• From the PCORI Washington, DC.
• University of Washington School of Medicine Seattle, WA.
• National Cancer Institute, National Institutes of Health, Rockville, MD.
• Fred Hutchinson Cancer Research Center, University of Washington, Seattle, WA.

BACKGROUND: Studies of prostate cancer progression are important for discovering risk factors that may increase the risk of prostate cancer-specific death; however, little is known about the validity of self-reported prostate cancer progression.

METHODS: We conducted a validation study of self-reported prostate cancer progression in the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial and in a prostate cancer cohort enrolled in a Fred Hutchinson Cancer Research Center (FHCRC)-based study. We calculated measures of validity for self-reported progression, including sensitivity, specificity, positive predictive value, and negative predictive value using medical records as the gold standard.

RESULTS: Our results suggest that ascertaining prostate cancer progression-related events (i.e., prostate-specific antigen elevation, recurrence, metastasis, and use of secondary treatment) through self-report may be a viable option for identifying men whose disease has progressed after diagnosis or initial therapy, particularly when multiple questions related to progression are included in the assessment (aggregate cluster of questions: sensitivity = 0.76 [PLCO]; 0.93 [FHCRC], specificity = 0.80 [PLCO]; 0.97 [FHCRC]). With an aggregate positive predictive value of 0.50 (PLCO), however, our PLCO results suggest that additional medical record verification of self-reported progression events may be necessary to rule out false positives. Most individuals reporting no evidence of progression-related events, however, were true negatives (aggregate negative predictive value = 0.92 [PLCO]; 0.98 [FHCRC]). Thus, there may be limited utility to investing resources in chart review to confirm self-reported nonevents.

CONCLUSION: Ascertaining prostate cancer progression through self-report provides an efficient and valid approach to enhancing existing cancer cohorts with updated data on progression status. See video abstract at, http://links.lww.com/EDE/B658.