The HIV Epidemic in Four Countries Seen through the Demographic and Health Surveys

Understanding the dynamics of the HIV epidemic in Africa requires information on the prevalence and incidence of infection and on associated behaviours. There are, however, few statistically representative surveys that can provide information at the national level on infection and/or behaviour. This paper examines a group of these surveys, namely the demographic and health surveys (DHS) for Kenya (1998), Tanzania (1996), Uganda (1995) and Zambia (1996). These surveys provide information on how people react to HIV/AIDS: knowledge acquisition; (self-declared) strategies for avoiding HIV; age at first intercourse; monogamy; abstinence; having been tested and wanting to be tested. A subsample of respondents are marriage partners allowing the analysis of assortativeness in behaviour.The paper has two distinct but related goals. The first is to use the DHSs to understand whether and how people are changing their behaviour in response to the epidemic. The second goal is to look for shortcomings in these surveys and to suggest how they can be extended to provide a better understanding of the epidemic. When possible, DHS findings are related to the epidemiological literature. Throughout, attention is given to the internal consistency of the surveys and their consistency with epidemiological studies. Suggestions are made for the improvement of DHS-type surveys. Bibliogr., notes, ref., sum. [Journal abstract, edited]

Title: The HIV Epidemic in Four Countries Seen through the Demographic and Health Surveys
Author: Gersovitz, Mark
Year: 2005
Periodical: Journal of African Economies
Volume: 14
Issue: 2
Period: June
Pages: 191-246
Language: English
Geographic terms: Kenya
Tanzania
Uganda
Zambia
External link: http://jae.oxfordjournals.org/content/14/2/191.full.pdf
Abstract: Understanding the dynamics of the HIV epidemic in Africa requires information on the prevalence and incidence of infection and on associated behaviours. There are, however, few statistically representative surveys that can provide information at the national level on infection and/or behaviour. This paper examines a group of these surveys, namely the demographic and health surveys (DHS) for Kenya (1998), Tanzania (1996), Uganda (1995) and Zambia (1996). These surveys provide information on how people react to HIV/AIDS: knowledge acquisition; (self-declared) strategies for avoiding HIV; age at first intercourse; monogamy; abstinence; having been tested and wanting to be tested. A subsample of respondents are marriage partners allowing the analysis of assortativeness in behaviour.The paper has two distinct but related goals. The first is to use the DHSs to understand whether and how people are changing their behaviour in response to the epidemic. The second goal is to look for shortcomings in these surveys and to suggest how they can be extended to provide a better understanding of the epidemic. When possible, DHS findings are related to the epidemiological literature. Throughout, attention is given to the internal consistency of the surveys and their consistency with epidemiological studies. Suggestions are made for the improvement of DHS-type surveys. Bibliogr., notes, ref., sum. [Journal abstract, edited]