Pregnancy and childbirth after repair of obstetric fistula in sub-Saharan Africa: Scoping Review

Authors


Abstract

Objective

To synthesise the evidence on pregnancy and childbirth after repair of obstetric fistula in sub-Saharan Africa and to identify the existing knowledge gaps.

Methods

A scoping review of studies reporting on pregnancy and childbirth in women who underwent repair for obstetric fistula in sub-Saharan Africa was conducted. We searched relevant articles published between 1 January 1970 and 31 March 2016, without methodological or language restrictions, in electronic databases, general Internet sources and grey literature.

Results

A total of 16 studies were included in the narrative synthesis. The findings indicate that many women in sub-Saharan Africa still desire to become pregnant after the repair of their obstetric fistula. The overall proportion of pregnancies after repair estimated in 11 studies was 17.4% (ranging from 2.5% to 40%). Among the 459 deliveries for which the mode of delivery was reported, 208 women (45.3%) delivered by elective caesarean section (CS), 176 women (38.4%) by emergency CS and 75 women (16.3%) by vaginal delivery. Recurrence of fistula was a common maternal complication in included studies while abortions/miscarriage, stillbirths and neonatal deaths were frequent foetal consequences. Vaginal delivery and emergency C-section were associated with increased risk of stillbirth, recurrence of the fistula or even maternal death.

Conclusion

Women who get pregnant after repair of obstetric fistula carry a high risk for pregnancy complications. However, the current evidence does not provide precise estimates of the incidence of pregnancy and pregnancy outcomes post-repair. Therefore, studies clearly assessing these outcomes with the appropriate study designs are needed.

Abstract

Objectif

Synthétiser les données sur la grossesse et l'accouchement après réparation de la fistule obstétricale en Afrique subsaharienne et identifier les lacunes dans les connaissances existantes.

Méthodes

Une analyse de la portée des études faisant état de la grossesse et de l'accouchement chez les femmes ayant subi une réparation de la fistule obstétricale en Afrique subsaharienne a été menée. Nous avons cherché des articles pertinents publiés entre le 1er janvier 1970 et le 31 mars 2016, sans restrictions méthodologiques ou linguistiques, dans les bases de données électroniques, les sources Internet générales et la littérature grise.

Résultats

16 études ont été incluses dans la synthèse narrative. Les résultats indiquent que de nombreuses femmes en Afrique subsaharienne restent désireuses de tomber enceinte après la réparation de leur fistule obstétricale. La proportion globale des grossesses après réparation estimée dans 11 études était de 17,4% (allant de 2,5% à 40%). Parmi les 459 accouchements pour lesquelles le mode d'accouchement a été rapporté, 208 femmes (45,3%) ont accouché par césarienne élective, 176 (38,4%) par césarienne d'urgence et 75 (16,3%) par voie vaginale. La récurrence de la fistule était une complication maternelle courante dans les études incluses, alors que les avortements/fausses couches, mortinaissances et décès néonatals étaient des conséquences fœtales fréquentes. L'accouchement vaginal et par césarienne d'urgence ont été associés à un risque accru de mortinatalité, la récurrence de la fistule ou même le décès maternel.

Conclusion

Les femmes qui tombent enceinte après réparation de la fistule obstétricale portent un risque élevé de complications de la grossesse. Toutefois, les données actuelles ne fournissent pas d'estimations précises sur l'incidence de la grossesse et les résultats de grossesse post-réparation. Par conséquent, des études évaluant clairement ces résultats avec des concepts appropriés sont nécessaires.

Abstract

Objetivo

Resumir la evidencia sobre el embarazo y parto después de la reparación de la fístula obstétrica en África subsahariana, e identificar las brechas que existen actualmente en el conocimiento.

Métodos

Se realizó una revisión de la evidencia más relevante de estudios que reportan embarazos y partos en mujeres que habían tenido una reparación de fístula obstétrica en África subsahariana. Buscamos artículos relevantes publicados entre el 1 de Enero de 1970 y el 31 de Marzo del 2016, sin restricciones metodológicas o de idioma, en bases de datos electrónicas, fuentes generales de internet y literatura gris.

