Author(s): Lee WC
Abstract Share this page
Abstract To characterize exposure-disease association in human populations, epidemiologists have long relied upon such indices as "relative risk' and/or 'attributable risk'. However, the relative risk is not in a common unit which permits comparison across different exposures or different diseases and the attributable risk may not adequately catch and describe the variation of disease risks in populations. The present paper discusses the possibility of using the summary index of the Lorenz curve, the Gini index, as an alternative measure of exposure-disease association. It is found that this index can be interpreted in several ways (as the coefficient of deviation in disease risk or relative risk, the information content of the exposure, the impact fraction of an exposure-lowering programme, and the averaged impact fraction) and is a promising alternative as a fundamental measure in epidemiology. Further studies are warranted to investigate its statistical properties.
This article was published in Stat Med
and referenced in Journal of Biometrics & Biostatistics