2007 JASN IMPACT FACTOR 7.111	HOME   AUTHOR INFO   EDITORIAL BOARD   SUBSCRIBE   FEEDBACK   ALERTS   HELP
advanced	CURRENT ISSUE	ARCHIVES	JASN Express	ONLINE SUBMISSION

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Keshaviah, P. Articles by Thorpe, K. E. Search for Related Content PubMed PubMed Citation Articles by Keshaviah, P. Articles by Thorpe, K. E.

Survival Comparison between Hemodialysis and Peritoneal Dialysis Based on Matched Doses of Delivered Therapy

Prakash Keshaviah^*, Allan J. Collins, Jennie Z. Ma, David N. Churchill and Kevin E. Thorpe

*Department of Physiology, Himalayan Institute Hospital Trust, Dehradun, U.P., India; Minneapolis Medical Research Foundation and University of Minnesota, Minneapolis, Minnesota; University of Tennessee, Memphis, Tennessee; St. Joseph’s Hospital, McMaster University, Hamilton, Ontario, Canada.

Correspondence to Dr. Allan Collins, Minneapolis Medical Research Foundation and University of Minnesota, 914 South 8th Street, Suite D206, Minneapolis, MN 55404. Phone: 612-347-5811; Fax: 612-347-5878; E-mail: acollins{at}nephrology.org

Abstract

ABSTRACT. Several studies have recently confirmed that hemodialysis (HD) and continuous ambulatory peritoneal dialysis (CAPD) survival is highly associated with delivered therapy Kt/V_urea. A direct comparison of equivalently dosed CAPD and HD has not previously been performed. A total of 968 incident HD patients at the Regional Kidney Disease Program from 1987 to June 1995 were studied, and these results were compared with those of the Canadian-United States prospective trial (CANUSA) consisting of 680 incident CAPD patients from September 1990 to December 31, 1992, with follow-up through December 31, 1993. All patients had quantitation of urea nitrogen for a total delivered dialysis session. On HD, in vivo, 2-pool, pre- and post-blood urea nitrogen kinetic modeling was performed with residual renal function determined every 6 mo. Patients were characterized by age, gender, race, renal diagnosis, and comorbid conditions. A Cox proportional hazards model was used to evaluate the effect of the individual comorbid conditions and the effect of dialysis therapy in the time-dependent method. The mean total Kt/V, both residual renal function and dialytic therapy in the HD patients, was 1.59. The CANUSA-delivered weekly Kt/V was 2.38 at the beginning of the baseline period and 1.99 after 24 mo of follow-up. When the peak concentration hypothesis was used, a Kt/V of 1.59 on HD was equivalent to a weekly CAPD dose of 2.1 to 2.2. A 1-unit increase in Kt/V was associated with 7% lower risk of death on HD and with a similar 8% lower risk of death while on CAPD. Patients with diabetes aged 46 to 60 yr had virtually identical 2-yr survival estimates on HD (83 to 90%), compared with CAPD (83 to 89%), with Kt/V ranges from 0.84 to 1.70 in HD and from 1.6 to 2.2 weekly Kt/V on peritoneal dialysis. Comparisons between HD and CAPD in older patients with diabetes yielded comparable results. Patient survival is highly influenced by delivered dialysis in both HD and peritoneal dialysis. Carefully matching of the therapies with delivered Kt/V demonstrates little differences in the survival outcome of HD and peritoneal dialysis patients, in contrast to some previous reports.

Several studies (1–9) have confirmed that for both hemodialysis (HD) and continuous ambulatory peritoneal dialysis (CAPD), patient survival improves with increased doses of delivered therapy. For HD, the Kt/V_urea index and for peritoneal dialysis (PD), the Kt/V_urea indices, as well as the weekly creatinine clearance per 1.73 m², are significant predictors of patient survival. Yet most of the survival comparisons between these two modalities have not been controlled for the delivered dose of dialysis. These studies have controlled for other significant predictors of survival (e.g., age, diabetic status, history of cardiovascular disease [CVD]) but have ignored the solute removal in the populations being compared. Such survival comparisons must therefore be interpreted cautiously because the differences in survival elicited by the analyses may not be related to the therapy modality per se but may be a consequence of differences in the delivered dose of dialysis.

