Purpose: We undertook this study to characterize the epidemiology of acute kidney injury (AKI) in Canadian critical care units. We aimed to identify predictors of mortality for patients diagnosed with AKI.
Methods: We conducted a prospective cohort study of consecutive patients admitted to critical care units at five Canadian hospitals over a 30-day period. Each patient was followed until hospital discharge or for a maximum of 30 days. The serum creatinine criteria for the Acute Kidney Injury Network (AKIN-SCr) system were used to identify, classify, and characterize patients who developed AKI. We used multivariable logistic regression to predict 30-day mortality among patients with AKI.
Results: We identified 603 patients, 161 (26.7%) of whom developed AKI. Compared to patients without AKI, those with AKI were more likely to die (29.2% vs 8.6%, P < 0.001). The risk of death increased with increasing AKIN-SCr stage (P < 0.001). In all, 19 patients (11.8% of those with AKI) commenced dialysis a median of one day (interquartile range, one to two days) after AKI diagnosis. At AKI diagnosis, the blood urea nitrogen (BUN) level (adjusted odds ratio [OR] 1.68, 95% confidence interval [CI] 1.01 to 2.79/10 mmol·L(-1)) and serum bicarbonate (adjusted OR 0.88, 95% CI 0.81 to 0.95/1 mmol·L(-1)) were associated with 30-day mortality and predicted death with an area under the receiver-operating characteristic curve of 0.79 (95% CI 0.71 to 0.86).
Conclusions: Acute kidney injury is a common complication of critical illness in Canada. The development of even the mildest stage of AKI is associated with a substantially higher risk of death. At AKI diagnosis, routine clinical data may be helpful for predicting adverse outcomes.