Resultados

Se incluyeron 16 estudios en la síntesis narrativa. Los hallazgos indican que muchas mujeres en África subsahariana aun desean quedarse embarazadas después de la reparación de su fístula obstétrica. La proporción general de embarazos después de una reparación, según lo calculado en 11 estudios, era del 17.4% (rangos entre 2.5% y 40%). Entre los 459 partos para los que se reportaba la forma de alumbramiento, 208 mujeres (45.3%) dieron a luz mediante una cesárea electiva, 176 mujeres (38.4%) mediante una cesárea de emergencia y 75 mujeres (16.3%) mediante parto vaginal. La recurrencia de la fístula era una complicación materna común en los estudios incluidos mientras que los abortos espontáneos, mortinatos y muertes neonatales eran consecuencias fetales frecuentes. El parto vaginal y la cesárea de emergencia estaban asociados con un mayor riesgo de mortinato, recurrencia de la fístula e incluso muerte materna.

Conclusión

Las mujeres que se quedan embarazadas después de una reparación de fístula obstétrica tienen un alto riesgo de complicaciones del embarazo. Sin embargo, la evidencia actual no provee un cálculo preciso de la incidencia del embarazo y resultados del embarazo después de la reparación. Por lo tanto, se requieren estudios que evalúen claramente estos resultados con un diseño de estudio apropiado.

Introduction

Obstetric fistula (OF) is a preventable maternal condition whose eradication remains a major challenge in developing countries, especially in sub-Saharan Africa, where health systems are weak and adequate emergency obstetric care services are lacking [1, 2]. A review by Adler et al. [3] and a meta-analysis by Maheu-Girou et al. [4] reported similar estimates of the lifetime prevalence of OF in sub-Saharan Africa: between 1.0 and 1.57 per 1000 in women aged 15–49 years. Women living with fistula may suffer not only from other physical impediments but also psychologically and socially (divorce, isolation or rejection by relatives and community) [5, 6].

In the past decade, an increasing commitment to holistic care of fistula has occurred in sub-Saharan Africa [2, 7]. After international funding of fistula care programmes worldwide [8-10], many countries have developed national plans for eliminating obstetric fistula, including a preventive component to strengthen emergency obstetric care but also the training and equipment of health facilities for the treatment of fistula. The latter contributed to achieving satisfactory post-operative closure rates for an increased number of women [11-15]. More research on fistula has resulted in a better understanding of determinants of both fistula development and surgical outcomes [16-18].

As reported in the literature, the fistula occurring during childbirth is, in the majority of cases (over 80%), associated with stillbirth [15, 18-20]. In addition, a recent review shows that the fulfilment of their social roles is a key factor for rehabilitation of women after OF repair [21]. In many countries where fistula incidence is high, becoming a mother is an important component of fulfilling these social roles [20, 22-24]. However, for a repaired woman, becoming pregnant and giving birth might increase the risk of fistula recurrence [22-24]. Also, repair after previous attempts to close the fistula is often technically difficult for the surgeon and exhausting for the woman [25, 26].

As efforts to increase access to treatment of fistula in low-income countries continue [2], pregnancy and childbirth after OF repair are an emerging concern because repaired women remain at high risk of developing either a recurrence of fistula or a complication during pregnancy and childbirth. It is usually recommended to perform an elective caesarean section for subsequent delivery of women repaired for fistula [23, 27, 28], but there is little evidence whether and to what extent this recommendation is followed and implemented. The few data available on pregnancy and childbirth after OF repair and the factors that influence its outcome have not been synthesised in order to guide fistula care programmes and to identify research priority areas. The objectives of this review were to [1] identify and describe studies examining pregnancy and childbirth after repair of OF in sub-Saharan Africa; [2] synthetise the evidence on main outcomes, that is pregnancy and childbirth and factors associated with these outcomes; and [3] identify the existing knowledge gaps along with future research priorities.

Methods

Search strategy

A scoping review of studies reporting on subsequent pregnancy and childbirth in women who underwent repair for obstetric fistula in sub-Saharan Africa was conducted using the guidance for conducting scoping reviews developed by the Joanna Briggs Institute [29]. Our aim was to get a broad overview of the existing evidence, which is consistent with scoping studies [29, 30].

A literature search for relevant articles published between 1 January 1970 and 15 March 2016, without methodological or language restrictions, was carried out in electronic databases PubMed (1970–2016), Popline (1970–2016), Embase (1970–2016), Web of Science (1973–2016), Global Library-African Index Medicus (2006–2016), general Internet sources (Google Scholar), grey literature and reference lists for primary studies on pregnancy and childbirth after obstetric fistula repair using the following search terms: ‘fistula’, ‘pregnancy’, ‘delivery’, ‘childbirth’, ‘birth’ and ‘reintegration’. The search strategies were designed according to the specifications of each database. There was no difference between the databases in terms of results yielded. We did not include ‘sub-Saharan Africa’ as search term to allow a broad search that would include all papers published on the study topic. The last search was carried out on 15 March 2016.