HD is an intermittent therapy with peaks and valleys of small solute concentrations over the duration of the week; concentrations fall steeply during the period of dialysis, then rise gradually over the interdialytic period as a result of solute generation within the body. In the case of CAPD, such wide swings of solute concentration are not typically observed because of the continuous nature of the therapy. The solute concentration profiles are relatively flat over the duration of the week, with solute removal being matched by solute generation. The peak concentration hypothesis (10) has provided a basis for matching an intermittent therapy such as HD with a continuous therapy such as CAPD by determining the dose of dialysis for the continuous therapy that results in a steady-state solute concentration that matches the peak solute concentration of the intermittent therapy rather than the time-averaged concentration.

We compared the survival of patients on HD and CAPD, controlling for dose of delivered dialysis; the doses of dialysis were matched on HD and CAPD by means of the peak concentration approach. We matched the Kt/V_urea per HD session with the weekly Kt/V_urea for CAPD. For each dialysis modality, we selected a relatively large number of dialysis patients who had actual measurements of delivered therapy (inclusive of residual renal function) and that had achieved patient survival results that were documented to be better than those seen in North America.

Materials and Methods

For this matched-dose comparison of patient survival, we selected the database of the Regional Kidney Disease Program (RKDP) in Minneapolis, Minnesota, for HD and the Canadian-United States prospective trial (CANUSA) database (3) for CAPD. In the RKDP database, we included 968 incident HD patients who had entered the HD dialysis program between January 1, 1987, and December 31, 1994, and who were followed until June 30, 1995. Only patients with measurements of Kt/V_urea on dialysis, residual renal function, and serum albumin were included in the analysis. Delivered single-pool (Kt/V)_urea was quantitated by the Daugirdas equation (11) with laboratory-measured values pre- and post-blood urea nitrogen. Urine was collected over the interdialytic period for quantifying the residual urea clearance and its contribution to total delivered Kt/V_urea. The quantity of Kt/V_urea contributed by the residual renal function in HD was determined by the Gotch method (11).

The CANUSA database included a prospective cohort of 680 patients commencing continuous PD in 14 centers in Canada and the United States between September 1, 1990, and December 31, 1992, with follow-up until December 31, 1993. In these patients, the dose of dialysis was quantitated by 24-h collection of effluent dialysate and urine, with representative aliquots from these collections assayed for urea nitrogen to determine the total delivered Kt/V_urea. To compare the CAPD and HD survival at each level of Kt/V_urea, estimated survival curves were developed that were based on specific patient profiles.

Each database of HD (RKDP) and PD (CANUSA) patients was analyzed separately. The patient and therapy variables were analyzed in a Cox regression model in a stepwise manner. The independent variables included age, gender, race, comorbidity, and the 6-mo data on mean Kt/V_urea from both the dialysis therapy and residual renal function. A time-dependent analysis was used to define the effect of dialysis therapy and residual renal function on mortality risk.

The Cox proportional hazards model (12) was used to compare patient survival in the two therapy modalities. In this comparison, age, diabetes mellitus as primary renal disease, history of CVD, serum albumin, and Kt/V were the selected covariates. Age was stratified into 3 categorical groups: 45 yr or less, 46 to 60 yr, and 61 yr or more. Kt/V and serum albumin were treated as continuous variables. In the survival predictions, the population average serum albumin was used. The doses of delivered dialysis, measured as total Kt/V_urea, were matched on the basis of the peak concentration hypothesis.

