Supplementary Material for: A Comparison of Traditional and Novel Definitions (RIFLE, AKIN, and KDIGO) of Acute Kidney Injury for the Prediction of Outcomes in Acute Decompensated Heart Failure
datasetposted on 25.02.2013 by Roy A.K., Mc Gorrian C., Treacy C., Kavanaugh E., Brennan A., Mahon N.G., Murray P.T.
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
Aims: To determine if newer criteria for diagnosing and staging acute kidney injury (AKI) during heart failure (HF) admission are more predictive of clinical outcomes at 30 days and 1 year than the traditional worsening renal function (WRF) definition. Methods: We analyzed prospectively collected clinical data on 637 HF admissions with 30-day and 1-year follow-up. The incidence, stages, and outcomes of AKI were determined using the following four definitions: KDIGO, RIFLE, AKIN, and WRF (serum creatinine rise ≥0.3 mg/dl). Receiver operating curves were used to compare the predictive ability of each AKI definition for the occurrence of adverse outcomes (death, rehospitalization, dialysis). Results: AKI by any definition occurred in 38.3% (244/637) of cases and was associated with an increased incidence of 30-day (32.3 vs. 6.9%, χ2 = 70.1; p < 0.001) and 1-year adverse outcomes (67.5 vs. 31.0%, χ2 = 81.4; p < 0.001). Most importantly, there was a stepwise increase in primary outcome with increasing stages of AKI severity using RIFLE, KDIGO, or AKIN (p < 0.001). In direct comparison, there were only small differences in predictive abilities between RIFLE and KDIGO and WRF concerning clinical outcomes at 30 days (AUC 0.76 and 0.74 vs. 0.72, χ2 = 5.6; p = 0.02) as well as for KDIGO and WRF at 1 year (AUC 0.67 vs. 0.65, χ2 = 4.8; p = 0.03). Conclusion: During admission for HF, the benefits of using newer AKI classification systems (RIFLE, AKIN, KDIGO) lie with the ability to identify those patients with more severe degrees of AKI who will go on to experience adverse events at 30 days and 1 year. The differences in terms of predictive abilities were only marginal.