Abstract
Background and Objectives: Early identification of patients with acute pancreatitis (AP) who will require operative intervention remains a major clinical challenge. Traditional severity scores and delayed inflammatory biomarkers offer limited early predictive accuracy. This study aimed to evaluate the combined predictive value of IL-6, necrotic burden, and organ failure parameters for determining the need for surgical management in AP. Materials and Methods: We performed a retrospective cohort study of 325 consecutive patients with AP admitted to Floreasca Emergency Hospital between January 2020 and December 2024. Clinical, laboratory, and imaging parameters were collected, including IL-6 at 24 h, CRP at 48-72 h, SOFA and BISAP scores, and morphologic complications. The primary endpoint was the requirement for operative intervention. Univariate and multivariable logistic regression analyses were conducted, and model performance was assessed using ROC curves and calibration testing. Results: Sixty patients (18.5%) underwent surgery for pancreatitis-related complications. Operated patients demonstrated higher IL-6 levels (median 120 vs. 55 pg/mL), necrosis >30% (68.3% vs. 14%), infected necrosis (55% vs. 2.3%), and persistent organ failure (46.7% vs. 9.8%) (all p < 0.001). In multivariable analysis, IL-6 at 24 h (OR 1.35 per 50 pg/mL), necrosis > 30% (OR 4.80), infected necrosis (OR 6.20), persistent organ failure ≥48 h (OR 2.60), SOFA score (OR 1.18 per point), and CRP at 48-72 h (OR 1.09 per 50 mg/L) independently predicted surgery. The final model showed excellent discrimination (AUC 0.88) and good calibration. Conclusions: IL-6-driven inflammatory escalation, combined with necrotic burden and persistent organ dysfunction, provides a robust early predictive framework for identifying AP patients likely to require surgical intervention. Integration of these parameters may improve triage, timing of intervention, and multidisciplinary decision-making in acute pancreatitis.