Fig. 2
A-F Decision curve analyses of each of the 6 models designed to estimate patient survival at (A) 1 year, (B) 2 years, (C) 3 years, (D) 4 years, (E) 5 years, and (F) 6 years after treatment or surgery for skeletal-related events due to bone metastasis from prostate cancer. These results suggest that all the models (dotted line) should be used rather than assuming all patients (continuous line) or no patients (thick continuous line) will survive longer than the period of each predictive model