Canadian Society of Nephrology Guidelines for the Management of Patients With ESRD Treated With Intensive Hemodialysis

Transcription

1 CSN Guideline Canadian Society of Nephrology Guidelines for the Management of Patients With ESRD Treated With Intensive Hemodialysis Gihad E. Nesrallah, MD, 1 * Reem A. Mustafa, MD, MPH, 2 * Jennifer MacRae, MD, 3 Robert P. Pauly, MD, 4 David N. Perkins, MD, 5 Azim Gangji, MD, MSc, 6 Jean-Philippe Rioux, MD, 7 Andrew Steele, MD, 8 Rita S. Suri, MD, 9 Christopher T. Chan, MD, 10 Michael Copland, MD, 11 Paul Komenda, MD, 12 Philip A. McFarlane, MD, PhD, 13 Andreas Pierratos, MD, 14 Robert Lindsay, MD, 15 and Deborah L. Zimmerman, MD, MSc 16 * Intensive (longer and more frequent) hemodialysis has emerged as an alternative to conventional hemodialysis for the treatment of patients with end-stage renal disease. However, given the differences in dialysis delivery and models of care associated with intensive dialysis, alternative approaches to patient management may be required. The purpose of this work was to develop a clinical practice guideline for the Canadian Society of Nephrology. We applied the GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach for guideline development and performed targeted systematic reviews and metaanalysis (when appropriate) to address prioritized clinical management questions. We included studies addressing the treatment of patients with end-stage renal disease with short daily ( 5 days per week, 3 hours per session), long (3-4 days per week, 5.5 hours per session), or long-frequent ( 5 days per week, 5.5 hours per session) hemodialysis. We included clinical trials and observational studies with or without a control arm (1990 and later). Based on a prioritization exercise, 6 interventions of interest included optimal vascular access type, buttonhole cannulation, antimicrobial prophylaxis for buttonhole cannulation, closed connector devices, and dialysate calcium and dialysate phosphate additives for patients receiving intensive hemodialysis. We developed 6 recommendations addressing the interventions of interest. Overall quality of the evidence was very low and all recommendations were conditional. We provide detailed commentaries to guide in shared decision making. The main limitation was the very low overall quality of evidence that precluded strong recommendations. Most included studies were small single-arm observational studies. Three randomized controlled trials were applicable, but provided only indirect evidence. Published information for patient values and preference was lacking. In conclusion, we provide 6 recommendations for the practice of intensive hemodialysis. However, due to very low-quality evidence, all recommendations were conditional. We therefore also highlight priorities for future research. Am J Kidney Dis. 62(1): by the National Kidney Foundation, Inc. BACKGROUND Longer and more frequent ( intensive ) hemodialysis has emerged as desirable alternative treatment strategies for patients with end-stage renal disease (ESRD). Although there is significant variability in prescription practices, 3 pragmatically defined categories encompass the majority of intensive dialysis prescriptions used in Canada and elsewhere: short daily ( 3.0 hours, 5-7 days per week), long (often nocturnal; 5.5 hours, 3-4 sessions per week), and long-frequent (usually nocturnal; 5.5 hours, 5-7 sessions per week) hemodialysis. Intensive hemodialysis has been associated with improvements across a wide range of physiologic and patient-important outcomes. 1 A number of studies, predominately of an observational design, have sug- From the 1 Division of Nephrology, Humber River Regional Hospital, University of Western Ontario, Toronto; 2 Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario; 3 Division of Nephrology, Foothills Hospital, University of Calgary, Calgary; 4 Division of Nephrology and Transplant Immunology, University of Alberta, Edmonton, Alberta; 5 Division of Nephrology, Credit Valley Hospital and Trillium Centre, Credit Valley; 6 Division of Nephrology, McMaster University, Hamilton, Ontario; 7 Division of Nephrology, Hopital du Sacre-Coeur de Montreal, Montreal, Quebec; 8 Division of Nephrology, Lakeridge Health, Queens University, Oshawa; 9 Kidney Clinical Research Unit, Division of Nephrology, University of Western Ontario, London; 10 Division of Nephrology, University Health Network, University of Toronto, Toronto, Ontario; 11 Division of Nephrology, Vancouver Hospital and Health Sciences Centre, University of British Columbia, Vancouver, British Columbia; 12 Division of Nephrology, Seven Oaks Hospital, University of Winnipeg, Winnipeg, Manitoba; 13 Division of Nephrology, St Michaels Hospital, University of Toronto; 14 Division of Nephrology, Humber River Regional Hospital, University of Toronto, Toronto; 15 Kidney Clinical Research Unit, Division of Nephrology, University of Western Ontario, London; 16 Kidney Research Centre of the Ottawa Hospital Research Institute, Division of Nephrology, University of Ottawa, Ottawa, Ontario, Canada. * G.E.N., R.A.M., and D.L.Z. contributed equally to this work. Originally published online April 15, Address correspondence to Deborah L. Zimmerman, MD, MSc, Ottawa Hospital Riverside Campus, 1967 Riverside Dr, Ottawa, ONK1H 7W9, Canada. dzimmerman@toh.on.ca 2013 by the National Kidney Foundation, Inc /$ Am J Kidney Dis. 2013;62(1):

