Published in:
01-08-2017 | Reports of Original Investigations
Early rise in postoperative creatinine for identification of acute kidney injury after cardiac surgery
Published in: Canadian Journal of Anesthesia/Journal canadien d'anesthésie | Issue 8/2017
Login to get accessAbstract
Purpose
Acute kidney injury (AKI) is a potentially serious complication of cardiac surgery. Treatment strategies are unlikely to prove efficacious unless patients are identified and treated soon after the onset of injury. In this observational study, we determined and validated the ability of an early rise in postoperative serum creatinine to identify patients who suffer AKI during cardiac surgery.
Methods
The relationship between an early rise in creatinine (immediate postoperative / preoperative creatinine) and AKI (> 50% increase in creatinine by postoperative calendar days 1or 2) was determined by logistic regression modelling. Existing databases were used for model development (n = 4,820; one institution) and validation (n = 6,553; 12 institutions).
Results
Acute kidney injury occurred in 9.1% (n = 437) and 9.8% (n = 645) of patients in the development and validation sets, respectively. An early rise in creatinine was related to AKI (P < 0.001), with an area under the receiver operating characteristic curve of 0.78 (95% confidence interval [CI], 0.75 to 0.80) in the development set and 0.77 (95% CI, 0.75 to 0.79) in the validation set. Using a threshold ratio of > 1.30 (n = 127), the sensitivity, specificity, positive, and negative predictive values for AKI in the development set were 20% (95% CI, 16 to 24), 99% (95% CI, 99 to 99), 68% (95% CI, 59 to 76), and 93% (95% CI, 92 to 93), respectively.
Conclusion
In patients undergoing cardiac surgery with cardiopulmonary bypass, an early rise in postoperative creatinine is a useful marker for the early identification of AKI patients. This could allow inclusion of such patients in clinical trials of promising therapeutic strategies that need to be initiated soon after the onset of injury.