Trends of stillbirth, early neonatal and perinatal mortality rates and sociodemographic differentials of perinatal mortality rates in 17 low- and middle-income countries: A modeling study

BACKGROUND: Stillbirths and early neonatal deaths account for a substantial proportion of global child mortality. Reliable measurement of these outcomes remains challenging in settings where civil registration systems are incomplete and household surveys are the primary data source. However, survey-based estimates are often affected by underreporting and misclassification between stillbirths and early neonatal deaths. We aimed to apply a statistical adjustment approach to improve estimates of stillbirth, early neonatal, and perinatal mortality trends and to assess socioeconomic inequalities in countries with repeated Demographic and Health Surveys (DHS). METHODS AND FINDINGS: We analyzed 93 DHS surveys with reproductive calendar data from 17 LMICs conducted between 1990 and 2023, covering 1,019,253 pregnancies reported by 733,181 women. Pregnancies of at least 7 months' gestation were included to align with the standard definition of stillbirth. Age-specific risks of stillbirth and early neonatal death were estimated using life-table methods and a modified Gompertz-Makeham (mGM) model that adjusts for common survey data quality issues, including underreporting, misclassification between stillbirths and deaths on days 0-1, and age-at-death heaping, i.e., the tendency to systematically misreport age of neonatal death often at 7 days at the expenses of neighboring days. Survey-specific estimates were used to evaluate trends over time and to examine associations with maternal and household characteristics through mGM models and the pooled etimates random-effects meta-regression; all covariates were used for adjusting the perinatal mortality rates and also taking the survey sampling design into consideration. A sensitivity analysis using alternate assumptions confirmed the robustness of the findings at country level. Across countries, stillbirth, early neonatal, and perinatal mortality rates generally declined over time, although trends varied substantially between settings. The pooled annual decline was -0.60 deaths per 1,000 births for stillbirths (95% CI [-0.78, -0.41]), -0.50 for early neonatal mortality (95% CI [-0.66, -0.35]), and -1.08 for perinatal mortality (95% CI [-1.40, -0.76]). Lower perinatal mortality was associated with urban residence, higher maternal education, wealthier households, and higher national Human Development Index levels, whereas maternal age ≥30 years was associated with higher risk. Limitations include potential residual underreporting of pregnancy losses and lack of information on causes of death, in addition to our model assumptions relying on constant perinatal mortality metrics across different settings. CONCLUSIONS: Adjusting for survey data quality issues may potentially improve the use of household survey data for estimating stillbirth and early neonatal mortality. Although most countries experienced declines in perinatal mortality, progress remains uneven and significant socioeconomic disparities persist. Strengthening maternal and newborn health services and improving data systems will be essential to accelerate progress toward global targets for reducing perinatal mortality.