Prospective Validation of ExCARE Model for 30-Day Postoperative Mortality

Prospective Validation of a Risk Stratification Model for Mortality Up to 30 Days After Surgery

This prospective observational cohort study aims to validate the ExCARE risk model, derived from four preoperative variables (age, ASA-PS classification, procedure urgency, and surgical magnitude), for predicting in-hospital mortality up to 30 days post-non-cardiac surgery in a Brazilian tertiary hospital. A secondary validation of the SORT (Surgical Outcome Risk Tool) model will also be performed for comparison. The study involves no interventions and focuses on risk stratification to improve perioperative care allocation.

Approximately 300 million major surgical procedures are performed worldwide annually, with postoperative complications increasing costs and mortality. This study addresses the need for better perioperative risk stratification in Brazil, following WHO recommendations for universal health coverage. The ExCARE model, developed from 13,581 patients at Hospital de Clínicas de Porto Alegre (HCPA), showed good discriminative capacity (repeated in a 7,254-patient validation cohort). This external prospective validation at Hospital Nossa Senhora da Conceição will assess its performance in a new setting. Patients are stratified into four risk classes: Class I (<2% mortality), Class II (2-5%), Class III (5-10%), Class IV (>10%). Data will be analyzed using logistic regression, AUROC for discrimination, and Hosmer-Lemeshow test for calibration. Expected enrollment: 3,000 patients over 6 months.

Inclusion Criteria:

• Age >18 years; Scheduled for elective, urgent, or emergency surgeries at Hospital Nossa Senhora da Conceição; For multiple surgeries in one admission, only the major procedure considered.

Exclusion Criteria:

• Diagnostic procedures; Sedation-only or local anesthesia; Hepatic, pulmonary, or cardiac transplants; Brain-dead patients for organ donation.