High prostate cancer mortality in Norway evaluated by automated classification of medical entities

Abstract

The new standard of cause of death classification is an automated selection of the underlying cause of death using the international software Automated Classification of Medical Entities (ACME). Norwegian mortality rates are, however, based on manual classification of deaths. The aim of this study was to investigate how the use of ACME would influence Norwegian prostate cancer mortality rates. A previously described cohort of Norwegian prostate cancer patients deceased during 1996 was applied. Multiple causes of death data based on information from death certificates, autopsies and queries was coded according to ACME specifications, thereby ACME selected the underlying cause of death. In addition, the underlying cause of death that originally was manually classified for the official mortality statistics was retrieved from Statistics Norway in all cases. Age-standardized prostate cancer mortality rates (world population) per 100,000 person-years were calculated. A total of 2012 cases were included. On the basis of ACME classification, the age-standardized prostate cancer mortality rate in Norway for 1996 would have been 24.4 (95% confidence interval: 22.9-26.0) as compared with the rate based on manual classification for the official mortality statistics of 24.9 (95% confidence interval: 23.4-26.5). Thus, automated classification by ACME does not significantly influence the age-adjusted Norwegian prostate cancer mortality rate for the year 1996. There is reason to assume that the use of manual classification of deaths is not a major explanation of the high prostate cancer mortality rates in Norway.