A community-based long-term follow up of women undergoing obstetric fistula repair in rural Ethiopia
Dr HS Nielsen, The Fertility Clinic, Rigshospitalet, Copenhagen University Hospital, DK-2200 Copenhagen Ø, Denmark. Email firstname.lastname@example.org
Objectives To assess urinary and reproductive health and quality of life following surgical repair of obstetric fistula.
Design Follow-up study.
Setting A newly established fistula clinic (2004) at Gimbie Adventist Hospital, a 71-bedded district general hospital in West Wollega Zone, in rural Western Ethiopia.
Population Thirty-eight women (86%) of 44 who had undergone fistula repair were identified in their community.
Methods Community-based structured interviews 14–28 months following fistula repair, using a customised questionnaire addressing urinary health, reproductive health and quality of life.
Main outcome measures Urinary health at follow up was assessed as completely dry, stress or urge incontinence, or fistula. King’s Health Questionnaire was modified and used for the quality-of-life assessment.
Results At follow up, 21 women (57%) were completely dry, 13 (35%) suffered from stress or urge incontinence and three (8%) had a persistent fistula. Surgery improved quality of life and facilitated social reintegration to a level comparable to that experienced before fistula development for both women who were dry and those with residual incontinence (P = 0.001). For women still suffering from fistula no change was seen (P = 0.1). Four women became pregnant following their surgery, among which there was one maternal death, three stillbirths and one re-occurrence of fistula.
Conclusion Community-based, long-term follow up after fistula repair succeeded in Western rural Ethiopia. Despite one-third still suffering stress or urge incontinence, the women reported improved quality of life and social reintegration after fistula closure.
Obstetric fistula, an abnormal opening between the vagina and bladder or rectum, is one of the most devastating complications of childbirth. The condition is extremely rare in the developed world, where the majority of women receive skilled obstetric care and early recourse to caesarean section prevents prolonged obstructed labour. In contrast, obstetric facilities in many developing countries are woefully inadequate to meet the needs of the population, and many women cannot access skilled obstetric care when needed. As a consequence approximately 2 million women worldwide are suffering from untreated obstetric fistula.1 In Ethiopia, it is estimated that 9000 women develop an obstetric fistula each year, of which only 1200 are surgically repaired.2
Despite many reports on the primary closure rates of fistula surgery,3–5 there is a paucity of data on long-term outcomes following fistula repair, in particular a lack of data on urinary symptoms, reproductive health and quality of life among the treated populations. Reasons cited for this lack of population-based follow up include the remote and inaccessible residence of most women with fistula, poor roads, the lack of telephones and postal services, and the relatively high cost of traveling.4,6–8
Here we present a community-based long-term follow-up study that assessed reproductive and urinary health, quality of life among women who underwent fistula repair at the fistula clinic at Gimbie Adventist Hospital (GAH) in rural Western Ethiopia.
Gimbie Adventist Hospital is a 71-bedded district general hospital in West Wollega Zone, in rural Western Ethiopia. The hospital is a ‘not-for-profit’ private institution that provides 24-hour medical, surgical, paediatric, obstetric and gynaecological services for a population in excess of 500 000. Since 2004, the hospital has been working in partnership with Maternity Worldwide, a non-governmental organisation with the primary objective to reduce maternal mortality and morbidity in developing countries. The Safe Motherhood Program includes the provision of essential obstetric care services, community education and women’s empowerment through a network of women’s groups providing micro-credit facilities. Additionally, financial barriers to obstetric services have been minimised through the establishment of a community managed voucher scheme, the Safe Birth Fund, which entitles women to receive highly subsided reproductive health care at GAH and affiliated clinics. In December 2004, Maternity Worldwide established a fistula clinic at GAH as part of the Safe Motherhood Program.
