Development and validation of a predictive model for the early occurrence of acute kidney injury in patients with acute pancreatitis

Abstract

Abstract Background: Acute pancreatitis (AP) is associated with a high incidence of acute kidney injury (AKI), which has a high mortality rate. Currently, there is no clinically useful tool for predicting AKI in AP patients. Therefore, this study aimed to develop a predictive nomogram of the early onset of AKI in AP patients admitted to the intensive care unit (ICU).Method: Data were extracted from the Medical Information Mart for Intensive Care IV version 1.0 (MIMIC-IV version 1.0) database. Eligible patients were randomly divided into training and validation cohorts. The training cohort was used to construct the model, while the validation cohort was used to validate the model. The independent prognostic factors for the early (within seven days of admission) development of AKI in AP patients were determined using the all-subsets regression method. Subsequently, a nomogram was constructed to predict the early occurrence of AKI in AP patients. After that, multiple regression analysis was used to validate the predictive factors. Finally, we determined the area under the receiver operating characteristic curve (AUC) values, developed calibration curves and conducted decision curve analysis (DCA).Results: Seven independent prognostic factors, including age, ethnicity, total bilirubin, activated partial thromboplastin time, need for mechanical ventilation, use of vasoactive drugs, and sepsis, were identified as predictive factors for early onset AKI in AP patients. The constructed nomogram of the training cohort had an AUC value determined at a 95% confidence interval (95% CI) of 0.795(0.758-0.832). However, the nomogram for the validation cohort had an AUC value of 0.772(0.711-0.832, 95% CI).The AUC values of the nomogram were higher than those of the BISAP, Ranson, APACHE II scores, indicating that the nomogram had a good differentiation and discriminative ability. Further, the calibration curve revealed that the predictions had a high agreement with the actual observations. Finally, the DCA curves showed that the nomogram had a good clinical applicability value.Conclusion: The constructed nomogram showed a good predictive ability for determining the early occurrence of AKI in AP patients. The findings of this study are clinically useful in guiding clinicians in averting the development of AKI in AP patients.