Effects of non-severe acute kidney injury on clinical outcomes in critically ill patients

Acute kidney injury (AKI) is a serious medical condition estimated to affect more than ten million people around the world annually [1]. AKI results in a 1.7- to 6.9-fold increased risk of hospital mortality, and risk-adjusted rates of AKI and mortality appear similar across the world [2]. Patients who develop AKI also have worse kidney function at hospital discharge and thus far more risk for chronic kidney disease [2]. However, these risks are most clearly apparent for patients with very severe AKI, such as those who receive renal replacement therapy (RRT). Milder forms of AKI are less clearly associated with adverse outcomes and thus it is unclear whether mild AKI is in the causal pathway for morbidity and mortality. For example, in a large international study of patients cared for in the intensive care unit (ICU), stage 2 or 3 AKIs were strongly associated with mortality even after risk adjustment, whereas for patients incurring only stage 1 AKI the association was attenuated and no longer significant (odds ratio 1.68, 95 % confidence interval (CI) 0.89–3.17; P = 0.11) [2]. For purposes of this review, we define “non-severe AKI” as AKI that is not managed by RRT. We acknowledge that this is an imperfect definition because the decision to commence RRT is often a difficult one and there exists considerable heterogeneity across centers and among physicians, even experts. Thus, the same patient could be counted as non-severe if cared for by one clinician but severe if cared for by another. Although differences exist at the individual patient-provider level, rates of RRT for critically ill patients with AKI in general, though increasing, are relatively consistent around the world [2, 3]. For example, a decade ago, Uchino and colleagues found that in 23 countries on four continents 4.2 % of patients admitted to the ICU received RRT for AKI and this rate was not different between world regions (95 % CI 4.0–4.4 %) [3]. Similarly, Hoste and colleagues recently found that in 33 countries on five continents 13.5 % of all patients admitted to the ICU (excluding patients with end-stage renal disease) received RRT for AKI; again, the range was rather narrow (95 % CI 12.0–15.1 %) [2]. Approximately 25 % of critically ill patients with AKI receive RRT; this article is about the remaining 75 %.