Inclusion/Exclusion criteria

The review included studies conducted in sub-Saharan Africa, between 1 January 1970 and 15 March 2016 which examined pregnancies and childbirth in women after repair of an OF. We excluded case studies and studies on non-obstetric fistula.

Screening

After removing duplicates and following to the review protocol developed a priori, two authors (AD and BU) independently reviewed titles and abstracts of all articles for their relevance and compliance with the selection criteria followed by full-text screening of selected articles. Discrepancies were resolved by discussion between two reviewers with the assistance of an experienced researcher (TD). Reference Manager® version 12.01 (Thomson Reuters) was used for screening/checking for duplicates and examining titles and abstracts. We also reported on limitations of the included studies.

Data collection

A standardised data collection form was used to extract study data, and evidence tables were generated to describe and summarise the characteristics of included studies, study outcomes and potential factors associated with study outcomes. Study characteristics including author, year of publication, time period covered by the study, country, setting (whether rural or urban, hospital based or population based or both), study design, sample size and study population were collected. Study outcomes on the maternal side comprised the occurrence of pregnancy (yes/no), pregnancy outcomes for the mother (abortion/miscarriage, term pregnancy with delivery, ongoing pregnancy at the end of the study and maternal death) and recurrence of the fistula. On the child side, study outcomes were live birth, stillbirth or neonatal death. The potential factors associated with pregnancy (or getting pregnant) included the reported desire for children, reported fear of fistula recurrence, resuming and timing of sexual intercourse, marital status after repair, use of family planning methods. Finally, factors associated with maternal and neonatal outcomes such as the use of antenatal care during pregnancy and delivery mode (at home or at hospital, vaginal delivery or planned vs. emergency C-section) were collected.

Results

Description of the studies included

The initial literature search yielded 2717 studies and 29 additional studies were identified through other sources. After removing duplicates, we screened the titles of the remaining 1947 studies and excluded 1767 irrelevant studies. A total of 133 of 180 were excluded based on the abstracts. Of the 47 full texts assessed for eligibility, 31 were excluded and 16 studies were included in the qualitative synthesis. Figure 1 shows the PRISMA flow chart of the study selection process [31].

Figure 1.

PRISMA flow chart of the studies selection process for the scoping review of pregnancy and childbirth after obstetric fistula repair in sub-Saharan Africa (1970–2016) [[31]].

The 16 studies that met our inclusion criteria were published between 1978 and 2016, mostly in the past 10 years. Studies were conducted in only seven sub-Saharan African countries: five in Nigeria [27, 28, 32-34], four in Ethiopia [5, 22-24], two in Malawi [35, 36], one in Niger [37], one in Mali [38], one in both Mali and Niger [39], one in Sierra Leone [40] and one in Eritrea [41] (Table 1).

Table 1. Characteristics of 16 studies reporting on pregnancy and childbirth after repair of obstetric fistula in sub-Saharan Africa
Study (Author and Year)Location and settingStudy designStudy period or follow-up timeStudy population and Sample sizeStudy objectivesOutcomes assessed
PregnancyChildbirth
  1. a

    Denominator: number of women discharged from hospital with the fistula successfully closed (required for the calculation of the proportion of pregnancies).

Drew et al., 2016 [35]Malawi (Urban and Rural, community based)Qualitative study1–2 years after fistula repair

20 women from seven health districts

Aged between 18 and 76 years

To assess long-term outcomes among women who underwent obstetric fistula repairYesYes
Lawani, Iyoke & Ezeonu, 2015 [34]Nigeria (Urban hospitals)Prospective cohort study24 months follow-up per woman (January 1, 2011 to December 31, 2013)188 women followed up among 211 included (23, 10.9% loss to follow-up)To determine the contraceptive practices of women after successful surgical repair of obstetric fistula and to assess determinants of uptake and the pregnancy rate in the first year of follow-up.YesNo
Tembely et al., 2014 [38]

Mali

(Urban referral hospital)

Prospective cohort study13 months (June 2008–June 2009)Not providedTo study the sexuality of women victims of obstetric urogenital fistula after surgical managementYesNo
Ford, Narrainen-Poulle & Mansaray, 2013 [39]

Sierra Leone

(Urban hospital)

A retrospective review of electronic and paper databases2010–2013 (32 months)

82 babies from previous VVF repaired patients among 2735 deliveries.