Participants and procedures
Between December 2004 and July 2006, three separate fistula surgery projects were conducted at GAH. The first project was undertaken in collaboration with Addis Ababa Fistula Hospital (AAFH), and the two subsequent projects were initiated solely by the fistula clinic at GAH. All patients received free treatment. Prior to each project, all women who suffered from constant and continuous leakage of urine following childbirth were invited to come for assessment. Information was sent to communities through community and women’s group leaders and local churches and mosques. Additionally, relevant government departments and all health facilities in the Zone were informed of each project and asked to identify women in their communities suspected of having obstetric fistula. Announcements were also made on the local radio station. In excess of 200 women came forward for assessment across the three projects. A gynaecologist undertook screening to identify those eligible for surgery. Fifty-eight women with fistula were identified, of these 14 were deemed too complicated for our facility and were referred to AAFH. Forty-four women with vesico-vaginal fistula were operated at GAH. Four women in the second project underwent a repeat surgical procedure in the third project. Hence, a total of 48 operations were performed on 44 women (20 procedures in the first project, 16 in the second project and 12 in the third project). Baseline information on all 44 women is presented in Table 1.
Table 1. Baseline data on 44 patients treated for obstetric fistula
|Mode of transport to hospital|
|Cause of fistula|
|Obstructed labour||43 (98%)|
|Mean age when fistula developed||24 (range 13–72)|
|Mean number of years living with fistula|| 3 (range 0.2–30)|
|Mean number of previous live births||1 (range 0–6)|
|Mean number of previous stillbirths||1 (range 0–6)|
|Previous fistula surgery at other institution|
|Stress incontinence also present|
|Urge incontinence also present|
All women were admitted to hospital prior to surgery for a high calorie diet and relevant treatment of urinary tract infection, nematode infections, malaria and syphilis. All fistula repairs were performed using a vaginal approach under spinal anaesthesia. The fistula(e) was identified and widely mobilised for the repair. Preferably, a two-layer tension-free closure was performed using absorbable sutures. For larger fistulae, and those involving the urethro-vesical junction and urethra a Martius pedicle graft of fibrofatty tissue from the right labium majus was used to secure and increase vascularity. A dye test was performed prior to closure of the vaginal wall. Forty-one women (93%) had a negative dye test following the primary surgical procedure. In the second project, a negative dye test was followed by a Marshall stress test for incontinence. For those with symptoms of stress incontinence, those with almost absent urethras or with circumferential fistulae, a pubococcygeal sling or urethral tape procedure was performed (n = 16).
Following surgery, postoperative antibiotics were administered to all women. Continuous bladder drainage was maintained for a minimum of 14 days on all women. The average length of postoperative stay was 20 days (range 11–27). On discharge from hospital following the primary surgical procedure, 35 (80%) women were completely dry, three (7%) suffered from stress incontinence, two (5%) suffered from urge incontinence and four (9%) continued to have a fistula. Three of the four women with incomplete fistula closure were re-operated at GAH. Two had a negative dye test following repeat surgery and were dry upon going home; for the third the fistula could not be closed and she was subsequently referred to AAFH.
Before discharge all women were counselled about the risks of future pregnancy and were advised to seek antenatal care and to attend hospital approximately 2 weeks prior to expected delivery in any future pregnancies. Contraception was provided if required. Where relevant, women were also given a ‘Safe Birth Fund’ voucher, which entitled them to subsidised medical treatment at GAH for any future pregnancies.
We devised a questionnaire to assess urinary and reproductive health as well as quality of life among an illiterate audience within a rural setting. The questionnaire was translated into the Afaan Oromo language and then translated back into English by a different individual to ensure accuracy of translation. The questionnaire was then piloted on six female hospital staff at GAH and two fistula patients who had been treated at AAFH to test comprehension and ease of completion. As a result of the pilot, some questions were simplified and a grading scale used in the questionnaire was modified to a scale of 1–3 instead of 1–5. The final questionnaire consisted of: five questions on urinary health; six questions concerning reproductive health; two questions concerning remembrance of advice given for hospitalisation in future pregnancies; and four questions concerning patients’ understanding of the cause of fistula, how it can be prevented, and awareness of others suffering from fistula in their community. Quality of life was assessed using 17 questions. The questions regarding urinary health and quality of life was based on the King’s Health Questionnaire, a validated assessment tool for the quality of life of women with urinary incontinence. (The questionnaire is available in Supporting Information.)9 At follow up, the respondents were asked to evaluate three different time periods: before developing a fistula, during the time when they had a fistula, and the current time.
