Review of a randomised controlled trial (RCT)
Apart from the explicit guidelines for early antibiotic initiation, the type and timing of organ support in the Intensive Care Unit (ICU) remains a much debated and controversial topic. Acute kidney injury (AKI) is a frequent complication of critical illness, occurring in 5-20% (1-3) of all ICU patients, an incidence that rises to 50% in sepsis (4). Renal replacement therapy (RRT) in various modes, is widely used to compensate for the renal failure and the sequelae of AKI. The absolute indications for establishing RRT are well defined (severe hyperkalaemia, metabolic acidosis, fluid overload, uraemia) but failing those, the optimal timing of introducing renal support remains unknown. After a systematic review based predominantly on retrospective observational studies suggested that early initiation of RRT in patients with AKI might be associated with improved survival (5), a number of randomised controlled trials have tried to clarify this very important topic (6-8).
The RCT published by Gaudry et al. in May 2016 in the New England Journal of Medicine (9) attempted to do just that. It included 620 critically patients with stage 3 AKI that required at least one more organ support (mechanical ventilation, catecholamine infusion or both) without immediate need for RRT. They were randomised to early (within 6 hours of the diagnosis of AKI) and late initiation of RRT, when the above absolute indications plus persistent oligo-anuria occurred. The duration of the follow-up was 60 days, with the primary outcome being overall survival measured from the day of randomisation until death or day 60, and a number of secondary outcomes including days alive and without support, ICU and hospital length of stay and adverse incidents relating to AKI or RRT. The authors reported no significant difference in 60-day mortality (48.5% in the early vs. 49.7% in the late RRT group, p=0.79) or in any other secondary outcome, apart from the number of renal replacement-free days (17 in the early vs. 19 in the late strategy group, p<0.001) and the rate of catheter-related bloodstream infections (31 in the early vs. 16 in the late RRT group, p=0.03).
Gaudry et al. attempt to provide answers to an important and under-researched topic. The study is a well-designed, randomised, multi-centred, prospective trial with a pre-published protocol (9), which addresses a very pertinent question in the critical care setting. When to start RRT has been identified as one of the top research priorities (10) and this study is the first RCT to be published in the field. It is methodologically robust and the lack of statistically significant difference in mortality it reports will inform intensivists’ decisions on starting RRT in ICU patients with severe AKI.
Nonetheless, there are a few considerations when interpreting the trials results, with the power of the study being one of the most important. The authors have estimated a sample size of 546 patients to demonstrate a 14% decrease in mortality (from 55 to 41%) between the two RRT strategies, a reduction that was grossly overestimated, as the actual mortality difference in the 2 groups was only 1.2%. This means that the trial is significantly under-powered, increasing the type two error and thus decreasing the possibility to detect an association between timing of RRT initiation and mortality. Unfortunately, to conduct a study able to detect such a small mortality difference a sample size of more than 70,000 patients would be required, a fact recognised by the authors. Another limitation is the choice of diuresis as a marker of renal function recovery that led the authors to report a more rapid resolution of AKI in the delayed strategy group. AKI does not invariably present with oliguria: non-oliguric AKI occurs in 28–45% of the general ICU population, but may be up to 50% of cases depending on the definition of AKI used, severity of illness and local practice patterns, and has been associated with higher in-hospital mortality (11). Thus, having urine output as the sole marker of renal recovery might actually be misleading.
Lastly, even though 61% of the patients in the late strategy group that were alive at day 60 had not received RRT, the authors correctly conclude that ‘a wait and see’ approach cannot be advised for all. These patients were significantly more acidotic, required more RRT sessions and had significantly higher SOFA scores at the time of RRT initiation than the early strategy ones. They had also received more diuretics, more sodium bicarbonate and higher vasopressor support (the latter difference was not statistically significant). Furthermore, in an unplanned analysis presented in the study’s supplementary appendix, it appears that the patients assigned to the late strategy group had the highest mortality (61.8%) compared to the ones who received RRT early (48.5%) and the ones that were never supported (37.1%), a difference that became non-significant after adjusting for baseline severity of illness.
In conclusion, this very important study by Gaudry et al. highlights the importance of large, international, multi-centred studies that will be able to detect a mortality difference in critically ill patients with AKI. It also emphasises the need of a tailored approach to these patients, in order to identify the ones that will benefit from a delayed initiation of RRT, without compromising their already borderline physiology. The additional awaited trials (6,7) and the future incorporation of sensitive renal biomarkers will enable the timely application of RRT in this group of ICU patients.
Article review submitted by ESICM Journal Review Club member Victoria Metaxa on behalf of the AKI Section.
1. Uchino S, Kellum JA, Bellomo R et al. Acute renal failure in critically ill patients: a multinational, multicenter study. JAMA 2005; 294:813
2. Bagshaw SM, Uchino S, Bellomo R, et al. Timing of renal replacement therapy and clinical outcomes in critically ill patients with severe acute kidney injury. J Crit Care 2009, 24:129
3. Joannidis M, Metnitz PG. Epidemiology and natural history of acute renal failure in the ICU. Crit Care Clin 2005, 21:239
4. Quenot JP, Binquet C, Kara F et al. The epidemiology of septic shock in Franch intensive care units: the prospective multicenter cohort EPISS study. Crti Care 2013; 17:R65
5. Seabra VF, Balk EM, Liangos O, Sosa MA, Cendoroglo M, Jaber BL: Timing of renal replacement therapy initiation in acute renal failure: a meta-analysis. Am J Kidney Dis 2008, 52:272
6. Barber SD, Binquet C, Monchi M et al. Impact on mortality of the timing of renal replacement therapy in patients with severe acute kidney injury (the ELAIN-Trail): study protocol for a randomized controlled trial. Trials 2016; 17:148
7. Wald R, Adhikari NK, Smith OM et al. Comparison of standard and accelerated initiation of renal replacement therapy in acute kidney injury. Kideny Int 2015; 88:897
8. Gaudry S, Hajage D, Schortgen F et al. Comparison of two strategies for initiating renal replacement therapy in the intensive care unit: study protocol for a randomized controlled trial (AKIKI). Trials 2015; 16:170
9. Gaudry S, Hajage D, Schortgen F et al. Initiation Strategies for Renal-Replacement Therapy in the Intensive Care Unit. N Engl J Med 2016; 375:122
10. Brochard L, Abroug F, Brenner M et al. An official ATS/ERS/ESICM/SCCM/SRLF statement: prevention and management of acute renal failure in the ICU patients: an international consensus conference in intensive care medicine. Am J Respir Crit Care Med 2010; 181:1128
11. Liangos O, Rao M, Balakrishnan VS, Pereira BJ, Jaber BL. Relationship of urine output to dialysis initiation and mortality in acute renal failure. Nephron Clin Pract 2005; 99(2): c56