Sowmya R. Rao, Gary Shapiro, and Theresa Diaz

Health surveys are important tools for studying the incidence and prevalence of diseases, health expenditures and use of services, effects of policy decisions, and association of important risk factors to various disease outcomes.

Conducting health surveys in developed countries can be difficult, but conducting health surveys in developing countries poses enormous challenges—for example, when a survey is carried out in a country trying to revive itself after a long civil war. The population may live mostly in rural and inaccessible areas, and a great deal of population migration might be taking place. It is difficult to compile a list of addresses to create a sampling frame (list of households or people in the population of interest) unless there has been a recent census, and so one has to choose an alternative way to get the information. Getting a representative sample is important, especially when one wants to obtain estimates of disease prevalence/incidence in a population, and a good sampling frame can help ensure this. There are several common sampling strategies (e.g., stratified, systematic, multi-stage cluster) to draw representative samples, but most of them are based on having access to a sampling frame.

Some content is only viewable by ASA Members. Please login or become an ASA member to gain access.