Interobserver variability in prostate cancer specific survival.

We evaluated the reliability of disease-specific survival (DSS) as an outcome measure in patients with carcinoma of the prostate (CaP). The records of 50 patients had a diagnosis of CaP and had expired were selected from the hospital tumor registry. Records were reviewed by six individuals and each individual was asked to specify cause of death as due to CaP or some other cause. DSS curves were generated based on the determinations of each reviewer. Although the DSS curves were generally parallel, a high degree of variability was seen at various intervals, leading us to conclude that DSS is dependent upon the individual reviewer. Published by Elsevier Science Inc.