In the RKDP population, history of CVD included atherosclerotic heart disease, defined as a history of angina, previous acute myocardial infarction detected by electrocardiogram or enzymes, history of coronary artery bypass operation, percutaneous transluminal coronary angioplasty, or angiographic evidence; congestive heart failure; or peripheral vascular disease, defined as bruits in vessels, steal syndromes from vascular accesses, aneurysms, poor extremity pulses, and amputations. The CANUSA definition of CVD was less broad and only included patients with one of the following: angina, myocardial infarctions, New York Heart Classifications grades III to IV congestive heart failure, and amputation as a result of vascular disease. Results of the analyses were compared both with and without peripheral vascular disease to evaluate the effect of this condition.

Results

The two populations being compared are described in Table 1. The mean age of the HD population is approximately 6 yr older, with a higher percentage of patients with diabetes as primary renal disease. The much larger percentage of patients in the RKDP population with CVD may be related to the broader definition of CVD in the RKDP database. The mean serum albumin is comparable in the two populations. The mean total Kt/V per session is 1.59 in the RKDP HD population, and in the CANUSA group, it varies between a weekly value of 2.38 at baseline and 1.99 after 24 mo of follow-up.

View this table: [in this window] [in a new window]   Table 8. Predicted 2-yr percent survival (mean ± SEM) versus Kt/V: Age 45 yr, diabetes mellitus not primary disease, cardiovascular disease absent

The Cox proportional hazards model clearly indicates that age, history of CVD, serum albumin, and Kt/V are significant predictors of patient survival for both HD and CAPD. Although the magnitude of the risks is different, the direction and significance of the risks are the same. A more direct comparison of risks was not possible because of the observational nature of the study. Differences, if any, may reflect severity of disease or acceptance rates. After adjustment for other risk factors, this study shows that diabetes mellitus as primary renal disease does not appear to pose any additional risk in either HD or CAPD, particularly when albumin and dialysis therapy are taken into consideration. We confined our analysis to patients with diabetes as a primary disease, excluding those with other etiologies of renal disease.

The 2-yr survival comparisons demonstrate that, although Kt/V is a significant predictor of patient survival, once Kt/V is controlled for, 2-yr patient survivals are similar in the 2 modalities being compared, independent of patient age, diabetic status, history of CVD, and the absolute differences in the value of the other risks. This appears to be the first time these observations have been demonstrated. Previous analyses by Bloembergen et al. (7) have suggested a 19% higher risk of death in CAPD relative to HD, this difference being even more pronounced in patients with diabetes, who had a 38% higher risk of death on CAPD. There are three major differences between our analysis and that of Bloembergen et al. We analyzed an incident dialysis population, whereas Bloembergen et al. analyzed point-prevalent patients with varying time on dialysis ("vintage"). Because dialysis dose and residual renal function vary over vintage, this issue should be more carefully considered. The dose of dialysis was not quantitated or controlled for in the Bloembergen et al. analysis, nor did the analysis adjust for comorbidity.

To resolve the results of the Bloembergen et al. (7) analysis with our results, one would have to account for the influence of vintage, dose, and comorbidity in the Bloembergen et al. analysis. Fenton et al. (14) have shown that for an incident population of 11,970 patients with end-stage renal disease from the Canadian Organ Replacement Register, the mortality rate ratio for CAPD-continuous cycling PD relative to HD was 0.73, based on a Poisson regression. The lower mortality on PD relative to HD was applicable to all subgroups defined by age and the presence of diabetes. However, when the analysis was repeated using prevalent patients, CAPD no longer had a lower mortality risk ratio (rate ratio = 1.01). These same observations were also confirmed by Collins et al. (15) in a US population of HD and PD patients.

