- Email: [email protected]
The Effect of Increasing Dialysis Dose in Overweight Hemodialysis Patients on Quality of Life: A 6-Week Randomized Crossover Trial
The Effect of Increasing Dialysis Dose in Overweight Hemodialysis Patients on Quality of Life: A 6-Week Randomized Crossover Trial Wenjie Wang, MD, PhD, FRCPC,1 Marcello Tonelli, MD, MSc, FRCPC,2,3 Brenda Hemmelgarn, MD, PhD, FRCPC,1,4 Song Gao, MSc,1 Jeffrey A. Johnson, PhD,2 Ken Taub, MD, MBA, FRCPC,1 and Braden Manns, MD, MSc, FRCPC,1,2,4 on behalf of the Alberta Kidney Disease Network Background: Using standard hemodialysis regimens, overweight patients often do not reach Kidney Disease Outcomes Quality Initiatives (KDOQI) Kt/V targets, and this has been associated with lower health-related quality of life (HRQL). Whether increasing dialysis adequacy in large patients not achieving KDOQI targets improves HRQL is unknown. Study Design: Randomized blinded crossover study. Setting & Participants: Overweight (⬎80 kg) underdialyzed patients from 6 dialysis units in 2 Canadian dialysis programs. Interventions: Six-week treatment periods with a standard dialysis regimen (4 hours 3 times weekly) and 3 augmented regimens: 4.5 hours of hemodialysis, 4 hours of hemodialysis with increased dialysate flow, and 4 hours of hemodialysis with 2 dialyzers in parallel. Outcomes & Measurements: The End-Stage Renal Disease Symptom domain of the Kidney Disease Quality-of-Life Short-Form questionnaire (primary outcome) and the Health Utilities Index Mark 2 (secondary outcome). Results: We enrolled 18 patients (mean weight, 109.7 ⫾ 16.2 [SD] kg); 12 completed all 4 regimens. Mean Kt/Vs during the study were 1.27 (95% confidence interval [CI], 1.19 to 1.35), 1.41 (95% CI, 1.32 to 1.50), 1.31 (95% CI, 1.22 to 1.39), and 1.41 (95% CI, 1.33 to 1.49) for patients receiving standard dialysis, 4.5 hours of hemodialysis, hemodialysis with increased dialysate flow, and hemodialysis with 2 dialyzers, respectively. Kidney Disease Quality-of-Life End-Stage Renal Disease Symptom domain and Health Utilities Index Mark 2 scores were 75.9 (95% CI, 70.7 to 81.2) and 0.69 (95% CI, 0.56 to 0.81) for patients receiving standard dialysis, respectively. These did not differ when patients received the 3 augmented dialysis regimens (P ⫽ 0.2 and P ⫽ 0.5, respectively). Limitations: Small sample size and inability to fully blind patients to the treatment they were receiving. Conclusion: Improving hemodialysis adequacy for large underdialyzed patients did not lead to improved HRQL. Our findings suggest that augmentation of the dialysis regimen is not required for these patients in the absence of overt uremic symptoms. Am J Kidney Dis 51:796-803. © 2008 by the National Kidney Foundation, Inc. INDEX WORDS: Overweight; dialysis adequacy; 2 dialyzers in parallel; Health Utilities Index Mark 2 (HUI2); Kidney Disease Quality of Life Short Form (KDQOL-SF).
he prevalence of obesity is increasing in patients with end-stage renal disease (ESRD),1 mirroring well-described trends in the general population. Kidney Diseases Outcomes Quality Initiatives (KDOQI) guidelines for dialysis adequacy recommend that delivered single-pool Kt/V be at least 1.2 in patients receiving thrice-weekly hemo-
T
dialysis.2 For the increasing number of obese patients on dialysis therapy, achieving this adequacy target is challenging.3,4 Observational studies indicated that underdialysis (ie, single-pool Kt/V ⬍ 1.2) was associated with lower health-related quality of life (HRQL)5 and other adverse health outcomes, including greater mortality.6-11
From the 1Department of Medicine, Division of Nephrology, University of Calgary, Foothills Medical Center, Calgary; 2Institute of Health Economics;3Department of Medicine, Division of Nephrology, University of Alberta, Edmonton; and 4Department of Community Health Sciences, University of Calgary, Foothills Medical Center, Calgary, Alberta, Canada. Received July 25, 2007. Accepted in revised form December 11, 2007. Originally published online as doi: 10.1053/j.ajkd.2007.12.031 on March 28, 2008.
Address correspondence to Braden J. Manns, MD, MSc, FRCPC, Foothills Medical Center, 1403-29th St NW, Calgary, Alberta, Canada T2N 2T9. E-mail: braden. [email protected] © 2008 by the National Kidney Foundation, Inc. 0272-6386/08/5105-0012$34.00/0 doi:10.1053/j.ajkd.2007.12.031
796
American Journal of Kidney Diseases, Vol 51, No 5 (May), 2008: pp 796-803
Dialysis Adequacy and HRQL
It is uncertain whether improving Kt/V from less than 1.2 to greater than 1.2 will result in improvement in clinical outcomes, including HRQL. In the largest prospective study to date, the Hemodialysis (HEMO) Study investigators did not find a mortality benefit comparing a standard dialysis regimen (single-pool Kt/V of 1.25) and a high-dose dialysis regimen (singlepool Kt/V of 1.65).12 In the HEMO Study, investigators also measured HRQL and did not find clinically significant differences between the standard-dose and high-dose groups.13 However, patients in the HEMO Study were smaller (mean weight, 69 kg), received shorter dialysis runs on average at baseline, and, to be eligible for the study, had to be able to achieve a minimum equilibrated Kt/V of 1.3 during patient screening. Moreover, mean baseline Kt/V for patients in the HEMO Study was 1.43, and very heavy patients were excluded on the basis of these eligibility criteria.12 For underdialyzed patients, less information is available about the impact of improving Kt/V on HRQL. In a prospective unblinded crossover study of 14 large patients, improving Kt/V from 1.16 to 1.34 by using 2 dialyzers in parallel was associated with improved HRQL, measured by using the 36-Item Short-Form Health Survey (SF-36).14 There are several potential methods for improving Kt/V in patients who remain underdialyzed despite optimizing their dialysis care within a 4-hour thrice-weekly regimen. Dialysate flow can be increased to greater than 500 mL/min,15,16 dialysis time can be extended to longer than 4 hours, or patients can be dialyzed using 2 dialyzers in parallel14,17 or 2 dialyzers in serial.17,18 Increasing dialysate flow rates is simple, but its impact on dialysis adequacy is limited.15 Extending the dialysis session is widely practiced in many dialysis centers; however, this often creates difficulty given the current shortage of nurses and dialysis stations in Canada and the requirement to provide 3 dialysis sessions daily in most dialysis units.19 Use of 2 dialyzers either in parallel or serial can improve delivered Kt/ V.14,17 Importantly, the question of whether improving Kt/V by using any of these methods leads to improved HRQL requires additional study.
797
Given these uncertainties, we sought to determine whether improving small-solute clearance by using any of these strategies was associated with improved HRQL in large hemodialysis patients with a history of inadequate dialysis adequacy.
METHODS Patients We performed a randomized crossover study in 6 dialysis units affiliated with the University of Calgary and University of Alberta. Inclusion criteria were as follows: patients on a stable regimen of hemodialysis for 3 months or longer, dry weight of 80 kg or greater, Kt/V of 1.2 or less on 2 occasions in the previous 6 months or requirement for longer than 12 h/wk of hemodialysis because of a history of inadequate dialysis, and blood flow rate of 350 mL/min or greater through a well-functioning access. Patients were excluded if they had renal transplantation scheduled in the next 6 months, had a contraindication to intradialytic anticoagulation, required longer than 4½ hours of hemodialysis 3 times weekly for the purpose of volume removal because of excessive interdialytic weight gains or intradialytic hypotension/ cramping, and failure to provide informed consent. All patients who met inclusion/exclusion criteria and were not already receiving a hemodialysis regimen consisting of 4 hours 3 times weekly were switched to this modality for 4 weeks before entry into the active treatment part of the study.
