Small area estimation of HIV prevelance using Botswana AIDS impact survey data, BAIS IV 2013; married people aged between 15 years to 49 years: case of Gaborone

Abstract

Sample survey data are extensively used to provide reliable direct estimates of parameters of interest for the whole population and for domains of different kinds and sizes. When the domains were not originally planned, they usually are poorly represented in the sample or even not represent at all. These domains are called small areas. Small area estimation (SAE) is a topic of great importance due to the growing demand for reliable small area statistics even when only very small samples are available for these areas. Over the years, there has been increasing demand in most countries for reliable data not only at the national level but also at sub-national levels or domains (Rao, 2003). This is due, among other things, to their growing use in formulating policies and programs, and in the allocation of government funds and in regional planning. The main objectives of the present study is to find out which factors have significant effects on the prevalence of the epidemic among married people aged 15-26 years, 27-38 years, 39-49 years and 15-49 years residing in Gaborone and also estimate the prevalence of HIV/AIDS among the same groups by applying small area technique methods on the data extracted from 2013 BAIS VI since the sample sizes for these age groups were small and direct estimates will not be reliable. Two small area estimation methods, the ratio estimator and the logistic regression methods, were used to estimate the HIV prevalence for married people aged 15-49 years in Gaborone. The methods yielded HIV prevalence rate 15.22% and 15.446% respectively. The direct HIV prevalence rate using sampled data from BAIS IV is at 19.0% for the married people aged 15-49 years residing in Gaborone. The HIV prevalence for two age intervals, 15-26 and 27-38 years were estimated at 0% by all the methods used, this could have been due to very low participation rate in HIV testing by married people in the two age brackets. Married people in the age bracket 39-49 years residing in Gaborone are found to be the most affected with the three methods, the direct method, the ratio estimator method and the logistic regression method estimating the prevalence at 33%, 35.232% and 49.975% respectively. Females in the age bracket 15-49 years residing in Gaborone are 1.598 times more likely to be HIV positive than males in the same age bracket. (Odd Ratio (OR) =1.598) The results also show that those without education are 52.623 times more likely to be HIV positive than those with education higher than senior secondary (reference level). Those who ended at primary and junior secondary level are 3.722 and 3.634 times more likely to be positive than those in the reference level respectively. Those who completed senior secondary are almost equally likely to be HIV positive compared to the reference level, OR of 0.935 The results also show that those who had taken alcohol in the last 12 months were 1.233 times more likely to be HIV positive than those who had not.