Abstract
BACKGROUND: Acute pancreatitis (AP) can lead to severe complications like pancreatic/peripancreatic necrosis and sepsis. Existing scoring systems for predicting severity and mortality have limitations. This study aimed to characterize the demographic and etiological aspects of AP and compare the predictive accuracy of the Acute Physiology and Chronic Health Evaluation II (APACHE-II), modified computed tomography severity index (mCTSI), and Modified Marshall Score (MMS) for assessing severity and mortality in a large cohort from India. METHODS: This retrospective observational study included consecutive AP patients admitted to a tertiary care hospital in northern India between January 2020 and December 2023. Patients aged ≥15 years were enrolled based on established criteria. Data on demographics, clinical outcomes, and scoring systems (APACHE-II, mCTSI, MMS) were analyzed using Chi-square tests and receiver operating characteristic (ROC) curve analysis for predictive accuracy. Statistical significance was defined as p < 0.05. RESULTS: Among 645 patients, 77% (n=501) were male, with a median age of 36 years. Mild and moderate disease occurred in 30.7% (n=198) and 42% (n=271) of cases, respectively, while severe AP was present in 27.3% (n=176). An APACHE-II score ≥8 (AUC: 0.941) at admission had higher diagnostic accuracy (91%) for severe AP compared to mCTSI ≥8 (AUC: 0.787), which had a diagnostic accuracy of 73.7%. APACHE-II (AUC: 0.857) was the most accurate for predicting infected necrosis, while mCTSI (AUC: 0.742) was better for predicting the need for intervention. Seventy-seven patients died (case fatality rate: 11.9%). The APACHE-II score (AUC:0.947) was a superior predictor of mortality compared to mCTSI (AUC: 0.790), with diagnostic accuracies of 83.1% vs. 67.6%, respectively. CONCLUSION: The APACHE-II score more accurately predicts the severity, infected necrosis, and mortality of acute pancreatitis, while mCTSI is better for assessing intervention needs, and both scoring systems are complementary in clinical management.