To identify women for follow-up announcements were made as for the original fistula projects. The lead investigator met with the leaders of all villages from which women had been treated. A list of names was also distributed to local health facilities and the local government departments of Health, Women’s Affairs and Social Affairs. Twenty-two women were interviewed in their homes and 16 attended GAH for follow-up interview. Transport and accommodation costs were provided for the women attending interview at GAH. The mean length of time between surgery (final episode for those who had two procedures) and follow-up interview was 21 months (range 14–28). Seven nurses and midwives at GAH were trained by the lead investigator to conduct patient interviews using the questionnaire. The lead investigator was present at all follow-up interviews that took place between April and July 2007. During the interviews, all questions were asked exactly as written in the questionnaire in order to secure a standardise approach.
There is no formal ethics process in Ethiopia. The Gimbie Adventist Hospital Management Committee and the Zonal Health Department approved the study. The women were informed of the nature of the study and consented orally to participate in the follow-up interview.
Quality-of-life scores and their differences were not normally distributed. Accordingly, Wilcoxon’s signed rank test for paired data was used to compare the quality-of-life score before fistula developed with the score while living with fistula, and the score at follow up with the score while living with a fistula.
A total of 38 (86%) women were identified for follow up, though only 37 completed a follow-up interview as one had died from obstructed labour. Information about this woman was provided by her family.
At the time of follow-up interview, 21 (57%) were completely dry, 13 (35%) had different degrees of incontinence and three (8%) had a fistula. Two had been wet since failed operation and one woman had experienced a new obstructed labour causing her eighth stillbirth and re-occurrence of the fistula. For the subgroup of women with a sling/tape procedure performed 15 attended follow up, eight (53%) were completely dry.
Quality of life
Median quality-of-life scores are given in Table 2. Quality of life was significantly lower during the time living with a fistula than before fistula developed for all women (P = 0.001). The lowest scores were on those indicators, which assessed social interaction such as visiting friends, travelling by bus or going to the market. Surgery significantly increased the quality of life among not only those women who remained dry but also those who continued to suffer from stress and/or urge incontinence (P = 0.001). The women who had a fistula at follow up reported no improvement in quality-of-life measures as compared to the time with the fistula (P = 0.1).
Table 2. The median quality-of-life (QoL) score as assessed at follow up regarding the time before fistula, while living with a fistula and following fistula surgery, categorised by urinary health at follow up
|Dry (21)||Prefistula||49|| |
|Incontinence (13)||Prefistula||50|| |
|Still fistula (3)||Prefistula||51|| |
Although not directly asked, five (14%) women volunteered the information that they had attempted suicide while suffering from fistula.
Marital status and reproductive health
Only two (5%) women were divorced while suffering from a fistula. At follow up, 27 (71%) women remained married and sexually active, seven (26%) were amenorrhoeic, 11 (41%) used contraception and five (33%) became pregnant. Table 3 provides details on the outcome of pregnancy for the five women who conceived. One woman died during obstructed labour.
Table 3. Pregnancy outcomes for five women who conceived following fistula surgery
|1||No||Maternal death during labour. In labour for over 24 hours|
|2||Yes, after 48 hours of labour||Stillbirth. Absent fetal heart beat on admission to hospital. Delivered by Caesarean section. Fistula recurred|
|3||Yes, after >24 hours in labour||Stillbirth. Absent fetal heartbeat on admission to hospital. Delivered by craniotomy. No subsequent urinary problems|
|4||Yes, admitted in the antenatal period to await delivery||Live birth delivered by elective caesarean section. Fistula still present from unsuccessful primary procedure|
|5||Ongoing pregnancy||Ongoing pregnancy|
Recollection of advice given regarding future pregnancy
Only 22 (59%) recalled receiving advice to attend hospital for future delivery.
