**Background:** Acute pancreatitis (AP) incidence is rising globally (33-74 cases per 100,000/year), with overall mortality of 5-6% in leading centers but 20-45.4% in acute necrotizing pancreatitis (ANP) complicated by purulent-septic complications. Infected pancreatic necrosis occurs in 40-80% of ANP cases and is an absolute indication for surgery. Modern management increasingly uses minimally invasive procedures (MIP) with a step-up approach. Enhanced Recovery After Surgery (ERAS) protocols have shown benefit in other surgical fields but their role in ANP is not well established.
**Methods:** A two-center case-control study was conducted on 148 patients with ANP at Kharkiv National Medical University from January 1, 2015 to October 30, 2022. Patients were divided into a main group (n=95, 2017-2022) receiving step-up approach with ERAS principles, and a comparison group (n=53, 2015-2016) receiving step-up approach without ERAS. Inclusion criteria: proven ANP in early (<1 week) or late (>1 week) phase, any MIP used, age 18-70 years. Exclusion criteria: postoperative AP, refusal to participate. Severity was assessed using BISAP, SOFA, and APACHE II scores. The ERAS protocol included minimized ICU stay, multimodal analgesia (epidural anesthesia at ThVII-ThVIII), early mobilization from postoperative day 1, physiotherapy, early oral intake of clear liquids, and early enteral nutrition. Postoperative complications were classified by Clavien-Dindo. Statistical analysis used STATISTICA 13.3 with χ² tests and ROC curve analysis.
**Key Results:** Baseline characteristics were similar between groups. Median age was 53 (IQR 29-69) in comparison group vs 57 (IQR 36-70) in main group (p=0.062). BISAP score was 4 (IQR 2-5) in both groups (p=0.664). Infected ANP occurred in 22.6% of comparison group vs 33.8% of main group (p=0.385). Pancreatic infection was documented in 92/148 patients (62.2%). Of 314 microbial strains isolated, 222 (70.7%) were gram-negative and 92 (29.3%) gram-positive. Among gram-negative organisms, Pseudomonas aeruginosa (65 strains, 29.3%) and Klebsiella spp. (46 strains, 20.7%) predominated. Among gram-positive organisms, Enterococcus faecalis (34 strains, 37%) was most common. Antibiotic sensitivity varied: Pseudomonas aeruginosa showed 100% sensitivity to ceftazidime/avibactam but only 35.4% to imipenem and 40% to meropenem. Klebsiella spp. showed 89.1% sensitivity to meropenem and 69.6% to imipenem. Enterococcus faecalis showed 100% sensitivity to tigecycline and linezolid but only 47.1% to vancomycin.
ICU stay was significantly shorter in the main group (median 3 days, IQR 1-8) vs comparison group (median 11 days, IQR 1-18) (p=0.004). Clavien-Dindo I-IIIa complications were significantly reduced in the main group (9.5% vs 52.8%, p<0.001). Total hospital stay was significantly shorter in the main group (median 23 days, IQR 10-29) vs comparison group (median 34 days, IQR 12-44) (p=0.003). Life-threatening complications (Clavien-Dindo IVa-IVb) were 3.7% lower in the main group but not statistically significant (p=0.269). 30-day mortality was 18.9% in the main group vs 22.6% in the comparison group (p=0.081, not significant). ROC analysis showed MOF before surgery had good predictive value for mortality (AUC=0.814, 95% CI 0.728-0.896, sensitivity 81.2%, specificity 74.6%), and MOF after surgery had excellent predictive value (AUC=0.931, 95% CI 0.892-0.994, sensitivity 89.2%, specificity 94.1%).
**Clinical Implications:** This study demonstrates that implementing ERAS principles within a step-up surgical approach for ANP can significantly reduce ICU stay, overall hospital stay (by an average of 11 days), and minor-to-moderate postoperative complications. However, ERAS did not significantly impact mortality, which was primarily driven by the presence and persistence of multiple organ failure. The microbiological data highlight the predominance of gram-negative pathogens and the emergence of antibiotic-resistant hospital strains during treatment, emphasizing the need for targeted antibiotic therapy based on local epidemiology. The strong predictive value of MOF for mortality (AUC 0.931 post-surgery) suggests that preventing and managing organ failure should be a primary therapeutic goal. Limitations include the non-randomized design, potential selection bias, and single-country setting which may limit generalizability.