2 Nesrallah et al Box 1. Summary of Recommendations for Clinicians These recommendations apply to adult ( 18 years) patients with ESRD receiving intensive hemodialysis: 1. For adult ESRD patients receiving intensive home hemodialysis with an AVF, we suggest the use of rope-ladder cannulation over buttonhole cannulation unless topical antimicrobial prophylaxis is used (See Recommendation 2). (Conditional recommendation; very low-quality evidence QŒŒŒ). 2. For adult ESRD patients using buttonhole cannulation for intensive home hemodialysis, we suggest the use of mupirocin antibacterial cream to reduce the risk of infection. (Conditional recommendation; very low-quality evidence QŒŒŒ). 3. For adult ESRD patients receiving intensive hemodialysis we suggest the use of arteriovenous access (AVF or AVG) over tunneled CVC for vascular access. (Conditional recommendation; very low-quality evidence QŒŒŒ). 4. For adult ESRD patients receiving intensive hemodialysis using a CVC for access we suggest the use of closed connector devices over usual care. (Conditional recommendation, very low-quality evidence QŒŒŒ). 5. For adult ESRD patients treated with long or long-frequent hemodialysis, we suggest using a dialysate calcium of 1.50 mmol/l or higher, to maintain a neutral or positive calcium balance, while avoiding predialysis hypercalcemia and oversuppression of PTH. (Conditional recommendation, very low-quality evidence QŒŒŒ). 6. For adult ESRD patients treated with long or long-frequent hemodialysis, we suggest using a phosphate dialysate additive to maintain the predialysis phosphate in the normal range if hypophosphatemia persists after stopping phosphate binders and liberalizing the diet. (Conditional recommendation, very low-quality evidence QŒŒŒ). Note: For the purpose of this CPG, intensive hemodialysis refers to any of the following: (1) short daily ( 3 hours, 5-7 d/wk), (2) long (often nocturnal, 5.5 hours, 3-4 sessions per week), and (3) long-frequent (usually nocturnal, 5.5 hours, 5-7 sessions per week) hemodialysis. The majority of studies used to inform these CPGs were small, observational, lacked parallel control groups, and included case reports and case series. Very few published randomized controlled trials were applicable and only, when available, provided indirect evidence. Published information on patient values and preference was lacking. The quality of the body of evidence and the strength of the recommendations were graded according to GRADE (Grading of Recommendations Assessment, Development and Evaluation approach). 11,12 Quality of evidence denoted as follows: QŒŒŒ, very low (equivalent to grade D in other systems); QQŒŒ, low (grade C); QQQŒ, moderate (grade B); QQQQ, high (grade A). The strength of the recommendation (conditional or strong) is based on the committee s confidence that adherence to the recommendation will do more good than harm. The strength of recommendation incorporates the quality of the supporting body of evidence, balance between benefits and harms, patients values and preferences, and, when appropriate, resource implications. For the purposes of this CPG, we were unable to identify any studies that directly evaluated patients values and preferences; instead they were inferred based on the collective experience of the CPG panel. Although the CPG panel discussed the resource implications of the CPGs, the final recommendations were not affected by cost. A strong recommendation (phrased as we recommend... ) implies that most patients in a given situation would want the recommended course of action and only a small proportion would not. A conditional recommendation (phrased as we suggest... ) implies that the majority of patients would still wish to follow the recommendation but many patients would not, such that clinicians will need to be prepared to assist patients in making decisions that are consistent with their values and preferences. Abbreviations: AVF, arteriovenous fistula; AVG, arteriovenous graft; CVC, central venous catheter; CPG, clinical practice guideline; ESRD, end-stage renal disease; PTH, parathyroid hormone. gested improved quality of life and patient survival with longer, more frequent hemodialysis. 2,3 A recent randomized controlled trial demonstrated improved quality of life and regression of left ventricular hypertrophy with short daily hemodialysis. 1 Other outcomes improved by more intensive hemodialysis include blood pressure control and a reduction in intradialytic hypotension. 4-6 Enhanced clearance of several solutes, including urea, 2 -microglobulin, and phosphate, has also been reported. 7,8 Patients are able to discontinue fluid, sodium, and phosphate restrictions (including phosphate binders). In addition, intensive dialysis has been shown to reduce overall cost in North American health care systems. 9,10 However, given differences in dialysis physiology and models of care delivery, intensive hemodialysis regimens require different approaches compared to conventional hemodialysis. Domains of interest include patient selection criteria, vascular access type, cannulation techniques, dialysate composition, and mineral metabolism management. In a recently conducted needs assessment survey (unpublished work), members of the Canadian Society of Nephrology (CSN) identified clinical practice guidelines related to more intensive hemodialysis as a priority. PURPOSE OF THE GUIDELINES Our objective was to develop a clinically useful guideline (summarized in Box 1) to assist physicians and allied health care practitioners in the management of patients with ESRD who have chosen more intensive hemodialysis as their treatment modality; it is not intended to replace clinical judgment. GUIDELINE PANEL COMPOSITION To form the guideline panel (Box 2), physicians with an established clinical interest in intensive hemodialysis prescriptions were identified with representation from across Canada. Many of these physicians also have a background in clinical investigation, research methods, guideline development, and knowledge translation. 188 Am J Kidney Dis. 2013;62(1):

