“Counting those never bubble over” Estimated prevalence and its determinants of obstetric stula among childbearing women in 14 African countries: Multilevel Analysis

Obstetric stula is a leakage between genital tract and urinary tract and/or between genital tract and rectum. The commonest cause of obstetric stula is prolonged labour which magnify in the areas of poor prenatal and emergency obstetric care. In Africa, there is poor of quality of obstetric care and poor social support for those who faced stula. Obstetric stula shatters the life of the women and the consequence is nasty while multicounty level estimate on the magnitude and determinates of stula were nil. Multicounty level of estimate of the magnitude of stula is important to design and ll the gaps of quality of obstetric care and design the appropriate corrective intervention mechanisms of obstetric stula. Therefore, this study aimed the estimate the magnitude of obstetric stula and its determinants among childbearing women in 14 Africa countries based on recent demographic and health survey data.

The magnitude of obstetric stula in 14 African countries were high as compared with the world health organization estimate. Maternal age, residence, educational status, husband's educational status, sex of household head, age at rst birth, history of terminating pregnancy and awareness on obstetric stula were the determinants identi ed in this study. Therefore, health interventions that reduce the occurrence of obstetric stula could be designed to address the women who lives in rural area, no formal education, male-headed household, husbands who never attended formal education, and women who had terminated pregnancy should be addressed in advance. Policies and programs of stula should be tailored the women which characterized as living in rural area, non-educated, young age at rst birth and no awareness on stula as well as male headed households. Evidence based multicounty interventions were highly recommended to eliminate obstetric stula and to achieve sustainable development goal.

Background
Obstetric stula is a leakage between genital tract and urinary tract or leakage between genital tract and rectum commonly caused by prolonged or obstructed labor especially in resource poor countries. It is preventable and treatable while it is ashamed and isolated from the communities. Nearly 2 million of women's globally severed from this preventable disease [1][2][3][4]. Fistula is the serious medical disorder, which results continuous leakage of stool or urine. Obstetric stula is the less intentioned and hidden condition in developing countries. The most marginalized women like poor, young, illiterate and women in remote area are mostly affected one. Obstetric stula is one of the most visible maternal health indicator and is still exist due to the health system failed to give accessible and quality of maternal health care. It is a public health concern, which has an attention in national and international government in recent time [5][6][7]. Even though there is no real estimate on obstetric stula nearly two million women were suffered from this problem in developing countries and fty to hundred thousand of new cases will occur annually [8].
In sub Saharan Africa this disorder is sever and at least 33,000 new cases were recorded annually, which is huge number, which highly contributes maternal mortality in the region. From the women who give birth annually in sub-Saharan Africa 30,000 to 130,000 of them develop stula due to obstructed or prolonged labour. In addition stula is sever in the place where a limited access and use of obstetric care. Obstetric stula are more prevalent in the area where poor or limited prenatal and intrapartum care[6, [9][10][11][12]. The prevalence of obstetric stula may vary from country to country or from area to area, in all region of the world the pooled prevalence of stula were 0.29 per 1000 women. By region 1.57 per 100 women in sub Saharan Africa and South Asia, 1.60 per 1000 women of reproductive age in sub Saharan Africa and 1.20 per 1000 in South Asia [13]. The prevalence of stula in Ethiopia was 0.4% among reproductive age women [14], in other demographic data analysis the prevalence of stula in Ethiopia were 0.42% [15],in other study 6 per 10,000 reproductive age women experienced obstetric stula [16]. Other study showed 0.12% of the women experienced obstetric stula [17], in Afghanistan the prevalence of obstetric stula was 0.4%[18], 1.06% experienced obstetric stula in their lifetime [12], in study conducted in Bangladesh the estimated prevalence of obstetric stula was 1.69 per 1000 ever married women [19], The incidence of stula in rural region of sub-Saharan Africa approximately 33,451 new case annually [20]. In two world region (sub-Saharan Africa and South Asia) there are over 6000 new cases per year [13].
While there is no uniform classi cation of obstetric stula. Obstetric cause of stula can be classi ed as Vesicovaginal stula (VVF) and rectovaginal stula (RVF) and there is iatrogenic types of stula [6,7,21]. The three main cause of obstetric stula are ischemia of tissue between vagina and urinary tract or rectum, direct tearing of tissues and elective abortion. Even though the causes of stula are not mutually exclusive, it may occur in complication of delivery and uterine evacuation [2,22,23].
The risk factors of obstetric stula originated from cultural, social, economic and biological that contributes for obstructed and prolonged labor. Sexual violence also results vaginal stula [24][25][26].
In addition there is health facility related factors like delay in reaching care, delay in decision to seek care, and delay in receiving adequate care [33,[35][36][37]. Literatures showed that obstetric stula may results from in different initiators like labor more than 24 hours [24], not attending antenatal care [38], and home delivery [39] while this isnot found in low prevalence areas. Additional factors associated with stula in previous literatures were early age at pregnancy, short stature, illiteracy, poverty, not attending antenatal care and rural place of residence or living far away (>3 km) from a health facility [23,40,41], duration of labour of >24 hours, seeking delivery services after 6 hours of labour onset, taking more than 2 hours to reach a health facility, having none or primary education and being referred to another facility for emergency obstetrics services [42], women who gave birth 10 or more and residence [12] were factors associated with obstetric stula. The three delays which contributes for obstetric stula were delay in the making the decision to seek care, delay in arrival at a health facility; and delay in the provision of adequate care [43].
Fistula has a multidimensional consequence on the women who experienced it such as health consequences like incontinence of urine or faecale or both [21], damage of Vulva or thighs[6], social consequences like stigma and marginalization by community [30,44], divorce [35,45], sexual and reproductive health consequences like loss of fertility or amenorrhea [24,37], absence of sexual intercourse [30,35,45], and mental health related consequences like depression [45,46] were some of the consequences of stula. Obstetric stula drop many women's live in dark situation and shatter many women life [7].
The number and/or the magnitude of stula in the region is the direct re ection of level and quality of perinatal care given by health care system in that speci c area. In the area of poor or absent maternal health care, obstetric stula is high and the treatment is unlikely [1,5,31]. Universal access of emergency obstetric care gives the best solution to the problem [7].Therefore, prevention of obstetric stula is the good ways of saving the women's life since in Africa treating obstetric stula is like "taking a serpent by the tail". Most of the African countries lacking the quality of stula care and perinatal care. Therefore understanding the current prevalence is important to design appropriate intervention mechanisms and bringing evidence based solution to those resource-limited country. Identifying the factors associated with obstetric stula is essential to design appropriate intervention approaches and tackle the factors in multicountry situation. In addition, it is important to design appropriate prevention policies and programs that can be all embracing. Therefor this study aimed to estimate the prevalence of obstetric stula and its determinants among childbearing women in 14 African countries based on the recent demographic and health survey data.

