Association of the Revised Cardiac Risk Index with 1-year postoperative mortality: A single-center retrospective study.

Objective: To explore risk factors for 1-year postoperative mortality and to identify its association with the Revised Cardiac Risk Index (RCRI).

Methods: This was a retrospective cohort study involving 54,933 patients aged 18 years and above who were surgically treated under general or regional anesthesia in a tertiary hospital in Singapore. Independent risk factors for 1-year postoperative mortality were identified by univariate Cox regression analysis. The association between the RCRI and 1-year postoperative mortality was assessed by the Kaplan-Meier estimator and multivariate Cox regression analysis and was further validated in subgroup analyses stratified by the sex, age, and type of anesthesia.

Results: A total of 54,933 eligible patients were enrolled in this study that included 23,922 patients classified as RCRI Class I, 25,979 as Class II, 3700 as Class III, and 1332 as Class IV. Cox regression analysis demonstrated that male sex, age, higher American Society of Anesthesiologists (ASA) physical status classification level, regional anesthesia, emergency surgery, degree of anemia, and increased RCRI were significantly associated with the increased risk of 1-year postoperative mortality (HR > 1, all P < 0.001). The significant association between RCRI and 1-year postoperative mortality still existed after adjusting for confounding factors. An RCRI Class IV was associated with a mortality risk greater than two-fold larger than that observed at an RCRI Class I (adjusted HR 2.14, 95 % CI 1.78 to 2.56, p < 0.001). Subgroup analyses revealed that the 1-year postoperative mortality was significantly higher in patients with RCRI Class IV than that of Classes I-III regardless of the sex, age, and type of anesthesia.

Conclusion: RCRI is significantly correlated with 1-year postoperative mortality regardless of sex, age, and type of anesthesia. Further studies to validate these findings are warranted.