Study Design During the study period, each patient received 4 different dialysis modality regimens in a randomized crossover fashion, each lasting 6 weeks. A 6-week duration for each strategy was chosen for convenience and because it previously was suggested that HRQL can change during the course of only 4 weeks in patients receiving increased dialysis adequacy.14 Although a formal washout period was not included between study periods, because each treatment was administered for 6 weeks and outcomes were measured in the final week of the treatment, it could be considered that we allowed time for washout to occur between treatments. To permit blinding of patients to dialysis treatment modality, all dialysis sessions were 4.5 hours; for the 3 strategies that included active dialysis for only 4 hours, dialysate flow was stopped and no dialysis occurred for the last ½ hour of the run. The standard hemodialysis prescription consisted of 4 hours of hemodialysis 3 times weekly. Hemodialysis was undertaken using a high-flux high-efficiency polysulfone dialyzer (ie, F80A dialyzer; Fresenius Inc, Walnut Creek, CA, or equivalent) with a blood flow of 350 to 400 mL/min and dialysate flow of 500 mL/min. During the increased dialysate flow treatment, patients received standard dialysis for 4 hours 3 times weekly, but dialysate flow rate was 800 mL/min, rather than 500 mL/min. The third modality was increased dialysis time at 4½ hours 3 times weekly. Last, in the final dialysis modality consisting of 2 dialyzers in parallel, patients received hemodialysis for 4 hours 3 times
798 weekly with 2 F80A dialyzers connected by a Y-connector in a parallel configuration during each run by using the method described by Powers et al.14 Dialysate flow was 800 mL/ min. Previous research showed that by using this setup, blood and dialysate flows were split almost equally between dialyzers.14 During all study dialysis sessions, dialysis machines were sequestered behind a curtain in an attempt to blind patients to the dialysis strategy they were receiving. Blinding was believed to be important because perceived HRQL is subjective and might be influenced by knowledge of which dialysis type a patient was receiving. Patients unable to reach their target weight within 4 hours because of excessive weight gain and/or intradialytic hypotension/cramping were permitted to receive ultrafiltration during the last half hour of the run (reassessed each run). This was unlikely to significantly improve small-solute clearance because the expected increase in delivered Kt/V would be less than 0.5%. To assess the extent to which blinding was successful, participants were asked to state which type of dialysis they believed they received after each 6-week period.
Study Measures For sessions during which Kt/V was measured, postdialysis blood samples were drawn 2 minutes after decreasing the pump speed to 50 mL/min (ie, at 4 hours for 3 of the strategies and at 4½ hours for the increased dialysis time strategy). Predialysis and postdialysis urea samples were drawn from the arterial needle. All measurements were done on the midweek dialysis sessions. Urea measurements were obtained from 2 central laboratories (Calgary Laboratory Services, Calgary; and Provincial Laboratory, Edmonton, Canada). Single-pool Kt/V was calculated as: Kt ⁄ V ⫽ {⫺ln(ureapost ⁄ ureapre) ⫹ 3[(weightpre ⫺ weightpost) ⁄ weightpost]} ⁄ [1 ⫺ 0.01786(timedialysis)] The primary outcome of this study was HRQL. Consistent with published guidelines,20 HRQL was assessed using a spectrum of instruments. The Kidney Disease Quality-ofLife Short-Form (KDQOL-SF) questionnaire includes questions targeted at particular health-related concerns for individuals on dialysis therapy.5,21,22 Scores for each KDQOL-SF dimension range from 0 to 100, with higher scores reflecting better HRQL. The primary outcome of this study is the ESRD Symptom domain of the KDQOL-SF measured at the end of each 6-week treatment session. This was used as the primary outcome because it comprehensively assesses typical uremic symptoms and we previously showed in an observational study that patients achieving adequate Kt/V (ⱖ1.3 versus ⬍1.3) have significantly higher scores on the ESRD Symptom domain of the KDQOL.5 The KDQOL-SF also contains the SF-36, a generic HRQL profile used in patients with ESRD14,23 that includes 8 multi-item scales describing general HRQL issues.24,25 As with the KDQOLSF, scores for each dimension can range from 0 to 100, with higher scores reflecting better HRQL. Another secondary HRQL outcome was the Health Utilities Index Mark 2 (HUI2), used to estimate utility scores (a measure of overall quality of life). The HUI2 focuses on 7
Wang et al dimensions of health, including such dimensions relevant to dialysis patients as mobility, self-care, emotion, cognition, and pain. The HUI2 is based on health state valuations elicited from the general public and ranges from ⫺0.03 to 1.0.26 We used the HUI2 rather than other indirect measures of utility because it assesses more dimensions of HRQL and, as such, may be more sensitive to smaller increments in HRQL.26 The HUI2 was shown to be valid, reliable, and responsive to changes in health and is easy to administer.26 HUInc granted permission for use of the copyrighted HUI2 questionnaire used in this study. All questionnaires were self-administered with the assistance of a trained administrator if necessary (required by only 1 visually impaired patient). The protocol received ethical approval from The Conjoint Health and Research Ethics Board (University of Calgary and University of Alberta).
Statistical Analysis Sample size calculations were based on the primary hypothesis that HRQL for patients treated with a strategy to increase dialysis adequacy will be similar to HRQL when treated with standard dialysis. Locally, in an observational study of 133 hemodialysis patients, we found that as Kt/V increased, there were clinically and statistically significant increases in the KDQOL ESRD Symptom score and Euroqol EQ-5D utility scores.5 We assumed that Kt/V would improve by 0.2 with each of the augmented strategies and were interested in testing for a difference in KDQOL ESRD Symptom score of 10 points (a clinically significant change of moderate size27,28). Assuming ␣ ⫽ 0.05 and  ⫽ 0.2 and acknowledging our crossover design (enabling paired analysis), we estimated that 18 patients were needed in total. This seemed reasonable given that Powers et al,14 in a nonrandomized study of 14 patients, were able to show statistically and clinically significant improvements in 6 of 8 health domains in the SF-36 after only 4 weeks of dialysis using 2 dialyzers in parallel. Baseline characteristic values are presented as mean and 95% confidence interval (CI) or medians and interquartile range (IQR), when appropriate, for continuous variables and proportion for dichotomous variables. To determine the association between dialysis treatment regimens and quality of life, we used a linear mixed-effects model (SAS Proc Mixed; SAS, version 9.1.3; SAS Institute Inc, Cary, NC), which takes into account the correlation of data caused by repeated measurements for each subject given the crossover design of the study.29,30 This analysis provides an assessment of within-subject treatment comparisons (to determine the effect of the treatment) and also permits assessment of between-subject sequence comparisons (to determine the influence of the order of the treatments) and subject-bysequence effects (to assess for potential carry-over effects). Treatment, treatment period, and treatment sequence were modeled as fixed effects with subjects modeled as random effects. Least-squares means were computed for each treatment, adjusting for other effects in the model. There was no evidence of period or carry-over effects; therefore, all comparisons are presented relative to standard hemodialysis without the need for consideration of carry-over effects. Bonferroni adjustments were used to account for multiple comparisons. Given the non-normal distribution of the out-
Dialysis Adequacy and HRQL
799
come variables (quality of life and Kt/V), we also conducted the analysis using the GLIMMIX procedure, which fits statistical models to data when the response is not normally distributed. The results obtained were essentially identical to those obtained from the more familiar generalized mixedeffects models. Therefore, those results are presented here.
RESULTS Baseline Characteristics A total of 18 patients were enrolled, 17 of whom were men (Table 1; Fig 1). Median body mass index was 32.0 kg/m2 (25th to 75th percentile, 30.0 to 36.0), and median time on dialysis therapy was 21.5 months (25th to 75th percentile, 16.0 to 30.0) at baseline. The majority of patients had diabetes (77.8%), and most patients had several comorbid conditions, shown by a median Charlson score of 4 (IQR, 3 to 7). Mean screening Kt/V for patients receiving 4 hours of hemodialysis 3 times weekly before enrollment was 1.20 ⫾ 0.05 (SD; n ⫽ 11). After enrollment, patients were subjected to a 4-week screening period while receiving 4 hours 3 times weekly of hemodialysis, and mean Kt/V was 1.26 ⫾ 0.15 (SD; n ⫽ 13; baseline Kt/V results were not available for 5 patients at 1 dialysis site).
Table 1. Baseline Patient Characteristics Characteristic
Age (y) Men White Body mass index (kg/m2) Duration on dialysis (mo) Screening Kt/V for patients receiving ⱕ4 h dialysis (n ⫽ 11) Baseline Kt/V for patients while receiving 4-h dialysis (n ⫽ 13) Cause of end-stage renal disease Diabetes Glomerulonephritis Polycystic kidney disease Transplant failure Charlson comorbidity score Diabetes mellitus Coronary artery disease Congestive heart failure Cerebral vascular disease Peripheral vascular disease
n ⫽ 18
56.0 (44-66) 17 (94.4) 13 (72.2) 32.0 (30.0-36.0) 21.5 (16.0-30.0) 1.20 ⫾ 0.05 1.26 ⫾ 0.15 11 (61.1) 4 (22.2) 1 (5.6) 1 (5.6) 4 (3-7) 14(77.8) 8 (44.4) 5 (27.8) 3 (16.7) 4 (22.2)
Note: Values expressed as median (interquartile range), number (percent), or mean ⫾ SD.
Figure 1. Trial flow chart.
