Project R-8134

The development and application of flexible multivariate models for current status data and count data with applications to HIV serodiscordance among couples and abortion, perinatal and infant mortality in Mozambique. (Research)

Mozambique is a low-income country in Southern Africa with about 54% of the population living below the poverty line. There has been promising progress in some of the health-related millennium development goals, while other health outcomes are still unsatisfactory (WHO). HIV prevention activities have been inadequate to curb the HIV prevalence trend. The high maternal and child mortality reflects the inability of women and children to access essential services due to inadequate geographical coverage of health services, inadequate financing, shortage of health professionals and essential medicines. By developing and applying innovative flexible, multivariate statistical models, this project aims at providing new insights in order to improve i) designs for HIV prevention plans, and ii) knowledge about hazardous signs for women during pregnancy. Based on, inter alia, data from the National Survey of Prevalence, Risk Behavioural and Information about HIV and AIDS (INSIDA), new statistical methods and parameterizations are proposed and studied to model HIV serodiscordance in hetero-sexual and homo-sexual couples in Mozambique. In a second part, multivariate models are developed to study how sociodemographic, cultural factors, level of knowledge about dangerous signs during pregnancy are associated with the number of abortions, number of dead children (post neonatal mortality) and number of stillbirths.

01 July 2017 - 31 October 2023