Correspondence

Effect of Dialysis Dose and Membrane Flux in Maintenance Hemodialysis

N Engl J Med 2003; 348:1491-1494April 10, 2003

Article

To the Editor:

We are concerned about the conclusions of the report by Eknoyan et al. on the multimillion-dollar Hemodialysis (HEMO) Study (Dec. 19 issue)1 that support the continued use of the current practice guidelines in the United States, which recommend a value of at least 1.2 for the single-pool Kt/V (where K represents the rate of urea clearance by the dialyzer in milliliters per minute, t the duration in minutes of the treatment session, and V the volume of distribution of urea in the patient in milliliters). There is now overwhelming evidence that such a dose is too low, resulting in morbidity and inadequate rehabilitation due to the persistence of the uremic syndrome. Far better results have been achieved with the use of longer sessions and especially with increases in the frequency of treatment.2,3 A criticism of the study is that the prescribed doses were too similar. This similarity speaks volumes about the upper limits on the dose provided by most dialysis centers in the 1990s, when the study design was formulated.

To achieve full rehabilitation and a better life expectancy for patients undergoing hemodialysis, the duration of the sessions, their frequency, or both must be increased.2,3 Any of these changes will minimize the rate of dialysis-related adverse events and make it easier to control hypertension with the use of the dry-weight method,4 the only method that has proved successful in patients undergoing hemodialysis. Uncontrolled hypertension is a major cause of illness and death in patients undergoing hemodialysis.5 One solution would be to restructure Medicare payments to provide financial incentives favoring more frequent dialysis and home-based hemodialysis, so that benefits can be provided to patients without increased costs.

Belding H. Scribner, M.D.
Christopher R. Blagg, M.D.
University of Washington, Seattle, WA 98195
blaggc@hotmail.com

5 References

1. 1

   Eknoyan G, Beck GJ, Cheung AK, et al. Effect of dialysis dose and membrane flux in maintenance hemodialysis. N Engl J Med 2002;347:2010-2019
   Full Text | Web of Science | Medline

2. 2

   Scribner BH, Oreopoulos DG. The hemodialysis product (HDP): a better index of adequate dialysis than Kt/V. Dial Transplant 2002;31:13-15
   Web of Science

3. 3

   Kooistra MP. Frequent prolonged home haemodialysis: three old concepts, one modern solution. Nephrol Dial Transplant 2003;18:16-18
   CrossRef | Web of Science | Medline

4. 4

   Charra B, Bergstrom J, Scribner BH. Blood pressure control in hemodialysis patients: importance of the lag phenomenon. Am J Kidney Dis 1998;32:720-724
   CrossRef | Web of Science | Medline

5. 5

   Levey AS, Beto JA, Coronado BE, et al. Controlling the epidemic of cardiovascular disease in chronic renal disease: what do we know? What do we need to learn? Where do we go from here? Am J Kidney Dis 1998;32:858-906
   CrossRef | Web of Science

To the Editor:

When compared with an overall first-year mortality rate of 21.7 percent for all patients undergoing hemodialysis in the United States (as reported by the U.S. Renal Data System),1 the first-year death rate of approximately 14 percent in the HEMO Study bespeaks obvious selection bias. It is precisely the patients who were excluded from the study (very old patients and those with multiple severe coexisting conditions) who might have benefited most from an increase in the dose of dialysis delivered. Caution is appropriate, therefore, in extrapolating the findings of the HEMO Study to the overall population of patients undergoing dialysis. A similar hazard has been noted in the drawing of inferences from the purported superiority of survival in Europe to that in the United States, since the treatment rate for end-stage renal disease in Europe is only half that in the United States because of the exclusion of sicker patients, who are most likely to die.2

Eli A. Friedman, M.D.
State University of New York Downstate Medical Center, Brooklyn, NY 11203-2098
elifriedmn@aol.com

2 References

1. 1

   Renal Data System. USRDS 2002 annual data report: atlas of end-stage renal disease in the United States. Bethesda, Md.: National Institute of Diabetes and Digestive and Kidney Diseases, 2002:Table I-22.
   

