Dose of dialysis: Key lessons from major observational studies and clinical trials

Abstract

Analyses based on the National Cooperative Dialysis Study (NCDS) provided the impetus for routine quantification of delivered dialysis dose in hemodialysis practice throughout the world, by suggesting minimum targets for small solute (urea) clearance. Morbidity and mortality in dialysis populations remain high despite many technological advances in dialysis delivery. A number of observational studies reported association between higher dose of dialysis as measured by Kt/V urea or urea reduction ratio with lower mortality risk. During the 1990s, a steady increase in dialysis dose and a modest reduction in mortality on dialysis were observed. However, observational studies only reveal associations and are limited by selection bias and confounding. The Kidney Disease Outcomes Quality Initiative guidelines on dialysis adequacy are based on results of observational studies and expert opinion. Since the NCDS, the HEMO Study was the first major randomized clinical trial designed to study the effect of dose of dialysis and dialyzer flux on patient outcomes. Despite adequate separation of dose and flux, however, results of the trial did not prove a beneficial effect of higher dose. The Dialysis Outcomes and Practice Patterns Study (DOPPS), in a major international effort designed to examine the effect of practice patterns on outcomes, has made significant contributions to the topic of dialysis dose. The following review critically examines data from observational studies, including the DOPPS, and from the HEMO Study, emphasizing important lessons from both, and discusses future paradigms for achieving dialysis adequacy to improve patient outcomes. Analyses based on the National Cooperative Dialysis Study (NCDS) provided the impetus for routine quantification of delivered dialysis dose in hemodialysis practice throughout the world, by suggesting minimum targets for small solute (urea) clearance. Morbidity and mortality in dialysis populations remain high despite many technological advances in dialysis delivery. A number of observational studies reported association between higher dose of dialysis as measured by Kt/V urea or urea reduction ratio with lower mortality risk. During the 1990s, a steady increase in dialysis dose and a modest reduction in mortality on dialysis were observed. However, observational studies only reveal associations and are limited by selection bias and confounding. The Kidney Disease Outcomes Quality Initiative guidelines on dialysis adequacy are based on results of observational studies and expert opinion. Since the NCDS, the HEMO Study was the first major randomized clinical trial designed to study the effect of dose of dialysis and dialyzer flux on patient outcomes. Despite adequate separation of dose and flux, however, results of the trial did not prove a beneficial effect of higher dose. The Dialysis Outcomes and Practice Patterns Study (DOPPS), in a major international effort designed to examine the effect of practice patterns on outcomes, has made significant contributions to the topic of dialysis dose. The following review critically examines data from observational studies, including the DOPPS, and from the HEMO Study, emphasizing important lessons from both, and discusses future paradigms for achieving dialysis adequacy to improve patient outcomes. THE CURRENT PARADIGM of thrice-weekly hemodialysis as renal replacement therapy is but a poor imitation of the native kidney’s continuous high-quality blood purification system, along with multiple other critical facets of native renal function. Despite impressive advances in dialysis technology, overall prevalent mortality rates for dialysis patients in the United States have improved only marginally (10%) since 1988, according to the 2003 annual data report of the United States Renal Data System.1U.S. Renal Data SystemUSRDS 2003 Annual Data Report: Atlas of End-Stage Renal Disease in the United States. National Institutes of Health. National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD2003Google Scholar Urea kinetic modeling introduced