Transient and Persistent Acute Kidney Injury and the Risk of Hospital Mortality in Critically Ill Patients: Results of a Multicenter Cohort Study.

Objective: To assess the prognostic impact of transient and persistent acute kidney injury in critically ill patients.
Design: Retrospective analysis of prospectively collected patient data
Setting: : Six hospital ICUs.
Patients: Critically-ill patients with ICU stay longer than three days.
Intervention: None.
Measurements and Main Results: Assessment of hospital survival with respect to acute kidney injury duration. A total of 447 patients were included in this study, including 283 patients (63.3%) with an acute kidney injury at admission (175 and 108 patients with persistent and transient acute kidney injury, respectively). Patients with persistent acute kidney injury more frequently had stage 3 acute kidney injury (42.9% vs 30.6%; p = 0.04). Hospital survival was 76.2% (n = 125) in patients without acute kidney injury, 70.4% (n = 76) in patients with transient acute kidney injury, and 61.1% (n = 107) in patients with persistent acute kidney injury. After adjustment for confounding factors, the factors associated with lower hospital survival were the need for vasopressors (odds ratio, 0.65; 95% CI, 0.43-0.98) and the presence of persistent acute kidney injury (odds ratio, 0.58; 95% CI, 0.36-0.95). When included in the final model, stage 3 acute kidney injury was independently associated with a lower hospital survival (odds ratio, 0.83; 95% CI, 0.70-0.98), and persistent acute kidney injury was no longer associated with outcome.
Conclusion: Two thirds of the critically ill patients with acute kidney injury have persistent acute kidney injury. Although mortality increased progressively with the duration of acute kidney injury, we found no independent association between this duration and patient outcome when the acute kidney injury severity is taken into account. Our results suggest that the classical "prerenal acute kidney injury" and "acute tubular necrosis" paradigm might be of limited interest from a pathophysiological or prognostic point of view.