Denominator not reported.a

To examine the obstetric outcomes of women who attended with pregnancies after fistula repairYesYes
Maulet & Macq, 2012 [41]Niger, Mali (community based)Mixed-methods cohort study18 months (2008–2009)93 women of reproductive age from 120 fistula patients (quantitative study) and 15 current and former fistula patients (qualitative study)To explore fistula patients’ reproductive health paths and perspectivesYesYes
Wilson et al., 2011 [36]

Malawi

(Rural)

Retrospective Cohort study2007–2009 with a minimum follow-up time of 6 months17 women repaired for OF out of 32 womenTo evaluate fertility and childbearing outcomes after obstetric fistula repair or in the absence of repairYesYes
Browning, 2009 [22]

Ethiopia

(Fistula referral hospital)

Retrospective cohort study

44 months (February 2005 to September 2008)

Average time from fistula repair of 37.5 months (range 7–288 months)

49 pregnant women following repair of OF.

Denominator not reported.a

To report on the experience of managing pregnancies from women who underwent repair for obstetric fistulaYesYes
Nielsen et al., 2009 [21]

Ethiopia

(Rural, community based)

Prospective cohort study14–28 months following fistula repair38 women who undergone fistula repairTo assess urinary and reproductive health and quality of life following surgical repair of obstetric fistulaYesYes
Browning & Menber, 2008 [23]Ethiopia (urban hospital)Prospective cohort study6 months240 returning women of 390 index casesTo quantify surgical and quality of life outcomes 6 months after obstetric fistula repairYesNo
Muleta et al., 2008 [5]

Ethiopia

(Rural setting)

Cross-sectional mixed-methods surveyNo follow-up time reported13 women with repaired fistula and 39 untreated among 22826 women of reproductive age (15–49 years)To assess health, social and psychological problems encountered by women with treated and untreated obstetric fistulaYesYes
Turan, Johson & Polan, 2007 [40]Eritrea (rural hospital)Qualitative study6–10 months (2004)15 women returning for follow-up after repair of 31 study participantsTo present the findings from qualitative interviews with new and returning fistula repair patients and their family membersYesNo
Emembolu, 1992 [32]Nigeria (urban hospital setting)Retrospective cohort study60 months (January 1986–December 1990)

75 women with repaired VVF are compared to 80 women with a non-repaired VVF.

Denominator not reported.a

To determine the factors associated with the improvements in the pregnancy outcome in the VVF patients following successful surgical correctionYesYes
Docquier, 1988 [37]Niger (urban hospital)Prospective cohort study96 months (1977–1984) with a minimal follow-up time of 2 months200 of reproductive age assessed from 230 returning women of 377 women successfully repaired.To describe the gynaecological and obstetrical status of women successfully repaired from obstetric fistulaYesYes
Otubu, Kumi & Ezem, 1981 [27]Nigeria (urban hospital)Retrospective cohort study35 months (February 1, 1977–December 30, 1979)110 pregnant women of 138 following successful repair of VVF. Denominator not reported.aTo examine the complications of pregnancy and labour in pregnant VVF patients following successful repairYesYes
Evoh & Akinla, 1978 [33]Nigeria (Urban hospital)Retrospective cohort study

10 years

(1966–1976)

148 women with successful repair of obstetric fistulaTo assess how Nigerian women with successfully repaired vesico-vaginal fistulae of obstetric origin fulfilled reproductive functionYesYes
Aimakhu, 1974 [28]Nigeria (Urban hospital)Retrospective cohort study

10 years

(1957–1966)

246 women successfully repairedTo study the reproductive functions after the successful repair of obstetric vesicovaginal fistulaeYesYes

Eleven studies were conducted in hospitals, with 10 studies in urban referral hospitals. Five studies were community based. In terms of methods used, 12 of the 16 were cohort studies (seven using a retrospective design and five a prospective design), two used only qualitative methods, one used a mixed-methods approach and one was a cross-sectional mixed-methods survey. Three studies (two qualitative and one cross-sectional mixed method) additionally reported on families and non-repaired women, and one documented pregnancies in women living with non-repaired fistula.

A sample size was available for 15 of 16 studies and ranged from 13 to 390 women. The length of follow-up time post-repair across all studies ranged from 6 months to 10 years, with a median follow-up time of 30 months (interquartile ranges of 12 and 60 months). The outcomes assessed across studies were the occurrence of pregnancy (all studies), mode of delivery (vaginal delivery, emergency or elective C-section), perinatal outcomes (12 studies), recurrence of fistula (nine studies), abortion/miscarriage (eight studies) (Tables 2 and 3). Maternal complications (during pregnancy and at delivery) were documented in nine studies. Ten studies reported on the predictors of pregnancy, 14 studies documented the outcome of the pregnancy for the mother (Table 2) and 10 reported on the factors associated with the birth outcomes (Table 3).