a measure of objectivity into the assessment of dialysis delivery in the 1980s following the landmark National Cooperative Dialysis Study (NCDS)2Gotch F.A. Sargent J.A. A mechanistic analysis of the National Cooperative Dialysis Study (NCDS).Kidney Int. 1985; 28: 526-534Crossref PubMed Scopus (1046) Google Scholar that randomized 160 patients into 4 treatment arms, with dose distributed over a wide range of single pool Kt/V (spKt/V) for urea. Subsequent analysis of the NCDS established the minimum targets for dialysis dose delivery (ie, spKt/V >1.0). This minimal target was 25% higher than the so-called “break point” at spKt/V = 0.8, at or below which the “relative probability of failure” (composite endpoint of death, hospitalization, and de novo uremic symptoms) of therapy was seen to be substantially higher. However, it was below the extrapolated spKt/V value of 1.3 per dialysis session, above which no further benefit was expected.3Gotch F.A. Levin N.W. Port F.K. et al.Clinical outcome relative to the dose of dialysis is not what you think The fallacy of the mean.Am J Kidney Dis. 1997; 30: 1-15Abstract Full Text PDF PubMed Scopus (66) Google Scholar The 1.3 limit came from the extrapolation of urea reduction ratio (URR) versus morbidity, which gave a more linear set of data points. Over the next 2 decades, data emerging from a number of observational studies showed that a higher dose of hemodialysis was associated with a lower mortality risk.4Port F.K. Ashby V.B. Dhingra R.K. et al.Dialysis dose and body mass index are strongly associated with survival in hemodialysis patients.J Am Soc Nephrol. 2002; 13: 1061-1066PubMed Google Scholar, 5Wolfe R.A. Ashby V.B. Daugirdas J.T. et al.Body size, dose of hemodialysis, and mortality.Am J Kidney Dis. 2000; 35: 80-88Abstract Full Text Full Text PDF PubMed Scopus (196) Google Scholar, 6Eknoyan G. Beck G.J. Cheung A.K. et al.The Hemodialysis (HEMO) Study GroupEffect of dialysis dose and membrane flux in maintenance hemodialysis.N Engl J Med. 2002; 347: 2010-2019Crossref PubMed Scopus (1513) Google Scholar, 7Depner T. Daugirdas J. Greene T. et al.Dialysis dose and the effect of gender and body size on outcome in the HEMO study.Kidney Int. 2004; 65: 1386-1394Crossref PubMed Scopus (146) Google Scholar, 8Owen Jr, W.F. Chertow G.M. Lazarus J.M. et al.Dose of hemodialysis and survival differences by race and sex.JAMA. 1998; 280: 1764-1768Crossref PubMed Scopus (172) Google Scholar, 9Owen Jr, W.F. Lew N.L. Liu Y. et al.The urea reduction ratio and serum albumin concentration as predictors of mortality in patients undergoing hemodialysis.N Engl J Med. 1993; 329: 1001-1006Crossref PubMed Scopus (1228) Google Scholar, 10Collins A.J. Ma J.Z. Umen A. et al.Urea index and other predictors of hemodialysis patient survival.Am J Kidney Dis. 1994; 23: 272-282PubMed Scopus (291) Google Scholar, 11Hakim R.M. Bryer J. Ismail N. et al.Effects of dose of dialysis on morbidity and mortality.Am J Kidney Dis. 1994; 23: 661-669PubMed Scopus (370) Google Scholar, 12Parker 3rd, T.F. Husni L. Huang W. et al.Survival of hemodialysis patients in the United States is improved with greater quantity of dialysis.Am J Kidney Dis. 1994; 23: 670-680PubMed Scopus (256) Google Scholar, 13Held P.J. Port F.K. Wolfe R.A. et al.The dose of hemodialysis and patient mortality.Kidney Int. 1996; 50: 550-556Crossref PubMed Scopus (493) Google Scholar, 14Li Z. Lew N.L. Lazarus J.M. Lowrie E.G. Comparing the urea reduction ratio and the urea product as outcome-based measures of hemodialysis dose.Am J Kidney Dis. 2000; 35: 598-605Abstract Full Text Full Text PDF PubMed Scopus (43) Google Scholar Since 1990, a trend toward a higher dose of dialysis and use of high-flux dialyzers was reported.15Port F.K. Orzol S.M. Held P.J. et al.Trends in treatment and survival for hemodialysis patients in the United States.Am J Kidney Dis. 1998; 32: S34-S38Abstract Full Text PDF PubMed Google Scholar, 162000 Annual Report ESRD Clinical Performance Measures Project. Department of Health and Human Services, Health Care Financing Administration, Office of Clinical Standards and Quality, Baltimore, MD:2000: 1-50Google Scholar This steady