Background: More than 75% of neonatal deaths occurred in the first weeks of life as a result of adverse birth outcomes. Low birth weight, preterm births are associated with a variety of acute and long-term complications. In Sub-Saharan Africa, there is insufficient evidence of adverse birth outcomes. Hence, this study aimed to determine the pooled prevalence and determinants of adverse birth outcomes in Sub-Saharan Africa. Method: Data of this study were obtained from a cross-sectional survey of the most recent Demographic and Health Surveys (DHS) of ten Sub-African (SSA) countries. A total of 76,853 children born five years preceding the survey were included in the final analysis. A Generalized Linear Mixed Models (GLMM) were fitted and an adjusted odds ratio (AOR) with a 95% Confidence Interval (CI) was computed to declare statistically significant determinants of adverse birth outcomes. Result: The pooled prevalence of adverse birth outcomes were 29.7% (95% CI: 29.4 to 30.03). Female child (AOR = 0.94, 95%CI: 0.91 0.97), women attended secondary level of education (AOR = 0.87, 95%CI: 0.82 0.92), middle (AOR = 0.94,95%CI: 0.90 0.98) and rich socioeconomic status (AOR = 0.94, 95%CI: 0.90 0.99), intimate-partner physical violence (beating) (AOR = 1.18, 95%CI: 1.14 1.22), big problems of long-distance travel (AOR = 1.08, 95%CI: 1.04 1.11), antenatal care follow-ups (AOR = 0.86, 95%CI: 0.83 0.86), multiparty (AOR = 0.88, 95%CI: 0.84 0.91), twin births (AOR = 2.89, 95%CI: 2.67 3.14), and lack of women involvement in healthcare decision-making process (AOR = 1.10, 95%CI: 1.06 1.13) were determinants of adverse birth outcomes. Conclusion: This study showed that the magnitude of adverse birth outcomes was high, abnormal baby size and preterm births were the most common adverse birth outcomes. This finding suggests that encouraging antenatal care follow-ups and socio-economic conditions of women are essential. Moreover, special attention should be given to multiple pregnancies, improving healthcare accessibilities to rural areas, and women’s involvement in healthcare decision-making.
The most recent Demographic and Health Surveys (DHS) of ten Sub-African (SSA) countries (Angola, Congo, Cote d’Ivoire, Gambia, Lesotho, Liberia, Madagascar, Nigeria, Rwanda, Togo) data were used to make analysis of this study. The DHS is a part of the measure DHS programs that collect national information on basic health measures such as mortality, morbidity, and maternal and child health service utilization. Using the Kids Record (KR file) dataset, all births in the preceding five years before the survey were the study population. In the selected enumeration areas (EAs) births that had data about birth weight, gestational age at birth, and perinatal death records were included in the study. During the measure DHS survey, a multi-stage (two-stage) stratified sampling technique was used to select study participants; children were nested within the enumeration areas. After the dataset was appended, the weighted sample size became 76,853 children and women who had given birth five years preceding the survey. The methodology section of the DHS report goes into great detail about the study participant selection and data collection [23]. The main outcome variable of this study was adverse birth outcomes, which is defined as the presence of at least one or more of the following conditions in recent pregnancy (low birth weight, macrosomia, preterm birth, or stillbirth) [13, 19]. The outcome variable was generated by composite low birth weight, macrosomia, stillbirth, and gestational age less than 37 weeks of pregnancy. Finally, the variable takes 1 if at least one of adverse birth outcomes reported which was labeled as “adverse birth outcome”, and 0 otherwise. Socio-demographic characteristics (residence, maternal education, husband education, maternal age, mother marital status, sex of the child, media exposure, household wealth index, and maternal working status), health service utilization and accessibility (women healthcare decision-making autonomy, ANC follow up, and distance to health facility), and obstetrics related characteristics (preceding birth interval, parity, type of birth, and delivery by CS) were explanatory variables identified after thorough review of literatures [13, 24–31] . Short birth interval is defined as the time between two births which is less than 24 months [32]. Also, women’s healthcare decision-making autonomy is the ability of the women to make decisions to use health care services and treatment options [25]. Finally, media exposure was defined as when a woman reads a newspaper or listens to the radio, or watches television at least three times per week. Before any statistical analysis, the data were weighted using sampling weight based on primary sampling unit, and strata to restore the representativeness of the survey and take sampling design when calculating standard errors and reliable estimates. Cross-tabulations and summary statistics were done using STATA software version 14 (StataCorp.2015. Stata Statistical Software: Release 14. College Station, TX: StataCorp LP). The pooled prevalence of adverse birth outcomes with a 95% Confidence Interval (CI) was reported using a forest plot. The DHS dataset has a hierarchical structure that failed to meet the standard logistic regression model assumptions of independent observation and equal variance. Meanwhile, the children were nested within a cluster household, and children from the same cluster were more similar than from other clusters. Therefore, a mixed effect logistic regression model (both fixed and random effect) was fitted to account for cluster variability by using the advanced models. The outcome variable of the study was binary, a standard logistic regression and Generalized Linear Mixed Models (GLMM) were fitted step by step. Because the models were nested, model fitness was compared using the Intra-class Correlation Coefficient (ICC), Likelihood Ratio (LR), Median Odds Ratio (MOR), and deviance (−2LLR) values. As a result, the mixed-effect logistic regression model with the lowest deviance value was selected as the most parsimonious model. (Shown on Supplementary file Table 1). In the bivariable analysis, variables with less than 0.2 p-values were selected and entered into the multivariable mixed-effect logistic regression model. Adjusted Odds Ratios (AOR) with a 95% CI were calculated in the multivariable model to see the strength of association between independent variables and adverse birth outcomes. Variables with a 0.05 p-value in the final model being used as a statistically significant determinant of adverse birth outcomes. Permission for data access was obtained from measure demographic and health survey through an online request from http://www.dhsprogram.com. The data used for this study were publicly available with no personal identifier.