Matt's email from : Wed, 31 Aug 2005 09:26:07

To determine exposure to study drug for purposes of reporting to the IDMC, I would suggest the following:

a) if the subject has a nonmissing date of last dose of study drug (TMTSPDT from the TMTLOG dataset), then you can determine exposure based on that (i.e., # days = TMTSPDT-TMTSTDT+1). Note that TMTLOG has 2 records per subject (data are coming from 2 separate pages) so be careful in obtaining the needed variables.

b) otherwise, for purposes of current reporting only, use an estimated date of last known clinic visit: use the latest of the visit dates for visits 3-10 (VSDT from the VISIT dataset restricted to SESS in the range from 3 to 10) and the dispensing dates (DISPDT from the DRUGDP dataset). Since the dispensing dates are on a running log, they may come in lagged behind the latest visit module; hence including both sets of data are adviseable.

Obviously, approach b) is an approximation, and will not be used for final reporting.