Table 2. Pregnancy, pregnancy outcomes and factors associated with getting pregnant in 16 studies reporting on pregnancy and childbirth after repair of obstetric fistula in sub-Saharan Africa
Study (Author, Year and Type)Number of women followed upNumber of women who became pregnantProportion of women who got pregnantPregnancy outcomeFactors associated with getting pregnant
Abortion/MiscarriageTerm pregnancy
  1. OF, obstetric fistula; FP, family planning methods.

Drew et al., 2016 [35]20 women3 women15.0%2/3 pregnancies1/3 pregnancy (ongoing)14 women (70%) were sexually active and 2 women using contraception (implant); 9 women desired another child but feared fistula recurrence in a future pregnancy
Lawani, Iyoke & Ezeonu, 2015 [34]188 women

13 women

(12 pregnancies among 118 women not using FP vs. 1 pregnancy in 70 women using)

6.9%

(10.2% in women not using FP vs. 1.4% in those using)

Not reportedNot reported

28% of the 188 women commenced sexual activity within 3 months post-discharge despite being counselled on the risk.

Family Planning use reduced the risk of pregnancy.

Tembely et al., 2014 [38]Not providedNot reported13%Not reportedNot reported62% of their sample had resumed sexual activity but 60% did not desire another children.
Ford, Narrainen-Poulle & Mansaray, 2013 [39]Not reported81 womenNot reported081/81 pregnanciesNot reported
Maulet & Macq, 2012 [41]93 women18 women19.4%6/18 pregnancies

12/18 pregnancies

(5 ongoing at the end of study)

Not reported
Wilson et al., 2011 [36]17 women6 women (in total 10 pregnancies)35.3%7/10 pregnancies3/10 pregnancies (2 ongoing)Not reported
Browning, 2009 [22]Not reported49 womenNot reported049/49 pregnanciesNot reported
Nielsen et al., 2009 [21]38 women5 women13.2%0/5 pregnancy

5/5 pregnancies

(1 ongoing)

27 (71%) were married and sexually active, 11 (41%) used contraception
Browning & Menber, 2008 [23]240 women6 women2.5%06 pregnanciesOnly 84/240 women had resumed sexual intercourse (35%)
Muleta et al., 2008 [5]13 women2 women15.4%02/2 pregnancies

7 had resumed sexual activity

Fear from fistula recurrence and incontinence identified as factors delaying the resuming of sexual activity

Turan, Johson & Polan, 2007 [40]15 women00.0%Not reportedNot reportedWomen who desired another child were sexually active while divorced and widowed women were not
Emembolu, 1992 [32]Not reported75 women (3 outcome unknown)Not provided3/72 pregnancies69/72Not reported
Docquier, 1988 [37]200 women

80 women

(Mean time between repair and pregnancy: 10 months)

40%5/80 pregnancies

75/80

(36 ongoing at the end of study)

87% had resumed with menstruation (were of reproductive age)
Otubu, Kumi & Ezem, 1981 [27]Not reported110 womenNo reported0110 pregnancies (100%)Among the 110 pregnant women, there were more women with juxta-urethral (39%) and mid-vaginal fistula (37%) than women with other types of fistula
Evoh & Akinla, 1978 [33]148 women

31 women

(in total 38 pregnancies)

Time to pregnancy varied from 2–24 months

20.9%1/38 pregnancies37/38There were more pregnant women with juxta-cervical (N = 18) and juxta-urethral (N = 12) fistula than with all other types of fistula (N = 8)
Aimakhu, 1974 [28]246 women

48 women

(in total 65 pregnancies)

19.5%5/65 pregnancies60/65Not reported
Total

1218 women

From 11 studies that reported on women followed up

212 women (with 240 pregnancies) from the 11 studies that reported the number of women followed up)

527 women (in total 555 pregnancies) from 15 studies

17.4%

(from the 11 studies reporting the number of women followed up)

29 cases

from 15 studies reporting the number of pregnancies

510 pregnancies (45 ongoing) from 15 studies reporting the number of pregnancies 
Table 3. Maternal and neonatal outcomes and factors associated with these outcomes in 16 studies reporting on pregnancy and childbirth after repair of obstetric fistula in sub-Saharan Africa
Study (Author, Year and Type)Number of deliveries (out of term pregnancies)Elective CSEmergency CSVaginal deliveriesMaternal and neonatal outcomesFactors associated with maternal and neonatal outcomes
Live birthsStillbirths/neonatal deathLive birthsStillbirths/neonatal deathLive birthsStillbirths/neonatal death
  1. CS, C-section; N/A, not applicable.