Blinding Success and Loss to Follow-up Overall, patients were unaware of their treatment allocation 57% of the time, ranging from 14% blinding success for dialysis using 2 dialyzers in parallel to 86% for dialysis using increased dialysate flow rate. Twelve of 18 patients completed all 4 treatment strategies, with 60 of 72 possible treatments (83.3%) completed overall. Results for the 12 patients with complete measurements were similar to those obtained for the 18 patients enrolled in the study. Reasons for early study withdrawal were as follows: 1 patient felt unwell with nausea and fatigue and refused to continue in the study (this occurred during standard dialysis treatment and was believed likely to be related to the prescribed dialysis regimen), 1 patient received a cadaveric transplant, 1 patient switched to home hemodialysis therapy, 2 patients dropped out of the study because of patient preference given the extra time requirement, and 1 patient experienced a serious adverse event that led to withdrawal while receiving dialysis using 2 dialyzers in parallel. This patient experienced severe intravascular hemolysis requiring hospitalization and transfusion related to kinked dialysis tubing required for the use of 2 dialyzers in parallel. Impact on Kt/V Hemodialysis regimens consisting of 2 dialyzers in parallel and increased dialysis time significantly
800
Wang et al Table 2. Difference in Single-Pool Kt/V by Dialysis Modality HD Modality*
Kt/V (95% CI)
Difference† (95% CI)
P‡
Standard HD Increased dialysate flow Increased dialysis time 2 Dialyzers in parallel
1.27 (1.19 to 1.35) 1.31 (1.22 to 1.39) 1.41 (1.32 to 1.50) 1.41 (1.33 to 1.49)
Reference 0.04 (⫺0.10 to 0.17) 0.14 (0.01 to 0.28) 0.15 (0.02 to 0.27)
0.9 0.04 0.03
Abbreviations: HD, hemodialysis; CI, confidence interval. *Standard HD, 4 hours of HD and ½ hour of ultrafiltration; Increased dialysate flow, 4 hours of HD with increased dialysate flow and ½ hour of ultrafiltration; Extended dialysis time, 4.5 hours of HD; and 2 dialyzers in parallel, 4 hours of HD with 2 dialyzers in parallel and ½ hour of ultrafiltration. †Compared with standard HD. ‡Overall P ⫽ 0.01.
improved Kt/V to 1.41 (95% CI, 1.33 to 1.49) and 1.41 (95% CI, 1.32 to 1.50) compared with the 4-hour standard hemodialysis regimen Kt/V of 1.27 (95% CI, 1.19 to 1.35), respectively (P ⫽ 0.03 and P ⫽ 0.04, respectively). There was no statistically significant improvement in Kt/V in patients receiving increased dialysis flow compared with standard hemodialysis (Table 2). There was no evidence of a carry-over effect (P ⫽ 0.1 for subject-by-sequence interaction) or a period effect (P ⫽ 0.4). Impact on HRQL: Primary and Secondary Outcomes There was no difference in the primary HRQL outcome assessed by using the KDQOL ESRD Symptom domain for any of the treatment regimens compared with standard hemodialysis (Table 3; P ⫽ 0.1 for overall compari-
sons). Moreover, using the HUI2 measures, no difference was found among the 4 dialysis regimens (Table 3; P ⫽ 0.5 for overall comparison). To ensure that results were not caused by differential loss to follow-up, we repeated the analysis using the 12 patients who completed all 4 dialytic modalities; results were unchanged (data not shown). There was no evidence of a period effect for either the KDQOL ESRD Symptom domain (P ⫽ 0.6) or the HUI2 (P ⫽ 0.5). Other HRQL Analyses When we looked at the other domains of the KDQOL-SF and SF-36, there was no obvious trend toward benefit for any of the HRQL domains with any of the augmented dialysis regimens (Appendix).
Table 3. Difference in Primary and Secondary HRQL Measures HD Modality*
Standard HD Increased dialysate flow Increased dialysis time 2 Dialyzers in parallel Standard HD Increased dialysate flow Increased dialysis time 2 Dialyzers in parallel
HRQL Measure
KDQOL ESRD Symptom Scale (95% CI) 75.9 (70.7 to 81.2) 70.7 (65.3 to 76.0) 79.2 (72.6 to 85.8) 78.7 (72.8 to 84.6) HUI2 (95% CI) 0.69 (0.56 to 0.81) 0.64 (0.52 to 0.76) 0.70 (0.57 to 0.84) 0.69 (0.57 to 0.82)
Difference† (95% CI)
P‡
Reference ⫺5.25 (⫺15.00 to 4.50) 3.27 (⫺7.55 to 14.09) 2.76 (⫺7.55 to 13.07)
0.5 0.9 0.9
Reference ⫺0.04 (⫺0.16 to 0.08)) 0.02 (⫺0.12 to 0.16) 0.01 (⫺0.11 to 0.13)
0.8 1.0 1.0
Note: Possible ranges for KDQOL-SF symptoms domain are 0 to 100, and for HUI2, ⫺0.03 to 1.0. In each case, higher score represents better HRQL. Abbreviations: HRQL, health-related quality of life; HD, hemodialysis; CI, confidence interval; HUI2, Health Utility Index Mark 2; KDQOL-SF, Kidney Disease Quality of Life Short Form; ESRD; end-stage renal disease. *Standard HD, 4 hours of HD and ½ hour of ultrafiltration; Increased dialysate flow, 4 hours of HD with increased dialysate flow and ½ hour of ultrafiltration; Extended dialysis time, 4.5 hours of HD; and 2 dialyzers in parallel, 4 hours of HD with 2 dialyzers in parallel and ½ hour of ultrafiltration. †Compared with standard HD. ‡Overall P ⫽ 0.1 (KDQOL ESRD Symptom) and P ⫽ 0.5 (HUI2).
Dialysis Adequacy and HRQL
We also examined the impact on the primary and secondary HRQL end points, controlling for a potential sequence effect. We did not find evidence of a carry-over effect (P ⬎ 0.05 for all combinations).
DISCUSSION We found that extending dialysis time by 30 minutes or using 2 dialyzers in parallel can increase Kt/V in large underdialyzed patients, confirming studies that showed Kt/V can be improved with the use of 2 dialyzers in parallel14,17,18 and by extending dialysis time. However, we found that use of these augmented dialysis regimens had no impact on HRQL, measured using several different indexes. Unlike the study by Hauk et al,15 but consistent with the findings of Ouseph and Ward,16 we did not find that increasing dialysate flow to 800 mL/min on its own resulted in a statistically or clinically significant difference in Kt/V. Our results contrast with those of Powers et al,14 who used 2 dialyzers in parallel to increase Kt/V from 1.16 to 1.34 and showed improvement in 6 of 8 domains of the SF-36. However, the study by Powers et al14 was not blinded, and it is possible that more careful attention to dialysis or other factors may have been responsible for the improvement in quality of life. It also is possible that subjects in the study by Powers et al14 were more underdialyzed than in the present study and had more significant uremic symptoms at baseline. Our results are consistent with and extend findings from the HEMO Study,12 which also showed that increasing single-pool Kt/V to greater than 1.25 in patients who predominantly weighed less than 80 kg did not improve quality of life. Taken together, results from these 2 randomized trials would suggest that regardless of patient size, increasing Kt/V within the range tested (and achievable by using conventional thrice-weekly dialysis) does not improve HRQL. For large hemodialysis patients such as those in our study, 4 hours of hemodialysis 3 times weekly may not provide enough dialysis to achieve target Kt/V greater than 1.2 on a consistent basis. Our findings suggest that augmentation of the dialysis regimen is not required in the absence of overt uremic symptoms. In addition, because mean delivered Kt/V at the beginning of
801
the study appeared to improve compared with that obtained at the screening visit (perhaps suggesting a Hawthorne effect), more careful attention to providing the standard dialysis regimen as prescribed may be sufficient to achieve recommended Kt/V targets in some cases. There are limitations to our study. First, our study had a small sample size, with only 18 patients enrolled and only 16 patients finishing most of the dialysis regimens. Using paired analysis possible with the crossover design in which patients served as their own controls and acknowledging that we had data for only 16 patients for most dialysis regimens, we estimate that we had 76% power to detect a clinically significant 10point difference in the ESRD-targeted symptom domain of the KDQOL. As such, we believe our study was powered to detect a quality-of-life difference of moderate size, although we cannot exclude the possibility that a smaller (but still clinically relevant) difference exists.28 However, we believe this is unlikely because there was no apparent trend to improvement in quality of life with the 3 augmented dialysis regimens. Second, the intent of the study is to determine the impact of augmented dialysis regimens on quality of life in underdialyzed patients. One could argue that the greater than expected baseline Kt/V means that our study predominantly included patients who were already meeting KDOQI targets. However, at the time these patients were enrolled, they were underdialyzed according to KDOQI criteria and thus were potential candidates for intensification of their dialysis regimen on clinical grounds. As such, we believe our results reflect the situation faced by clinicians in an everyday practice setting. Another limitation is that our study periods were only 6 weeks in duration. It is possible that a longer period would be required to impact on quality of life, although previous studies suggested that HRQL could change during the course of only 4 weeks in patients receiving increased dialysis adequacy.14 Finally, patients using 2 dialyzers in parallel were not adequately blinded, and this might have impacted on their assessment of their HRQL. However, this would be expected to have favorably influenced perceptions of HRQL and therefore is unlikely to have affected our conclusions given that we did not detect improvement in
802
Wang et al
HRQL for patients receiving treatment with 2 dialyzers in parallel. When considering whether to use 2 dialyzers in parallel, it is important to note that its use was associated with a treatment-related serious adverse event in this small study. This may have been caused by unfamiliarity with this technique by staff, although it occurred more than 1 year after initiating the study. In conclusion, large hemodialysis patients commonly fail to achieve KDOQI targets for smallsolute clearance, which may lead to a change in dialysis prescription. Although we found that use of augmented dialysis regimens can increase delivered Kt/V, we found no evidence that quality of life improved with such regimens. Our findings suggest that extended dialysis time, increased dialysate flow, or the use of 2 dialyzers in parallel provide no benefit for large asymptomatic patients who are having difficulty achieving delivered Kt/V greater than 1.2 with standard hemodialysis regimens, and in the absence of overt uremic symptoms, augmentation of the dialysis regimen is not required for these patients.
