PIRO Model: Predicting Acute Kidney Injury in Acute on Chronic liver Failure

Abstract

Background and Aim: There is limited data on predictors of acute kidney injury in acute on chronic liver failure. We developed a PIRO model (Predisposition, Injury, Response, Organ failure) for predicting acute kidney injury in a multicentric cohort of acute on chronic liver failure patients. Patients and Methods: Data of 2360 patients from APASL-ACLF Research Consortium (AARC) was analysed. Multivariate logistic regression model (PIRO score) was developed from a derivation cohort (n = 1363) which was validated in another prospective multicentric cohort of acute on chronic liver failure patients (n = 997). Results: Factors significant for P component were serum creatinine[(≥ 2 mg/dL), OR 4.52, 95% CI (3.67 - 5.30)], bilirubin [(< 12 mg/dL, OR 1) vs (12 - 30 mg/dL, OR 1.45, 95% 1.1 - 2.63) vs (≥ 30 mg/dL, OR 2.6, 95% CI 1.3 - 5.2)], serum potassium [(< 3 mmol/L, OR-1) vs (3 - 4.9 mmol/L, OR 2.7, 95% CI 1.05 - 1.97) vs (≥ 5 mmol/L, OR 4.34, 95% CI 1.67 - 11.3)] and blood urea (OR 3.73, 95% CI 2.5 - 5.5); for I component nephrotoxic medications (OR 9.86, 95% CI 3.2 - 30.8); for R component, Systemic Inflammatory Response Syndrome (OR 2.14, 95% CI 1.4 - 3.3); for O component, Circulatory failure (OR 3.5, 95% CI 2.2 - 5.5). The PIRO score predicted acute kidney injury with C-index of 0.95 and 0.96 in the derivation and validation cohort. The increasing PIRO score was also associated with mortality (p < 0.001) in both the derivation and validation cohorts. Conclusions: The PIRO model identifies and stratifies acute on chronic liver failure patients at risk of developing acute kidney injury. It reliably predicts mortality in these patients, underscoring the prognostic significance of acute kidney injury in patients with acute on chronic liver failure.

Publication
Development of predisposition, injury, response, organ failure model for predicting acute kidney injury in acute on chronic liver failure. Liver Int. 2017;00:1–11
Date

Know more about Predictive Nomogram

Click here for the Predictive Nomogram

comments powered by Disqus