Drew et al., 2016 [35]0/1000000Not reportedNo delivery was reported Miscarriage was frequent
Lawani, Iyoke & Ezeonu, 2015 [34]Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
Tembely et al., 2014 [38]Not providedNot providedNot providedNot providedNot providedNot providedNot providedNot providedNot provided
Ford, Narrainen-Poulle & Mansaray, 2013 [39]

81/81

(with 82 babies)

22/81 deliveriesNot reported57/81 deliveriesNot reported3/81 deliveriesNot reported

Foetal size at delivery was small (mean birth weight of 2.8 kg)

No recurrence of fistula

No maternal death

Elective CS improves maternal/neonatal outcomes
Maulet & Macq, 2012 [41]7/123/7 deliveries000

3/7 deliveries

(2 medically assisted and 1 not)

1/7 deliveries not medically assisted1 suspicion of fistula recurrenceDelivery without medical assistance appeared to be associated with worse maternal and neonatal outcomes
Wilson et al., 2011 [36]1/300001/1 delivery01 recurrence of fistulaPerinatal mortality high among women becoming pregnant post-repair. Recurrence of fistula is associated with vaginal delivery
Browning, 2009 [22]

49/49

(with 50 neonates)

5/49 deliveries041/49 deliveries1/49 deliveries1/49 deliveries2/49 deliveries

No recurrence of fistula

No maternal death

Neonatal death associated with vaginal deliveries and emergency CS when compared to planned elective CS; CS was found to reduce bad maternal outcomes
Nielsen et al., 2009 [21]

3/5

(and 1 death in labour)

1/3 deliveries

(Elective CS planned for all women)

001/3 deliveries01/3 deliveries

1 maternal death in labour

1 fistula recurrence (emergency CS)

Only 22 of 38 (59%) recalled receiving this advice.

Elective CS improved maternal/neonatal outcomes

Browning & Menber, 2008 [23]6/6Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
Muleta et al., 2008 [5]2/22/2 deliveriesNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot provided
Turan, Johson & Polan, 2007 [40]0/0Not reportedNot reportedNot reportedNot reportedNot reportedNot reported N/A
Emembolu, 1992 [32]69/69

19/69 deliveries

(51/75 pregnant women were planned for elective CS)

Not reported35/69 deliveriesNot reported15/69 deliveriesNot reported

9 perinatal deaths

4 women with anaemia; no maternal death but 8 recurrences of fistula

4 women who did not attend their elective appointment got early perinatal deaths

Obstructed labour, vaginal delivery and emergency CS associated with recurrence of fistula;

Early referral and acceptance of elective CS and multiparity had positive impact on the outcomes for mother and child.

Docquier, 1988 [37]39/7514/39 deliveriesNot reportedNot reportedNot reported25/39 deliveriesNot reportedNot reportedNot reported
Otubu, Kumi & Ezem, 1981 [27]110/110

75/110 deliveries

(Elective CS planned for all women)

5/110 deliveries14/110 deliveries8/110 deliveries3/110 deliveries6/110 deliveries

Antenatal complications: anaemia (11.8%) and urinary tract infection (15.4%).

1 uterus rupture and 1 maternal death;

5 recurrence of fistula

Delay/refusal in seeking car or refusal of elective CS led to more maternal/neonatal complications. Maternal/neonatal outcomes were worse with vaginal and emergency CS compared to elective CS
Evoh & Akinla, 1978 [33]37/37

30/37 deliveries

(CS planned for all women)

2/37 deliveriesNot reportedNot reported3/37 deliveries2/37 deliveries

5 recurrence of fistula after delivery and 1 uterine rupture

No maternal death but 2 women with stress incontinence after CS

Delay/refusal in seeking care or refusal of CS led to more negative maternal and neonatal outcomes

Husbands’ agreement increases acceptance of elective CS;

Resumption of coitus risk for recurrence of fistula

Aimakhu, 1974 [28]

59/60 deliveries

(and 1 death during labour)