ACKNOWLEDGEMENTS Support: This study was sponsored by the Kidney Foundation of Canada. Financial Disclosure: None.
REFERENCES 1. Kramer HJ, Saranathan A, Luke A, et al: Increasing body mass index and obesity in the incident ESRD population. J Am Soc Nephrol 17:1453-1459, 2006 2. National Kidney Foundation: KDOQI Clinical Practice Guidelines for Hemodialysis Adequacy: Update 2006. Am J Kidney Disease 48:S3-S29, 2006 (suppl 1) 3. Elangovan L, Shinaberger CS, Kraut JA, et al: HEMO equilibrated Kt/V goals are difficult to achieve in large male patients. ASAIO J 47:235-239, 2001 4. Ifudu O, Mayers JD, Matthew JJ, et al: Standardized hemodialysis prescriptions promote inadequate treatment in patients with large body mass. Ann Intern Med 128:451-454, 1998 5. Manns BJ, Johnson JA, Taub K, et al: Dialysis adequacy and health related quality of life in hemodialysis patients. ASAIO J 48:565-569, 2002 6. Wolfe RA, Ashby VB, Daugirdas JT, et al: Body size, dose of hemodialysis, and mortality. Am J Kidney Dis 35:80-88, 2000 7. Owen WF, Lew NL, Liu Y, et al: The urea reduction ratio and serum albumin concentration as predictors of mortality in patients undergoing hemodialysis. N Engl J Med 329:1001-1106, 1993
8. Bloembergen WE, Stannard DC, Port FK, et al: Relationship of dose of hemodialysis and cause-specific mortality. Kidney Int 50:557-565, 1996 9. Held PJ, Port FK, Wolfe RA, et al: The dose of hemodialysis and patient mortality. Kidney Int 50:550-556, 1996 10. Salahudeen AK: Obesity and survival on dialysis. Am J Kidney Dis 41:925-932, 2003 11. Salahudeen AK, Fleischmann EH, Bower JD: Impact of lower delivered Kt/V on the survival of overweight patients on hemodialysis. Kidney Int 56:2254-2259, 1999 12. Eknoyan G, Beck GJ, Cheung AK, et al: Effect of dialysis dose and membrane flux in maintenance hemodialysis. N Engl J Med 347:2010-2019, 2002 13. Unruh M, Benz R, Greene T, et al: Effects of hemodialysis dose and membrane flux on health-related quality of life in the HEMO Study. Kidney Int 66:355-366, 2004 14. Powers KM, Wilkowski MJ, Helmandollar AW, et al: Improved urea reduction ratio and Kt/V in large hemodialysis patients using two dialyzers in parallel. Am J Kidney Dis 35:266-274, 2003 15. Hauk M, Kuhlmann MK, Riegel W, et al: In vivo effects of dialysate flow rate on Kt/V in maintenance hemodialysis patients. Am J Kidney Dis 35:105-111, 2000 16. Ouseph R, Ward RA: Increasing dialysate flow rate increases dialyzer urea mass transfer-area coefficients during clinical use. Am J Kidney Dis 37:316-320, 2001 17. Fritz BA, Doss S, McCann LM, Wrone EM: A comparison of dual dialyzers in parallel and series to improve urea clearance in large hemodialysis patients. Am J Kidney Dis 41:1008-1015, 2003 18. Sridhar NR, Hurst C, Hayes P: Tandem dialyzers with two monitors to meet target Kt/V. ASAIO J 51:65-69, 2005 19. Manns BJ, Taub KJ. The economics of end-stage renal disease care in Canada: Incentives and impact on delivery of care. Int J Health Care Finance Econ 7:149-169, 2007 20. Canadian Coordinating Office for Health Technology Assessment: Guidelines for Economic Evaluation of Pharmaceuticals: Canada. Ottawa, Canada, The Canadian Coordinating Office for Health Technology Assessment, 2006 21. Hays RD, Kallich JD, Mapes DL, et al: Kidney Disease Quality of Life Short Form, version 1.3: A Manual for Use and Scoring. Report. Santa Monica, CA, RAND, 1995 22. Hays RD, Kallich JD, Mapes DL, et al: Development of the Kidney Disease Quality of Life (KDQOL) instrument. Qual Life Res 3:329-338, 1994 23. Korevaar JC, Jansen MA, Merkus MP, et al: Quality of life in predialysis end-stage renal disease patients at the initiation of dialysis therapy. The NECOSAD Study Group. Perit Dial Int 20:69-75, 2000 24. Hays RD, Sherbourne C, Mazel R, et al: User’s Manual for Medical Outcomes Study (MOS) Core Measures of HealthRelated Quality of Life. Santa Monica, CA, RAND, 1995 25. McHorney CA, Ware JE Jr, Lu JF, et al: The MOS 36-Item Short-Form Health Survey (SF-36): III. Tests of data quality, scaling assumptions, and reliability across diverse patient groups. Med Care 32:40-66, 1994 26. Furlong WJ, Feeny DH, Torrance GW, et al: The Health Utilities Index (HUI) system for assessing health-related quality of life in clinical studies. Ann Med 33:375-384, 2001
Dialysis Adequacy and HRQL
803
27. Norman GSJ, Wyrwich KTW. Interpretation of changes in health-related quality of life: The remarkable universality of half a standard deviation. Med Care 41:582592, 2003 28. Wyrwich KTW, Babu AN, Kroenke K, et al: A comparison of clinically important differences in health-
related quality of life for patients with chronic lung disease, asthma, or heart disease. Health Serv Res 40:577-591, 2005 29. McCallagh P, Nelder JA: Generalized Linear Models (ed 2). London, UK, Chapman & Hall, 1989 30. Nelder JA, Wedderburn RWM: Generalized linear models. J R Stat Soc 135:370-384, 1972
APPENDIX. DIFFERENCES IN ALL REMAINING HRQL MEASURES BY DIALYSIS MODALITY HRQL
Standard HD (95% CI)
KDQOL kidney disease-targeted scales Symptoms 75.9 (70.7-81.2) Effects of kidney disease 61.4 (53.2-69.2) Burden of kidney disease 36.5 (25.3-47.7) Work status 18.8 (0.8-36.7) Cognitive function 76.1 (61.7-90.5) Quality of social interaction 66.4 (59.1-73.7) Sexual dysfunction 67.5 (50.4-84.5) Sleep 52.2 (41.2-63.2) Social support 80.1 (70.1-90.1) Dialysis staff encouragement 86.3 (78.8-93.8) Patient satisfaction 79.4 (66.4-92.3) SF-36 Health Scale Physical Function 48.5 (32.9-64.0) Role–Physical 35.1 (14.2-55.9) Pain 61.3 (46.8-75.8) General Health Perception 41.4 (32.7-50.2) Emotional Well-Being 69.3 (60.8-77.8) Role–Emotional 63.3 (43.1-83.6) Social Function 66.2 (51.9-80.6) Energy/Fatigue 44.8 (29.9-59.7)
Increased Dialysate Flow (95% CI)
Increased Dialysis Time (95% CI)
2 Dialyzers in Parallel (95% CI)
Overall P
70.7 (65.3-76.0) 53.7 (45.3-62.1) 40.0 (28.6-51.3) 18.2 (0-38.1) 82.8 (68.2-97.3) 70.0 (62.5-77.4) 66.4 (49.4-83.5) 49.9 (38.6-61.1) 78.0 (67.9-88.2) 82.8 (75.2-90.5) 77.4 (64.2-90.5)
79.2 (72.6-85.8) 62.2 (52.5-71.8) 43.5(31.2-55.9) 20.0 (0-46.1) 87.3 (72.2-102.4) 73.5 (64.9-82.2) 67.1 (50.1-84.2) 48.3 (35.4-61.1) 78.8 (66.2-91.4) 90.2 (81.6-98.8) 77.3 (62.8-91.8)
78.7 (72.8-84.6) 64.2 (55.4-73.1) 40.5 (28.8-52.2) 31.3 (0-57.0) 79.2 (64.5-93.9) 73.7 (65.7-81.6) 66.5 (49.4-83.5) 53.8 (42.0-65.6) 79.8 (68.6-91.0) 84.7 (76.7-92.8) 79.5 (65.8-93.1)
0.1 0.1 0.6 0.6 0.1 0.3 0.3 0.9 1.0 0.3 1.0
44.3 (28.6-60.0) 25.1 (3.7-46.4) 53.6 (38.9-68.4) 44.2 (35.3-53.2) 67.7 (59.0-76.4) 51.0 (30.3-71.7) 55.1 (40.5-69.7) 46.3 (31.2-61.3)
43.3 (26.7-59.9) 55.4 (30.5-80.2) 71.0 (54.8-87.1) 43.9 (33.5-54.3) 80.4 (70.6-90.1) 89.1 (64.7-113.5) 69.7 (52.9-86.4) 50.7 (34.8-66.6)
44.7 (28.7-60.7) 55.6 (33.0-78.2) 58.4 (43.1-73.6) 45.0 (35.6-54.5) 72.5 (63.4-81.6) 63.6 (41.5-85.7) 66.5 (51.1-81.9) 48.8 (33.4-64.1)
0.8 0.04 0.1 0.9 0.04 0.05 0.1 0.7
Note: Possible ranges for KDQOL-SF and SF-36 dimensions are 0 to 100; in each case, higher score represents better HRQL. Abbreviations: HRQL, health-related quality of life; CI, confidence interval; HD, hemodialysis; KDQOL-SF, Kidney Disease Quality of Life Short Form; SF-36, 36-Item Short Form Health Survey. *Standard HD, 4 hours of HD and ½ hour of ultrafiltration; Increased dialysate flow, 4 hours of HD with increased dialysate flow and ½ hour of ultrafiltration; Extended dialysis time, 4.5 hours of HD; and 2 dialyzers in parallel, 4 hours of HD with 2 dialyzers in parallel and ½ hour of ultrafiltration.