3 CSN Guidelines Box 2. Working Group Membership Chair of the Guidelines Group and the Mineral Metabolism Group: Deborah Zimmerman Chair of the Vascular Access Group: Reem A. Mustafa Chair of the Buttonhole Cannulation Group: Gihad E. Nesrallah Members of the Mineral Metabolism Group: Christopher T. Chan, Michael Copland, Paul Komenda, Philip A. McFarlane Members of the Vascular Access Group: Azim Gangji, Jean-Philippe Rioux, Andrew Steele, Rita S. Suri Members of the Buttonhole Cannulation Group: Jennifer MacRae, Robert P. Pauly, David N. Perkins Final Review: Andreas Pierratos, Robert Lindsay, Brenda Hemmelgarn, and the CSN Clinical Practice Guideline committee METHODS Target Population and Interventions Intensive hemodialysis was defined as any hemodialysis schedule that included an increase in frequency and/or an increase in session duration compared to conventional hemodialysis (ie, 3 times per week, 3-5 hours per session). Short daily hemodialysis was defined as 5 or more dialysis sessions per week with fewer than 3 hours per session. Long hemodialysis was defined as greater than or equal to 5.5 hours per session, 3-4 times per week, and long-frequent hemodialysis was defined as greater than or equal to 5.5 hours per session for 5 or more sessions per week. Guideline Development Methods The CSN has adopted the GRADE (Grading of Recommendations Assessment, Development and Evaluation; org) approach in evaluating the quality of evidence and developing recommendations. 11,12 Accordingly, we used the GRADE approach in developing this guideline. Initially, the panel members developed questions of interest in this area using PICO (population, intervention, comparison, and outcome) format. 13 The questions covered 4 main areas, including patient selection, vascular access, dialysis prescription, and mineral metabolism. Panel members then rated the questions using a unipolar 9-point adjectival scale and retained the top-ranking questions (Item S1, available as online supplementary material) to be addressed in this clinical practice guideline. The panel also developed an exhaustive list of outcomes that should be considered when addressing each of the questions. Each outcome was rated from 1-9 based on their importance to patients, with only the important (score 4-6) and critical (score 7-9) outcomes for decision making retained for the guideline development (Item S2). Given the broad range of topics identified, the panel elected to focus the first guideline document on the 6 highest ranking questions, which focused specifically on aspects of clinical management. An experienced health information specialist conducted a large pragmatic literature search encompassing all 6 domains. The panel was divided into 3 subcommittees (each addressing 2 related questions) based on interest and to ensure balance in clinical and methodological expertise. After completion of the systematic reviews, 14,15 we summarized the evidence and rated the quality of evidence based on the 5 domains of GRADE: risk of bias, imprecision, indirectness, inconsistency, and publication bias. 16 We summarized the available body of evidence, its quality, and judgments about each of the domains in GRADE evidence profile tables using GRADE Profiler software. 17 After summarizing the evidence in GRADE evidence profiles, each subcommittee drafted an evidence to recommendations table, which summarized the overall quality of the body of evidence, balance between risks and benefits, and patients values and preferences. Although resource implications can also be considered in the GRADE framework, the guideline committee decided a priori that this would not be considered in this iteration of the guidelines. The final draft of the guideline was reviewed by 2 clinical content experts (A.P. and R.L.) who were not involved in the data abstraction and recommendations development phases. After approval by the entire guidelines panel, the document was submitted to the CSN Clinical Practice Guidelines Committee for external peer review and approval. Knowledge Translation and Monitoring Strategy The Canadian Kidney Knowledge Translation and Generation Network (CANN-NET) Knowledge Translation (KT) Committee will develop an integrated knowledge translation and communication strategy for this guideline, based on the priorities of and with input from CANN-NET knowledge users (heads of renal programs across Canada). In conjunction with the Canadian Organ Replacement Register (CORR), the CANN-NET KT group will also develop and implement a monitoring strategy that will evaluate the effectiveness