Signal generation and clarification: use of case-control data.

Case-control surveillance systems are useful for 'signal' generation, i.e., signaling potential previously unidentified adverse effects of drugs. Two systems currently in operation, the Slone Epidemiology Unit's Case-Control Surveillance and the Birth Defects Study, have monitored drug effects since 1976. With extensive information on the diagnoses and covariates, the systems have the capacity to carry out in-depth analyses in which the outcome measure is more specifically defined and in which confounding is controlled, thus reducing the possibility of false alarms.