2. 2

   Friedman EA. International comparisons of survival on dialysis: are they reliable? Hemodial Int 2003;7:59-66
   CrossRef | Medline

To the Editor:

In his editorial discussing the findings of the HEMO Study, Himmelfarb1 warns against the reduction of the duration of treatment sessions for patients in whom a high value for Kt/V is achieved. The use of Kt/V to characterize the efficiency of treatment inevitably favors the removal of small, fast-diffusing molecules such as urea, and the use of high flow rates for blood and dialysis fluid leads to a reduction in the duration of treatment — an approach favored by patients that also provides a cost-effective method of using the existing infrastructure in an era of financial constraint. However, its clinical application may preclude the achievement of dry weight and optimal control of blood pressure.2 In the short term, provided that a constant dose of dialysis is maintained, a reduction in the duration of treatment does not appear to be detrimental to the patient.3

The highest long-term survival rates among patients undergoing dialysis treatment three times weekly have been reported by groups that have used high doses and long treatment times.4,5 In one such group, in Tassin, France, the patients are also subject to strict control of sodium intake. This control may have a larger role in the outcome than the actual duration of treatment.

Nicholas A. Hoenich, Ph.D.
University of Newcastle, Newcastle upon Tyne NE2 4HH, United Kingdom
nicholas.hoenich@ncl.ac.uk

5 References

1. 1

   Himmelfarb J. Success and challenge in dialysis therapy. N Engl J Med 2002;347:2068-2070
   Full Text | Web of Science | Medline

2. 2

   Katzarski KS, Charra B, Luik AJ, et al. Fluid state and blood pressure control in patients treated with long and short haemodialysis. Nephrol Dial Transplant 1999;14:369-375
   CrossRef | Web of Science | Medline

3. 3

   Hartley GH, Goodship TH, Hoenich NA, et al. Is decreased treatment time in hemodialysis patients harmful if solute clearance is maintained? Int J Artif Organs 2002;25:844-851
   Web of Science | Medline

4. 4

   Covic A, Goldsmith DJ, Venning MC, Ackrill P. Long-hours home haemodialysis -- the best renal replacement therapy method? QJM 1999;92:251-260
   CrossRef | Medline

5. 5

   Innes A, Charra B, Burden RP, Morgan AG, Laurent G. The effect of long, slow haemodialysis on patient survival. Nephrol Dial Transplant 1999;14:919-922
   CrossRef | Web of Science | Medline

To the Editor:

The HEMO Study failed to demonstrate any major effect of the dose of dialysis and the use of a high-flux membrane on the survival of patients undergoing hemodialysis. These disappointing results are not wholly unexpected1 and might be partially attributable to the demographic and dialytic characteristics of the participants.

First, the study sample was not representative of the population of patients undergoing hemodialysis in the United States.2 In fact, participants were younger than the average patient (mean [±SD] age, 57.6±14.0 years), were less malnourished (a serum albumin concentration of <2.6 mg per deciliter was a criterion for exclusion), and had been receiving dialysis for a relatively long period at base line (mean, 3.7 years).

Second, at base line, high-flux membranes were used in 60 percent of participants, and the mean equilibrated Kt/V value was 1.43±0.21. Hence, it is likely that after randomization a carryover effect occurred, with some of the patients previously treated with high-flux membranes being randomly assigned to the low-flux group and some of those with high Kt/V values at base line being randomly assigned to the standard-dose group. In addition, the practice of reuse of dialyzers might have affected the performance of high-flux membranes, thus interfering with the comparison between high-flux membranes and low-flux membranes.

Third, participants in the HEMO Study were patients who were already undergoing hemodialysis. Thus, their history of dialysis before randomization and selection of the patients with longer survival might have affected the results. To rule out any effect from previous treatment schedules, my colleagues and I included in the Membrane Permeability Outcome study3 only patients who were just beginning to undergo dialysis; in this European study, we aim to evaluate the effect of flux on outcome.

