Statistical methods for analysing complex survey data : an application to HIV/AIDS in Ethiopia.
Year: 2013
Type: dissertation
Abstract: The HIV/AIDS pandemic is currently the most challenging public health matter that \n \nfaces third world countries, especially those in Sub-Saharan Africa. Ethiopia, in East \n \nAfrica, with a generalised and highly heterogeneous epidemic, is no exception, with \n \nHIV/AIDS affecting most sectors of the economy. The first case of HIV in Ethiopia \n \nwas reported in 1984. Since then, HIV/AIDS has become a major public health con \n \ncern, leading the Government of Ethiopia to declare a public health emergency in \n \n2002. In 2011, the adult HIV/AIDS prevalence in Ethiopia was estimated at 1.5%. \n \nApproximately 1.2 million Ethiopians were living with HIV/AIDS in 2010. \n \nSurveys are an important and popular tool for collecting data. Analytical use of survey \n \ndata especially health survey data has become very common, with a focus on the association of particular outcome variables with explanatory variables at the population \n \nlevel. In this study we used the data from the 2005 Ethiopian Demographic and Health \n \nSurvey, (EDHS 2005), and identified key demographic, socioeconomic, sociocultural, \n \nbehavioral and proximate determinants of HIV/AIDS risk factor. Usually most survey \n \nanalysts ignore the complex survey design issues like clustering, stratification and unequal probability of selection (weights). This study deals with complex survey design \n \nand takes the design aspect into account, because failure to do so leads to bias parameters estimates and standard error, wide confidence intervals and statistical tests \n \nwill be incorrect. \n \nIn this study, three statistical approaches were used to analyse the complex survey \n \ndata. The first approach was a survey logistic regression used to model the binary \n \noutcome (HIV serostatus) and set of explanatory variables (the dependence of the \n \nHIV risk factors). The difference between survey logistic regression and the ordinary \n \nlogistic regression is that survey logistic regression approach takes the study design \n \ninto account during analysis. The second approach was a multilevel logistic regression \n \nmodel, that assumed that the data structure in the population was hierarchical, and \n \nthat individual within household was selected from clusters that were randomly selected \n \nfrom a national sampling frame. We considered a three-level model for our analysis. \n \nThis second approach considered the results from Frequentist and a Bayesian multilevel \n \nmodels. Bayesian methods can provide accurate estimates of the parameters and the \n \nuncertainty associated with them. The third approach used was a Spatial models \n \napproach where model parameters were estimated under the Integrated Nested Laplace \n \nApproximation (INLA) paradigm.
Author: Mohammed O. M. Mohammed
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2133469243
