A method for relating a set of information about an individual to one or more other sets, collected at different times and often different places, with the aim of demonstrating a connection, such as effects that are remote in time and place from their underlying causes, e.g., cancer with long latency between exposure and onset of disease. An essential prerequisite is a reliable and consistent way to identify individuals, using tested and proven methods to do so. One method that works with very high reliability is to use a sequence of numbers and letters for the individual's birth date, sex, birth place, birth order in cases of multiple births, personal name, and mother's family name at the time of her birth. The odds of repeating the same sequence for two individuals, and thus confusing their identities, are less than one in a trillion if this system is adhered to. In some nations, record linkage studies are conducted in a national census office, where the officials (statisticians, epidemiologists) have taken an oath of secrecy to preserve personal privacy and confidentiality. In some nations, record linkage studies require the informed consent of the individuals to be studied, which is obviously impossible when the individuals are deceased.
Subjects: Public Health and Epidemiology.