RESUMO
BACKGROUND: Infected pancreatic necrosis (IPN) accounts for 30% mortality in severe acute pancreatitis (SAP). Early prediction of IPN occurrence is critical for prophylactic measures to be taken. This study aimed to evaluate the predicting value for IPN of combined markers at early stages of SAP. METHODS: The clinical records of 324 SAP patients admitted within 48 h after disease onset were retrospectively analyzed. As potential predictors, the neutrophil to lymphocyte ratio (NLR), blood procalcitonin (PCT) concentration on the 1st, 4th, and 7th day post admission, as well as modified computerized tomography severity index (MCTSI) on the 5-7th day post admission, were extracted. Correlations between these features with IPN were analyzed using logistic regression, and predictive values were estimated using the receiver operating characteristic curve analyses. RESULTS: NLR, PCT, body mass index, and MCTSI were significantly higher in the IPN group (p < 0.001) compared to the control, among which NLR, PCT, and MCTSI were identified as independent predictors for IPN in logistic regression model. Combination of these parameters yielded significant predicting values with an area under curve of 0.92, sensitivity of 97.2%, and specificity of 77.2% in receiver operating characteristic curve analysis. CONCLUSION: Combination of NLR, PCT, MCTSI might facilitate the prediction of IPN occurrence in SAP patients.