Sequential Medical Plans

I N CLINICAL MEDICAL EXPERIMENTATION the disease under study may be relatively rare and it may be necessary to continue the study over months or even years in order to acquire a respectable series of patients. In the last few years statistical techniques have been developed which apply to situations where observations are gathered over time. Althouigh such techniques seem well suited to the needs of clinical research (at least in principle), it is necessary to adapt these methods to the thorny special problems of medical research. The procedures discussed in this paper are an attempt to make this adaptation-an attempt which is necessarily a preliminary one. The presentation is intended for medical research workers.