Search for a command to run...
Abstract Background Pregnancy loss, including miscarriage and stillbirth, is a major public health issue with major physical and psychological consequences for pregnant women. Prevalence estimates in low-resource settings remain scarce due to the lack of adequate data. This study assessed the prevalence, timing, and maternal characteristics associated with stillbirth and miscarriage using novel longitudinal data collected in five low- and middle-income countries (LMICs). Methods and Findings We analyzed longitudinal data from 5,755 pregnant women in Ethiopia, India, Kenya, South Africa, and Côte d’Ivoire. Women were enrolled during pregnancy and followed through delivery. Gestation-specific and cumulative risks of miscarriage and stillbirth were estimated using competing-risks survival analysis, adjusting for timing of enrollment. We examined associations with maternal age, education, wealth, and country using Fine and Gray sub-distribution hazard models. Among pregnancies surviving to 8 weeks, the cumulative risk of pregnancy loss by 28 weeks was 84 per 1,000 pregnancies (95% CI: 69–100) and from 28 to 44 weeks the risk was 19 per 1,000 (15–24), resulting in a total pregnancy loss risk after 8 weeks of gestation of 103 per 1,000 (88–119). Risks were highest in Côte d’Ivoire and lowest in South Africa. Losses peaked between 8 and 6 weeks of gestation, with a secondary rise after 36 weeks. Women aged ≥35 years had higher loss risk (HR 1.78, 95% CI: 1.27–2.48), whereas wealth and education showed no consistent association. Conclusions Pregnancy loss remains common across LMICs, with significant risk in both early and late gestation. Conventional estimates that do not account for delayed enrollment underestimate miscarriage rates. Enhanced surveillance and targeted interventions throughout pregnancy—especially during early gestation—are essential to reduce preventable fetal losses and meet associated global goals. Author’s Summary Evidence before this study Data on the prevalence of miscarriage and stillbirth in low- and middle-income countries (LMICs) are scarce. Previous studies have reported substantial variation in miscarriage and stillbirth rates across LMICs, but most relied on retrospective surveys, facility-based data, or late pregnancy enrollment. Few multi-country longitudinal studies have examined gestation-specific patterns of pregnancy loss. Added value of this study This study uses newly available longitudinal cohort data from five LMICs and applies a novel competing risks survival methods to account for delayed enrollment and loss to follow-up. We show that failure to capture early pregnancy losses and failure to account for delayed entry and competing risksleads to systematic underestimation of miscarriage risk. We show that early pregnancy loss is substantially more common than suggested by crude estimates and that risk remains elevated beyond the first trimester. Our multi-country design allows for direct comparison across diverse health system and socioeconomic contexts. What do these findings mean? Our findings indicate that pregnancy loss in LMICs remains a major and under-estimated public health burden, particularly during early gestation. Improved measurement systems that capture pregnancies earlier and follow women longitudinally are essential for monitoring progress. Interventions to improve early antenatal care, strengthen risk screening, and ensure timely obstetric management are needed to reduce preventable losses. Policies focused solely on late pregnancy and delivery may miss a large proportion of fetal deaths occurring earlier in gestation.