increase in dialysis dose may have been responsible, at least in part, for the reduction in mortality observed during the 1990s.15Port F.K. Orzol S.M. Held P.J. et al.Trends in treatment and survival for hemodialysis patients in the United States.Am J Kidney Dis. 1998; 32: S34-S38Abstract Full Text PDF PubMed Google Scholar, 162000 Annual Report ESRD Clinical Performance Measures Project. Department of Health and Human Services, Health Care Financing Administration, Office of Clinical Standards and Quality, Baltimore, MD:2000: 1-50Google Scholar Hence, Guideline 4 of the Kidney Disease Outcomes Quality Initiative (K/DOQ1): Minimum Delivered Dose of Hemodialysis, states that “the dialysis care team should deliver a spKt/V of at least 1.2 (single pool, variable volume) for both adult and pediatric hemodialysis patients. (spKt/V represents dialyzer Kt/V; when the postdialysis blood sample is drawn as specified by K/DOQI guidelines, immediately after shutting off the blood pump and after taking precautions to avoid errors from access recirculation, spK should equate to the average dialyzer clearance during the treatment.) For those using the URR, the delivered dose should be equivalent to an spKt/V of 1.2 (ie, an average URR of 65%). However, URR can vary substantially as a function of fluid removal.”17National Kidney FoundationK/DOQI Clinical Practice Guidelines for Hemodialysis Adequacy Update 2000.Am J Kidney Dis. 2001; 37: S7-S64Google Scholar The equivalent minimum prescribed dialysis dose in a double pool model would be an eKt/V of 1.05. It has been reported that a low dialysis prescription is a strong predictor of inadequate dialysis dose.18Sehgal A. Snow R.J. Sinder M.E. et al.Barriers to adequate delivery of hemodialysis.Am J Kidney Dis. 1998; 31: 593-601Abstract Full Text Full Text PDF PubMed Scopus (88) Google Scholar Furthermore, based on analysis of data from the HEMO Study, the 90% confidence interval for the spKt/V of 1.3 was 0.10 units and for URR was 4%.17National Kidney FoundationK/DOQI Clinical Practice Guidelines for Hemodialysis Adequacy Update 2000.Am J Kidney Dis. 2001; 37: S7-S64Google Scholar Nephrologists using the K/DOQI guidelines would therefore be advised to target a higher Kt/V or URR, as the case may be, to achieve the minimum target, as indicated earlier. Hence, the K/DOQI hemodialysis adequacy Guideline 5: Prescribed Dose of Hemodialysis states, “to prevent the delivered dose of hemodialysis from falling below the recommended minimum dose, the prescribed dose of hemodialysis should be spKt/V 1.3. A Kt/V of 1.3 corresponds to an average URR of 70%, but the URR corresponding to a spKt/V of 1.3 can vary substantially as a function of ultrafiltration.”17National Kidney FoundationK/DOQI Clinical Practice Guidelines for Hemodialysis Adequacy Update 2000.Am J Kidney Dis. 2001; 37: S7-S64Google Scholar It is recognized that several factors may result in a deficit in the achieved versus prescribed dialysis dose. Guideline 14: Inadequate Delivery of Hemodialysis presents an algorithm for detecting the cause(s) of inadequate delivery of hemodialysis and for implementing corrective action.17National Kidney FoundationK/DOQI Clinical Practice Guidelines for Hemodialysis Adequacy Update 2000.Am J Kidney Dis. 2001; 37: S7-S64Google Scholar The K/DOQI guidelines are based predominantly on evidence from national registry data and expert opinion; these sources suggest that higher dose of dialysis would result in better outcomes, based on a priori reasoning. Because observational studies mainly provide association rather than causality and because some of these4Port F.K. Ashby V.B. Dhingra R.K. et al.Dialysis dose and body mass index are strongly associated with survival in hemodialysis patients.J Am Soc Nephrol. 