of knowledge translation interventions that will accompany these recommendations. At the time of this writing, modifications to the CORR database are underway, which will enable relevant indicators to be tracked longitudinally. More information about CANN-NET KT initiatives can be found at RECOMMENDATIONS Recommendation 1 Recommendation 1 is provided in Box 3. Background Buttonhole or same-site cannulation has been in use since the 1970s 18 and has recently gained popularity among home hemodialysis programs. While the conventional rope-ladder or rotating-site cannulation technique allows for skin healing and new puncture sites with consecutive cannulation procedures, the buttonhole technique results in fibrotic track (buttonhole site) formation, which ultimately allows for conversion to blunt needles. Self-cannulation with the buttonhole technique is an important theoretical advantage that has likely contributed to the adoption of the technique by hemodialysis programs and the publication of enthusiastic anecdotal reports supporting its use Cannulation by a nurse in-center differs technically from self-cannulation at home. Differences in operator and treatment frequency, for example, may result in different outcomes. Because there were no studies that specifically addressed buttonhole cannulation in intensive in-center hemodialysis and the technique differs from buttonhole cannulation at home, this guideline did not address buttonhole cannulation in-center. Summary of Findings No randomized controlled trials were available to address this recommendation directly. The quality of Am J Kidney Dis. 2013;62(1):

4 Nesrallah et al Box 3. Recommendation 1 For adult ESRD patients receiving intensive home hemodialysis with an AVF, we suggest the use of rope-ladder cannulation over buttonhole cannulation, unless topical antimicrobial prophylaxis is used. (Conditional recommendation; very low-quality evidence QŒŒŒ). Underlying Values and Preferences: This recommendation places a high value on preventing systemic infections with buttonhole cannulation and a relatively low value on the avoidance of repeated needle cannulation and potentially shorter training times for home hemodialysis. Remarks: This recommendation pertains specifically to home intensive hemodialysis patients and does not address buttonhole cannulation in in-center intensive hemodialysis patients (see text). Patients with short usable segments, aneurysmal dilatation, or those unable to self-cannulate using rope-ladder cannulation may benefit from buttonhole cannulation. When buttonhole cannulation is used, we suggest using it in conjunction with topical antimicrobial prophylaxis (See Recommendation 2). Due to uncertainty regarding infection risk with buttonhole cannulation, clinicians should engage patients in shared decision-making and obtain informed consent prior to adopting this cannulation technique. Note: See Box 1 for explanation of the grading of the quality of the body of evidence and the strength of the recommendations. the available evidence from observational studies was very low. The absolute risk of Staphylococcus aureus bacteremia with buttonhole cannulation for patients receiving intensive home hemodialysis ranged from episodes per 1,000 patient-days across 4 studies However, in an indirect comparison between intensive home hemodialysis with buttonhole cannulation versus in-center conventional hemodialysis with rope-ladder cannulation, the relative risk for bacteremia ranged from in a single-center study of 56 consecutive home hemodialysis patients using buttonhole cannulation. Importantly, the majority of buttonhole-related systemic infections in this home intensive hemodialysis cohort were with S aureus, which was associated with metastatic complications including septic arthritis, septic pulmonary emboli, vertebral osteomyelitis, and death. 26 Increased rates of bacteremia and metastatic infection have also been reported with buttonhole cannulation in patients undergoing conventional in-center hemodialysis. 27,28 Benefits Ease of cannulation and reduced pain have been anecdotally reported as potential benefits of the buttonhole (vs rope-ladder) method. 29,30 However, we were unable to find any literature specific to intensive hemodialysis patients to support this assertion. Given the potential differences between self-cannulation at home and buttonhole cannulation by a nurse in-center, we did not attempt to draw inferences related to cannulation pain from the in-center hemodialysis literature. It is noteworthy that one study of 33 patients failed to detect a difference in pain scores using a visual analogue scale among patients converting from rope-ladder to buttonhole cannulation for home conventional hemodialysis. 31 Two studies in conventional hemodialysis populations suggested that buttonhole cannulation resulted in less aneurysm formation or even healing of aneurysmal segments. 27,32 No similar study has been reported in home intensive hemodialysis patients. We found no published data addressing the impact of buttonhole cannulation on home dialysis training time. Our collective experience suggests that buttonhole cannulation may be of value to some patients who must self-cannulate, and that the adoption of the buttonhole technique may enable such patients to undergo dialysis at home instead of in-center or with a catheter. However, we found no published evidence to support this either. Harms Although the absolute risk of bacteremia with buttonhole cannulation is less than 0.3 episodes per patient-year, the large relative risk compared with rope-ladder cannulation is of major concern and is comparable to bacteremia rates observed with tunneled central venous dialysis catheters. 