Position Paper - 2003 Haemodialysis Prescription: Interpreting the HEMO Study
David Johnson, Kym Bannister, Peter Kerr

A consensus statement produced for and by the Dialysis Nephrology Transplant subcommittee of the KHA and ANZSN.

Background

• Mortality rates for patients treated with haemodialysis in Australia remain high.
  
• A number of observational cohort studies have reported significant improvements in morbidity and mortality with increased delivery of dialysis dose. The NKF/K-DOQI clinical practice guidelines recommend a minimum targeted dose, expressed as single pool Kt/V (spKt/V), of 1.3 per dialysis treatment administered 3 times a week. The CARI guidelines currently recommend a spKt/V target of 1.2. Subsequent observational studies have suggested that the optimal spKt/V is in the range of 1.4-1.8 or even greater.
  
• Observational studies have also suggested that high flux membranes are associated with improved outcomes.

The HEMO Study (N Engl J Med 347:2010-9, 2002)

The Hemodialysis (HEMO) study did not confirm these earlier studies.

The HEMO study was a randomized, open-label controlled trial. Prevalent patients on haemodialysis > 3months were assigned in a 1:1 ratio with a two-by-two factorial design to either a standard dose (eKt/V 1.05, equivalent approximately to spKt/V 1.25 or URR 65%) or high dose goal (eKt/V 1.45, equivalent approximately to spKt/V 1.65 or URR 75%) and to dialysis with either a low flux or high flux membrane (ß2-microglobulin clearance <10 vs >20 ml/min, respectively). HEMO investigators concluded that patients undergoing thrice weekly haemodialysis appear to derive no major benefit from a higher dialysis dose than that recommended by current US (or Australian) guidelines or from the use of a high flux membrane. Subgroup analysis identified possible benefits for high dose goals in females and high flux membranes in patients on dialysis > 3.7 years.

Limitations in Applying HEMO to the Australian Haemodialysis Population

• The study excluded high risk patients (serum albumin <26 g/L or patients > 80 yo) and heavyweight individuals who could not achieve an eKt/V > 1.3 within 4.5 hours (97% weighed < 100 kg).
  
• Approximately 63% of the study population was African-American.
  
• The study recruited prevalent patients (mean time on dialysis 3.7 years) raising the possibilities of survivor and ascertainment biases.
  
• At baseline, the mean eKt/V was 1.43 and 60% of patients used high flux membranes. A carry-over effect of these factors into the trial, thereby reducing a differential effect of dialysis prescription on outcomes, cannot be excluded (mean follow-up was 2.84 years).
  
• The practice of dialyser re-use might have affected the performance of high flux membranes.
  
• Dialysis frequency was restricted to 3 times per week and session duration was restricted to between 2.5 and 4.5 hours. A beneficial effect of higher dialysis doses achieved by longer or more frequent dialysis cannot be excluded.
  
• Similarly, a beneficial effect of higher ß2-microglobulin clearances achievable with haemodialfiltration or sorbent techniques cannot be excluded.
  
• The reported beneficial effects of high dose targets for females and high flux membranes for patients on dialysis > 3.7 years should be interpreted with caution because of the risk of false positive results (type 1 statistical errors) from multiple sub-group analyses.

Recommendations on Haemodialysis Prescription in Australia

• The assessment of haemodialysis adequacy should involve many factors in addition to small solute clearance, including clinical assessment of well being, fluid removal and the impact of treatment on an individual's life.
  
• The HEMO study results should not be taken as a justification for targeting a spKt/V of 1.25 in all patients.
  
• Clinical evidence of uraemia should be considered an indication for increasing dialysis dose, regardless of Kt/V.
  
• Increasing spKt/V > 1.25 or URR > 65% does not appear to be justified in patients who are clinically well.
  
• The HEMO study was not specifically designed to study treatment time. This aspect of dialysis prescription, as well as enhanced frequency haemodialysis, warrant further study. Clinicians should not conclude from the HEMO study that it is safe to reduce treatment time if Kt/V is adjusted to meet a spKt/V of 1.25 or greater.
  
• There is currently no conclusive evidence that high flux membranes offer a major benefit to haemodialysis patients. The results of the European Membrane Permeability Outcome (MPO) study may further clarify their role.
  
• Possible beneficial effects of higher dose or high flux haemodialysis in certain sub-groups (females and long-term haemodialysis patients, respectively) require confirmation by additional studies.