2002; 13: 1061-1066PubMed Google Scholar, 5Wolfe R.A. Ashby V.B. Daugirdas J.T. et al.Body size, dose of hemodialysis, and mortality.Am J Kidney Dis. 2000; 35: 80-88Abstract Full Text Full Text PDF PubMed Scopus (196) Google Scholar had suggested that a URR of greater than 70% to 75% was associated with further reduction in relative risk of mortality, proof of concept was clearly required by a large randomized clinical trial. Moreover, results of observational studies could be biased in favor of healthier patients having superior outcomes, especially if they were able to receive a higher dose of dialysis or if those less healthy could not receive the higher dose for reasons such as a poorly functioning vascular access. Thus, with equipoise and a firm focus on need to study interventions aimed at reducing the persistently high mortality rate among hemodialysis patients in the United States, the HEMO Study, a randomized clinical trial sponsored by the National Institutes of Health, was undertaken.19Depner T.A. How will the results of the HEMO Study impact dialysis practice?.Semin Dial. 2003; 16: 8-11Crossref PubMed Scopus (16) Google Scholar The HEMO Study randomized 1,846 adult patients (18–80 years of age) from 15 centers (72 dialysis units) to standard dose versus high dose and to low-flux versus high-flux dialyzers, using a 2 × 2 factorial design. Patients needed to be on dialysis for at least 3 months and were excluded if the residual urea clearance (by urine collection) exceeded 1.5 mL/min per 35 L volume of urea distribution, if serum albumin was <2.6 g/dL (26 g/L), or if there was a failure to achieve the high target dialysis dose in ≤4.5 hours on 2 of 3 consecutive monitored dialysis sessions. Other exclusion criteria included the presence of serious comorbid conditions, including active malignancy or infection, unstable angina or end-stage cardiac, pulmonary, or hepatic disease, or being scheduled for a living-donor kidney transplant. The standard-dose group achieved currently prescribed K/DOQI guidelines with a URR of 66.3 ± 2.5 (spKt/V = 1.32 ± 0.09 and eKt/V = 1.16 ± 0.08). The high-dose group achieved a URR of 72.5% ± 2.5%, spKt/V of 1.71 ± 0.11, and eKt/V of 1.53 ± 0.09. Flux, quantified by beta-2 microglobulin clearance, was 3 ± 7 mL/min and 34 ± 11 mL/min in the low- versus high-flux groups, respectively. Overall, the relative risk (RR) of mortality in the high-dose group was 0.96 (confidence interval 0.84–1.1, P = 0.53), implying a statistically nonsignificant 4% reduction in RR in the high-dose group. Similarly, the RR of death in the high-flux arm was 0.92 (confidence interval 0.81–1.05, P = 0.23). A secondary analysis suggested a benefit of high-flux membranes in those who were on hemodialysis for >3.7 years.6Eknoyan G. Beck G.J. Cheung A.K. et al.The Hemodialysis (HEMO) Study GroupEffect of dialysis dose and membrane flux in maintenance hemodialysis.N Engl J Med. 2002; 347: 2010-2019Crossref PubMed Scopus (1513) Google Scholar Higher dose was shown to be associated with a trend toward reduction in relative risk of mortality for women but not men; this was found to be nonsignificant after applying the Bonferroni correction for multiple comparisons.7Depner T. Daugirdas J. Greene T. et al.Dialysis dose and the effect of gender and body size on outcome in the HEMO study.Kidney Int. 2004; 65: 1386-1394Crossref PubMed Scopus (146) Google Scholar Although the HEMO study is a significant landmark in the history of dialysis therapy, several limitations need to be considered while interpreting its results. Mean age (57.6 ± 14.0 years) was lower than that of the general US dialysis population. Blacks (who have better outcomes on dialysis) were overrepresented when compared with the US population (62% versus 37%). Heavier patients had to be excluded because of the requirement to deliver a high dose in a limited time. Thus, 97% of patients who underwent randomization weighed less than 100 kg. Mean albumin was 3.6 ± 0.4 mg/dL (36 ± 4 g/L), and severely malnourished patients were excluded. Patients were, in general, receiving a high dose of dialysis at baseline and a significant proportion were being treated with high-flux dialyzers; a carryover effect into the trial has been suggested that could have contributed to the negative outcome.20Locatelli F. How will results of the HEMO Study impact dialysis practice?.Semin Dial. 2003; 16: 20-21Crossref Google Scholar There were no protocol-driven guidelines to achieve dry weight, an important parameter in the optimal delivery of hemodialysis. However, this