he prevalence of obesity is increasing in patients with end-stage renal disease (ESRD),1 mirroring well-described trends in the general population. Kidney Diseases Outcomes Quality Initiatives (KDOQI) guidelines for dialysis adequacy recommend that delivered single-pool Kt/V be at least 1.2 in patients receiving thrice-weekly hemo-
T
dialysis.2 For the increasing number of obese patients on dialysis therapy, achieving this adequacy target is challenging.3,4 Observational studies indicated that underdialysis (ie, single-pool Kt/V ⬍ 1.2) was associated with lower health-related quality of life (HRQL)5 and other adverse health outcomes, including greater mortality.6-11
From the 1Department of Medicine, Division of Nephrology, University of Calgary, Foothills Medical Center, Calgary; 2Institute of Health Economics;3Department of Medicine, Division of Nephrology, University of Alberta, Edmonton; and 4Department of Community Health Sciences, University of Calgary, Foothills Medical Center, Calgary, Alberta, Canada. Received July 25, 2007. Accepted in revised form December 11, 2007. Originally published online as doi: 10.1053/j.ajkd.2007.12.031 on March 28, 2008.
Address correspondence to Braden J. Manns, MD, MSc, FRCPC, Foothills Medical Center, 1403-29th St NW, Calgary, Alberta, Canada T2N 2T9. E-mail: braden. [email protected] © 2008 by the National Kidney Foundation, Inc. 0272-6386/08/5105-0012$34.00/0 doi:10.1053/j.ajkd.2007.12.031
796
American Journal of Kidney Diseases, Vol 51, No 5 (May), 2008: pp 796-803
Dialysis Adequacy and HRQL
It is uncertain whether improving Kt/V from less than 1.2 to greater than 1.2 will result in improvement in clinical outcomes, including HRQL. In the largest prospective study to date, the Hemodialysis (HEMO) Study investigators did not find a mortality benefit comparing a standard dialysis regimen (single-pool Kt/V of 1.25) and a high-dose dialysis regimen (singlepool Kt/V of 1.65).12 In the HEMO Study, investigators also measured HRQL and did not find clinically significant differences between the standard-dose and high-dose groups.13 However, patients in the HEMO Study were smaller (mean weight, 69 kg), received shorter dialysis runs on average at baseline, and, to be eligible for the study, had to be able to achieve a minimum equilibrated Kt/V of 1.3 during patient screening. Moreover, mean baseline Kt/V for patients in the HEMO Study was 1.43, and very heavy patients were excluded on the basis of these eligibility criteria.12 For underdialyzed patients, less information is available about the impact of improving Kt/V on HRQL. In a prospective unblinded crossover study of 14 large patients, improving Kt/V from 1.16 to 1.34 by using 2 dialyzers in parallel was associated with improved HRQL, measured by using the 36-Item Short-Form Health Survey (SF-36).14 There are several potential methods for improving Kt/V in patients who remain underdialyzed despite optimizing their dialysis care within a 4-hour thrice-weekly regimen. Dialysate flow can be increased to greater than 500 mL/min,15,16 dialysis time can be extended to longer than 4 hours, or patients can be dialyzed using 2 dialyzers in parallel14,17 or 2 dialyzers in serial.17,18 Increasing dialysate flow rates is simple, but its impact on dialysis adequacy is limited.15 Extending the dialysis session is widely practiced in many dialysis centers; however, this often creates difficulty given the current shortage of nurses and dialysis stations in Canada and the requirement to provide 3 dialysis sessions daily in most dialysis units.19 Use of 2 dialyzers either in parallel or serial can improve delivered Kt/ V.14,17 Importantly, the question of whether improving Kt/V by using any of these methods leads to improved HRQL requires additional study.
797
Given these uncertainties, we sought to determine whether improving small-solute clearance by using any of these strategies was associated with improved HRQL in large hemodialysis patients with a history of inadequate dialysis adequacy.
METHODS Patients We performed a randomized crossover study in 6 dialysis units affiliated with the University of Calgary and University of Alberta. Inclusion criteria were as follows: patients on a stable regimen of hemodialysis for 3 months or longer, dry weight of 80 kg or greater, Kt/V of 1.2 or less on 2 occasions in the previous 6 months or requirement for longer than 12 h/wk of hemodialysis because of a history of inadequate dialysis, and blood flow rate of 350 mL/min or greater through a well-functioning access. Patients were excluded if they had renal transplantation scheduled in the next 6 months, had a contraindication to intradialytic anticoagulation, required longer than 4½ hours of hemodialysis 3 times weekly for the purpose of volume removal because of excessive interdialytic weight gains or intradialytic hypotension/ cramping, and failure to provide informed consent. All patients who met inclusion/exclusion criteria and were not already receiving a hemodialysis regimen consisting of 4 hours 3 times weekly were switched to this modality for 4 weeks before entry into the active treatment part of the study.
Study Design During the study period, each patient received 4 different dialysis modality regimens in a randomized crossover fashion, each lasting 6 weeks. A 6-week duration for each strategy was chosen for convenience and because it previously was suggested that HRQL can change during the course of only 4 weeks in patients receiving increased dialysis adequacy.14 Although a formal washout period was not included between study periods, because each treatment was administered for 6 weeks and outcomes were measured in the final week of the treatment, it could be considered that we allowed time for washout to occur between treatments. To permit blinding of patients to dialysis treatment modality, all dialysis sessions were 4.5 hours; for the 3 strategies that included active dialysis for only 4 hours, dialysate flow was stopped and no dialysis occurred for the last ½ hour of the run. The standard hemodialysis prescription consisted of 4 hours of hemodialysis 3 times weekly. Hemodialysis was undertaken using a high-flux high-efficiency polysulfone dialyzer (ie, F80A dialyzer; Fresenius Inc, Walnut Creek, CA, or equivalent) with a blood flow of 350 to 400 mL/min and dialysate flow of 500 mL/min. During the increased dialysate flow treatment, patients received standard dialysis for 4 hours 3 times weekly, but dialysate flow rate was 800 mL/min, rather than 500 mL/min. The third modality was increased dialysis time at 4½ hours 3 times weekly. Last, in the final dialysis modality consisting of 2 dialyzers in parallel, patients received hemodialysis for 4 hours 3 times
798 weekly with 2 F80A dialyzers connected by a Y-connector in a parallel configuration during each run by using the method described by Powers et al.14 Dialysate flow was 800 mL/ min. Previous research showed that by using this setup, blood and dialysate flows were split almost equally between dialyzers.14 During all study dialysis sessions, dialysis machines were sequestered behind a curtain in an attempt to blind patients to the dialysis strategy they were receiving. Blinding was believed to be important because perceived HRQL is subjective and might be influenced by knowledge of which dialysis type a patient was receiving. Patients unable to reach their target weight within 4 hours because of excessive weight gain and/or intradialytic hypotension/cramping were permitted to receive ultrafiltration during the last half hour of the run (reassessed each run). This was unlikely to significantly improve small-solute clearance because the expected increase in delivered Kt/V would be less than 0.5%. To assess the extent to which blinding was successful, participants were asked to state which type of dialysis they believed they received after each 6-week period.
