PURPOSE OF REVIEW: Due to its long natural history, prostate cancer illustrates best the need for tools that adequately predict life expectancy. We reviewed the actual tools available for clinicians involved in therapeutic decisions in newly diagnosed prostate cancer and examined their accuracy to provide individual life expectancy. RECENT FINDINGS: Life tables, comorbidity indices, and multivariate prognostic models can assist clinicians for life expectancy predictions. However, the accuracy of life tables (60.9%) and comorbidity indices (accuracy unknown) may be as weak as clinician-derived life expectancy predictions (69%). Actually, statistical models provide the highest accuracy (69-84.3%). To date, Walz et al. developed the most accurate model (84.3%), predicting the risk of death of nonprostate cancer-related causes within 10 years of definitive therapy. SUMMARY: Clinicians need the most accurate estimates of life expectancy in situations in which there is uncertainty regarding the need for aggressive local therapy. As the accuracy of clinician-derived life expectancy prediction is relatively modest, clinicians may benefit from assisted life expectancy prediction by life tables and statistical tools in their daily clinical practice. This would enhance the accuracy of the life expectancy predictions of individual candidates to definitive therapy for prostate cancer. Actually, nomograms provide the most accurate health-adjusted life expectancy prognostication.
PURPOSE OF REVIEW: Due to its long natural history, prostate cancer illustrates best the need for tools that adequately predict life expectancy. We reviewed the actual tools available for clinicians involved in therapeutic decisions in newly diagnosed prostate cancer and examined their accuracy to provide individual life expectancy. RECENT FINDINGS: Life tables, comorbidity indices, and multivariate prognostic models can assist clinicians for life expectancy predictions. However, the accuracy of life tables (60.9%) and comorbidity indices (accuracy unknown) may be as weak as clinician-derived life expectancy predictions (69%). Actually, statistical models provide the highest accuracy (69-84.3%). To date, Walz et al. developed the most accurate model (84.3%), predicting the risk of death of nonprostate cancer-related causes within 10 years of definitive therapy. SUMMARY: Clinicians need the most accurate estimates of life expectancy in situations in which there is uncertainty regarding the need for aggressive local therapy. As the accuracy of clinician-derived life expectancy prediction is relatively modest, clinicians may benefit from assisted life expectancy prediction by life tables and statistical tools in their daily clinical practice. This would enhance the accuracy of the life expectancy predictions of individual candidates to definitive therapy for prostate cancer. Actually, nomograms provide the most accurate health-adjusted life expectancy prognostication.
Authors: M N Agathokleous; E Nena; D Chadolias; A Zissimopoulos; N Polyzos; E Jelastopoulou; T C Constantinidis Journal: Hippokratia Date: 2016 Apr-Jun Impact factor: 0.471
Authors: Sophie Knipper; David Pröwrock; Zhe Tian; Hans Heinzer; Derya Tilki; Pierre Karakiewicz; Markus Graefen Journal: World J Urol Date: 2019-03-04 Impact factor: 4.226
Authors: David A Alter; Dennis T Ko; Jack V Tu; Therese A Stukel; Douglas S Lee; Andreas Laupacis; Alice Chong; Peter C Austin Journal: J Gen Intern Med Date: 2012-05-02 Impact factor: 5.128
Authors: Richard M Hoffman; Tatsuki Koyama; Peter C Albertsen; Michael J Barry; Timothy J Daskivich; Michael Goodman; Ann S Hamilton; Janet L Stanford; Antoinette M Stroup; Arnold L Potosky; David F Penson Journal: J Gen Intern Med Date: 2015-02-13 Impact factor: 5.128