Time lines and computer-based visual editing: new techniques for assessingexposure in epidemiologic studies

We describe a time-line-based methodology for collecting exposure data for epidemiologic studies and for processing these data for statistical analysi s with readily available software for the personal computer. The four compo nents to this approach ale: (1) collecting data in a memory-enhancing time- line format: (2) entering data from time lines into a computer database and editing them: (3) making a quantitative estimate of exposure, intake, or d ose for each exposure event: and (4) creating analysis datasets by 'slicing ' the quantified time lines based on desired exposure intervals or disease latent periods. Compared with fixed-format interviews, time-line-based inte rviews help subjects organize remembered events, thereby reducing confusion . They do not restrict responses to predetermined categorical exposure resp onses. The time-line methodology also facilitates the collection of supplem entary data necessary for computing doses for complex exposures and the pac kaging of quantified exposures into analysis datasets for any time period o f interest.