Study Measures For sessions during which Kt/V was measured, postdialysis blood samples were drawn 2 minutes after decreasing the pump speed to 50 mL/min (ie, at 4 hours for 3 of the strategies and at 4½ hours for the increased dialysis time strategy). Predialysis and postdialysis urea samples were drawn from the arterial needle. All measurements were done on the midweek dialysis sessions. Urea measurements were obtained from 2 central laboratories (Calgary Laboratory Services, Calgary; and Provincial Laboratory, Edmonton, Canada). Single-pool Kt/V was calculated as: Kt ⁄ V ⫽ {⫺ln(ureapost ⁄ ureapre) ⫹ 3[(weightpre ⫺ weightpost) ⁄ weightpost]} ⁄ [1 ⫺ 0.01786(timedialysis)] The primary outcome of this study was HRQL. Consistent with published guidelines,20 HRQL was assessed using a spectrum of instruments. The Kidney Disease Quality-ofLife Short-Form (KDQOL-SF) questionnaire includes questions targeted at particular health-related concerns for individuals on dialysis therapy.5,21,22 Scores for each KDQOL-SF dimension range from 0 to 100, with higher scores reflecting better HRQL. The primary outcome of this study is the ESRD Symptom domain of the KDQOL-SF measured at the end of each 6-week treatment session. This was used as the primary outcome because it comprehensively assesses typical uremic symptoms and we previously showed in an observational study that patients achieving adequate Kt/V (ⱖ1.3 versus ⬍1.3) have significantly higher scores on the ESRD Symptom domain of the KDQOL.5 The KDQOL-SF also contains the SF-36, a generic HRQL profile used in patients with ESRD14,23 that includes 8 multi-item scales describing general HRQL issues.24,25 As with the KDQOLSF, scores for each dimension can range from 0 to 100, with higher scores reflecting better HRQL. Another secondary HRQL outcome was the Health Utilities Index Mark 2 (HUI2), used to estimate utility scores (a measure of overall quality of life). The HUI2 focuses on 7
Wang et al dimensions of health, including such dimensions relevant to dialysis patients as mobility, self-care, emotion, cognition, and pain. The HUI2 is based on health state valuations elicited from the general public and ranges from ⫺0.03 to 1.0.26 We used the HUI2 rather than other indirect measures of utility because it assesses more dimensions of HRQL and, as such, may be more sensitive to smaller increments in HRQL.26 The HUI2 was shown to be valid, reliable, and responsive to changes in health and is easy to administer.26 HUInc granted permission for use of the copyrighted HUI2 questionnaire used in this study. All questionnaires were self-administered with the assistance of a trained administrator if necessary (required by only 1 visually impaired patient). The protocol received ethical approval from The Conjoint Health and Research Ethics Board (University of Calgary and University of Alberta).
Statistical Analysis Sample size calculations were based on the primary hypothesis that HRQL for patients treated with a strategy to increase dialysis adequacy will be similar to HRQL when treated with standard dialysis. Locally, in an observational study of 133 hemodialysis patients, we found that as Kt/V increased, there were clinically and statistically significant increases in the KDQOL ESRD Symptom score and Euroqol EQ-5D utility scores.5 We assumed that Kt/V would improve by 0.2 with each of the augmented strategies and were interested in testing for a difference in KDQOL ESRD Symptom score of 10 points (a clinically significant change of moderate size27,28). Assuming ␣ ⫽ 0.05 and  ⫽ 0.2 and acknowledging our crossover design (enabling paired analysis), we estimated that 18 patients were needed in total. This seemed reasonable given that Powers et al,14 in a nonrandomized study of 14 patients, were able to show statistically and clinically significant improvements in 6 of 8 health domains in the SF-36 after only 4 weeks of dialysis using 2 dialyzers in parallel. Baseline characteristic values are presented as mean and 95% confidence interval (CI) or medians and interquartile range (IQR), when appropriate, for continuous variables and proportion for dichotomous variables. To determine the association between dialysis treatment regimens and quality of life, we used a linear mixed-effects model (SAS Proc Mixed; SAS, version 9.1.3; SAS Institute Inc, Cary, NC), which takes into account the correlation of data caused by repeated measurements for each subject given the crossover design of the study.29,30 This analysis provides an assessment of within-subject treatment comparisons (to determine the effect of the treatment) and also permits assessment of between-subject sequence comparisons (to determine the influence of the order of the treatments) and subject-bysequence effects (to assess for potential carry-over effects). Treatment, treatment period, and treatment sequence were modeled as fixed effects with subjects modeled as random effects. Least-squares means were computed for each treatment, adjusting for other effects in the model. There was no evidence of period or carry-over effects; therefore, all comparisons are presented relative to standard hemodialysis without the need for consideration of carry-over effects. Bonferroni adjustments were used to account for multiple comparisons. Given the non-normal distribution of the out-
Dialysis Adequacy and HRQL
799
come variables (quality of life and Kt/V), we also conducted the analysis using the GLIMMIX procedure, which fits statistical models to data when the response is not normally distributed. The results obtained were essentially identical to those obtained from the more familiar generalized mixedeffects models. Therefore, those results are presented here.
RESULTS Baseline Characteristics A total of 18 patients were enrolled, 17 of whom were men (Table 1; Fig 1). Median body mass index was 32.0 kg/m2 (25th to 75th percentile, 30.0 to 36.0), and median time on dialysis therapy was 21.5 months (25th to 75th percentile, 16.0 to 30.0) at baseline. The majority of patients had diabetes (77.8%), and most patients had several comorbid conditions, shown by a median Charlson score of 4 (IQR, 3 to 7). Mean screening Kt/V for patients receiving 4 hours of hemodialysis 3 times weekly before enrollment was 1.20 ⫾ 0.05 (SD; n ⫽ 11). After enrollment, patients were subjected to a 4-week screening period while receiving 4 hours 3 times weekly of hemodialysis, and mean Kt/V was 1.26 ⫾ 0.15 (SD; n ⫽ 13; baseline Kt/V results were not available for 5 patients at 1 dialysis site).
Table 1. Baseline Patient Characteristics Characteristic
Age (y) Men White Body mass index (kg/m2) Duration on dialysis (mo) Screening Kt/V for patients receiving ⱕ4 h dialysis (n ⫽ 11) Baseline Kt/V for patients while receiving 4-h dialysis (n ⫽ 13) Cause of end-stage renal disease Diabetes Glomerulonephritis Polycystic kidney disease Transplant failure Charlson comorbidity score Diabetes mellitus Coronary artery disease Congestive heart failure Cerebral vascular disease Peripheral vascular disease
n ⫽ 18
56.0 (44-66) 17 (94.4) 13 (72.2) 32.0 (30.0-36.0) 21.5 (16.0-30.0) 1.20 ⫾ 0.05 1.26 ⫾ 0.15 11 (61.1) 4 (22.2) 1 (5.6) 1 (5.6) 4 (3-7) 14(77.8) 8 (44.4) 5 (27.8) 3 (16.7) 4 (22.2)
Note: Values expressed as median (interquartile range), number (percent), or mean ⫾ SD.
Figure 1. Trial flow chart.
Blinding Success and Loss to Follow-up Overall, patients were unaware of their treatment allocation 57% of the time, ranging from 14% blinding success for dialysis using 2 dialyzers in parallel to 86% for dialysis using increased dialysate flow rate. Twelve of 18 patients completed all 4 treatment strategies, with 60 of 72 possible treatments (83.3%) completed overall. Results for the 12 patients with complete measurements were similar to those obtained for the 18 patients enrolled in the study. Reasons for early study withdrawal were as follows: 1 patient felt unwell with nausea and fatigue and refused to continue in the study (this occurred during standard dialysis treatment and was believed likely to be related to the prescribed dialysis regimen), 1 patient received a cadaveric transplant, 1 patient switched to home hemodialysis therapy, 2 patients dropped out of the study because of patient preference given the extra time requirement, and 1 patient experienced a serious adverse event that led to withdrawal while receiving dialysis using 2 dialyzers in parallel. This patient experienced severe intravascular hemolysis requiring hospitalization and transfusion related to kinked dialysis tubing required for the use of 2 dialyzers in parallel. Impact on Kt/V Hemodialysis regimens consisting of 2 dialyzers in parallel and increased dialysis time significantly
800
Wang et al Table 2. Difference in Single-Pool Kt/V by Dialysis Modality HD Modality*
Kt/V (95% CI)
Difference† (95% CI)
P‡
Standard HD Increased dialysate flow Increased dialysis time 2 Dialyzers in parallel
1.27 (1.19 to 1.35) 1.31 (1.22 to 1.39) 1.41 (1.32 to 1.50) 1.41 (1.33 to 1.49)
Reference 0.04 (⫺0.10 to 0.17) 0.14 (0.01 to 0.28) 0.15 (0.02 to 0.27)
0.9 0.04 0.03
Abbreviations: HD, hemodialysis; CI, confidence interval. *Standard HD, 4 hours of HD and ½ hour of ultrafiltration; Increased dialysate flow, 4 hours of HD with increased dialysate flow and ½ hour of ultrafiltration; Extended dialysis time, 4.5 hours of HD; and 2 dialyzers in parallel, 4 hours of HD with 2 dialyzers in parallel and ½ hour of ultrafiltration. †Compared with standard HD. ‡Overall P ⫽ 0.01.
