Abstract
Purpose Prostate cancer patients tend to be older with multiple comorbidities and are thus at increased risk for postoperative cardiovascular events after radical prostatectomy (RP). Thus, proper patient selection strategies are essential to decide for or against a surgical approach. We aimed to derive a prostatectomy specific index (PSI) for patients undergoing RP and compare its performance to universally used indices. Methods The cohort was derived from National Surgical Quality Improvement Program database between 2005 and 2012. The primary outcome was incidence of major adverse cardiovascular events at 30 days post-surgery including: death, myocardial infarction, or stroke. A multivariable logistic regression model was constructed, performance and calibration were evaluated using a ROC analysis and the Hosmer-Lemeshow test, the PSI index was derived and compared to the RCRI and AUB-HAS2 indices. Results A total of 17,299 patients were included in our cohort, with a mean age of 62 +/- 7.4 years. Seventy three patients had a cardiac event post RP. The final PSI index encompassed six variables: history of heart disease, age, anemia, American society of anesthesiology class, surgical approach, and hypertension. The PSI ROC analysis provided C-statistic = 0.72, calibration R-2 = 0.99 and proper goodness of fit. In comparison, the C-statistics of RCRI and AUB-HAS2 were found to be 0.57 and 0.65, respectively (p value < 0.001). Conclusion The PSI model is a procedure tailored index for prediction of major cardiovascular events post RP. It was calibrated using a large national database aiming to optimize treatment selection strategies for prostate cancer patients.