Academia.eduAcademia.edu

Maternal mortality: a cross-sectional study in global health

2015, Globalization and Health

Abstract

Background: Although most of maternal deaths are preventable, maternal mortality reduction programs have not been completely successful. As targeting individuals alone does not seem to be an effective strategy to reduce maternal mortality (Millennium Development Goal 5), the present study sought to reveal the role of many distant macrostructural factors affecting maternal mortality at the global level. Methods: After preparing a global dataset, 439 indicators were selected from nearly 1800 indicators based on their relevance and the application of proper inclusion and exclusion criteria. Then Pearson correlation coefficients were computed to assess the relationship between these indicators and maternal mortality. Only indicators with statistically significant correlation more than 0.2, and missing values less than 20% were maintained. Due to the high multicollinearity among the remaining indicators, after missing values analysis and imputation, factor analysis was performed with principal component analysis as the method of extraction. Ten factors were finally extracted and entered into a multiple regression analysis.