Background: Long-term survival is an important metric in assessing procedural value. We previously confirmed that the Society of Thoracic Surgeons predicted risk of mortality score (PROM) accurately predicts 30-day mortality in Israeli patients. The present study investigated the ability of the PROM to reliably predict long-term survival.
Methods: Data on 1279 patients undergoing cardiac surgery were prospectively entered into our database and used to calculate PROM. Long-term mortality was obtained from the Israeli Social Security Database. Patients were stratified into five cohorts according to PROM (A: 0-0.99%, B: 1.0-1.99%, C: 2.0-2.99%, D: 3.0-4.99% and E: ≥ 5.0%). Kaplan-Meier estimates of survival were calculated for each cohort and compared by Wilcoxon signed-rank test. We used C-statistics to assess model discrimination. Cox regression analysis was performed to identify predictors of long-term survival.
Results: Follow-up was achieved for 1256 (98%) patients over a mean period of 62 ± 28 months (median 64, range 0-107). Mean survival of the entire cohort was 95 ± 1 (95% CI 93-96) months. Higher PROM was associated with reduced survival: A-104 ± 1 (103-105) months, B-96 ± 2 (93-99) months, C-93 ± 3 (88-98) months, D-89 ± 3 (84-94) months, E-74 ± 3 (68-80) months (p < 0.0001). The Area Under the Curve was 0.76 ± 0.02 indicating excellent model discrimination. Independent predictors of long-term mortality included advanced age, lower ejection fraction, reoperation, diabetes mellitus, dialysis and PROM.
Conclusions: The PROM was a reliable predictor of long-term survival in Israeli patients undergoing cardiac surgery. The PROM might be a useful metric for assessing procedural value and surgical decision-making.
Keywords: Logistic regression; Long-term survival; Risk prediction.
© 2022. The Author(s).