Calcium and pH value might predict persistent renal failure in acute pancreatitis in the early phase

Objectives Persistent renal failure (PRF) increases morbidity and mortality in acute pancreatitis (AP). Traditional scoring systems achieve good diagnostic value of AP but not PRF alone. Our study aimed to determine PRF predictors in AP patients for early intervention in the disease development. Methods In the prospective observational study, we consecutively recruited AP patients from October 2013 to October 2016. Complete clinical characteristics on admission were collected. The 2012 revision of the Atlanta classification diagnosed AP, and the Modified Marshall scoring system defined organ failures. We used univariate and multivariate analyses to select risk factors, and plotted survival curves of different groups and ROC curves of parameters to analyze PRF predictors in AP. Results A total of 29 AP patients with PRF and 280 AP patients without PRF were included. Severity scoring and ICU admission rate were higher in the former group. The PRF group's mortality was 10-fold higher than without PRF (20.7% versus 2.1%, p < .001). Most relevant kidney metabolism indicators and excretion have significant differences (p < .05) between the two groups. Serum calcium (Ca) and pH value (pH) were independent risk factors of PRF (p < .05). ROC curve analysis indicated Ca and pH might predict PRF in AP with areas under the curves (AUCs) of 0.758 and 0.809. Conclusions AP patients with PRF had higher morbidity and mortality rate. Our study showed that Ca < 1.94 mmol/L and pH < 7.37 when patients on admission could be used to predict PRF in AP.