Mortality rate estimation models for patients with prostate cancer diagnosis

Abstract

Lithuanian men deaths. Malignancy of the prostate gland in Lithuania is the most common oncological illness among the men. The mortality from this disease in other European countries tends to be decreasing but Lithuania still experiences annual increase in prostate cancer mortality. Article presents models to assess the long-term mortality risk among patients with diagnosed malignant tumor of the prostate gland. The latter models complement popular decision support methods in clinical practice used so far, gives significant information about prostate malignant tumors and their aggressiveness potential. Cox proportional hazards regression and Fine-Gray competing risk regression comparison showed that Cox method excels not only in higher accuracy and AUC value, but also in higher sensitivity which is particularly important in clinical trials.