The methodology of HIV/AIDS impact studies: a review of current practices

Abstract

This review of studies on the socio-economic impact of HIV/AIDS shows that diversity in methodological design, which often is a result of practical considerations and resource constraints rather than of poor design, is the norm. This limits the comparability of research findings. More detailed reporting on method, which is not the norm, can go some way towards facilitating such comparison. Furthermore, the review underlines the importance of exploring intervention issues in more detail. Researchers need to employ results in answering specific policy questions. Scope remains for more impact studies to be conducted in developing countries in general and in certain high prevalence countries in specific, i.e. Southern Africa. Studies that explore the urban/rural dynamics of and clients' perceptions and behavior in seeking care and support are necessary to better understand the epidemic. The role of community-based organizations, non-governmental organizations and other stakeholders in studies of this nature can be expanded. Larger studies generally have more statistical power, but smaller, in-depth studies can be equally valuable. A careful stratification of sample populations can enhance the quality of cross-sectional studies. Qualitative methods should be used to complement the current reliance on survey-based methods of data collection. More longitudinal studies are required to explore the long-term impacts of HIV/AIDS. HIV/AIDS training for fieldworkers should be standard in studies of this nature, while cognizance should be taken of the dangers of employing local people as fieldworkers in studies of such sensitive nature. Scope remains for the further empirical analysis of data from impact studies, which requires these data sets being made accessible to more researchers. In the longer term, an attempt at standardizing core modules in impact studies can help to improve our understanding of the impact of HIV/AIDS in different settings.