33 Conclusions and Research Needs Buttonhole cannulation, particularly in the absence of topical antimicrobial prophylaxis, is associated with an increased risk of systemic infection that may have devastating consequences. Uncontrolled studies suggest that buttonhole cannulation can restore function in structurally abnormal fistulas and promotes self-cannulation, which may facilitate home therapy. For patients who continue to use the buttonhole technique as their preferred form of cannulation, studies to identify mechanisms to reduce the risk of infection are needed. Studies properly assessing patients values and preferences, impact on quality of life, and complexity of training with buttonhole versus rope-ladder cannulation are also needed. Such studies should consider potential differences between treatment settings (home vs center) and operator (self vs health care professional) in affecting outcomes with buttonhole cannulation. Recommendation 2 Recommendation 2 is provided in Box 4. Background With the popularization of buttonhole cannulation over the last decade, there have been several studies 190 Am J Kidney Dis. 2013;62(1):

5 CSN Guidelines suggesting an increased risk of bacteremia with this technique. 27,28 Even when a rigorous infection prevention program is adopted for in-center hemodialysis patients using buttonhole cannulation that did not include antimicrobial prophylaxis, the infection risk remains elevated compared to rope-ladder cannulation. 28 Topical antimicrobial prophylaxis has demonstrated efficacy in reducing episodes of catheterrelated bacteremia and, to some extent, peritoneal dialysis catheter related infections. Hence, this approach may be of benefit to patients using buttonhole cannulation. Summary of Findings Only one study was identified that specifically addressed this recommendation. This study compared topical antimicrobial prophylaxis with none in patients undergoing long and long-frequent hemodialysis at home with buttonhole cannulation. 26 No infections were observed in any patient using the intervention, while a total of 12 (odds ratio, 35; 95% confidence interval [CI], 2-627) were observed in patients not using the intervention. When analyzed by intention to treat, the odds ratio was 6.4 (95% CI, ) because 2 infections occurring during the treatment period in non treatment-adherent patients were attributed to the intervention. Study quality was limited by pre-post design. The study did not evaluate other important or critical outcomes. Benefits Mupirocin applied to buttonhole cannulation sites posthemodialysis appears to significantly reduce the Box 4. Recommendation 2 For adult ESRD patients using buttonhole cannulation for intensive home hemodialysis, we suggest the use of mupirocin antibacterial cream to reduce the risk of infection. (Conditional recommendation; very low-quality evidence QŒŒŒ). Underlying Values and Preferences: This recommendation places a high value on preventing infection associated with the buttonhole cannulation technique and a relatively low value on the potential emergence of mupirocin-resistant bacteria, patient inconvenience, and cost associated with the application of the cream. Remarks: This recommendation pertains specifically to home intensive hemodialysis patients and does not address buttonhole cannulation in in-center intensive hemodialysis patients (see text). Local factors, such as mupirocin resistance rates and institutional policies should be considered when deciding to use mupirocin cream. Other topical antimicrobial agents may provide similar efficacy, but studies directly evaluating their use in the setting of buttonhole cannulation have not been done. Note: See Box 1 for explanation of the grading of the quality of the body of evidence and the strength of the recommendations. risk of infection associated with this cannulation technique. 26 Harms There is an increased risk of mupirocin resistance that has been reported in studies of other populations. 34,35 Patients may also be inconvenienced by the additional requirement of antibacterial cream application, as well as the cost of mupirocin cream. Conclusions and Research Needs Compared with rope-ladder cannulation, buttonhole cannulation is associated with an increased risk of infection that may be reduced by the addition of mupirocin antibacterial cream. Overall quality of evidence favoring this approach was very low. A clinical trial comparing benefits and harms of buttonhole cannulation with antimicrobial cream versus ropeladder cannulation in the home hemodialysis population would be useful in informing practice. Outcomes of interest may include patient satisfaction, home dialysis training time, mechanical complications, and infection rates. Recommendation 3 Recommendation 3 is provided in Box 5. Background Arteriovenous fistulas (AVFs) and grafts (AVGs) are generally accepted as the optimal vascular access for hemodialysis. To date, these assertions have been based on studies of patients treated