A New Method for Deriving Global Estimates of Maternal Mortality

Area of Science:

• Global Health
• Demography
• Epidemiology

Background:

• Maternal mortality is a critical indicator of population health and socioeconomic development.
• Global monitoring of maternal mortality relies on statistical models due to data limitations.
• United Nations Millennium Development Goals (MDGs) target a significant reduction in maternal mortality ratio (MMR).

Purpose of the Study:

• To introduce a novel statistical method for estimating global maternal mortality levels and trends.
• To provide updated United Nations (UN) estimates for maternal mortality.
• To analyze the sources of uncertainty in maternal mortality estimates.

Main Methods:

• Developed a parametric model to estimate maternal deaths, separating HIV/AIDS-related from non-HIV/AIDS-related deaths.
• Employed hierarchical linear regression with country and regional variable intercepts for non-HIV/AIDS maternal deaths.
• Utilized simulation methods to assess the uncertainty of the estimates, accounting for data and model input variability.

Main Results:

• The new method yielded estimates indicating a more rapid decline in the global maternal mortality ratio than previously reported.
• The approach was used to generate the latest UN maternal mortality estimates (published September 2010).
• Analysis revealed key components contributing to the uncertainty intervals of the estimates.

Conclusions:

• The refined estimation method provides stronger evidence for a faster decline in global maternal mortality.
• Comparison with other recent studies highlights the impact of methodological differences on MMR trends.
• Further research is needed to resolve discrepancies and enhance the accuracy of global maternal mortality monitoring.