Francesco Locatelli, M.D.
Ospedale A. Manzoni, 23900 Lecco, Italy
nefrologia@ospedale.lecco.it

3 References

1. 1

   Locatelli F, Marcelli D, Conte F, Limido A, Malberti F, Spotti D. Comparison of mortality in ESRD patients on convective and diffusive extracorporeal treatments. Kidney Int 1999;55:286-293
   CrossRef | Web of Science | Medline

2. 2

   Renal Data System. USRDS 2002 annual data report: atlas of end-stage renal disease in the United States. Bethesda, Md.: National Institute of Diabetes and Digestive and Kidney Diseases, 2002.
   

3. 3

   Locatelli F, Hannedouche T, Jacobson S, et al. The effect of membrane permeability on ESRD: design of a prospective, randomised, multicentre trial. J Nephrol 1999;12:85-88
   Web of Science | Medline

Author/Editor Response

The HEMO Study was designed to determine whether an increased dose of dialysis or the use of high-flux membranes reduces morbidity and mortality among patients undergoing hemodialysis thrice weekly in accordance with current practice in the United States, which usually includes the reuse of dialyzers.1,2 Thus, as Drs. Scribner and Blagg and Dr. Hoenich note and as we point out in our report, our results do not rule out benefits of more intensive therapies such as daily treatment (or six times per week) or very long dialysis sessions (more than six hours each). In fact, the absence of substantial effects of the interventions used in the HEMO Study supports the importance of future trials of such intensive therapies, which are fundamentally distinct from those tested in our study.

We disagree with the contention of Dr. Friedman and Dr. Locatelli that selection bias is likely to have prevented the detection of beneficial effects. The randomized design of our study ensured that the treatment groups were balanced with respect to base-line risk, which allowed us to avoid the difficulties of interpretation that, as Friedman notes, arise in comparisons of outcomes between populations in different countries with different risk profiles. The age limit of 80 years and higher enrollment of black patients from urban centers did result in a death rate in our study that was moderately lower than that in the general population of patients undergoing dialysis in the United States. However, for patients younger than 80 years of age with the racial distribution of our study population, the projected death rate of approximately 18 percent from the U.S. Renal Data System3 is similar to the observed death rate of 16.6 percent in our study. Thus, the investigators in our study enrolled patients with a representative risk profile, after accounting for these two factors.

Moreover, the relative risk of death among nonblack patients in the high-dose group as compared with those in the standard-dose group (1.11 [95 percent confidence interval, 0.89 to 1.37]) and the relative risk of death among nonblack patients in the high-flux group as compared with those in the low-flux group (1.04 [95 percent confidence interval, 0.84 to 1.26]) exceeded 1.00, suggesting that increased enrollment of nonblack patients would not have increased the likelihood that benefits would be detected. Subgroup analyses did not suggest benefits of the dose or flux interventions among elderly or other high-risk subgroups.

Finally, our data do not support the contention that a benefit of treatment with high-flux membranes could have been detected had we restricted enrollment to patients who had not previously undergone dialysis. Rather, there was some suggestion of a greater benefit of the use of high-flux membranes among patients who had had more years of dialysis before the study began. Effects of flux could be obscured by residual kidney function in a study of patients with a new need for dialysis. The possibility that treatment effects may be reduced early in the follow-up period because of emerging effects of preexisting conditions or treatments applies to both patients who have previously undergone dialysis and those who have not, and it was incorporated into the study's calculations of statistical power.2

Tom Greene, Ph.D.
Cleveland Clinic Foundation, Cleveland, OH 44195

Alfred K. Cheung, M.D.
University of Utah, Salt Lake City, UT 84112

Garabed Eknoyan, M.D.
Baylor College of Medicine, Houston, TX 77030

for the Hemodialysis (HEMO) Study Group

3 References

1. 1

   Eknoyan G, Levey AS, Beck GJ, et al. The Hemodialysis (HEMO) Study: rationale for selection of interventions. Semin Dial 1996;9:24-33
   CrossRef | Web of Science

2. 2

   Greene T, Beck GJ, Gassman JJ, et al. Design and statistical issues of the Hemodialysis (HEMO) Study. Control Clin Trials 2000;21:502-525
   CrossRef | Medline

3. 3

   Renal Data System. USRDS 2000 annual data report: atlas of end-stage renal disease in the United States. Bethesda, Md.: National Institute of Diabetes and Digestive and Kidney Diseases, 2000:370-1, 605.
   