in-center. Given the potentially large differences in other prognostic variables, outcomes for patients who undergo dialysis at home may not vary with access type to the extent that has been observed with in-center treated populations. Moreover, the convenience associated with catheter use may be an important facilitating factor for home hemodialysis adoption and may be preferred by some patients and physicians. Finally, the impact of longer and more frequent treatment on access function is a potential concern. This was recently highlighted in the Frequent Hemodialysis Network (FHN) Daily Trial, in which an increased number of interventions were observed in the short daily arm. 1 These factors highlight the many areas of uncertainty surrounding vascular access type for home intensive hemodialysis. Summary of Findings We were unable to find any randomized controlled trials that specifically addressed this recommendation. We identified observational studies, none of which adjusted for differences in baseline prognostic characteristics among the groups using different access types; Am J Kidney Dis. 2013;62(1):

6 Nesrallah et al Box 5. Recommendation 3 For adult ESRD patients receiving intensive hemodialysis we suggest the use of arteriovenous access (AVF or AVG) over tunneled CVC for vascular access. (Conditional recommendation; very low-quality evidence QŒŒŒ). Underlying Values and Preferences: This recommendation places a high value on preventing the potential increased risk of infection with CVCs and a relatively low value on the avoidance of repeated needle cannulation and potentially shorter training times for home hemodialysis. Remarks: A strong patient preference for CVC over AVF or AVG should not be a barrier to receiving intensive hemodialysis. Note: See Box 1 for explanation of the grading of the quality of the body of evidence and the strength of the recommendations. this resulted in a serious risk of bias within this body of evidence. There is very low-quality evidence to support the reduced risk of infection with AVFs/AVGs for patients who are receiving intensive dialysis. Benefits We found no published comparisons of mortality rates with different vascular access types in patients receiving intensive hemodialysis. We identified 9 reports that evaluated the effect of access type on infections, interventions and loss in patients receiving intensive hemodialysis. 6,36-43 Bacteremia rates in AVFs/AVGs were 0-11 per 100 patient-years, while they were reported to be 0-19 per 100 patient-years for central venous catheters (CVCs) Additionally, we identified 2 studies 38,39 that reported increased rates of access loss in CVCs ( per 100 patient-years) compared with rates of access loss in AVGs (11 per 100 patient-years) 38 and AVFs (range, 2-14 per 100 patient-years). 38,40-43 We did not consider differential effects of access type on small-solute clearance in developing this recommendation. However, the increased blood flow rates that are possible with an AVF/AVG are potentially important for short daily hemodialysis, but of lesser concern for patients treated with long or long-frequent dialysis, in which the blood pump speed is typically lower. Harms Theoretically, the more frequent cannulation associated with an AVF/AVG may be associated with an increased risk of access complications or interventions. We identified 3 studies that assessed the rates of access intervention. These rates ranged from per 100 patient-years inavfs, 38,44,45 while they were reported to be per 100 patient-years in AVGs. 45 Reporting of CVC intervention rates was difficult to interpret because some studies reported access loss, but did not count catheter exchanges as interventions. This may have resulted in an underestimate of the rate of complications with CVCs in these studies ,43-47 In multiple studies, complication rates were reported without specifying the type or follow-up time, which precluded weighing benefits and harms based on patient important outcomes. Conclusions and Research Needs The available evidence supports the use of AVFs/ AVGs over the use of CVCs for intensive dialysis, but it is of very low quality. In one study of patients treated with nocturnal hemodialysis, the use of CVCs was not associated with the composite outcome of technique failure and mortality in multivariate analysis, suggesting that the patient population treated with intensive hemodialysis may not experience the same negative consequences with CVCs as patients treated with in-center hemodialysis. 48 However, more definitive studies evaluating the benefits and harms of different access types in intensive hemodialysis patients are needed. Differentiating between AVFs and AVGs in these studies will be essential. Studies properly assessing patients values and preferences about different access types are also needed to better inform clinical practice in this area. Recommendation 4 Recommendation 4 is provided in Box 6. Background In this document, we use the phrase closed connector device to describe any device or catheter extension that is designed to decrease the risk of accidental disconnection from a CVC. These devices include InterLink (BD), Tego Needlefree Hemodialysis Connector (ICU Medical, Inc), and Swan-Lock (Codan Medical Inc). These devices were originally introduced for central venous access in intensive care units and since then have been applied to different types of catheters, including hemodialysis CVCs. Summary of Findings We were unable to find any randomized controlled trials that directly addressed the comparative effectiveness of closed connector devices over standard luer lock fittings. Ouwendyk et al 37 reported 4 events in which patients inadvertently allowed a small amount of air to enter the catheter at the time of connect or disconnect. Consequently, the investigators initiated the use of InterLink closed connector device to prevent future such events. 