ARF Trial Network

Intensity of Renal Support in Critically Ill Patients with Acute Kidney Injury

VA/NIH Acute Renal Failure Trial Network, NEJM 2008;359(1):7-20. doi:10.1056/NEJMoa0802639

Clinical Question

  • In critically ill patients with acute kidney injury (AKI) does intensive compared with ‘less intensive’ renal-replacement therapy (RRT) improve mortality?


  • Randomised controlled trial
  • Stratified by site, SOFA cardiovascular score, and oliguria
  • Non-blinded
  • Intention to treat analysis
  • Power calculation: 1164 patients required to detect a decrease in 60 day mortality from 55% to 45%, assuming a 10% loss to follow up
    • 10% probability of false negative result
    • 5% probability of false positive result


  • 27 Veterans Affairs and University Affiliated Medical Centres, North America
  • November 2003 – July 2007


  • Inclusion criteria:
    • adult patients with AKI consistent with diagnosis of acute tubular necrosis, defined as:
      • clinical setting of ischaemic or nephrotoxic injury
      • oliguria (<20ml/hr) for 24hrs or increase in serum Creatinine of >177μmol/l in males and 133μmol/l in females over a period of up to 4 days
    • plan for RRT in critical care unit
    • one non-renal organ failure Sequential Organ Failure Assessment (SOFA) score of 2+ or the presence of sepsis
  • Exclusion:
    • Baseline creatinine >177 μmol/l in males or >133 μmol/l in females
    • prior kidney transplant
    • pregnancy
    • had received >1 session of haemodialysis or low-efficiency dialysis, or >24 hours of continuous RRT before randomisation
  • 1124 patients randomised


  • Intensive haemodialysis – either:
    • Intermittent haemodialysis or low-efficiency dialysis 6 times per week, or
    • Continuous venovenous haemodiafiltration at 35ml/kg/hr (mean delivered dose 35.8ml/kg/hr)


  • Less-intensive haemodialysis – either:
    • Intermittent haemodialysis or low-efficiency dialysis 3 times per week, or
    • Continuous venovenous haemodiafiltration at 20ml/kg/hr (mean delivered dose 22ml/kg/hr)
In both intervention and control groups:
  • Haemodynamic stability defined as SOFA cardiovascular score of 0-2
    • Haemodynamically stable patients underwent intermittent haemodialysis
    • Haemodynamically unstable patients underwent continuous venovenous haemodiafiltration or sustained low-efficiency dialysis according to local practice.
  • Intermittent haemodialysis was prescribed as Kt/Vurea of 1.2 to 1.4 per session
    • Mean delivered Kt/Vurea was 1.32
  • RRT provided for up to 28 days post randomisation, or until recovery of kidney function, discharge from acute care, withdrawal of life-sustaining therapy, or death


  • Primary outcome:
    • death from any cause by day 60 – no significant difference
      • 53.6% in intensive therapy vs. 51.5% with less intensive
  • Secondary outcome:
    • in-hospital death – no significant difference
    • recovery of kidney function by day 28 – no significant difference
      • defined as
        • complete if serum Creatinine no more than 44μmol/l above baseline
        • partial if serum Creatinine remained at 44μmol/l or more above baseline but patient was not dialysis dependent
    • complications
      • Any serious adverse event – no significant difference
      • Hypotension requiring vasopressor support – significantly more in intensive group
      • Hypokalaemia – significantly more in intensive group
      • Hypophosphataemia – significantly more in intensive group

Authors’ Conclusions

  • Intensive renal support in critically ill patients with acute kidney injury did not decrease mortality, improve recovery of kidney function, or reduce the rate of non-renal organ failure as compared with less-intensive therapy


  • Randomised
  • Multi-centre
  • Achieved target dose of RRT in both intervention and control groups
  • Minimal loss to follow-up


  • Timing of initiation of RRT not standardised
  • Excluding patients with chronic kidney disease may limit generalisability
  • Use of intermittent haemodialysis is not standard practice in the UK
  • Multiple modes of RRT were used and therefore difficult to know whether this influenced outcome

The Bottom Line

  • In critically ill patients with renal failure, intensive renal replacement therapy, compared with less intensive renal replacement therapy, did not improve mortality or recovery of renal function. There was significantly more hypotension and electrolyte disturbances in the intensive therapy group. Therefore the standard treatment should be the less intensive strategy.
  • However, observational studies have shown that the target dose of RRT achieved is <70% (SCCM) compared with 110% in this study and therefore you may want to target a higher dose in order to achieve the doses given in the less intensive group.

External Links


Summary author: @davidslessor
Summary date: 22th September 2014
Peer-review editor: @DuncanChambler

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.