Epidemiologic research using existing databases of environmental measures.

When detailed information on individuals is not available, databases of environmental measures are sometimes used in epidemiologic studies to provide data on human exposure. The use of such databases is based on the assumption that the data are representative of individual human exposure and are applicable to the study. Twenty-six databases that contained data on the environment pertaining to California were summarized and evaluated by categorical criteria of spatial and temporal precision, availability, amount of quantitative information, "machine-readability," and type of environmental measure.