Menu
Menu
Menu
Menu
Background: Home birth is a common contributor to maternal and neonatal deaths particularly in low and middle-income countries (LMICs). We generally refer to home births as all births that occurred at the home setting. In Benin, home birth is phenomenal among some category of women. We therefore analysed individual and community-level factors influencing home birth in Benin. Methods: Data was extracted from the 2017–2018 Benin Demographic and Health Survey females’ file. The survey used stratified sampling technique to recruit 15,928 women aged 15–49. This study was restricted to 7758 women in their reproductive age who had complete data. The outcome variable was home birth among women. A mixed effect regression analysis was performed using 18 individual and community level explanatory variables. Alpha threshold was fixed at 0.05 confidence interval (CI). All analyses were done using STATA (v14.0). The results were presented in adjusted odds ratios (AORs). Results: We found that 14% (n = 1099) of the respondents delivered at home. The odds of home births was high among cohabiting women compared with the married [AOR = 1.57, CI = 1.21–2.04] and women at parity 5 or more compared with those at parity 1–2 [AOR = 1.29, CI = 1.01–1.66]. The odds declined among the richest [AOR = 0.07, CI = 0.02–0.24], and those with formal education compared with those without formal education [AOR = 0.71, CI = 0.54–0.93]. Similarly, it was less probable for women whose partners had formal education relative to those whose partners had no formal education [AOR = 0.62, CI = 0.49–0.79]. The tendency of home birth was low for women who did not have problem in getting permission to seek medical care [AOR = 0.62, CI = 0.50–0.77], had access to mass media [AOR = 0.78, CI = 0.60–0.99], attained the recommended ANC visits [AOR = 0.33, CI = 0.18–0.63], belonged to a community of high literacy level [AOR = 0.24, CI = 0.14–0.41], and those from communities of high socio-economic status (SES) [AOR = 0.25, CI = 0.14–0.46]. Conclusion: The significant predictors of home birth are wealth status, education, marital status, parity, partner’s education, access to mass media, getting permission to go for medical care, ANC visit, community literacy level and community SES. To achieve maternal and child health related goals including SDG 3 and 10, the government of Benin and all stakeholders must prioritise these factors in their quest to promote facility-based delivery.
The present study made use of the women’s file of the 2017–2018 Benin Demographic and Health Survey (BDHS). The 2017–2018 BDHS was conducted in order to better operationalize and monitor the indicators of the Sustainable Development Goals (SDGs) for Benin. The National Institute of Statistics and Economic Analysis (INSAE) [23] carried out the survey in collaboration with the Ministry of Health. Assistance was obtained from Inner City Fund (ICF) through the international DHS (The Demographic and Health Survey) Program. The government of the Republic of Benin and the Agency of the United States for International Development (USAID) funded the 2017–2018 BDHS. The survey took place from November 6, 2017 to February 28, 2018. The main issues captured include fertility, fertility and infant and child mortality, contraceptive use, maternal health, children’s health, vaccination and other essential issues. The survey used a stratified sampling technique that was representative nationally. The 2017–2018 BDHS involved 14,156 households. Specifically, all women aged 15–49 in selected households and were present the night before the survey were eligible to be interviewed. This led to 16,233 eligible women, however 15,928 completed the interviews at a response rate of 98.1%. The current study was restricted to 7758 women aged 15–49 who had complete data. The dataset is publicly available at Measure DHS repository (https://dhsprogram.com/data/dataset/Benin_Standard-DHS_2017.cfm?flag=1) and details of the sampling processes are available in the 2017–2018 BDHS report [13]. The main outcome variable for the study was “home birth among women aged 15-49”. In the 2017–2018 BDHS, women were asked where they gave birth during their last childbirth which was posed as “Where did you deliver [name]?” accompanied by these responses: “home”, “other home”, “government hospital”, “government health centre/clinic”, “government health post/Community-based Health Planning and Services (CHPS)”, “other public”, private hospital/clinic”, “maternity homes”, and “others”. Following previous study [1], these responses were grouped into two responses and are “home birth” to denote every delivery that occurred outside health facility setting and “health facility delivery” to signify those that delivered in a health facility. “Home birth” was recoded as “1″ whereas “health facility delivery” was recoded as “0″. Eighteen explanatory variables were selected for the study. These are age, wealth status, religion, education, marital status, total children ever born, occupation, partner’s education, access to mass media, getting medical help for self; getting permission to go, getting medical help for help: getting money needed for treatment, getting medical help for self: distance to health facility, ANC visit and health decision making. All these constituted the individual-level factors. The community variables comprised