is likely to have been similar in all groups. The range of treatment times was relatively narrow and did not permit an examination of the independent effect of “t” in Kt/V. The separation of dose, although achieved by diligent adherence to protocol, may be in the zone in the dose response curve, where the law of diminishing returns prevails especially with a relatively fixed “t.” Thus, an alternative conclusion could be that the HEMO study could not detect a significant difference between high- and standard-dose groups under the fairly strict conditions of the study. It has also been reasoned that the results of the HEMO study are not inconsistent with the findings of large observational studies because the confidence interval of the relative risk of mortality was not significantly different from the much narrower confidence interval observed with the much larger “n” of observational studies, despite the limitations commented on earlier.21Port F.K. Wolfe R.A. How will results of the HEMO Study impact dialysis practice?.Semin Dial. 2003; 16: 13-16Crossref Google Scholar Results from the international Dialysis Outcomes and Practice Patterns Study (DOPPS) are consistent with prior observational studies insofar as the relationship between dialysis dose and mortality outcomes is concerned. The RR of mortality for spKt/V <1.2 was 1.16 (P = 0.025).22Port F.K. Pisoni R.L. Bragg-Gresham J.L. et al.DOPPS estimates of patient life years attributable to modifiable hemodialysis treatment practices in the United States.Blood Purif. 2004; 22: 175-180Crossref PubMed Scopus (92) Google Scholar Furthermore, recent analyses of both DOPPS data and the Centers for Medicare and Medicaid Services database have revealed that the benefit of eKt/V above 1.2 was seen for women but not for men,23Port F.K. Wolfe R.A. Hulbert-Shearon T.E. et al.High dialysis dose is associated with lower mortality among women but not among men.Am J Kidney Dis. 2004; 43: 1014-1023Abstract Full Text Full Text PDF PubMed Scopus (84) Google Scholar despite adjustment for body size. This is shown in Fig 1 and is consistent with a similar finding by sex from the HEMO Study.7Depner T. Daugirdas J. Greene T. et al.Dialysis dose and the effect of gender and body size on outcome in the HEMO study.Kidney Int. 2004; 65: 1386-1394Crossref PubMed Scopus (146) Google Scholar The possible reasons why this may be the case is a subject of intense speculation, but no clear explanation has been forthcoming. Variability in demographic and comorbid conditions explains only part of the difference in mortality between dialysis centers, both within countries and across continents. The remainder of the difference could potentially relate to practice patterns associated with dialysis.24Goodkin D.A. Bragg-Gresham J.L. Koenig K.G. et al.Association of comorbid conditions and mortality in hemodialysis patients in Europe, Japan, and the United States in the Dialysis Outcomes and Practice Patterns Study (DOPPS).J Am Soc Nephrol. 2003; 14: 3270-3277Crossref PubMed Scopus (610) Google Scholar Certain practices related to hemodialysis have been associated with an increased mortality risk in the DOPPS (Table 1).22Port F.K. Pisoni R.L. Bragg-Gresham J.L. et al.DOPPS estimates of patient life years attributable to modifiable hemodialysis treatment practices in the United States.Blood Purif. 2004; 22: 175-180Crossref PubMed Scopus (92) Google Scholar Results of a DOPPS estimate regarding patient life years gained if individuals outside guideline-based targets are brought within them was recently published (Table 2).22Port F.K. Pisoni R.L. Bragg-Gresham J.L. et al.DOPPS estimates of patient life years attributable to modifiable hemodialysis treatment practices in the United States.Blood Purif. 