30/58 deliveries0/58 deliveries17/58 deliveries2/58 deliveries5/58 deliveries4/58 deliveries2 recurrence of fistula and 2 maternal deaths (1 in labour and 1 CS)Elective CS increased neonatal survival; Vaginal delivery associated with bad maternal and neonatal outcomes, especially with recurrence of fistula
Total

463/510

in 14 studies

201 live births (in 13 studies)7 Stillbirths/neonatal death164/201 live births12 Stillbirths/neonatal death59/201 live births16 Stillbirths/neonatal death

23 recurrence of fistula

4 maternal death

2 uterine ruptures

 

Occurrence of pregnancy post-repair of obstetric fistula

The occurrence of pregnancy and the factors associated with getting pregnant across studies are summarised in Table 2. Fifteen of these 16 studies reported on numbers and recorded 555 pregnancies in 527 women with three studies including multiple pregnancies (28 additional pregnancies). Only 11 studies reported on the number of women followed up after repair. In these studies, 212 of 1218 followed up women became pregnant after repair (in total 240 pregnancies), corresponding to an overall proportion of 17.4% (range 2.5–40%). Of the 555 pregnancies reported in the included studies, the outcome of 539 pregnancies was known: 29 (5.4%) ended in abortion or miscarriage, 465 were term pregnancies (86.3%) and 45 pregnancies were ongoing at the time of the study (8.3%). Overall, 465 term pregnancies resulted in 463 deliveries with two reported maternal deaths.

Only six studies reported that women included were successfully repaired at discharge from hospital. None of the studies stratified data on pregnancies after repair and pregnancy outcomes by fistula repair status at discharge.

Factors associated with getting pregnant post-obstetric fistula repair

Ten studies reported on the predictors of pregnancy after repair (Table 2). As expected, married women and those women who said they wanted more children at discharge were more likely to resume sexual activity [22, 24, 35, 37]. More women with juxta-cervical, juxta-urethral or mid-vaginal fistula became pregnant post-repair than women with other types of fistula [27, 33]. Inversely, multiparous women, women who reported having a fear of fistula recurrence or those using family planning methods, were less likely to become pregnant [5, 34, 36]. In one study, for instance, there was only one pregnancy among 70 women using family planning (FP) methods (1.4%) while 12 pregnancies were recorded in 118 women who were not using a FP method (10.2%) after 24 months follow-up [34]. Even after being counselled on the risk of sexual activity at the time of their surgery, more than one-quarter of women resumed sexual activity within 3 months of repair [34].

Maternal and neonatal outcomes post-fistula repair

Maternal and neonatal outcomes as reported in the 16 included studies are summarised in Table 3. Overall 463 deliveries were recorded in the included studies but for only 459 the mode of delivery was reported. A total of 208 women (45.3%) delivered by elective caesarean section (CS), 176 (38.4%) by emergency CS and 75 (16.3%) by vaginal delivery.

In three studies [27, 32, 33], only 131 of 198 pregnant women (66.2%) appointed for an elective CS respected the appointment. The most common maternal complications were the recurrence of a fistula (23/463 deliveries; 5.0%), maternal death (four cases; 0.9%) and two uterine ruptures (0.4%). One study [27] included antenatal complications such as anaemia (13/110 women, 11.8%) and urinary tract infections (17/110 women; 15.4%).

Overall the proportion of stillbirths across studies was 7.6% (35/459 neonates). However, considering the perinatal outcome by mode of delivery across studies, there were more stillbirths associated with a vaginal delivery (16/75 neonates, 21.3%) than emergency CS (12/176 neonates, 6.8%) and elective CS (7/208 neonates, 3.4%).

Factors associated with maternal and neonatal outcomes post-fistula repair

Most of the studies (12 of 16) identified elective CS as the main mode of delivery resulting in a better maternal and neonatal outcome of pregnancies after successful repair of an obstetric fistula. Delays in seeking care or refusal to deliver with skilled birth attendants [23, 27, 32], non-use of antenatal care [27], were identified as factors that contributed to delaying delivery at a hospital. Vaginal delivery without medical assistance and emergency CS were listed as factors increasing the risk of recurrence of a fistula post-repair. One study [33] reported that the resumption of sexual activity within 3 months was associated with the recurrence of fistula.