Data source and population
This study was analyzed secondary data from the recent Demographic and Health Surveys which contained detailed obstetric stula for all interviewed childbearing women. Data were obtained and extracted from individual record (IR) le based on the objective and the previous literature nding variables. Demographic and health survey data were collected by a strati ed, multi-stage (cluster), random sampling design. The detailed method of data collection were accessed at demographic and health survey database. The source population were all childbearing age women in survey period across the fourteen African countries whereas the study populations were all reproductive age women were included in this survey in the selected Enumeration Areas (EA). The study included all childbearing agewomen found in the selected clusters at least one night before data collection period. All women's who did not give birth before the survey period were excluded from the study. A total of 232,050 reproductive age women were involved in this study.

Variables and measurement Dependent variable
The outcome variable is obstetric stula where it is a binary variable which a woman who experience obstetric stula is coded as '1' for Yes while not experienced obstetric stula is coded as '0' for No.

Independent variable
In this study, both the individual and community level variables were included. The independent variables were age, residence, educational status, husbands educational status, wealth status, sex of household head, media exposure, age at rst birth, total number of children, working status, history of terminating pregnancy,and awareness on stula Operational de nition and description of variables Obstetric stula Obstetric stula among women who had at least one birth in the 5-year period prior to the survey was the outcome of the study which coded as Not experienced obstetric stula "0" and Experienced obstetric stula "1".

Working status
Women working status was coded as No "if women who didn't have any work", coded as '0', and Yes "If a woman were working, she might be self-employed, professional or nonprofessional employee or government employed" coded as '1'.

