Background:
Older patients undergoing noncardiac surgery are at risk of postoperative cardiovascular events. Accurate cardiovascular risk assessment is important for informed decision-making.
Methods:
This prospective cohort study enrolled older patients undergoing elective major noncardiac surgery. A frailty index based on preoperative geriatric assessments (FI-PGA) was constructed using 32 health-related parameters. The primary outcome was the occurrence of any cardiovascular events within 30 days after surgery. The associations between the FI-PGA and outcomes were assessed using logistic regression models. The added predictive value was evaluated by comparing nested models using improvement in model fit, fraction of new predictive information, net reclassification improvement, and decision curve analysis. The predictive performance of the Clinical Frailty Scale was also evaluated.
Results:
A total of 1808 patients were included, with 316 (17.5%) patients experiencing the primary outcome. The FI-PGA was associated with increased odds of the primary outcome after adjustment for clinical predictors (odds ratio 1.56, 95% CI 1.33-1.82 per 0.1-point increment), and clinical predictors plus preoperative N-terminal pro–B-type natriuretic peptide (odds ratio 1.37, 95% CI 1.16-1.61 per 0.1-point increment). Integration of the FI-PGA in prediction models significantly improved model fit and provided new predictive information. Net reclassification improvement analysis showed that adding the FI-PGA to risk models improved risk estimation for patients who did not develop postoperative cardiovascular events, but did not improve risk estimation for those who experienced events. Decision curves showed the models containing the FI-PGA achieved higher net benefit. Improved model performance was also observed when the Clinical Frailty Scale was used for frailty assessment, although the added predictive values appeared lower.
Conclusions:
A frailty index derived from preoperative multidimensional geriatric assessment can improve cardiovascular risk prediction before noncardiac surgery, primarily by improving risk estimation for patients who will not develop postoperative cardiovascular events.