We estimated a prevalence of fistula of at least 30 per 100,000 women of reproductive age in WBG, suggesting that if the results are extrapolated, there are at least 619 (128–1807) untreated women in South Sudan. However, the findings from our small sample in WBG are unlikely to be generalisable to the whole country given the wide range of access to maternity services from state to state. In 2006 in WBG, 13.8% of women had a skilled attendant present at their birth, compared to the 4.9% in Eastern Equatoria [19] and 12% in the whole country [14]. The risk of obstructed labour and consequently of fistula may vary between regions. Therefore, we cannot be confident that this is a true reflection of the minimal prevalence across the country.
The quality and capacity of current fistula repair services are inadequate to meet the needs of the population. This is within a context of limited health services in general and particularly for women. The campaign system has been a good short term measure, however it has not provided consistent or equitable care and is insufficient to cope with the backlog of cases, with nearly as many cases on waiting lists as have been treated in the last six years. Clinicians’ and policy makers within South Sudan described the key barriers to providing consistent hospital-based services, which would require investment in human resources and facilities to overcome. There are currently clinicians who have been trained to carry out basic fistula repairs and these skills could be better utilised and supported to provide a service for South Sudan. Any resources used for fistula service development have to be carefully weighed up against the greater numbers of people who are without the most basic of health services. This is particularly true in light of the relatively low prevalence estimates obtained in our study. Nonetheless there is a need to facilitate the access of these women with an unmet need for services; awareness of fistula could be raised among the population as part of the overall package of obstetric-related health messages.
Our estimated prevalence of 30 per 100,000 is considerably lower than previous population-based estimates [8, 10, 11]. Given the poor maternal health indicators in South Sudan, one would expect at least as high a prevalence as other regions of Sub-Saharan Africa. There are several potential explanations for our finding.
Our estimate could be correct, with others being incorrect. Many of the prevalence estimates reported in the literature are not based on epidemiological studies and as a result, are likely to be inaccurate; if publications are advocating to address the issue, they may well have overestimated the prevalence. Alternatively, our estimates may be correct but unrepresentative as it is possible that our selected area was too close to Wau, and therefore the communities that we visited would have had greater access to maternal services and therefore a decreased risk of fistula. However, the communities were one hour’s drive with good vehicles from the city, which would not be possible for most women in the communities, and the roads used were only completed in 2010. Finally, it is also possible that as South Sudan reports having the highest maternal mortality in the world and has high rates of all cause mortality, the women with obstructed labour did not survive long enough to present with fistula.
Alternatively, it is possible we underestimate the prevalence for a number of reasons. First, the KIM may not be an effective method for complete ascertainment of cases of fistula, and we did not find all the eligible cases. KIM has been validated for use for other rare conditions including childhood disability and childhood blindness [12, 13], but while these are often stigmatizing conditions, it may be that mothers will bring children for assessment but not come themselves. Previous work on women with fistula has emphasised their isolation from family, friends and community [2, 20, 21]. This could mean that women were “hidden” and unable to be identified. Our key informants particularly the CBDs, were key members of their community and we feel it would be unlikely that in close communities women with fistula would be completely hidden. It is also possible that women with fistula were identified but refused to attend verification. Again, the training of the key informants was designed to ensure that women were supported, reassured about confidentiality and given information about treatment. The women with fistula who did come forward did not report stigma as a major issue.
Since women from outside the target area presented for verification, this suggested to us that messages were effectively communicated across a wide area and some women at least were willing to come forward. Additionally, subsequent research conducted in the same area by a team with a social anthropologist identified further women in areas outside of the one we studied, but no additional women were found in the region we studied.
Finally the prevalence could be influenced by inaccurately estimating the denominator, which was obtained from recent surveys conducted by the government. There is some debate about how accurate the household surveys were. However even if we overestimated the population by 10,000 people (23%), this would still have only resulted in an estimate of 40 per 100,000 women of reproductive age, which does not raise the prevalence by much.
This is the first time that KIM has been used to identify maternal morbidities, and we believe that it is a promising tool for the future. We have shown that it is possible to find women with fistula using a KIM. It has the advantage of being much faster and cheaper than household surveys, as well as benefiting from community involvement. However it has the limitation of having to rely on key informants of varying ability to find the women and the possibility of missing eligible cases. The next step is to use this method in further studies, in other regions of South Sudan and Sub-Saharan Africa. Complementary qualitative data that have been collected from the key informants and women with fistula will offer further insights to improve the key informant methodology. This methodology has the potential to become a practicable method for estimating the prevalence of women with fistula, which will be invaluable in planning future services and ensuring equitable access. Further use of this method in other regions of the world, possibly in contexts where other methods are used as well, would enable the production of comparable estimates, to identify whether the low prevalence found by this study in South Sudan is accurate or a result of methodological limitations. It has the added advantage of having a clinincal assessment which confirms the diagnosis. It is well understood that using women’s self-reported morbidity will over estimate the prevalence of rare conditions, unless the specificity of the questions is very, very high.
This study was a preliminary assessment of fistula in a region of South Sudan, identifying limited services and quantifying the potential unmet need. This provides important information for future policy and service development; to ensure that fistula is integrated with other services, existing skills are built on and women with fistula are proactively identified to receive services. Any service developed needs to be considered within the context of limited resources and that current demand is less than may have been expected.