Postoperative Fluid Overload is a Useful Predictor of the Short-Term Outcome of Renal Replacement Therapy for Acute Kidney Injury After Cardiac Surgery

To analyze the predictive value of postoperative percent fluid overload (PFO) of renal replacement therapy (RRT) for acute kidney injury (AKI) patients after cardiac surgery. Data from 280 cardiac surgery patients between 2005 January and 2012 April were collected for retrospective analyses. A receiver operating characteristic (ROC) curve was used to compare the predictive values of cumulative PFO at different times after surgery for 90-day mortality. The cumulative PFO before RRT initiation was 7.9% ± 7.1% and the median PFO 6.1%. The cumulative PFO before and after RRT initiation in intensive care unit (ICU) was higher in the death group than in the survival group (8.8% ± 7.6% vs 6.1% ± 5.6%, P = 0.001; −0.5[−5.6, 5.1]% vs 6.9[2.2, 14.6]%, P < 0.001). The cumulative PFO during the whole ICU stay was 14.3% ± 15.8% and the median PFO was 10.7%. The areas under the ROC curves to predict the 90-day mortality by PFO at 24 hours, cumulative PFO before and after RRT initiation, and PFO during the whole ICU stay postoperatively were 0.625, 0.627, 0.731, and 0.752. PFO during the whole ICU stay ≥7.2% was determined as the cut-off point for 90-day mortality prediction with a sensitivity of 77% and a specificity of 64%. Kaplan–Meier survival estimates showed a significant difference in survival among patients with cumulative PFO ≥ 7.2% and PFO < 7.2% after cardiac surgery (log-rank P < 0.001). Postoperative cumulative PFO during the whole ICU stay ≥7.2% would have an adverse effect on 90-day short-term outcome, which may provide a strategy for the volume control of AKI-RRT patients after cardiac surgery.