sex of household head, community literacy level and community socioeconomic status. For clarity of presentation, some of the explanatory variables were recoded. Age was recoded as “19 years and under”, “20–34 years” and “35 years and above”. Religion was recoded as “Non-religious” and “Religious.” Education was recoded into “Without formal education” and “Formal education”; marital status was recoded into “Never married”, “Married”, “Cohabiting”, “Widowed” and “Divorced”; considering fertility rate of Benin which is about 5.7 children per woman [13], total children ever born was recoded into “1–2 births”, “3–4 births”, and “5 births or more”. Occupation was also recoded as “Not working” and “Working”, partner’s education recoded into “Without formal education” and “Formal education”. Access to mass media was constructed from three prime variables: frequency of reading newspaper/magazine; frequency of listening to the radio; and frequency of watching television. Each of these media variables had three responses: ‘not at all’, ‘less than once a week’, and ‘at least once a week’. A composite variable was created whereby those that indicated ‘less than once a week’ and ‘at least once a week’ were categorised as having access to mass media whilst ‘not at all’ was considered as not having access to mass media. ANC visit was recoded into “Below recommended” for less than eight visits and “recommended” for at least eight ANC visits, health decision making was recoded into “Alone”, “Respondent and partner” and “Others”. Community literacy level was generated by decomposing community literacy into three categories: “Low”, “Medium” and “High” and similar procedure was followed to generate community socioeconomic status. All these variables were selected due to their theoretical significance to maternal healthcare utilisation, specifically home delivery [1, 43]. The study set forth to unravel individual and community-level factors that determine home birth among Benin women aged 15–49. Based on this aim, these procedures were followed to analyse the dataset. The weighting factor built in the dataset (v005/100000) and the “svy command” were applied to deal with over and under sampling biases and to gauge for the complex survey design and generalizability of the findings respectively. The proportion of women who delivered home or otherwise were calculated. This was followed with univariate descriptive computation of the explanatory variables to show the summary statistics of the data. Thereafter, a cross-tabulation computation of outcome variable across the explanatory variables was done and the results were presented in proportions and percentages. Additionally, a chi square test of independence was applied to assess the association between the outcome variable and the explanatory variables at 0.05 alpha threshold. The variance inflation factor (VIF) command was applied to interrogate the collinearity among the explanatory variables and the results (Additional file 1) showed no evidence of multicollinearity between them (Mean VIF = 1.50, maximum VIF = 2.44, minimum VIF = 1.04). At 95% confidence interval, four regression models were built. The first model was a null model (Model 0) and accounted for the variations in home births, which is attributable to the clustering of the primary sampling units (PSUs) without the effect of both individual and community-level factors. In the DHS, primary sampling units are equivalent to clusters or communities that houses a number of households [11]. Therefore, in this study, we considered clustering in the PSUs to be the same as clustering across communities. The second model (Model I) considered individual-level factors solely whereas the third model (Model II) considered the effects of community-level factors on home births alone. Finally, the last model (Model III) was a full model containing both individual and community-level factors. The results for the fixed effects were presented as adjusted odds ratio (AOR) whereby any odds less than one was interpreted as reduced likelihood to home births whilst an odds higher than 1 meant otherwise. Since the models were nested, the Akaike information criterion (AIC) and Bayesian information criterion (BIC) techniques were used to measure their fitness [1, 43]. The random effects which are measures of variation of home births across communities or clusters, were expressed in terms of Intra-Class Correlation (ICC) [1, 28, 43]. These were calculated to quantify the degree of variation of home delivery across clusters and the proportion of variance explained by successive models. The analyses were done using STATA version 14.0. The present study made use of an already existing dataset. Hence, authors of this article were not involved in the implementation of the original study. However, the request to use the dataset was sought from Measure DHS. Measure DHS assessed the intent of our request and subsequently granted us access to download the dataset. The dataset is available at the Measure DHS repository at https://dhsprogram.com/data/dataset/Benin_Standard-DHS_2017.cfm?flag=1. Measure DHS anonymised the dataset before making it available for public use. The 2017–2018 BDHS reported that all ethical considerations applicable to human research participation were followed. Details of the ethical considerations are available in the survey report [13].