Author/Editor Response

The HEMO Study demonstrated no major benefit in the primary outcome (mortality from any cause) or in major secondary outcomes, either with a higher dialysis dose or with the use of a high-flux dialysis membrane. Dr. Hoenich questions the recommendation in my accompanying editorial that the results of the HEMO Study not be used to systematically reduce the duration of dialysis treatments.

As Dr. Hoenich indicates, there are substantial data from observational studies suggesting a clinical benefit of longer treatment sessions with thrice-weekly maintenance hemodialysis. Longer dialysis sessions may allow optimal elimination of tissue edema and optimal blood-pressure control while minimizing the risk of hypotensive events during dialysis.1 A large observational study in a national random sample of patients demonstrated an inverse association between mortality and the duration of dialysis treatment.2 The longest duration of survival among patients undergoing thrice-weekly dialysis has been reported among patients with long-term treatment sessions.3 Although the restriction on salt intake to which these patients adhere may also have salutary effects, there are no convincing data that salt restriction alone improves the outcome for patients undergoing dialysis. Given the high mortality among patients undergoing maintenance hemodialysis, the data from observational studies suggesting a benefit of longer treatment sessions, and the lack of convincing evidence from the HEMO Study or other randomized clinical trials demonstrating that shorter treatment sessions are safe, prudence continues to dictate that systematic reduction of dialysis treatment times on the basis of the results of the HEMO Study not be considered safe practice.

Jonathan Himmelfarb, M.D.
Maine Medical Center, Portland, ME 04102
himmej@mmc.org

3 References

1. 1

   Katzarski KS, Charra B, Luik AJ, et al. Fluid state and blood pressure control in patients treated with long and short haemodialysis. Nephrol Dial Transplant 1999;14:369-375
   CrossRef | Web of Science | Medline

2. 2

   Held PJ, Levin NW, Bovbjerg RR, Pauly MV, Diamond LH. Mortality and duration of hemodialysis treatment. JAMA 1991;265:871-875
   CrossRef | Web of Science | Medline

3. 3

   Charra B, Calemard E, Ruffet M, et al. Survival as an index of adequacy of dialysis. Kidney Int 1992;41:1286-1291
   CrossRef | Web of Science | Medline

Citing Articles (6)

Citing Articles

1. 1

   DAVID W JOHNSON, CAROL A POLLOCK, IAIN C MACDOUGALL. (2007) Erythropoiesis-stimulating agent hyporesponsiveness (Review Article). Nephrology 12:4, 321-330
   CrossRef

2. 2

   Fatiu Arogundade, Rashad S Barsoum. (2005) Indices for assessment of hemodialysis adequacy: A comparison of different formulae. Hemodialysis International 9:4, 325-331
   CrossRef

3. 3

   Roberto Pecoits-Filho, Lucimary C. Sylvestre, Peter Stenvinkel. (2005) Chronic kidney disease and inflammation in pediatric patients: from bench to playground. Pediatric Nephrology 20:6, 714-720
   CrossRef

4. 4

   Gihad E Nesrallah, Rita S Suri, Amit X Garg, Louise M Moist, Christian Awaraji, Robert M Lindsay, . (2004) The International Quotidian Hemodialysis Registry: Rationale and methods. Hemodialysis International 8:4, 354-359
   CrossRef

5. 5

   Gihad E. Nesrallah, Louise M. Moist, Christian Awaraji, Robert M. Lindsay. (2004) DAILY HEMODIALYSIS-SELECTED TOPICS: An International Registry to Compare Quotidian Dialysis Regimens with Conventional Thrice-Weekly Hemodialysis: Why, How, and Potential Pitfalls. Seminars in Dialysis 17:2, 131-135
   CrossRef

6. 6

   Josephine P. Briggs. (2004) DAILY HEMODIALYSIS-SELECTED TOPICS: Evidence-Based Medicine in the Dialysis Unit: A Few Lessons from the USRDS and the NCDS and HEMO Trials. Seminars in Dialysis 17:2, 136-141
   CrossRef