Abstract

WERNECK, Guilherme L.  and  COSTA, Carlos H. N.. Use of census information to proxy individual socioeconomic characteristics: an empirical evaluation. Epidemiol. Serv. Saude [online]. 2005, vol.14, n.3, pp.143-150. ISSN 1679-4974.  http://dx.doi.org/10.5123/S1679-49742005000300002.

Gathering socioeconomic information at the individual level is not a simple task. Many researchers use census data as a proxy for individual characteristics. The objective of this report was to evaluate the validity of the use of census information as a proxy for individual socioeconomic characteristics. Using data from a population-based case-control study to investigate the impact of land occupation and environmental degradation on the incidence of visceral leishmaniasis in the Municipality of Teresina, Piauí State, Brazil, this report compares measures of effect for different risk factors, controlling for census variables and individual-level variables. Considering a relative discrepancy of 10% or less as a criterion of similarity between the odds ratios (OR) obtained by using individual level socioeconomic variables (ORi) and census data (ORc), some 47% of the estimates of ORc could be considered as a different estimate, compared to the ORi. Considering the ORi as the gold-standard, the specificity and sensitivity of the ORc were both approximately 60%. The use of census variables does not necessarily provide effective control for socioeconomic confounding. Census data should not be considered by themselves good substitutes for individual-level data as they represent different concepts that operate in another level of determination.

Keywords : socioeconomic factors; censuses; confounding factors; bias; ecological fallacy.

        · abstract in Portuguese     · text in Portuguese     · Portuguese ( pdf )