Regarding dose, one would have to postulate that if the typical CAPD patient is being less adequately dialyzed then the typical HD patient, one might expect a higher risk of death on CAPD, particularly if vintage on dialysis therapy is a surrogate for residual renal function. This postulate may not be completely unreasonable. Studies on the adequacy of CAPD are relatively recent—less than a decade old—whereas HD adequacy has been a subject of interest for almost three decades. This is not surprising when one considers that CAPD is a much newer therapy than HD. Clinicians have therefore not paid as much attention to CAPD therapy prescription practices as they have to HD prescriptions.

In addition, surveys (16) indicate that the majority of CAPD patients receive the standard prescription of 2-L exchanges 4 times a day without regard to body size or residual renal function. It has been calculated (17) that this standard regimen of CAPD cannot provide an adequate dose of dialysis in even average-sized patients without residual renal function. It has been recommended (18) that the volume of the CAPD exchange be tailored to the size of the patient, with average-sized patients using 2.5-L exchanges and larger patients using 3-L exchanges. If the patients in the Bloembergen et al. (7) population underwent dialysis with the standard 2-L, 4-times-a-day exchange regimen and had no significant residual renal function, it is likely that these patients were underdialyzed and therefore had a higher risk of death. The 19% higher risk of death in the Bloembergen et al. analysis may therefore have been a consequence not only of uncontrolled vintage but also of the dose of dialysis delivered rather than the modality.

The 2-yr survival data in younger (<45 yr), healthier (no diabetes mellitus, no CVD) patients indicate a relatively high survival rate (approximately 90%) with a consequently lesser effect of Kt/V on survival. This is probably a testimonial to the resilience of young, healthy patients. However, this should not be misinterpreted as suggesting that one should deliver less adequate dialysis in these patients.

The peak concentration hypothesis is approximately a decade old and is, by its very nature, not easy to test directly. The results of this analysis provide an indirect verification of this hypothesis and indicate that it may be a reasonable basis for comparing a continuous therapy with an intermittent one.

To our knowledge, this comparative analysis of HD and CAPD survival is the first to control for the delivered dose of dialysis. The analysis shows clearly that for both HD and CAPD, dose of dialysis as measured by the Kt/V index is a significant predictor of survival. Furthermore, our study demonstrates that when dose of dialysis is matched, HD and CAPD provide comparable 2-yr survival rates, independent of age, diabetic status, and history of CVD. These results are an indirect confirmation of the peak concentration hypothesis.

