Acute kidney injury in pancreatitis: association and predictive value of composite inflammatory indices

Objectives To investigate the associations of composite inflammatory indices with acute kidney injury (AKI) and hospital mortality in patients with pancreatitis, and to evaluate their incremental value in AKI prediction models. Methods This retrospective cohort study included patients with pancreatitis admitted to the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. The associations of composite inflammatory indices, including the systemic immune-inflammation index (SII), neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR), with AKI and hospital mortality were evaluated. An AKI prediction model was developed based on clinical variables and inflammatory indices, and its performance was assessed using receiver operating characteristic (ROC) curves, calibration analysis, and decision curve analysis. External validation was performed using the eICU Collaborative Research Database (eICU–CRD) cohort. Results A total of 646 patients from the MIMIC-IV cohort were included, with 319 (68.02%) developing AKI. Composite inflammatory indices showed limited predictive performance for AKI when used alone. The integrated prediction model combining inflammatory indices and clinical variables showed acceptable discrimination in the MIMIC-IV cohort (AUC = 0.654, 95% CI 0.602–0.706) and was externally validated in the eICU–CRD cohort (AUC = 0.658, 95% CI 0.608–0.708). RCS analysis showed no significant associations between inflammatory indices and AKI, while PLR showed a significant nonlinear association with hospital mortality only in the eICU–CRD cohort. Conclusions Composite inflammatory indices alone showed limited discrimination for AKI prediction, whereas their integration with clinical variables improved discrimination compared with inflammatory indices alone, although no significant incremental benefit was observed over clinical variables alone. The developed model may provide a useful tool for early identification of patients at higher risk of AKI in pancreatitis.