37 There were 2 observational studies that evaluated closed connector devices in conventional hemodialysis patients. One study 49 (n 23) reported access flows 250 ml/min without any clinical evidence of thrombotic occlusions or infections. The lack of infections was supported by a second crossover study 50 in which 50 pediatric hemo- 192 Am J Kidney Dis. 2013;62(1):

7 CSN Guidelines Box 6. Recommendation 4 For adult ESRD patients receiving intensive hemodialysis using a CVC for access, we suggest the use of closed connector devices over usual care (Conditional recommendation, very low-quality evidence QŒŒŒ). Underlying Values and Preferences: We placed a high value on preventing potentially fatal air embolism and bleeding, and a relatively low value on potential risks if the device is not connected properly. Remarks: If not connected properly, closed connector devices can also lead to accidental disconnection and related complications, including air embolism and bleeding. Appropriate training and extra care should be used. Patients who have difficulty using closed connector devices (eg, limited manual dexterity) should routinely receive assistance with making connections to ensure that they are using the closed connector device correctly. Note: See Box 1 for explanation of the grading of the quality of the body of evidence and the strength of the recommendations. dialysis patients had 7.8 infections per 1,000 patientdays during the pre Tego Needlefree Hemodialysis Connector period versus 3.62 infections per 1,000 patient-days postconversion to the Tego device. We surveyed 23 Canadian program directors regarding their facility practices pertaining to the use of closed connector devices. Sixteen of 23 program directors (70%) responded. Respondent programs had been in operation from with a range of patients per program ( 1,500 patients in total). The home hemodialysis schedules included conventional, short daily, and nocturnal hemodialysis of varying frequency. The use of CVCs ranged from 0%-70% (mean, 30%). Of the 15 programs that reported having patients with CVCs for access, 3 did not use any closed connector devices. Four cases of air embolism and 2 near misses were reported in patients not using closed connector devices, while at least 2 cases of air embolism were reportedly prevented in patients using closed connector devices. Four accidental hemorrhages were reported in patients using closed connector devices; 2 of these events were likely related to improper connection of the device. The FHN Nocturnal Trial reported one death due to air embolism. 6,51 After this death, all patients receiving hemodialysis at home with CVCs were required to use needle-free device extension sets to prevent blood loss or air entry. 51 Benefits The use of closed connector devices is expected to prevent air embolism and hemorrhage in the event of accidental disconnection or improper handling of clamps when connecting or disconnecting the CVC to or from the hemodialysis circuit; our survey did not identify any episodes of air embolism in patients using closed connector devices. Although the event rates appear to be very small, the consequences may be catastrophic. Although the panel decided not to consider resource implications in these recommendations, we believe it is worth highlighting that because of the potential for reuse of closed connector devices, it is unlikely that their use is associated with greater incremental costs. Harms There is a theoretical increased risk of infection with closed connector devices. However, the single relevant study that we identified does not support an increased risk of infection. 50 Conclusions and Research Needs Closed connector devices may be associated with a decreased risk of air embolism, but we identified 2 reported cases of hemorrhage likely related to improper connection or faulty diaphragms. There is a need for well-conducted randomized trials to assess the benefits, harms, and cost-effectiveness of different closed connector devices in intensive hemodialysis patients. However, given the rare occurrence of related complications, very large and likely infeasible trials would be required. Prospective observational studies assessing the cost-effectiveness and rates of complications of closed connector devices in this patient population should therefore be undertaken. Recommendation 5 Recommendation 5 is provided in Box 7. Background Overall calcium balance for patients with ESRD is determined by intake, absorption, diffusion, and loss. Dietary intake of calcium tends to be low secondary to dietary restriction, but the use of calcium-based phosphate binders may contribute significantly to calcium intake, the absorption of which will be affected by vitamin D status. Diffusion from dialysate to the patient will occur when the dialysate calcium concentration is higher than the plasma calcium concentration. Loss is via residual renal function and ultrafiltration. All these factors must be considered in determining calcium balance. One of the reported benefits for long and long-frequent hemodialysis (compared to short daily hemodialysis) is the discontinuation of calcium-based phosphate binders, hence the guideline panel elected to focus on these hemodialysis modalities in this recommendation. Am J Kidney Dis. 2013;62(1):