Educational status
The minimum level of education the mother achieve and it was coded as '0' for no formal education, as '1' for primary education, as '2' for secondary, and as '3' higher (college and above level of education) .

Wealth status
Categorized as; poor "if woman was in poorer and poorest household in the DHS database" coded as '0', middle where categorized as middle coded as '1', and rich "if woman was in richer and richest household" coded as '2'.

Husband's educational status
It was an education status of the husbands achieve and it was coded as '0' for no formal education, as '1' for primary education, as '2' for secondary education and as '3' for higher (college and above level of education).

Residence
The place of residence for the mother was coded as '0'for rural and as '1' for urban Media exposure a composite variable of frequency of listening radio, watching television and reading newspaper, in which households were said to have media exposure "if they have exposed to either of listening radio or watching television or reading newspaper at least one a week" coded as '1' and no "if they did not have exposure to all of the above media sources once per week" coded as '0'.

Sex of household head
The variable sex of household head was coded as '0' for male and as '1' for female in the dataset and we used without change.

Total number of children
The total number of children in the household was categorized as '0' for ≤3 child, as '1' for 4-6child, as '2' for 7-9child, and as '3' for ≥10 child

Health facility visit
The women who visit health facility in last 12 months were coded as '1' as Yes and who did not visit health facility were coded as '0' as No.

History of terminating pregnancy
The women who experienced of terminated pregnancy were coded as '1' for Yes and women who did not experienced terminated pregnancy were coded as '0' for No.

Awareness on stula
The women who aware on the stula were coded as '1' for Yes and the women who did not aware on stula were coded as '0' for No used as dataset coded.

Data analysis
The variables of the study were extracted from individual record (IR) le dataset using STATA version 15. Before any analysis, the data were weighted using sampling weight to account the sampling design. After the data were cleaned, categorized, coded and weighted by using STATA, the author explored the descriptive statistics by using the frequencies and percentages of data and presented by using tables. Intraclass Correlation Coe cient (ICC), proportional change in variance (PCV) and median odds ratio (MOR) were calculated for the appropriateness of multilevel logistic regression analysis and checking the presence of clustering. We used the ICC value greater than 5% to consider a variation of magnitude of stula across the cluster. Signi cant clustering was found therefore multilevel logistic regression were more appropriate. Information Criteria (AIC) is the best-tted model. Therefore, a model with a smallest Akaike Information Criteria (AIC) value was selected and all interpretations and inferences were made based on this model. After selecting the best-tted model, bivariable and multivariable multilevel logistic regression was done to determine the determinants of magnitude of stula in 14 African countries. Both bivariable and multivariable multilevel analysis was done accordingly. Variable in bivariable analysis with p-value<0.25 were selected in multivariable multilevel analysis. Finally, P-value ≤ 0.05 to declare statistically signi cant variables. The level of association were declared by odds ratio and 95% con dence interval.

Result
Characteristics of the study population

Random effect Analysis
The appropriateness of multilevel analysis of the study was done before the actual multilevel analysis of the study. Multilevel analysis is necessary because there is a signi cant clustering of stula in this DHS data. The intraclass correlation coe cient (ICC) was 28.33% means 28.33% of the variability in magnitude of stula among reproductive age women were attributed to the clusters. The median odds ratio (MOR) value of the null model 2.96 also indicates the presence of variation in unmet need for family planning between clusters. It means if we randomly select households from different clusters, those households at the cluster with higher stula had 2.96 times higher chance of having stula compared to their counter parts. As shown in Table 3 Table 3). In addition the proportional change in variance (PCV) increases from 0.77% (null model) to 22.1% (model 3), indicating that mode 3 were the best explains the variability of stula among the clusters. Therefore, this model is the best-tted model for the data because it has the smallest AIC as compared to the rest models. So interpretation and reports were made based on this model. The other important determinants of obstetric stula were awareness towards stula. The odds of obstetric stula was 65% [AOR=0.35; 95% CI: 0.26, 0.45] less likely among women who had awareness on stula than the women who had no any awareness on obstetric stula (Table 4).

