Multilevel modelling of determinants of perinatal mortality in East Africa: a pooled analysis of National health survey data

Abstract

Background

Perinatal mortality, which includes stillbirths and early neonatal deaths, is a critical indicator of maternal and newborn health, especially in developing countries. It highlights the effectiveness of healthcare systems and socioeconomic inequalities. Despite global efforts, such as the Sustainable Development Goals (SDGs), to reduce perinatal mortality, developing countries continue to experience high rates due to factors like inadequate access to quality healthcare, maternal health issues, and socioeconomic disparities. Since, there is limited evidence in the region, this study investigates perinatal mortality in East Africa, using data from Demographic and Health Surveys (DHS) to identify key determinants and inform policy interventions aimed at improving health outcomes.

Methods

This study utilized data from the DHS conducted in East Africa. A weighted sample of 101,728 children was included in the analysis using R-4.4.0 software. Descriptive data, including frequencies and texts, were performed. A multilevel modeling analysis was employed to analyze perinatal mortality, considering both individual-level factors and contextual factors. The multilevel model accounts for clustering within countries and allows for the examination of both fixed and random effects that influence perinatal mortality. For the multivariable analysis, variables with a p-value ≤ 0.2 in the univariate analysis were considered. The Adjusted Odds Ratio (AOR) with a 95% Confidence Interval (CI) and a p-value < 0.05 was reported to indicate statistical significance and degree of association in the final model.

Results

The overall pooled effect size of perinatal mortality is 3.67 (2.92, 4.59), with Tanzania having the highest rate and Comoros having the lowest rate. Women aged 25–34 years (AOR = 0.86, 95% CI: 0.81, 0.95), 35–49 years (AOR = 0.89, 95% CI: 0.79, 0.97), and 35–49 years (AOR = 0.89, 95% CI: 0.79, 0.97) compared to women aged 15–24 years, gave birth the first before the age of 20 (AOR = 1.09, 95% CI: 1.03, 1.28), have secondary or higher education (AOR = 0.76, 95% CI: 0.69, 0.81), not being married (AOR = 1.11, 95% CI: 1.05, 1.21), poorer (AOR = 0.94, 95% CI: 0.89, 0.98), and richest women (AOR = 0.95, 95% CI: 0.91, 0.97) compared to the poorest women, mass media exposure (AOR = 1.09, 95% CI: 1.03, 1.15), women with 3–5 children (AOR = 1.16, 95% CI: 1.08, 1.21), and with more than 5 children had even greater odds (AOR = 1.36, 95% CI: 1.29, 1.44), twin births (AOR = 3.62, 95% CI: 3.41, 3.79), modern contraceptive (AOR = 0.82, 95% CI: 0.81, 0.91), had history of abortion (AOR = 8.53, 95% CI: 8.29, 8.79), birth interval of 24–36 (AOR = 0.68, 95% CI: 0.65, 0.73), and 37–59 months (AOR = 0.61, 95% CI: 0.55, 0.67) compared to intervals of < 24 months respectively, having health insurance (AOR = 0.87, 95% CI: 0.82, 0.92), rural residence (AOR = 1.05, 95% CI: 1.02, 1.18), residing in low-income (AOR = 1.33, 95% CI: 1.28, 1.49), and higher literacy rates (AOR = 0.81, 95% CI: 0.79, 0.89) were statistically associated with perinatal mortality respectively.

Conclusions

The study reveals several significant factors associated with perinatal mortality in East Africa. Factors such as women who gave birth before the age of 20, not married, mass media exposure, having more children, twin births, history of abortion, residing in rural areas, and in low-income countries were linked to higher odds of perinatal mortality, however, being older age, better education, better wealth, modern contraception, longer birth intervals, have health insurance, and high literacy rate countries were linked to lower odds of perinatal mortality. To reduce perinatal mortality in East Africa, targeted interventions should focus on improving educational attainment for women, enhancing access to health insurance, and promoting the use of modern contraceptive methods. Additionally, policies aimed at supporting unmarried mothers, managing multiple births, and addressing rural healthcare disparities are essential.