improved Kt/V to 1.41 (95% CI, 1.33 to 1.49) and 1.41 (95% CI, 1.32 to 1.50) compared with the 4-hour standard hemodialysis regimen Kt/V of 1.27 (95% CI, 1.19 to 1.35), respectively (P ⫽ 0.03 and P ⫽ 0.04, respectively). There was no statistically significant improvement in Kt/V in patients receiving increased dialysis flow compared with standard hemodialysis (Table 2). There was no evidence of a carry-over effect (P ⫽ 0.1 for subject-by-sequence interaction) or a period effect (P ⫽ 0.4). Impact on HRQL: Primary and Secondary Outcomes There was no difference in the primary HRQL outcome assessed by using the KDQOL ESRD Symptom domain for any of the treatment regimens compared with standard hemodialysis (Table 3; P ⫽ 0.1 for overall compari-
sons). Moreover, using the HUI2 measures, no difference was found among the 4 dialysis regimens (Table 3; P ⫽ 0.5 for overall comparison). To ensure that results were not caused by differential loss to follow-up, we repeated the analysis using the 12 patients who completed all 4 dialytic modalities; results were unchanged (data not shown). There was no evidence of a period effect for either the KDQOL ESRD Symptom domain (P ⫽ 0.6) or the HUI2 (P ⫽ 0.5). Other HRQL Analyses When we looked at the other domains of the KDQOL-SF and SF-36, there was no obvious trend toward benefit for any of the HRQL domains with any of the augmented dialysis regimens (Appendix).
Table 3. Difference in Primary and Secondary HRQL Measures HD Modality*
Standard HD Increased dialysate flow Increased dialysis time 2 Dialyzers in parallel Standard HD Increased dialysate flow Increased dialysis time 2 Dialyzers in parallel
HRQL Measure
KDQOL ESRD Symptom Scale (95% CI) 75.9 (70.7 to 81.2) 70.7 (65.3 to 76.0) 79.2 (72.6 to 85.8) 78.7 (72.8 to 84.6) HUI2 (95% CI) 0.69 (0.56 to 0.81) 0.64 (0.52 to 0.76) 0.70 (0.57 to 0.84) 0.69 (0.57 to 0.82)
Difference† (95% CI)
P‡
Reference ⫺5.25 (⫺15.00 to 4.50) 3.27 (⫺7.55 to 14.09) 2.76 (⫺7.55 to 13.07)
0.5 0.9 0.9
Reference ⫺0.04 (⫺0.16 to 0.08)) 0.02 (⫺0.12 to 0.16) 0.01 (⫺0.11 to 0.13)
0.8 1.0 1.0
Note: Possible ranges for KDQOL-SF symptoms domain are 0 to 100, and for HUI2, ⫺0.03 to 1.0. In each case, higher score represents better HRQL. Abbreviations: HRQL, health-related quality of life; HD, hemodialysis; CI, confidence interval; HUI2, Health Utility Index Mark 2; KDQOL-SF, Kidney Disease Quality of Life Short Form; ESRD; end-stage renal disease. *Standard HD, 4 hours of HD and ½ hour of ultrafiltration; Increased dialysate flow, 4 hours of HD with increased dialysate flow and ½ hour of ultrafiltration; Extended dialysis time, 4.5 hours of HD; and 2 dialyzers in parallel, 4 hours of HD with 2 dialyzers in parallel and ½ hour of ultrafiltration. †Compared with standard HD. ‡Overall P ⫽ 0.1 (KDQOL ESRD Symptom) and P ⫽ 0.5 (HUI2).
Dialysis Adequacy and HRQL
We also examined the impact on the primary and secondary HRQL end points, controlling for a potential sequence effect. We did not find evidence of a carry-over effect (P ⬎ 0.05 for all combinations).
DISCUSSION We found that extending dialysis time by 30 minutes or using 2 dialyzers in parallel can increase Kt/V in large underdialyzed patients, confirming studies that showed Kt/V can be improved with the use of 2 dialyzers in parallel14,17,18 and by extending dialysis time. However, we found that use of these augmented dialysis regimens had no impact on HRQL, measured using several different indexes. Unlike the study by Hauk et al,15 but consistent with the findings of Ouseph and Ward,16 we did not find that increasing dialysate flow to 800 mL/min on its own resulted in a statistically or clinically significant difference in Kt/V. Our results contrast with those of Powers et al,14 who used 2 dialyzers in parallel to increase Kt/V from 1.16 to 1.34 and showed improvement in 6 of 8 domains of the SF-36. However, the study by Powers et al14 was not blinded, and it is possible that more careful attention to dialysis or other factors may have been responsible for the improvement in quality of life. It also is possible that subjects in the study by Powers et al14 were more underdialyzed than in the present study and had more significant uremic symptoms at baseline. Our results are consistent with and extend findings from the HEMO Study,12 which also showed that increasing single-pool Kt/V to greater than 1.25 in patients who predominantly weighed less than 80 kg did not improve quality of life. Taken together, results from these 2 randomized trials would suggest that regardless of patient size, increasing Kt/V within the range tested (and achievable by using conventional thrice-weekly dialysis) does not improve HRQL. For large hemodialysis patients such as those in our study, 4 hours of hemodialysis 3 times weekly may not provide enough dialysis to achieve target Kt/V greater than 1.2 on a consistent basis. Our findings suggest that augmentation of the dialysis regimen is not required in the absence of overt uremic symptoms. In addition, because mean delivered Kt/V at the beginning of
801
the study appeared to improve compared with that obtained at the screening visit (perhaps suggesting a Hawthorne effect), more careful attention to providing the standard dialysis regimen as prescribed may be sufficient to achieve recommended Kt/V targets in some cases. There are limitations to our study. First, our study had a small sample size, with only 18 patients enrolled and only 16 patients finishing most of the dialysis regimens. Using paired analysis possible with the crossover design in which patients served as their own controls and acknowledging that we had data for only 16 patients for most dialysis regimens, we estimate that we had 76% power to detect a clinically significant 10point difference in the ESRD-targeted symptom domain of the KDQOL. As such, we believe our study was powered to detect a quality-of-life difference of moderate size, although we cannot exclude the possibility that a smaller (but still clinically relevant) difference exists.28 However, we believe this is unlikely because there was no apparent trend to improvement in quality of life with the 3 augmented dialysis regimens. Second, the intent of the study is to determine the impact of augmented dialysis regimens on quality of life in underdialyzed patients. One could argue that the greater than expected baseline Kt/V means that our study predominantly included patients who were already meeting KDOQI targets. However, at the time these patients were enrolled, they were underdialyzed according to KDOQI criteria and thus were potential candidates for intensification of their dialysis regimen on clinical grounds. As such, we believe our results reflect the situation faced by clinicians in an everyday practice setting. Another limitation is that our study periods were only 6 weeks in duration. It is possible that a longer period would be required to impact on quality of life, although previous studies suggested that HRQL could change during the course of only 4 weeks in patients receiving increased dialysis adequacy.14 Finally, patients using 2 dialyzers in parallel were not adequately blinded, and this might have impacted on their assessment of their HRQL. However, this would be expected to have favorably influenced perceptions of HRQL and therefore is unlikely to have affected our conclusions given that we did not detect improvement in
802
Wang et al
HRQL for patients receiving treatment with 2 dialyzers in parallel. When considering whether to use 2 dialyzers in parallel, it is important to note that its use was associated with a treatment-related serious adverse event in this small study. This may have been caused by unfamiliarity with this technique by staff, although it occurred more than 1 year after initiating the study. In conclusion, large hemodialysis patients commonly fail to achieve KDOQI targets for smallsolute clearance, which may lead to a change in dialysis prescription. Although we found that use of augmented dialysis regimens can increase delivered Kt/V, we found no evidence that quality of life improved with such regimens. Our findings suggest that extended dialysis time, increased dialysate flow, or the use of 2 dialyzers in parallel provide no benefit for large asymptomatic patients who are having difficulty achieving delivered Kt/V greater than 1.2 with standard hemodialysis regimens, and in the absence of overt uremic symptoms, augmentation of the dialysis regimen is not required for these patients.
ACKNOWLEDGEMENTS Support: This study was sponsored by the Kidney Foundation of Canada. Financial Disclosure: None.