2004; 22: 175-180Crossref PubMed Scopus (92) Google Scholar This sort of analysis helps measure the relative impact of individual practice patterns (assuming causality) and emphasizes the fact that dialysis dose is only part of the overall care of the dialysis patient. Clearly, other practices such as achievement of dry weight, blood pressure control, and cardiovascular and infection prophylaxis have not yet been subjected to these kinds of analyses. The magnitude of potential savings in life years should encourage greater adherence to guidelines such as the K/DOQI and to practices that are significantly associated with better survival.Table 1Adjusted*Stratified by country and adjusted for age, race, years of ESRD, and 15 summary comorbid conditions. Relative Risk (RR) of Mortality by Guideline or Practice Pattern and Percent of Patients Outside Guideline or Practice PatternSelected Modifiable Practice PatternsMortality RR*Stratified by country and adjusted for age, race, years of ESRD, and 15 summary comorbid conditions. From DOPPS IPercent of Patients Outside Range From DOPPS II (US)RRP ValueDialysis doseSpKt/V <1.21.160.02512.1%Mineral metabolismPO4 >5.5 mg/dL1.110.00549.2%Anemia managementHemoglobin <11 g/dL1.14<0.000127.2%NutritionIDWG >5.7%1.220.00212.5%Albumin <3.5 g/dL1.38<0.000120.5%Vascular accessFacility catheter use >28% versus <7%1.230.0150.0%NOTE. To convert hemoglobin in g/dL to g/L, multiply by 10; albumin in g/dL to g/L, multiply by 10.Abbreviation: IDWG, interdialytic weight gain.Reprinted with permission from Port et al,22Port F.K. Pisoni R.L. Bragg-Gresham J.L. et al.DOPPS estimates of patient life years attributable to modifiable hemodialysis treatment practices in the United States.Blood Purif. 2004; 22: 175-180Crossref PubMed Scopus (92) Google Scholar © 2004 S. Karger AG, Basel.* Stratified by country and adjusted for age, race, years of ESRD, and 15 summary comorbid conditions. Open table in a new tab Table 2Projected Patient Years Gained, Based on Relative Risk for 100% Compliance With Guidelines and Other Practice Patterns (5 Years, United States)MeasureCurrent Statistics123456Total*Adjusted for the factors in columns 1–6 plus age, sex, race, years on HD, and 15 summary comorbid conditions. (Sum of 1–6)Kt/V ≥1.2PO4 ≤5.5Hgb ≥11 g/dLIDWG ≤5.7%Albumin ≥3.5 g/dLFacility Catheter Use ≤7%Annual death rate0.240.23050.22290.22640.22870.21720.21080.186Annual other loss rate†Other loss includes receiving a transplant and switching to peritoneal dialysis.0.0650.0650.0650.0650.0650.0650.0650.065Total loss rate0.300.29550.28790.29140.29370.28220.27580.251Patient years (PY), total1,746,4011,758,8471,780,1941,770,3111,763,8651,796,4801,815,0541,890,018Potential PY gained, 100% brought within targets—12,44633,79323,91017,46450,07968,653143,617‡Total is less than the sum of columns 1–6 since it takes into account correlated findings.§The total potential PY gain of 143,617 represents the difference between 1,890,018 and 1,746,401.Potential PY gained, 50% brought within targets—6,01116,32211,5488,43524,18833,15969,367§The total potential PY gain of 143,617 represents the difference between 1,890,018 and 1,746,401.NOTE. Each model was based on 313,000 prevalent patients at the start (projected for 2004) and added 116,477 per year, for 5 years. To convert hemoglobin in g/dL to g/L, multiply by 10; albumin in g/dL to g/L, multiply by 10.Abbreviations: Hgb, hemoglobin; IDWG, interdialytic weight gain.Reprinted with permission from Port et al,22Port F.K. Pisoni R.L. Bragg-Gresham J.L. et al.DOPPS estimates of patient life years attributable to modifiable hemodialysis treatment practices in the United States.Blood Purif. 2004; 22: 175-180Crossref PubMed Scopus (92) Google Scholar © 2004 S. Karger AG, Basel.* Adjusted for the factors in columns 1–6 plus age, sex, race, years on HD, and 15 summary comorbid conditions.† Other loss includes receiving a transplant and switching to peritoneal dialysis.‡ Total is less than the sum of columns 1–6 since it takes into account correlated findings.§ The total potential PY gain of 143,617 represents the difference between 1,890,018 and 1,746,401. Open table in a new tab NOTE. To convert hemoglobin in g/dL to g/L, multiply by 10; albumin in g/dL to g/L, multiply by 10. Abbreviation: IDWG, interdialytic weight gain. Reprinted with permission from Port et al,22Port F.K. Pisoni R.L. Bragg-Gresham J.L. et al.DOPPS estimates of patient life years attributable to modifiable hemodialysis treatment practices in the United States.Blood Purif. 