Predictive Value of MELD Score and Charlson Comorbidity Index in Thoracic Aortic Surgery Patients

Le résumé fourni par la source

Thoracic aortic aneurysms (TAAs) carry a high risk of fatal rupture, necessitating improved preoperative risk stratification. This study evaluates the predictive value of systemic risk scores—specifically the Model for End-Stage Liver Disease (MELD) and the Charlson Comorbidity Index (CCI)—for in-hospital mortality, length of stay, and one-year mortality in patients undergoing elective ascending aortic surgery. The study further compares MELD variants (MELD-Na and MELD-XI) for their prognostic performance in this context. This retrospective single-center study analyzed digital medical records of 500 patients undergoing elective surgery for ascending thoracic aortic disease between 2003 and 2023. MELD, MELD-Na (incorporating sodium), and MELD-XI (excluding INR for anticoagulated patients) were calculated from preoperative laboratory data. The CCI was derived from documented comorbidities. Outcomes included in-hospital mortality, length of stay (from admission to discharge), and one-year mortality assessed via outpatient follow-up. The study excluded patients undergoing emergency surgery for Stanford type A aortic dissection. MELD-Na incorporates serum sodium, while MELD-XI is a variant that excludes INR for patients with anticoagulation. The Charlson Comorbidity Index (CCI) was derived from patients’ medical histories prior to surgery. Length of stay was defined as total inpatient days between admission and discharge. One-year mortality was assessed via outpatient follow-up data. Loss to follow-up did not exceed 30%. Of 500 patients (median age 64 years, 72.8% male), the MELD-Na score showed the strongest ability to predict in-hospital mortality (AUC = 0.698), outperforming both the standard MELD (AUC = 0.690) and the age-adjusted CCI (AUC = 0.631). For one-year mortality (N = 355), MELD-Na again performed best (AUC = 0.732), while the unadjusted CCI showed minimal predictive value (AUC = 0.509). Predictive power for hospital length of stay was limited across all scores; the age-adjusted CCI achieved the highest, though modest, discrimination (AUC = 0.627). 1-year mortality was assessed in 355 patients with available follow-up data (29.0% lost to follow-up). Among these, non-survivors had significantly higher MELD scores (p < 0.001). MELD-Na demonstrated the strongest predictive performance (AUC = 0.732). The MELD score, particularly MELD-Na, demonstrated strong predictive ability for in-hospital and 1-year mortality, but showed limited value in estimating hospital stay duration. MELD-Na and the age-adjusted CCI provide valuable preoperative prognostic information for patients undergoing elective ascending aortic surgery. While not intended to replace established risk models, their simplicity and reliance on routine clinical data make them attractive tools for early triage, especially in older or multimorbid patients. Their integration into preoperative planning may enhance individualized risk assessment and resource allocation.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.