249: Early Predictors of Icu and 6-Month Mortality After Cardiac Surgery: A Costa Rican Cohort

Introduction: Despite advances in surgical and critical care, postoperative mortality following cardiac surgery remains a major challenge. Early identification of patients at higher risk may help guide monitoring strategies and resource allocation in the ICU setting. Methods: We conducted a retrospective cohort study of all adult patients undergoing cardiac surgery at Hospital México (San José, Costa Rica) during 2023 who were admitted postoperatively to the ICU. We collected data on demographics, comorbidities, EuroSCORE II, NT-proBNP levels, renal function, and early postoperative organ dysfunction. Outcomes included ICU, 28-day, and 6-month mortality. Survival analysis and multivariable Cox regression identified independent predictors of mid-term mortality. ROC curve analyses were used to evaluate the discriminative capacity of key predictors and define optimal cut-off points. Results: A total of 147 patients were included, with a median age of 61 years; 65.3% were younger than 65. ICU mortality was 4.8%, 28-day mortality 3.4%, and 6-month mortality 6.1%. Three early predictors were independently associated with increased 6-month mortality. Preoperative NT-proBNP >1800 pg/mL nearly doubled the risk (HR 1.92, 95% CI: 1.11–3.33), while a SOFA score ≥6 within the first 24 hours of ICU admission increased the hazard more than fivefold (HR 5.31, 95% CI: 1.06–26.60). Cardiac arrest during ICU stay was the most significant predictor (HR 15.95, 95% CI: 3.34–76.17). ROC curve analyses confirmed the strong predictive ability of NT-proBNP and SOFA score. An NT-proBNP cut-off of 1800 pg/mL yielded an area under the curve (AUC) of 0.77, with 77.8% sensitivity and 68.7% specificity for 6-month mortality. Similarly, a SOFA score ≥6 had an AUC of 0.82, with 85.7% sensitivity and 76.5% specificity. These findings support their use as early, non-invasive markers for risk stratification. Conclusions: NT-proBNP and SOFA score, assessed within the first 24 hours of ICU admission, demonstrated strong prognostic value for mid-term mortality following cardiac surgery. Their predictive performance, supported by robust ROC analysis, provides a practical framework for early identification of high-risk patients in critical care settings—especially where resources are limited.