2004; 22: 175-180Crossref PubMed Scopus (92) Google Scholar © 2004 S. Karger AG, Basel. NOTE. Each model was based on 313,000 prevalent patients at the start (projected for 2004) and added 116,477 per year, for 5 years. To convert hemoglobin in g/dL to g/L, multiply by 10; albumin in g/dL to g/L, multiply by 10. Abbreviations: Hgb, hemoglobin; IDWG, interdialytic weight gain. Reprinted with permission from Port et al,22Port F.K. Pisoni R.L. Bragg-Gresham J.L. et al.DOPPS estimates of patient life years attributable to modifiable hemodialysis treatment practices in the United States.Blood Purif. 2004; 22: 175-180Crossref PubMed Scopus (92) Google Scholar © 2004 S. Karger AG, Basel. Ultrafiltration rate during dialysis varies substantially among patients and, depending on patient tolerance for large swings in intravascular volume, may be responsible in large part for intradialytic hypotension and the resultant unstable treatments requiring extra attention from staff. That ultrafiltration rate may be an independent predictor of mortality and hospitalization outcome has only recently come to light, based on an analysis of the DOPPS data.25Saran R. Bragg-Gresham J.L. Combe C. et al.High ultrafiltration rates are associated with poor outcomes in hemodialysis patients The Dialysis Outcomes and Practice Patterns Study (DOPPS).J Am Soc Nephrol. 2003; 14 (abstr): 234ACrossref PubMed Scopus (53) Google Scholar Closely linked to ultrafiltration rate is the key issue of treatment time that has not yet received the attention it deserves in hemodialysis-related clinical trials. The National Cooperative Dialysis Study did suggest that there was a trend for benefit of the longer treatment time, but this was not statistically significant. DOPPS data suggest that, all else being equal (ie, after adjustments for multiple comorbidities, patient demographics, and Kt/V), treatment time of <3.5 hours is associated with a significantly higher mortality risk compared with >3.5 hours (relative ratio = 1.17, P = 0.02).25Saran R. Bragg-Gresham J.L. Combe C. et al.High ultrafiltration rates are associated with poor outcomes in hemodialysis patients The Dialysis Outcomes and Practice Patterns Study (DOPPS).J Am Soc Nephrol. 2003; 14 (abstr): 234ACrossref PubMed Scopus (53) Google Scholar Unfortunately, there is not a wide enough spread of dialysis times within the DOPPS to evaluate the beneficial effects, if any, of longer than 4.5 hours of dialysis, and the numbers of patients are too small to allow for statistically meaningful results. However, the data are supportive of the paradigm of longer dialysis time and need to be examined further. Preliminary data from the Australia and New Zealand Dialysis and Transplant Registry indicate that dialysis duration of >4.5 hours may be associated with a lower relative risk of mortality and duration <3.5 hours associated with higher mortality risk (M. Marshall, personal communication, July 2004). Recent evidence based on an analysis of European DOPPS data also suggests that hemodiafiltration is associated with a lower relative risk of mortality compared with conventional hemodialysis (relative ratio = 0.77, P = 0.02).26Canaud B. Bragg-Gresham J.L. Marshall M.R. et al.Patients receiving hemodiafiltration or hemofiltration have lower mortality risk than patients receiving hemodialysis without replacement fluid (HD) in Europe The Dialysis Outcomes and Practice Patterns Study (DOPPS).J Am Soc Nephrol. 2003; 14 (abstr): 31AGoogle Scholar After adjustment for Kt/V, however, this difference became statistically nonsignificant, suggesting that the higher small solute clearance associated with hemodiafiltration could possibly explain the difference in outcomes between the two modes of therapy. These results merit further examination by randomized clinical trials of hemodiafiltration versus conventional hemodialysis. The added benefits of convective clearance could be mediated via the improved clearance of middle and large molecules that are important uremic toxins. Moreover, convective therapies require the use of ultrapure dialysate (ie, sterile and therefore nonpyrogenic), which reduces the propensity to systemic microinflammation. This is a randomized controlled clinical trial currently in progress at 9 clinical centers in Europe27Locatelli F. Hannedouche T. Jacobson S. et al.The effect of membrane permeability on ESRD Design a prospective randomized multicentre trial.J Nephrol. 