References

1. Collins AJ, Ma JZ, Umen A, Keshaviah P: Urea index and other predictors of hemodialysis patient survival. Am J Kidney Dis 23: 272–282, 1994[Medline]
2. Hakim RM, Breyer J, Ismail N, Schulman G: Effects of dose of dialysis on morbidity and mortality. Am J Kidney Dis 23: 661–669, 1994[Medline]
3. Owen WF, Lew NL, Liu Y, Lowrie EG, Lazarus JM: The urea reduction ratio and serum albumin concentration as predictors of mortality in patients undergoing hemodialysis. N Engl J Med 329: 1001–1006, 1993[Abstract/Free Full Text]
4. Parker T, Husni L, Huang W, Lew N, Lowrie EG, Dallas Nephrology Associates: Survival of hemodialysis patients in the United States is improved with a greater quantity of dialysis. Am J Kidney Dis 23: 670–680, 1994[Medline]
5. Churchill DN, Taylor DW, Keshaviah PR: Adequacy of dialysis and nutrition in continuous peritoneal dialysis: Association with clinical outcomes. J Am Soc Nephrol 7: 198–207, 1996[Abstract]
6. Teschan BP, Schleifer CR, Brown J: Adequacy of continuous ambulatory peritoneal dialysis: Morbidity and mortality in chronic peritoneal dialysis. Am J Kidney Dis 24: 990–1001, 1994[Medline]
7. Bloembergen WE, Port FK, Mauger EA, Wolfe RA: A comparison of mortality between patients treated with hemodialysis and peritoneal dialysis. J Am Soc Nephrol 6: 177–183, 1995[Abstract]
8. Held PJ, Port FK, Turenne MN, Gaylin DS, Hamburger RJ, Wolfe RA: Continuous ambulatory peritoneal dialysis and hemodialysis: Comparison of patient mortality with adjustment for comorbid conditions. Kidney Int 45: 1163–1169, 1994[Medline]
9. Maiorca R, Vonesh EF, Cavilli P, De Vecchi A, Giangrande A, La Grecca G, Scarpioni L, Bragantini L, Cancarini G, Cantaluppi A, Castelnovo C, Castiglioni A, Poisetti P, Viglino G: A multicenter selection-adjusted comparison of patient and technique survivals on CAPD and hemodialysis. Perit Dial Int 11: 118–127, 1991
10. Keshaviah PR, Nolph KD, Van Stone JC: The peak concentration hypothesis: A urea kinetic approach to comparing the adequacy of continuous peritoneal dialysis (CAPD) and hemodialysis. Perit Dial Int 9: 257–260, 1989
11. Gotch FA: Kinetic modeling in hemodialysis.In: Clinical Dialysis, 2nd Ed, edited by Nissenson A, Fine R, Gentile D, Norwalk, CT, Appleton & Lange, 1990,pp 118–146
12. Cox D: Regression models and life-tables. J R Stat Soc 34: 187–201, 1972
13. National Kidney Foundation-Dialysis Outcomes Quality Initiative: Clinical practice guidelines for peritoneal dialysis adequacy. Am J Kidney Dis 30 [Suppl 2]: S70–S73, 1997
14. Fenton SSA, Schaubel DE, Desmeules M, Morrison HI, Mao Y, Copleston P, Jeffery JR, Kjellstrand CM: Hemodialysis versus peritoneal dialysis: A comparison of adjusted mortality rates. Am J Kidney Dis 30: 334–342, 1997[Medline]
15. Collins AJ, Hao E, Xia H, Ebben JP, Everson SE, Constantini EG, Ma JZ: Mortality risks of peritoneal dialysis and hemodialysis. Am J Kidney Dis 34: 1065–1074, 1999[Medline]
16. US Renal Data System: The USRDS dialysis morbidity and mortality study (wave 2).In: US Renal Data Systems 1997 Annual Data Report,edited by National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, 1997,pp 49–67
17. Nolph KD, Jensen RA, Khanna R, Twardowski ZJ: Weight limitations for weekly urea clearances using various exchange volumes in continuous ambulatory peritoneal dialysis. Perit Dial Int 14: 261–264, 1994
18. Keshaviah P, Emerson PF, Vonesh EF, Brandes JC: Relationship between body size, fill volume, and mass transfer area coefficient in peritoneal dialysis. J Am Soc Nephrol 4: 1820–1826, 1994[Abstract]

This article has been cited by other articles:

				J. Malyszko, J.S. Malyszko, P. Kozminski, K. Pawlak, and M. Mysliwiec Adipokines, Linking Adipocytes and Vascular Function in Hemodialyzed Patients, May Also Be Possibly Related to CD146, a Novel Adhesion Molecule Clinical and Applied Thrombosis/Hemostasis, July 1, 2008; 14(3): 338 - 345. [Abstract] [PDF]

				G. Viglino, L. Neri, S. Alloatti, G. Cabiddu, R. Cocchi, A. Limido, G. Marinangeli, R. Russo, U. Teatini, and F. P. Schena Analysis of the factors conditioning the diffusion of peritoneal dialysis in Italy Nephrol. Dial. Transplant., December 1, 2007; 22(12): 3601 - 3605. [Abstract] [Full Text] [PDF]

				G. M. Chertow A 43-Year-Old Woman With Chronic Renal Insufficiency JAMA, March 10, 2004; 291(10): 1252 - 1259. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Keshaviah, P. Articles by Thorpe, K. E. Search for Related Content PubMed PubMed Citation Articles by Keshaviah, P. Articles by Thorpe, K. E.