Deriving more up-to-date estimates of long-term patient survival

Statistics reflective of patient survival should be as up-to-date as possible. We illustrate that traditional methods for analysing survival have important limitations with regard to up-to-dateness of long-term cumulative survival estimates. We propose an alternative approach, denoted ''period analysis,'' that may be used to overcome or at least substantially reduce these limitations. The conceptual background and the computational realization of the new method are outlined and its application is illustrated by an analysis of long-term survival of patients with testicular cancer in Saarland, Germany. Copyright (C) 1997 Elsevier Science Inc.