Discussion
Obstetric stula has devastating consequences on the live of the women and it shatters the women's life. This problem eliminated from developed nation while it continuously affects the poorest of the poor [47,48]. Previously obstetric stula were not give a full attention in maternal and reproductive health arena [49][50][51][52][53]. Even though UNFPA plan to end stula [48] there is a limited study conducted with nationally representative data at multi-country level. Therefore, this study aimed to estimate the magnitude and its determinants of obstetric stula in 14 African countries. The current study nding showed that the magnitude of stula among childbearing women were 0.84 [95%CI: 0.79, 0.88]. According to world health organization estimate suggested that 3 out of 1000 women of reproductive age women were experienced stula which is lower than the current study estimate [54]. In addition, this nding is lower than the pooled analysis study in Sub Saharan Africa that indicated that the prevalence of stula was 1.60 and South Asia, 1.20 [13] and the study conducted in Ethiopia, which indicated that the lifetime prevalence of obstetric stula was 1.06 [12] and South India [17]. This variation could be due to the variation in the quality of maternal health care among the countries and the difference in the quality of obstetric care services in those regions with the current study area. This implies that universal access of quality of maternal care could eliminate obstetric stula. Prenatal care, skill birth attendants and access emergence obstetric care has a great effort to prevent obstetric stula [55].The other possible justi cation of this variation could be UNFPA and other national and international organization gives an emphasis to end stula which could be lower the current magnitude of the study. Speci cally the study conducted in South India was before the action of to end stula and before the start of sustainable development goal which focus on maternal health improvement and reduction of maternal mortality.
While the current study nding higher than the study conducted in Ethiopia [15][16][17], Afghanistan[18], Bangladesh [19], Rajasthan, India[56] and rural Maharashtra [57]. This could be due to variation in intervention and access of emergency obstetric care services. The current study was multi-country and nationally representative while previous study conducted in single area with small sample size, which may lower the magnitude. Speci cally in Maharashtra, India, more than half study participants had formal education and majority of them were economically active since stula are common among poor and illiterate women. The other possible reason could be there is variation in social support and burden of obstructed labor across the countries.
This study also identi ed the determinants of obstetric stula in African country. The determinants identi ed were maternal age, educational status of the women, residence, husband's educational status, sex of the household head, age at rst birth, history of terminating pregnancy and awareness on stula. The odds of obstetric stula among older women (age >=41) were 1.38 times more likely experienced stula as compared with the women age less than or equal to 20 years. This nding was comparable with the study conducted in Uganda which showed that the odds of obstetric stula were increased at the older age[58] and the study conducted in India which indicated that the occurrence of stula were more common above the age of 29 years [59]. This could be due to the women may bear macrosomic infant when age increases[60] which could contributes obstructed labor and stula. The other possible reason could be the women's at older age decreased health seeking behaviour and reluctant to take maternal health services like antenatal care and skill birth delivery.
The risk of experiencing obstetric stula among the women who had secondary and higher education were 41% and 60% less likely as compare with the women who never attended formal education respectively. This nding is comparable with the study conducted in Kenya which showed that women who never attended formal education were at risk of obstetric stula [59], in study conducted in India[56], in another study conducted rural community showed in poor educational level women obstetric stula were more prevalent [61]. In other study conducted in north-eastern Nigeria indicated that illiteracy was one of the major risk factors for obstetric stula [23] and in study conducted in Cameroon indicated that poor educational level women were associated with more experienced obstetric stula [22]. This could be due to maternal educational status had a net effect on maternal health services utilization[62] and had positive association with maternal health service use which can leads to reduce risk of obstetric stula.
The other possible reasons could be in less educated women maternal health service utilization is lower[63] which could elevate the risk of stula. This could be due to the increased women's health seeking behaviour and educated women could have an increased literacy level. The other possible reason could be educated women could be more in use of contraceptive, skilled delivery and deciding to take health services.
The odds of experiencing obstetric stula among urban residence were 31% less likely as compare with the women who live in rural area. This nding is comparable with the study conducted in Cameroon [22], in Ethiopia [15], Zambia [24], West Africa[64]), north-eastern Nigeria [23], and in Ethiopia[65] which indicated that rural place residence were more risk for obstetric stula. This could be due to maternal health service utilization is lower among women living in rural area [63] which could increase the risk of obstetric stula. Geographical access to health facility is one of the factor to address universal access of maternal health services[66] which could be contribute the risk of stula. In rural area, there could be long distance and low quality of service that lower maternal health service usage[67] which may accelerate the occurrence of stula. This could be due to the lower access to emergence obstetric care and other maternal health services like ANC and skilled delivery. The other possible reason could be traditional practices like FGM and sexual violence may common in rural areas. This implies that critically see the maternal health services and community value on women's health is important to save the women's from shattering of their live.
Husband's educational status were one of the determinate for obstetric stula. The odds obstetric stula among women whose husbands attended primary education were 20% less likely as compare the women's husband never attended formal education. The husbands contribute to the maternal health service use[68] which can improve the risk of experiencing obstetric stula. This could be partners/husbands educational status were less likely associated with obstructed labor[69] which is main risk for obstetric stula. This could be due to increased husbands support to take maternal health service, improve the intake of health services and support his wife to take skilled delivery and antenatal care. This implies that husband's education could improve the women's future life from abject. The risk of obstetric stula were lower among female-headed households as compare with male-headed household. The odds of obstetric stula among the household headed by female were 22% less likely as compared with the household headed by male. This could be due to female headed household could be more likely receive maternal health services than the male headed households do [70].This could be the women able to access and control resource which can cover the service and transport cost to take maternal health services. The other possible reasons for this could be the women's autonomy to decide for maternal health services are easy and they could reduce their risk of obstetric stula.
The odds of obstetric stula among the women who give their rst birth at the age of 16-20 years, 21-25 years, and >=26 years were 22%, 34% and 33% less likely as compare with the women who give birth less than or equal to 15 years respectively. This nding were comparable with the study conducted in Ethiopia[65],Tanzania [71], and India [59] which showed that early childbearing were signi cantly associated with obstetric stula. This could be due to physically maturity is the main factor for the occurrence of stula [59]. This could be due to younger and early adult age women were immature pelvis and more at risk of obstructed labor that results obstetric stula. Other possible reason could be those younger and early adult age girls and women had poor in maternal health service utilization.
The women who had history of terminating pregnancy were 1.51 times more likely at risk of developing obstetric stula as compare to the women who never experience obstetric stula. This nding were comparable with the study conducted in Uganda showed that induced abortion is one of the attributes of stula [72]. This could be due to low access of obstetric care and poor quality of care that results to obstetric stula. The other possible reasons could be due to poor/limited access of prenatal care of the women. The other determinants of obstetric stula were awareness towards stula. The odds of experiencing stula among aware women were 65% less likely as compare with those women who had no awareness on stula. This could be better knowledge on the problem able to prevent from the occurrence [2]. This could be due to understanding of the women regarding stula consequences that could set early prevention method for it.