Understanding of fistula and knowledge of others with fistula in community
Only ten (27%) correctly stated that fistula could be prevented by avoidance of prolonged labour and attendance at a healthcare facility for delivery. Six (16%) women reported that they had shared this knowledge with members of their community. Eight (22%) women reported that they knew at least one other woman in their community with symptoms suggestive of fistula. Four reported that they had encouraged these other women to seek treatment, while three had not done so because they were afraid to discuss the problem. The final woman reported that her surgery was unsuccessful and she did not wish to recommend surgery.
All the women had some form of occupation when the fistula developed. Thirty-four (92%) were farmers, two (5%) were domestic workers and one (3%) was a potter. While living with a fistula eight (22%) stopped working and remained at home. Six (75%) of these women returned to their former occupation following surgery.
Community-based, long-term follow up after fistula repair was found to be feasible in this area of rural west Ethiopia. Despite one-third still suffering stress or urge incontinence, the women reported improved quality of life and social reintegration years after fistula closure.
Strengths and weaknesses
Community-based follow up of women who have undergone fistula repair surgery has been stated to be extremely difficult.4,6–8 To the best of our knowledge, only one previous study have succeeded in such follow up.10 We have high follow-up rates, with information on 86% of women. Our success can be attributed to a number of factors including: the location of the fistula clinic in the rural area where the women live, the close cooperation of village leaders, partner health facilities and local government sectors to support the fistula clinic and assist with identifying women for follow up, and the provision of transport and accommodation costs.
The low number of women limits the significance of our results. Data on reproductive outcome following surgery in particular are too sparse for statistical analysis.
Standard urological questions were used to distinguish different types of incontinence. Given that these were based on the King’s Health Questionnaire, this study is limited by the necessary modifications of the questions to secure relevance in a rural area among an illiterate population. Obviously, this modified questionnaire needs to be validated in larger study populations. On the other hand, our results demonstrate that the questionnaire was robust for detecting significant differences in quality of life. As the questionnaire was developed for the follow-up interview, we cannot rule out recall bias. Also a wish to satisfy the interviewer may have influenced the answers. However, those women who continued to suffer from a fistula at the time of follow up reported no improvements on any of the quality-of-life measures and hence did not seem influenced by a desire to please the interviewer in the answers given.
Comparison to other studies
As has been reported in other series, our patients reported high levels of mental distress when living with a fistula.11 The issue causing the greatest distress and frustration was the lack of social life. Nearly all women suffered from isolation, shame and stigma and dared not go to public places due to the risk of being insulted or simply ignored. Many explained that people in the community used to hold their nose, laugh or talk about the smell when they approached. The magnitude of the depression caused by the fistula was confirmed by the fact that five (14%) of the women reported having attempted suicide while living with the fistula. A recent prospective study found that surgical treatment of obstetric fistula results in marked improvements in mental health as evaluated 2 weeks after surgery.11 Our study demonstrates that such improvements in mental health continue in the long term, even when some urinary symptoms (such as stress or urge incontinence) remain.
When compared to patients undergoing fistula surgery at other centres, some differences in the impact of fistula on socio-economic wellbeing are apparent. Firstly, the majority (79%) of women in our series kept their job. Secondly, in our series only two (5%) woman were divorced because of her fistula, compared to 52 and 74% reported by others.8,12 Another important difference is that in our series only 11 (25%) women were primiparous compared to 63% of patients treated at the AAFH.12 The differences may be explained by the location of our clinic in the rural area where patients live. In contrast, the AAFH treats patients from across Ethiopia, some of whom travel a distance of over 1000 km to reach the hospital. It is possible that only those with the greatest disadvantage such as divorce and loss of income are prepared to make the journey to Addis Ababa for treatment, hence the AAFH may see a higher proportion of the most disadvantaged women.
