The Kidney Disease Improving Global Outcome (KDIGO) in comparison with Risk, Injury, Failure, Loss and End Stage (RIFLE) Criteria as a predictor of clinical outcomes among critically ill patients with acute kidney injury

Background: Acute kidney injury (AKI) is a major clinical complication in the hospital setting particularly in the intensive care unit. Staging of AKI is as important as the diagnosis since it can predict the need for renal replacement therapy (RRT). There are several risk prediction models currently available. However, the data on predicting the risk for death and need for RRT among critically ill patients with AKI are still lacking in our local setting. This study aims to compare two risk prediction models: the Risk, Injury, Failure, and Loss of Kidney function (RIFLE) and the Kidney Disease: Improving Global Outcome (KDIGO).

Methods: This is a retrospective cohort study involving critically ill patients with AKI admitted at an intensive care unit of a tertiary hospital. Review of medical records was done to collect data and variables needed for the prediction models. Statistical and area under the curve analysis were used to calculate and predict mortality and need for renal replacement therapy outcomes of the two risk scoring systems.

Results: A total of 517 patients were included in the study. In predicting mortality among high risk patients, RIFLE had 14.9% sensitivity and 93.4% specificity while KDIGO had 19.2% sensitivity and 89.6% specificity. In predicting RRT among high risk patients, RIFLE had 16.0% sensitivity and 93.8% specificity while KDIGO had 30.8% sensitivity and 95.2% specificity.

Conclusion: The incidence of AKI in critically ill patients varied according to the criteria used. The two criteria can still be used for prediction of mortality and need for renal replacement therapy. We recommend a bigger sample population that is more representative of patients with acute kidney injury.