Vinsonneau, 2006

Web site designed and maintained by Chris Cairns © SICS EBM Group 2004

CVVH versus Intermittent Dialysis in MODS

Intermittent haemodialysis can be is as effective as CVVH in the treatment of patients with acute renal failure as part of multi organ dysfunction syndrome.

Level of Evidence: 1^-(RCT with a high risk of bias)

Citation/s: Vinsonneau C, Camus C, Combes A, et al, on behalf of the Hemodiafe Study Group. Continuous venovenous haemodiafiltration versus intermittent haemodialysis for acute renal failure in patients with multiple organ dysfunction syndrome: a multicentre randomised trial. Lancet 2006; 368: 379-85.

Lead author's email: Dr. C Vinsonneau. Christophe.vinsonneau@cch.aphp.fr

Three-part Clinical Question:

Patients: ICU patients with ARF and MODS

Interventions: CVVHF vs intermittent dialysis

Outcomes: Primary: Survival. Secondary: LOS, renal replacement days.

Search Terms: 1. exp hemo/haemofiltration (2268); 2. exp renal dialysis (26105); 3. 1 or 2 (27259); 4. exp Acute Kidney Failure (6208); 5. 3 and 4 (1356); 5. limit 5 to human / adult / RCT (64).

The Study: Non-blinded randomised controlled trial with intention-to-treat.

The Study Patients: 360 patients from 21 medical or multi-disciplinary intensive care units in France, with acute renal failure (Urea >36 mmol/l, Creatinine>310 mmol/l), requiring renal replacement therapy as part of MODS (logistic organ dysfunction score >6). Median SAPS II score 64.5, median LOD score 10. Due to low inclusion rates, further criteria of oliguria added (urine output < 320 ml over 16 hour period despite fluid loading). Exclusion criteria: pregnancy, age <18 years, CRF, ARF of vascular or obstructive aetiology, continuing ACEI therapy, coagulopathy, thrombocytopenia (plt < 30,000) uncontrolled haemorrhage, SAPS II <37, moribund state or survival expectancy < 8 days. All patients randomised to CVVH or intermittent haemdialysis once criteria met.

Control group (N = 176; 175 analysed): CVVH (PRISMA machine, polyacrylonitrile membrane). N=176, analysed 175 (1 withdrawal). Protocol: Flows of 120ml/min or more, dialysate flow 500ml/hr or more, UF flow of 1000ml/hr or more and membranes changed every 48 hours. Regimen tailored to maintain urea <30mmol/l.

Experimental group (N = 184; 184 analysed): Intermittent haemodialysis (machine type not specified, polyacrylonitrile membrane) N= 184, analysed = 184. Protocol: blood flow 250ml/min or more, dialysate flow 500ml/min with Na concentration 150mmol/l, isovolaemic connection and low temperature dialysate for at least 4 hours, administered every 48 hours if anuria or oliguria present. Regimen tailored to achieve a urea reduction ratio of > 65% for each session. Dose of dialysis not recorded.

The Evidence:

Outcome	Time to Outcome	CER	EER	RRR	ARR	NNT
60 day survival		0.324	0.315	3%	0.009	NS
60 day survival	95% Confidence Intervals:			NS	NS	NS
90 day survival		0.284	0.272	4%	0.012	NS
90 day survival	95% Confidence Intervals:			-NS	NS	NS
Adverse events		0.841	0.870	-3%	-0.029	NS
Adverse events	95% Confidence Intervals:			NS	NS	NS

Non-Event Outcomes	Control group	Experimental group	P-value
Length of ICU stay (days)	19 (15-22)	20 (16-23)	0.73
Renal support duration (days)	11 (8-14)	11(8-13)	0.84
Hypothermia	31	10	0.0005

EBM Comments:

1) The methods do allow testing of the hypothesis? No. There are several concerns. Firstly, the study is under powered (480 patients required rather than the 360 recruited). Secondly, doses of CVVH and IHD are not compared. Thirdly, there was a significant switch of patients from CVVH group to IHD group. Lastly, eligibility criteria altered during the recruitment period.

2) Do the statistical tests correctly test the results to allow differentiation of statistically significant results? See above.

3) Are conclusions valid in light of the results? The patient numbers were insufficient to determine whether there is a treatment benefit with either strategy. It does seem that, with strict guidelines, most critically ill patients can tolerate intermittent haemodialysis.

4) Did any results get omitted, and why? No patients lost to follow-up.

5) Did they suggest areas of further research? No.

6) Did they make any recommendations based on the results and were they appropriate?Yes & Yes. ”strict guidelines to improve tolerance and metabolic control” for intermittent haemodialysis.

7) Is this study relevant to my clinical practice? Perhaps: only if you have a choice of delivering both CVVH and IHD.

8) Whet level of evidence does this study represent: 1^-(RCT with a high risk of bias)

9) What grade of recommendation can I make on this result alone? D

10) What grade of recommendation can I make when this study is considered along with other available evidence? D

11) Should I change my practice because of these results? No. Reasons other that mortality benefit, for example local expertise, equipment availability or specific patient requirements, should be used to determine which mode of renal replacement therapy is employed.

12) Should I audit my current practise because of these results? It may be useful to compare your IHD protocol and rates of intolerance with those in the study.

Appraised by: Dr Chris Cairns (Consultant), Dr Rachel Kearns (SHO), Stirling Royal Infirmary, UK, 10 September 2006
Email: rkearns79@hotmail.com

Kill or Update By: September 2011

Reviewed and edited by Ewan Jack & Bruce Taylor.

Citation: EBM Critical Appraisals. Scottish Intensive Care Society EBM Group. 2006 & JICS 2006 Vol7(3);72-73. Cairns CJS, Kearns R. Vinsonneau C, Camus C, Combes A, et al, on behalf of the Hemodiafe Study Group. Continuous venovenous haemodiafiltration versus intermittent haemodialysis for acute renal failure in patients with multiple organ dysfunction syndrome: a multicentre randomised trial. Lancet 2006; 368: 379-85.

Printer friendly view