1999; 12: 85-88PubMed Google Scholar that has been designed to prospectively evaluate the long-term effect of membrane permeability (ie, flux) on clinical outcomes such as mortality, morbidity, vascular access survival, and nutritional status. Incident dialysis patients are included to avoid the effect of prior treatment on outcomes; the expected sample size is 660 patients and duration of follow-up will be 3 to 6 years. No reuse will be allowed in order to eliminate the possibility of reduction in flux because of reuse. The field of dialysis has seen very few randomized controlled clinical trials. It is to be hoped that this trend is changing as the renal community increasingly recognizes the challenges of very high mortality and morbidity among dialysis patients. It is quite clear, however, that the paradigm of thrice-weekly dialysis is faced with diminishing returns, with the possible exception of long dialysis sessions. This extended approach may allow for slower ultrafiltration and perhaps better control of blood pressure and middle molecular clearance, while avoiding the hemodynamic stress of short dialysis that could be responsible for repetitive micro-ischemic insults to various organs, eventually contributing to organ dysfunction and mortality.28Charra B. Chazot C. Jean G. et al.Long 3 × 8 hr dialysis A three-decade summary.J Nephrol. 2003; 16: S64-S69PubMed Google Scholar The paradigms of more frequent dialysis sessions as short daily or slow nocturnal dialysis are increasingly viewed as possible advances in the near future. There are several groups that have reported encouraging results in uncontrolled studies at single centers.29Lindsay R.M. Blake P.G. How will results of the HEMO Study impact dialysis practice?.Semin Dial. 2003; 16: 16-19Crossref Google Scholar However, unless this paradigm is subjected to the rigors of a randomized controlled clinical trial, acceptance is unlikely by the nephrology community at large. Therefore, the National Institute of Diabetes and Digestive and Kidney Diseases has recently funded a large multicenter study using these approaches; it is hoped that such a study will commence by early 2005. The nephrology community will eagerly await these results, which should provide many useful insights into the management of uremia. Furthermore, the role of convective therapies such as hemofiltration and hemodiafiltration is unresolved and deserves to be rigorously examined in randomized clinical trials. Dose of dialysis is a crucial element in the overall management of patients with end-stage renal disease. The NCDS provided the first evidence that monitoring small solute clearance was an effective means of quantification of dialysis dose. Subsequently, the duration of dialysis sessions was not given the importance it deserved and has not yet received scrutiny in randomized controlled clinical trials. There is some preliminary evidence from the DOPPS and the Australia and New Zealand Dialysis and Transplant Registry that treatment time may be independently associated with outcomes among dialysis patients. Observational studies have consistently shown a decreased mortality risk associated with higher achieved URR or Kt/V. However, observational studies are subject to selection bias and confounding. The multicenter randomized HEMO Study found no overall survival benefit when high-dose dialysis was compared with standard dose (based on K/DOQI guidelines) and high flux compared with low flux. Secondary analyses suggest a beneficial effect of high dose in women but not men, a finding supported both by DOPPS and Centers for Medicare and Medicaid Services data. The HEMO Study did not lend itself to the examination of duration of the dialysis session as a predictor of outcomes. The newer paradigms of more frequent or prolonged slow dialysis and convective therapies await critical evaluation by randomized controlled clinical trials.