A comparison of cause of death determination in men previously diagnosed with prostate cancer who died in 1985 or 1995

PURPOSE: We quantified the agreement between the underlying cause of death determination from information in hospital medical records and on death certificates, and determined whether the frequency of assigning death from prostate cancer had changed since the introduction of testing for prostate specific antigen. MATERIALS AND METHODS: We retrospectively reviewed and analyzed the information in hospital medical records and on death certificates for men previously diagnosed with prostate cancer who died in 1985 or 1995. RESULTS: The underlying cause of death determinations from a review of 201 hospital medical records agreed with those from information on part 1 of the death certificate in 87% of cases and with those using the International Classification of Diseases-9 system coding rules in 80%. Agreement was higher in men who were older than those who were younger at the time of death, and higher in those diagnosed with prostate cancer several years before death than in those diagnosed shortly before death. CONCLUSIONS: There was a high level of agreement concerning the underlying cause of death after a review of the information in hospital medical records and on death certificates for men with prostate cancer when cause of death was viewed as a dichotomous variable. The International Classification of Diseases-9 coding rules concerning the underlying cause of death favor overreporting rather than underreporting prostate cancer deaths compared with a review of hospital medical records. Cause of death determination does not appear to have changed after the introduction of prostate specific antigen testing.