Environmental impacts on human health: the agenda for long-term research and development. Draggan S, Cohrssen JJ, Morrison RE, eds. New York: Praeger, 1987 Jun; :139-164
The usefulness of health registries in gathering data concerning disease as it relates to environmental pollutant exposure was discussed. A registry by definition is not a study, but rather a collection of records. In discussing how well information taken from a registry compares with information taken from other sources, the questions of selection bias, information bias, confounding, the role of chance and statistical power, and the influence of causal and noncausal relationships must be considered. The conditions under which registries are of value in monitoring pollutant related disease include the perspective with which registries are established, maintained, and utilized. The importance of realizing the epidemiologic strengths and weaknesses of specific registries was emphasized. The full use of the registry as a source of information on the impacts of pollutant related diseases was considered as it relates to researchers, administrators, local governments, registrants, and society. The use of the registry to monitor precursor conditions and effectiveness of interventions was considered. Several research needs and recommendations were offered including specific areas of research such as study design, frequency outcomes, self selection bias, quantitative assessment of exposure, interactions, tracing and follow up of registrants, and registries as part of a national pollution control effort.