Development and validation of a novel prediction model for predicting renal function recovery after diversion in patients with obstructive uropathy.

ABSTRACT

OBJECTIVE: To develop and validate a novel prediction model predicting renal function recovery following diversion in patients with obstructive uropathy (OU) to the emergency department (ED). METHODS: After a systematic literature search, a novel prediction model called PGIMER Obstructive Uropathy Score (POUS) was constructed including five variables age (<60 or >60 years), duration of symptoms (<4 or >4 weeks), presence of solitary functioning kidney, baseline hemoglobin levels and venous blood pH. This model was then validated in a prospective, observational single-center study of patients presenting with OU caused by various etiologies. Patients with OU and raised serum creatinine (>2 mg/dL) presenting to our ED were included. Renal function recovery was defined as creatinine value <1.5 mg/dL at 4 weeks following diversion. RESULTS: In this study, 174 consecutive patients with OU were recruited, and 74 (42.5%) patients had renal function recovery. All the variables included in the POUS were noted to be statistically significant on univariate analysis. On multivariate logistic regression analysis, only POUS was identified as an independent predictor of renal function recovery. On receiver operating curve analysis, the area under the curve for POUS was 0.832 for predicting recovery. A POUS of 5 or more had specificity and sensitivity of 83% and 73.6%, respectively, in predicting renal function recovery. The goodness of fit and calibration plots showed good concordance of the predicted values with the observed values. CONCLUSIONS: The POUS model is an accurate and simple-to-use tool for predicting renal function recovery. POUS model requires external validation prior to clinical use in different populations.