REFERENCES 1. Kramer HJ, Saranathan A, Luke A, et al: Increasing body mass index and obesity in the incident ESRD population. J Am Soc Nephrol 17:1453-1459, 2006 2. National Kidney Foundation: KDOQI Clinical Practice Guidelines for Hemodialysis Adequacy: Update 2006. Am J Kidney Disease 48:S3-S29, 2006 (suppl 1) 3. Elangovan L, Shinaberger CS, Kraut JA, et al: HEMO equilibrated Kt/V goals are difficult to achieve in large male patients. ASAIO J 47:235-239, 2001 4. Ifudu O, Mayers JD, Matthew JJ, et al: Standardized hemodialysis prescriptions promote inadequate treatment in patients with large body mass. Ann Intern Med 128:451-454, 1998 5. Manns BJ, Johnson JA, Taub K, et al: Dialysis adequacy and health related quality of life in hemodialysis patients. ASAIO J 48:565-569, 2002 6. Wolfe RA, Ashby VB, Daugirdas JT, et al: Body size, dose of hemodialysis, and mortality. Am J Kidney Dis 35:80-88, 2000 7. Owen WF, Lew NL, Liu Y, et al: The urea reduction ratio and serum albumin concentration as predictors of mortality in patients undergoing hemodialysis. N Engl J Med 329:1001-1106, 1993
8. Bloembergen WE, Stannard DC, Port FK, et al: Relationship of dose of hemodialysis and cause-specific mortality. Kidney Int 50:557-565, 1996 9. Held PJ, Port FK, Wolfe RA, et al: The dose of hemodialysis and patient mortality. Kidney Int 50:550-556, 1996 10. Salahudeen AK: Obesity and survival on dialysis. Am J Kidney Dis 41:925-932, 2003 11. Salahudeen AK, Fleischmann EH, Bower JD: Impact of lower delivered Kt/V on the survival of overweight patients on hemodialysis. Kidney Int 56:2254-2259, 1999 12. Eknoyan G, Beck GJ, Cheung AK, et al: Effect of dialysis dose and membrane flux in maintenance hemodialysis. N Engl J Med 347:2010-2019, 2002 13. Unruh M, Benz R, Greene T, et al: Effects of hemodialysis dose and membrane flux on health-related quality of life in the HEMO Study. Kidney Int 66:355-366, 2004 14. Powers KM, Wilkowski MJ, Helmandollar AW, et al: Improved urea reduction ratio and Kt/V in large hemodialysis patients using two dialyzers in parallel. Am J Kidney Dis 35:266-274, 2003 15. Hauk M, Kuhlmann MK, Riegel W, et al: In vivo effects of dialysate flow rate on Kt/V in maintenance hemodialysis patients. Am J Kidney Dis 35:105-111, 2000 16. Ouseph R, Ward RA: Increasing dialysate flow rate increases dialyzer urea mass transfer-area coefficients during clinical use. Am J Kidney Dis 37:316-320, 2001 17. Fritz BA, Doss S, McCann LM, Wrone EM: A comparison of dual dialyzers in parallel and series to improve urea clearance in large hemodialysis patients. Am J Kidney Dis 41:1008-1015, 2003 18. Sridhar NR, Hurst C, Hayes P: Tandem dialyzers with two monitors to meet target Kt/V. ASAIO J 51:65-69, 2005 19. Manns BJ, Taub KJ. The economics of end-stage renal disease care in Canada: Incentives and impact on delivery of care. Int J Health Care Finance Econ 7:149-169, 2007 20. Canadian Coordinating Office for Health Technology Assessment: Guidelines for Economic Evaluation of Pharmaceuticals: Canada. Ottawa, Canada, The Canadian Coordinating Office for Health Technology Assessment, 2006 21. Hays RD, Kallich JD, Mapes DL, et al: Kidney Disease Quality of Life Short Form, version 1.3: A Manual for Use and Scoring. Report. Santa Monica, CA, RAND, 1995 22. Hays RD, Kallich JD, Mapes DL, et al: Development of the Kidney Disease Quality of Life (KDQOL) instrument. Qual Life Res 3:329-338, 1994 23. Korevaar JC, Jansen MA, Merkus MP, et al: Quality of life in predialysis end-stage renal disease patients at the initiation of dialysis therapy. The NECOSAD Study Group. Perit Dial Int 20:69-75, 2000 24. Hays RD, Sherbourne C, Mazel R, et al: User’s Manual for Medical Outcomes Study (MOS) Core Measures of HealthRelated Quality of Life. Santa Monica, CA, RAND, 1995 25. McHorney CA, Ware JE Jr, Lu JF, et al: The MOS 36-Item Short-Form Health Survey (SF-36): III. Tests of data quality, scaling assumptions, and reliability across diverse patient groups. Med Care 32:40-66, 1994 26. Furlong WJ, Feeny DH, Torrance GW, et al: The Health Utilities Index (HUI) system for assessing health-related quality of life in clinical studies. Ann Med 33:375-384, 2001
Dialysis Adequacy and HRQL
803
27. Norman GSJ, Wyrwich KTW. Interpretation of changes in health-related quality of life: The remarkable universality of half a standard deviation. Med Care 41:582592, 2003 28. Wyrwich KTW, Babu AN, Kroenke K, et al: A comparison of clinically important differences in health-
related quality of life for patients with chronic lung disease, asthma, or heart disease. Health Serv Res 40:577-591, 2005 29. McCallagh P, Nelder JA: Generalized Linear Models (ed 2). London, UK, Chapman & Hall, 1989 30. Nelder JA, Wedderburn RWM: Generalized linear models. J R Stat Soc 135:370-384, 1972
APPENDIX. DIFFERENCES IN ALL REMAINING HRQL MEASURES BY DIALYSIS MODALITY HRQL
Standard HD (95% CI)
KDQOL kidney disease-targeted scales Symptoms 75.9 (70.7-81.2) Effects of kidney disease 61.4 (53.2-69.2) Burden of kidney disease 36.5 (25.3-47.7) Work status 18.8 (0.8-36.7) Cognitive function 76.1 (61.7-90.5) Quality of social interaction 66.4 (59.1-73.7) Sexual dysfunction 67.5 (50.4-84.5) Sleep 52.2 (41.2-63.2) Social support 80.1 (70.1-90.1) Dialysis staff encouragement 86.3 (78.8-93.8) Patient satisfaction 79.4 (66.4-92.3) SF-36 Health Scale Physical Function 48.5 (32.9-64.0) Role–Physical 35.1 (14.2-55.9) Pain 61.3 (46.8-75.8) General Health Perception 41.4 (32.7-50.2) Emotional Well-Being 69.3 (60.8-77.8) Role–Emotional 63.3 (43.1-83.6) Social Function 66.2 (51.9-80.6) Energy/Fatigue 44.8 (29.9-59.7)
Increased Dialysate Flow (95% CI)
Increased Dialysis Time (95% CI)
2 Dialyzers in Parallel (95% CI)
Overall P
70.7 (65.3-76.0) 53.7 (45.3-62.1) 40.0 (28.6-51.3) 18.2 (0-38.1) 82.8 (68.2-97.3) 70.0 (62.5-77.4) 66.4 (49.4-83.5) 49.9 (38.6-61.1) 78.0 (67.9-88.2) 82.8 (75.2-90.5) 77.4 (64.2-90.5)
79.2 (72.6-85.8) 62.2 (52.5-71.8) 43.5(31.2-55.9) 20.0 (0-46.1) 87.3 (72.2-102.4) 73.5 (64.9-82.2) 67.1 (50.1-84.2) 48.3 (35.4-61.1) 78.8 (66.2-91.4) 90.2 (81.6-98.8) 77.3 (62.8-91.8)
78.7 (72.8-84.6) 64.2 (55.4-73.1) 40.5 (28.8-52.2) 31.3 (0-57.0) 79.2 (64.5-93.9) 73.7 (65.7-81.6) 66.5 (49.4-83.5) 53.8 (42.0-65.6) 79.8 (68.6-91.0) 84.7 (76.7-92.8) 79.5 (65.8-93.1)
0.1 0.1 0.6 0.6 0.1 0.3 0.3 0.9 1.0 0.3 1.0
44.3 (28.6-60.0) 25.1 (3.7-46.4) 53.6 (38.9-68.4) 44.2 (35.3-53.2) 67.7 (59.0-76.4) 51.0 (30.3-71.7) 55.1 (40.5-69.7) 46.3 (31.2-61.3)
43.3 (26.7-59.9) 55.4 (30.5-80.2) 71.0 (54.8-87.1) 43.9 (33.5-54.3) 80.4 (70.6-90.1) 89.1 (64.7-113.5) 69.7 (52.9-86.4) 50.7 (34.8-66.6)
44.7 (28.7-60.7) 55.6 (33.0-78.2) 58.4 (43.1-73.6) 45.0 (35.6-54.5) 72.5 (63.4-81.6) 63.6 (41.5-85.7) 66.5 (51.1-81.9) 48.8 (33.4-64.1)
0.8 0.04 0.1 0.9 0.04 0.05 0.1 0.7
Note: Possible ranges for KDQOL-SF and SF-36 dimensions are 0 to 100; in each case, higher score represents better HRQL. Abbreviations: HRQL, health-related quality of life; CI, confidence interval; HD, hemodialysis; KDQOL-SF, Kidney Disease Quality of Life Short Form; SF-36, 36-Item Short Form Health Survey. *Standard HD, 4 hours of HD and ½ hour of ultrafiltration; Increased dialysate flow, 4 hours of HD with increased dialysate flow and ½ hour of ultrafiltration; Extended dialysis time, 4.5 hours of HD; and 2 dialyzers in parallel, 4 hours of HD with 2 dialyzers in parallel and ½ hour of ultrafiltration.