Rev Assoc Med Bras (1992). 2026 Jun 29;72(3):e20251821. doi: 10.1590/1806-9282.20251821. eCollection 2026.
ABSTRACT
INTRODUCTION: Surgical procedures are associated with an excess of postoperative cardiovascular events. Risk scores are used for cardiovascular stratification but lack proper validation across diverse populations.
OBJECTIVE: The aim of this study was to assess the accuracy of the Revised Cardiac Risk Index, American College of Physicians score, and intuitive clinical judgment in predicting postoperative cardiovascular complications.
METHODS: In this prospective, longitudinal study, adults assessed for cardiovascular risk undergoing elective non-cardiac surgery were included. Perioperative clinical variables, Revised Cardiac Risk Index, American College of Physicians, and clinical judgment were evaluated, as well as in-hospital and 30-day postoperative outcomes. Data were analyzed using descriptive statistics, and score performance was assessed by C-statistics.
RESULTS: A total of 420 participants, with a mean age of 64.1±12.6 years, were studied, of which 47.1% were female and 71.2% were Black/Hispanic. Among the studied patients, 74.3% had hypertension, 32.9% were diabetics, 13.1% had heart failure, 11.2% had coronary artery disease, and 7.9% had stroke. Preoperative risk levels (low/medium/high) were as follows: Revised Cardiac Risk Index 88%/10.5%/1.4%; American College of Physicians 71.2%/27.4%/1.4%; and clinical judgment 64.5%/33.1%/2.4%. During hospitalization, there were 11 deaths, 3 myocardial infarctions, and 1 acute pulmonary edema. The risk scores showed poor discrimination for death and combined in-hospital cardiovascular outcomes, with area under the receiver operating characteristic curve <0.70. Cardiologist judgment performed slightly better than the scores. Results were similar for major and non-major surgeries, and when evaluating only patients aged 50 years or older.
CONCLUSION: Cardiac risk assessment demonstrated low accuracy in predicting in-hospital postoperative cardiovascular outcomes, regardless of surgical magnitude.
PMID:42384973 | DOI:10.1590/1806-9282.20251821