现在的位置: 首页研究点评, 进展交流>正文
[Blue Journal]: 渥太华大学推荐阅读(1/3)
2017年05月29日 研究点评, 进展交流 暂无评论

Recommended Reading from the University of Ottawa Nephrology Fellows

David Massicotte-Azarniouch, Syed Obaid Amin, Caitlin Hesketh and Edward G. Clark

AJRCCM Articles in Press. Published on 02-May-2017 as 10.1164/rccm.201611-2375RR

Gaudry S, et al.; Initiation strategies for renal-replacement therapy in the intensive care unit. N Engl J Med (1)

Reviewed by David Massicotte-Azarniouch

Meta-analyses of mainly observational studies have shown relatively earlier initiation of Renal Replacement Therapy (RRT) for Acute Kidney Injury (AKI), defined according to clinical and laboratory findings at the time of initiation, is associated with a survival benefit (2, 3). Until recently, no large, multi-center randomized controlled trials (RCTs) had addressed this important and controversial topic.

The Artificial Kidney Initiation in Kidney Injury (AKIKI) study (1) was a multi-center (31 ICUs throughout France), open-label RCT of 620 adult patients admitted to the ICU with KDIGO Stage 3 AKI (serum creatinine ≥ 3 times baseline or ≥ 4 mg/dl, and/or urine output < 0.3 ml per kilogram for ≥ 24 hours or anuria for ≥ 12 hours) due to ischemic or toxic acute tubular necrosis. Patients were randomized in a 1 to 1 ratio to ‘early’ vs ‘delayed’ initiation of RRT. For those in the ‘early’ arm, RRT was started within 6 hours of KDIGO Stage 3 AKI criteria being met. In the ‘delayed’ arm, RRT was only initiated if oliguria persisted for 72 hours or once a clear indication developed, such as hyperkalemia, severe acidosis, uremia or pulmonary edema refractory to diuretics. The average age of enrolled patients was 66 years. Eighty-six percent of patients were mechanically ventilated and 85% were receiving vasopressors. The mean SOFA score was 11 and 80% of patients had sepsis. Baseline characteristics were nearly identical in both groups. The primary outcome, mortality at 60 days, was similar between ‘early’ and ‘delayed’ groups (48.5% vs 49.7% respectively, P = 0.79). In the ‘delayed’ group, only 157 patients (51%) underwent RRT. Patients in the ‘early’ group received RRT a median of 2 hours after randomization versus a median of 57 hours amongst those who received RRT in the ‘delayed’ group. Recovery of urine output, a proxy for improved kidney function, occurred earlier in the ‘delayed’ vs ‘early’ group (P<0.001). There were also less catheter-related bloodstream infections in the ‘delayed’ vs ‘early’ group (5% vs 10%, P=0.03). Sub-group analysis showed the lowest mortality in the subset of ‘delayed’ arm patients who never required initiation of RRT (37.1%), intermediate mortality (48.5%) in ‘early’ arm patients (all of whom received RRT soon after randomization), and the highest mortality (61.8%) in ‘delayed’ arm patients who ultimately required RRT (p<0.001). These differences became non-significant after adjustment for severity of illness suggesting that ‘delayed’ RRT is not harmful unto itself.

This study has several important limitations. In particular, the study was only powered to demonstrate a 15% absolute survival benefit with ‘delayed’ vs ‘early’ initiation at 60 days. As some expert commentaries have highlighted (4, 5), a 15% survival benefit greatly exceeds that of most interventions shown to be effective in the ICU. Thus, despite being the largest study to date, it was likely still underpowered to demonstrate a clinically meaningful benefit due to ‘delayed’ initiation. Nonetheless, the AKIKI study indicates that, for acutely ill, septic patients with severe AKI, a strategy of delayed RRT initiation is not associated with worse outcomes. It highlights that many patients with severe AKI recover kidney function without ever needing RRT and, as such, a strategy of delayed initiation might be favored. These findings should be considered in the context of the ELAIN study (6), another recent RCT that addressed the question of initiation timing with conflicting results (reviewed next).

给我留言

您必须 [ 登录 ] 才能发表留言!

×
腾讯微博