Diagnostic Accuracy of the Modified CT Severity Index for Severe Acute Pancreatitis Using the Revised Atlanta Classification as Reference Standard

Abstract

OBJECTIVES

This study aimed to determine the diagnostic accuracy of the Modified Computed Tomography Severity Index (mCTSI) for identifying severe acute pancreatitis using the Revised Atlanta Classification as the reference standard.

METHODOLOGY

This cross-sectional validation study included 155 adults aged 20–60 years with acute pancreatitis who underwent contrast-enhanced computed tomography at Fauji Foundation Hospital, Rawalpindi, from 22 February to 22 May 2025. mCTSI was calculated from computed tomography performed after 48 hours of admission, while severe disease, by Revised Atlanta Classification, was defined by persistent organ failure for more than 48 hours using the Modified Marshall score. Diagnostic performance was evaluated using a 2×2 table and reported as sensitivity, specificity, predictive values, and overall accuracy.

RESULTS
Mean age was 42.8 ± 11.6 years, and 94 (60.6%) patients were male. Severe acute pancreatitis was classified in 46 (29.7%) patients by mCTSI and in 41 (26.5%) by Revised Atlanta criteria. There were 37 true-positive, 105 true-negative, nine false-positive, and four false-negative results. mCTSI demonstrated sensitivity of 90.2%, specificity of 92.1%, positive predictive value of 80.4%, negative predictive value of 96.3%, and overall diagnostic accuracy of 91.6%. Performance remained broadly consistent across age, sex, and symptom-duration strata.

CONCLUSION

CTSI showed high diagnostic accuracy for severe acute pancreatitis compared with the Revised Atlanta Classification. Its high negative predictive value supports its usefulness for radiological risk stratification when interpreted alongside clinical assessment and serial organ-failure evaluation and may support timely escalation of monitoring and supportive clinical care decisions.