Conclusion
The magnitude of obstetric stula were 0.84% higher as compare to world health organization estimate which showed the occurrence of stula were 3 per 1000 women. The determinants of obstetric stula identi ed in this study were maternal age >=41, residence, women who attended secondary education, women who attended higher education, female household head, husbands who attended primary education, women who give their rst birth 16-20years, 21-25 years, ≥26 years, history of terminating pregnancy and awareness on stula. The sun should never rise twice in one glob for the women therefore national and international organization could work to decrease the risk of stula. Programs and policy should be design to address those women older age, younger age at rst birth male-headed households, women who never attended formal education and had no awareness on stula. In the long run high level of education of women, increasing age at rst birth, and increased awareness on stula could reduce the risk of experiencing obstetric stula. Ethics approval and consent to participant A waiver of written informed consent was secured from Demographic and Health Surveys (DHS) program data archivists to download the dataset for this study. After the data were downloaded from the measure DHS website the research of this study will maintain the con dentiality of the data. Since it was based on secondary data, which was publicly available, ethical approval was not required. However, we accessed the data set from the DHS website (https://dhsprogram.com/ ) through registering or online requesting. The DHS data were conducted in accordance with the Declaration of Helsinki.

Data Availability statement
The datasets analyzed during the current study are available from the demographic and health survey dataset with the reasonable request. The author included all result-based data within the manuscript and the data set can be accessed online from www.measuredhs.com/data

Competing interest
The author declare that they have no competing interests.

Funding
The author did not receive any fund for this research.
Authors contribution MS involved in the conception of the study. MS involved in the conceptualized, design, data processing and accuracy, analysis and interpretations of the ndings. MS prepared the draft of the manuscript. The author read and approved the nal manuscript.