Abstract 15896: Predictive Models for Kidney Disease: Improving Global Outcomes (KDIGO) Defined Acute Kidney Injury
Acute Kidney Injury (AKI) is a common and severe complication of cardiac surgery. Preoperative identification of patients at increased risk of AKI is an important consideration in clinical management and selection of patients for consideration in clinical trials of novel renoprotective interventions. Existing scores are limited in that they predict only severe AKI requiring renal replacement therapy. The aim of this study was to develop a novel risk score that predicts all patients at risk of AKI in an unselected, multicentre cardiac surgery cohort.
Prospective routinely collected clinical data from 3 UK cardiac surgical centres was linked to institutional laboratory databases. AKI (Stages 1-3) were defined as per the Kidney Disease: Improving Global Outcomes (KDIGO) Guidelines, which reconciles important differences between earlier consensus definitions. Model building was performed on two datasets. The third dataset was used to externally validate the model. Diagnostic utility was also compared to the Cleveland Clinic Score and the Mehta Score.
In the entire patient cohort (n= 28,187) increasing severity of AKI as determined by KDIGO criteria was associated with incremental increases in morbidity, resource use and mortality. Two risk scores; one for any stage AKI (Stage 1-3), and one for Stage 3 AKI demonstrated excellent discrimination in both development and external validation datasets, with areas under the ROC curve in the external validation sample of 0.74 (95% CI 0.72, 0.76) for the Any Stage, and 0.78 (95% CI 0.74, 0.81) for the Stage 3 AKI models. Comparison with two established risk score, the Cleveland Clinic Score and the Mehta Score has demonstrated better discrimination for Any Stage AKI, and equivalent discrimination for Stage 3 AKI.
This is the first risk score that accurately identifies patients at risk of Any Stage AKI. We suggest that this will have utility in the perioperative management of high risk patients as well as in clinical trial design.
- © 2013 by American Heart Association, Inc.