstetric fistula repair or even on reproductive capabilities after repair. This study examines 49 pregnancies following obstetric fistula repair and the management and outcome of delivery.


Obstetric fistula is a severe and debilitating condition occurring throughout the developing world. Although there are no accurate data on the incidence and prevalence of the condition, it is thought that globally there are at least two million women still waiting treatment, most of whom live in Africa.1

It has often been stated that obstetric fistula patients suffer ‘more than a hole in the bladder’. Women are often divorced because of their condition, ostracised by their communities, suffer depression and their reproductive capabilities are damaged.2

In the cultures where this condition is prevalent, women are usually disempowered, have little decision-making capability and their significance as a member of their community is often linked with their ability to bear children.

A full recovery from their condition should be aimed not only at regaining continence but also at enabling reintegration into their communities again. Part of this reintegration is a return of reproductive capacity.

There is very little written about reproductive outcomes following fistula repair; evidence points to only 19–20% of women achieving a further successful term pregnancy.3,4

Likewise, there is very little written about the optimal way to deliver a patient who previously has had a severe obstructed labour, obstetric fistula and fistula repair. One article pointed out that only 50% of patients were selected to undergo a trial of vaginal delivery and of those 50% needed a caesarean; of those delivering vaginally, 27% had their fistula reopen during the delivery.5

This study reports on the experience of managing pregnancies from the Barhirdar Hamlin Fistula Centre, a dedicated obstetric fistula unit situated in the north of Ethiopia.


All patients admitted to the Barhirdar Hamlin Fistula Hospital with obstetric fistula receive treatment. As part of the discharge protocol after obstetric fistula repair, all patients are counselled to return to the unit if they become pregnant again. They are told to present at 7–8 months pregnant and wait within the unit for their delivery.

Patients who had previously been treated in the Addis Ababa Fistula Hospital were given the same information and if they live closer to the unit in Barhirdar, they are told to present there if it is more accessible.

All patients presenting between February 2005 and September 2008 pregnant and nearing term or suspected of nearing term were admitted and were included in this study.

On admission, they underwent routine antenatal investigations including ultrasound scan.

Our protocol is that all patients should be delivered by a timely caesarean section. Patients thus await caesarean section, which is preformed in the government hospital adjacent to the fistula unit. Nearly all patients are from very poor socio-economic backgrounds and are illiterate; the date of the last menstrual period is usually unknown. The lack of neonatal facilities coupled with the uncertainty of dates, has resulted in an agreement with the obstetricians within the government hospital that women undergo an emergency caesarean at the onset of labour. The intention is to reduce the risk of performing a caesarean section prior to term.

During the time of the study, 61 deliveries were performed.

The records of all patients were sought and information entered into an Excel spreadsheet. Information collected included, parity, number of live children, date of fistula occurrence, date of repair, type of fistula, outcome of fistula surgery, date of delivery, outcome of delivery, any complications related to the delivery.


Of the 61 deliveries, only 50 notes were retrieved, the remaining notes having not been returned from the government hospital.

Of the 50 notes retrieved, one patient had a delivery after a repair of a recto-vaginal fistula, which was secondary to sexual intercourse at a prepubescent age. Hence, this delivery was not following an obstetric fistula and was thus excluded. 49 notes were therefore available for analysis.

Of the 49 patients, there were 115 deliveries prior to the index delivery, which resulted in the fistula. Of these deliveries, there were only 33 live children (28.7%). Seventy-five of previous deliveries had ended in a stillbirth (65.2%) with five miscarriages, one neonatal death and one childhood death.

Following their fistula repairs, ten still suffered some degree of stress urinary incontinence, two of whom were using a urethral plug to maintain continence, one had a secondary stress incontinence operation after the fistula repair and was subsequently cured. One patient had urinary retention and was dry while self catheterising and there was only one patient who had a combined rectovaginal and vesico-vaginal fistula; both fistulae were cured.

One patient presented pregnant with an unrepaired 2.5cm juxtacervical vesico-vaginal fistula. She was delivered by caesarean before the fistula was repaired.

Twenty-five of the women had their fistula repair in the Barhirdar Hamlin fistula Centre while the remainder had their repairs in other centres.

There were three pregnancies within the group following the fistula repair and prior to the supervised deliveries in question. One ended in a miscarriage, one delivered on the road on her way to a hospital and had a stillbirth after 18 hours of labour, one delivered by assisted vaginal delivery in another hospital, presenting after 2 days of labour, had a repeat of her fistula and a stillbirth.

The average time from fistula repair to delivery was 37.5 months (range 7–288 months).

Five patients had a planned elective caesarean as they were quite certain of their period dates. All of these had a live child.

Forty-one patients had an ‘emergency caesarean’; two were preformed following premature rupture of membranes (PROM) and the remainder at the onset of labour. One patient had twins (both were live born) and one patient had a still birth. The fetus was alive at the onset of labour, but there was a delay in instituting the caesarean delivery within the government hospital by 6 hours and the child was stillborn with the umbilical cord wrapped three times around the child’s neck.

There were three vaginal deliveries, one was a precipitous labour, delivering a live born child within 30 minutes of the onset of labour. The second was a mother at 26 weeks gestation with severe malaria and premature delivery. The neonate died within 30 minutes of delivery; the mother fully recovered. The third was a spontaneous delivery at approximately 30 weeks gestation. The neonate died within 1 hour of delivery. There are no facilities for premature infants within the country.

There were no recurrences of the obstetric fistula in any of the reported deliveries.

Thus, from 49 deliveries, there were 50 neonates (one twins) with the loss of three; two from prematurity and one from a delayed institution of a caesarean section.


It is difficult to make comparisons and conclusions from such uncontrolled data, but during the same study period, we have had one report of a woman whose family refused to take her to a hospital for a delivery following fistula repair, she died during labour. During the same time period, there have also been 24 women presenting following successful repair with repeat fistula arising from having another delivery. Of these 24, eight were by assisted vaginal delivery in a health centre or hospital, nine were vaginal deliveries at home, six were by caesarean section in a hospital after 2 days of labour (range 1.5–3 days), one by caesarean hysterectomy after 2 days of labour. All were stillbirths and as stated, all had a recurrence of their obstetric fistula. The reasons for not coming to the unit and waiting for a caesarean were that their husband did not let them, they lived too far away, were fearful of an operation or they were intending to come, but the labour started too quickly.

It is unknown how many other women may have delivered in the countryside after fistula repair and have either died, had a stillbirth or perhaps even had a live child. It is clear that only those who have survived and encountered a problem after their delivery returned for further care. It does, however, illustrate that unsupervised delivery following obstetric fistula repair is hazardous at best. This has been confirmed by other reports, which also recommend caesarean section as the best management following obstetric fistula repair.3–5

This study has pointed out, however, that all units treating fistula patients should strongly counsel their patients to deliver in an institution with ready access to a caesarean delivery. For those places where women live too far from emergency obstetric help, policy makers and fistula units should provide waiting areas for these women to ensure they do not run the risks of repeated morbidity or even mortality. Without the provision of waiting areas, the patients will stay at home and assistance during labour will be delayed, often with disastrous consequences.


The optimal way of managing labour following obstetric fistula repair is to provide a waiting area for women as they approach term and perform a timely caesarean section. This should be at around term in those where dates are known with some confidence or otherwise at the onset of labour.

Disclosure of interests

There are no conflicts of interest.

Contribution to authorship

The author (AB) managed the patients, collected and reviewed the notes and wrote the article.

Details of ethics approval

There is no ethics committee in the institution where this report was carried out. Approval was obtained from the director of the institution.


There was no specific funding.