In other aspects, the majority of our women were similar to those reported in previous larger studies, as the fistulae were almost exclusively caused by obstructed labour resulting in stillbirths, the majority of women were illiterate, farmers, most had a BMI < 19 and a high proportion were of short stature of (≤150 cm). Our primary closure rate, going home dry rate and re-operation rate are also comparable to larger series.8,12
Our study demonstrates that only one-third of the women had understanding of the cause of obstetric fistula and how to prevent it, and only half of these had shared this knowledge with other members of their community.
A further important finding of the study relates to reproductive health and pregnancy outcome following fistula surgery. Patients with obstetric fistula are at high risk of future obstruction, ruptured uterus and possible maternal and fetal death should they become pregnant again. Therefore, it is extremely important that patients who undergo fistula repair surgery and their families receive advice regarding future pregnancy and the need to deliver within a health facility that can offer caesarean section. All women received such advice, although40% did not recall receiving this information. Despite the provision of advice and a voucher (to receive highly subsidised obstetric care), there was only one successful pregnancy outcome in the four who delivered following fistula surgery. Indeed the pregnancy outcomes of the remaining three were tragic, with one maternal death, three fetal deaths and one fistula recurrence. The two women with stillbirth and the family of the woman who died revealed that the reluctance to come to hospital lay with the patients’ husbands, who were reluctant to pay for transportation, and their hope that this delivery would be less complicated than previous. These findings emphasise the need to embed fistula repair services within a broader safe motherhood strategy, working with husbands, families and communities as a whole to ensure timely access to maternal health services, especially for those in greatest need. Greater emphasis needs to be given to removing the stigma associated with fistula and to improve understanding of fistula cause and prevention.
This study was performed within a rural area, close to the homes of those suffering from fistula. The majority of fistulae arise in rural areas where access to health services is lacking. In contrast, most fistula clinics are located in large urban centres. Indeed prior to the fistula clinic at GAH, the closest fistula service for the population of West Wollega was the AAFH, which required a journey in excess of 1–2 days to reach. The majority of women who underwent surgery at GAH reported that they would not have travelled to Addis Ababa for treatment because the journey seemed more terrifying than living with a fistula, they could not leave their dependents for weeks or months, and could not afford travel expenses. Our project demonstrates that fistula surgery can be conducted in remote settings, working through partner health facilities.
This study shows that long-term follow up of patients following fistula repair surgery is indeed possible. Patient numbers in this study were small and hence it is important that similar follow-up studies are replicated in larger populations and in other settings.
Disclosure of interests
None of the authors has any personal, financial, political, intellectual or religious interests to declare.
Contribution to authorship
H.S.N. conceived and designed the study; participated in setting up the fistula service and the first project, created the questionnaire, organised the follow up, analyzed the data and drafted the manuscript. H.S.N. is the guarantor. L.L. performed the pilot testing, trained interviewers, oversaw all interviews and analyzed the data. U.N. designed the study, created the questionnaire, and analyzed the data. H.A. collected data in the first project. O.L.J. collected data in the third project. B.S. collected data in the second project. M.R. collected data in the second project. M.C. collected data in the second project. R.L. collected data in the first project, organised the follow up locally, analyzed the data. S.D. collected data in the third project, created the questionnaire. All authors interpreted the data, critically revised the draft for important intellectual content, and gave final approval of the version to be published.
Details of ethics approval
There is no formal ethics process in Ethiopia. The Gimbie Adventist Hospital Management Committee and the Zonal Health Department approved the study. The patients were informed of the nature of the study and consented orally to participate in the follow-up interview.
The Danish Association of Obstetrics and Gynecology FIGO foundation and Maternity Worldwide supported the study financially. The funding sources had no involvement in the study design, data collection, analysis and interpretation, neither in the writing and decision to submit for publication.