Discussion

Our scoping review on pregnancy and childbirth post-repair of an obstetric fistula in sub-Saharan Africa found that many women still desire to become pregnant after the repair of their OF [22, 24, 33, 36, 41]. However, the incidence of pregnancies was not reported in the reviewed studies, and the overall incidence could not be estimated in this review. The overall estimate of the proportion of women who became pregnant during the follow-up was low but varied widely across studies. Even though evidence is scarce [42], return to fertility has been reported to be difficult in women after OF repair in a number of studies [5, 28, 30, 43]. It has been stated that women with a previous traumatic birth experience had fewer pregnancies post-repair than women without such an experience [44]. Women treated for OF are usually advised to wait for 3–6 months before resuming sexual intercourse [45, 46] to allow for complete healing. In addition, the fear of fistula recurrence might delay the resuming of sexual activities [43, 45, 47]. Divorce and mental trauma, but also urinary infections or iatrogenic amenorrhoea, which are common among fistula women, might reduce the probability of getting pregnant [27, 28, 30].

The low proportion of pregnancies and the wide range across studies observed in this review might be explained by (i) the limited number of studies reporting the number of women followed up over time; (ii) the heterogeneity of follow-up time across studies (two-fifth of the studies included had a follow-up time of <2 years); (iii) the use of purposive and small samples using qualitative or cross-sectional designs which are not representative of all repaired women; (iv) the lack of information on the proportion of women of reproductive age across one-third of the included studies and (v) the fear of recurrence of the repaired fistula or formation of a new fistula following a pregnancy [33].

Women who get pregnant after repair of OF carry a high risk of pregnancy and childbirth complications [22]. Across studies, the mode of delivery was an important determinant of pregnancy outcomes. Stillbirths, recurrence of the fistula or even maternal death were more likely to occur with vaginal delivery and emergency CS than with a scheduled CS [22, 27, 28, 32, 33, 36, 39, 40]. Therefore, delivery with a skilled birth attendant, especially through elective C-section (CS) was identified across all studies as a key factor to achieve good maternal and child health outcomes.

Our review reveals that many women are encouraged by healthcare providers to attend antenatal care and deliver in a hospital by elective CS. However, for various reasons, none of the women could follow these recommendations [27, 32, 33]. The reasons were poverty, refusal due to cultural reasons, husband or family opposition [22]. The experience of women who fear that their husbands refuse to send them to the hospital if they become pregnant has been reported by Donnely et al. [47]. Studies generally described women suffering from obstetric fistula being poor and living in rural areas with insufficient access to emergency obstetric care [45, 46]. It is therefore important to ensure transportation and free care are offered as facilitators of access to women planned to deliver in health facilities [46]. If barriers to a delivery in a health facility are not removed, access to elective C-section for women pregnant after successful repair of OF might be insufficient.

Our findings suggest that recurrence of fistula is a common maternal complication of pregnancy after the repair of OF; abortions/miscarriage, stillbirths and neonatal deaths are frequent foetal consequences. As reported in some studies on fistula outcomes [14, 18, 43], for up to 90% of women who develop an OF, the delivery ends in a stillbirth. This emphasises the need to strengthen and extend follow-up of women post-repair in order to preserve the health of the mother and her future babies. Post-repair counselling about fistula and the risk factors for OF recurrence, such as early resuming of sexual intercourse, as well as the implementation of a community-based follow-up care have been recommended [21, 33, 47]. However, questions remain on how these recommendations can be realised and integrated into routine maternal and child health programmes that involve husbands and families [22].

Finally, although a few studies specified if women were successfully repaired for their OF, in most studies information was missing whether women followed up had their fistula closed or were continent at discharge from the repair hospital. In addition, none of the included studies reported on pregnancies and pregnancy outcomes by fistula repair status (closed vs. continent at discharge).

Our review has some limitations: (i) missing information in studies did not permit good estimates of the study outcomes; (ii) it included qualitative studies that did not focus on providing the estimates of study outcomes; (iii) the geographical distribution of studies included was rather narrow and might therefore affect the external validity of the findings; (iv) there was heterogeneity of recruitment strategies and follow-up of women included in studies. Nonetheless, this review forms the basis for more research on the life of women successfully repaired for OF in sub-Saharan Africa.

Conclusion

Pregnancy and childbirth after repair of obstetric fistula carry a high risk of adverse outcomes in sub-Saharan Africa. Current evidence does not provide precise estimates of the incidence of pregnancy and pregnancy outcomes post-OF repair. We need to set up a data collection and reporting system for women after OF repair that could be integrated into the existing system, and studies assessing outcomes by repair status and providing more precise estimates of their incidence.

Acknowledgement

We thank Dominique Dubourg, Abdoulaye II Toure, Mark Barone, Steven Arrowsmith, Joseph Ruminjo, Vera Frajzyngier, Bethany Cole and Armando Seuc.

Ancillary