Continuous venovenous hemofiltration versus extended daily hemofiltration in patients with septic acute kidney injury: a retrospective cohort study

Sun. Critical Care 2014; 18:R70. doi:10.1186/cc13827

Clinical Question

  • In patients with acute kidney injury (AKI) and severe sepsis or septic shock, does extended daily haemofiltration (EDHF) compared with continuous venovenous haemofiltration (CVVHF) reduce time to renal recovery or mortality at 60 days?


  • Retrospective, observational cohort study
  • Allocation to EDHF or CVVHF at clinician’s discretion
  • Prospective data recording but retrospective analysis
  • No blinding of analysts
  • Statistical control for patho-physiological variables


  • One teaching hospital in China
  • July 2009 to May 2013


  • Inclusion: patients with septic shock or severe sepsis according to standard international criteria and AKI according to RIFLE criteria
  • Exclusion: non-sepsis AKI, failure to meet CVVHF or EDHF criteria as determined by consultant nephrologist, death or cessation of RRT within 72 hours
  • 583 patients admitted for RRT of which 145 met the above criteria:
    • 65 included in CVVHF analysis
    • 80 included in EDHF analysis


  • EDHF
    • Haemofiltration for 8–12 hours daily


    • Haemofiltration continuously with replacement of the filter every 24 hours routinely
    • 39% changed to EDHF or intermittent haemodialysis (IHD) after 72 hours due to a change in clinical state (increased stability)
Comparison of groups, as EDHF vs CVVHF:
  • Patient characteristics:
    • Mean age – 69 years vs 68 years
    • Mean APACHE II score – 31 vs 31
    • Severe sepsis : septic shock ratio – 81:19 vs 68:32 (p=0.06)
    • Mean arterial pressure – 74 mmHg vs 65 mmHg (p=0.037)
    • RIFLE criteria as % risk:injury:failure – 16:39:45 vs 18:38:43 (p=0.936)
  • RRT characteristics:
    • Mean blood flow – 231 ml/min vs 213 ml/min
    • Mean ultrafiltrate – 242 ml/h vs 149 ml/h
    • Mean replacement per hour – 40 ml/kg/h vs 29 ml/kg/h

Both groups were:

  • Buffered with bicarbonate replacement fluid in 100% pre-dilutional mode
  • anticoagulated with unfractionated heparin, low molecular weight heparin, citrate or argatroban within the extracorporeal circuit
  • stopped when urine output was greater than 1000 ml/day and renal biochemistry was improving or normal


  • Primary outcome:
    • Renal recovery at 60 days: EDHF 32.50% vs CVVHF 50.77% (p=0.026)
    • All-cause mortality at 60 days: EDHF 46.25% vs CVVHF 44.62% (p=0.844)
  • Secondary outcome:
    • Median time to renal recovery: EDHF 25.46 days vs CVVHF 17.26 days (p=0.039)
    • Occurrence of adverse events:
      • Bleeding: EDHF 13% vs CVVHF 22% (p=0.145)
      • Hypotension: 26% vs 15% (p=0.112)
      • Hypokalaemia: 39% vs 51% (p=0.147)
      • Hypophosphataemia: 29% vs 55% (p=0.001)

Authors’ Conclusions

  • CVVHF was associated with better rates of renal recovery and earlier recovery, but did not change mortality rates at 60 days. “[CVVHF] is beneficial for haemodynamically unstable patients in the acute stages of septic AKI”


  • Data collected prospectively
  • Univariate and multivariate analysis conducted to confirm raw findings


  • Retrospective observational study: suggests association not causation
  • Single centre and small numbers for observational study
  • Arbitrary decision to perform EDHF or CVVHF depending on clinician’s opinion

The Bottom Line

  • This study does not identify any significant reasons to consider EDHF as more beneficial than the conventional CVVHF.
  • A large randomised, controlled trial will be needed to explore this further if the clinical question remains relevant (roughly 120 patients per group will be needed – at 80% power and alpha level of 0.05 based upon the observed primary outcome values in this study)

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.