Abstract
Background: Maternal mortality is widely regarded as a key indicator of population health and social and economic development. Its rates and trends are closely monitored by the United Nations and others, inspired in part by the Sustainable Development Goals (SDGs), which call for a three-quarter reduction in the maternal mortality ratio by 2015. Unfortunately, the empirical basis for such monitoring is still quite weak, requiring the use of models. statistics to get estimates for most countries.
Objective: In this paper we describe a new method for estimating global rates and trends in maternal mortality. For countries that lack sufficient data for direct estimation of estimates, we use a parametric model that separates maternal deaths from deaths from HIV/AIDS.
Materials and methods: For maternal deaths not associated with HIV/AIDS, the model consists of a hierarchical linear regression with three predictors and variable intercepts for country and region. Estimated uncertainty is assessed by simulating the estimation process, taking into account the variability in both the data and other model inputs. This method was used to derive the most recent set of PBB estimates, published in September 2010. Here, we provide a brief description and description of the approach, including a novel analysis of the components of variability reflected in the uncertainty interval.
Results: Final estimates provide evidence of a more rapid decline in the global maternal mortality ratio than suggested by previous studies, including another study published in April 2010. We compared the findings of the two most recent studies and discussed topics for further research to help resolve differences.