8 Nesrallah et al Box 7. Recommendation 5 For adult ESRD patients treated with long or long-frequent hemodialysis, we suggest using a dialysate calcium of 1.50 mmol/l or higher, to maintain a neutral or positive calcium balance, while avoiding predialysis hypercalcemia and oversuppression of PTH. (Conditional recommendation, very low-quality evidence QŒŒŒ). Underlying Values and Preferences: This recommendation places a high value on preventing the potential adverse consequences of a negative calcium balance. Remarks: This guideline requires that clinicians also consider factors that are important in determining calcium balance, including ongoing use of calcium-based phosphate binders, frequency of hemodialysis, and ultrafiltration volumes. An increasing bone alkaline phosphatase and PTH suggest that higher dialysate calcium may be required. Although we did not identify published evidence of harm, clinicians should recognize the theoretical risks associated with prolonged positive calcium balance. Note: See Box 1 for explanation of the grading of the quality of the body of evidence and the strength of the recommendations. Summary of Findings Very few studies directly compared long or longfrequent hemodialysis with and without increased dialysate calcium; no published studies addressed any of the critical and most of the important outcomes. However, there are data, although limited, for the biochemical markers related to mineral metabolism In the study reported by Lindsay et al, 52 patients on long-frequent hemodialysis demonstrated an increase in alkaline phosphatase and parathyroid hormone (PTH) levels that was reversed by increasing the dialysate calcium concentration from 1.25 to 1.75 mmol/l. Similar results were reported for a cohort of patients receiving long hemodialysis those who were able to discontinue their calcium-based phosphate binders had an increase in alkaline phosphatase and PTH levels that subsequently responded to an increase in dialysate calcium concentration 1.25 mmol/ L. 2 These clinical results are supported by a massbalance study in which the dialysate calcium required to maintain a neutral calcium balance was found to be 1.5 mmol/l (the use of calcium-based phosphate binders was not considered in the calcium balance analysis). 55 Only one study maintained patients on long hemodialysis on a 1.25-mmol/L dialysate bath without a reported increase in PTH levels. 56 The participants in this study were able to decrease their calcium-based phosphate binders but not discontinue them entirely. For the critical outcomes (Item S2) of fragility fractures, peripheral vascular disease, and coronary artery disease, only indirect evidence (surrogate outcome) was identified. Bone mineral density (BMD) has been reported in several studies. In long-frequent hemodialysis, dialysate calcium concentration was increased from 1.5 to 1.75 mmol/l in response to a declining BMD in 11 patients. 53 Median t and z scores improved over the following year. This finding is consistent with a case report of a patient on longfrequent hemodialysis in which BMD at the femur increased and BMD at the L-spine remained stable on a dialysate calcium concentration of mmol/l and high-dose calcitriol. 57 In a study by Van Eps et al, 58 patients receiving long hemodialysis had no measurable change in BMD at the femur or L-spine over 12 months on 1.6 mmol/l of dialysate calcium compared with baseline values measured at the time of conversion from conventional hemodialysis. There was very little published information related to peripheral vascular disease among patients receiving intensive hemodialysis. In 2 different case reports of patients receiving long-frequent hemodialysis, one patient did not experience any obvious change in vascular calcification, whereas the second patient experienced resolution of iliac and tibial arterial stenosis bilaterally despite using a dialysate calcium concentration 1.5 mmol/l. 57,59 For a cohort of patients receiving long hemodialysis with 1.6 mmol/l of dialysate calcium, the combination of vascular and ectopic calcification improved in 8 patients, remained stable in 6 patients, and worsened in 2 patients after 12 months of treatment. 58 There was no information that suggests that altering the dialysate calcium concentration has an impact on coronary artery disease. However, in one study of 38 patients receiving longfrequent hemodialysis, coronary calcification did not worsen after 1 year of intensive hemodialysis treatment despite increasing dialysate calcium concentration (compared to conventional hemodialysis). 60 For the important outcomes (Item S2) of calcific uremic arteriolopathy, mortality, and intradialytic hypotension, no studies were identified that directly addressed the impact of dialysate calcium concentration. Although intradialytic hypotension was reduced with long-frequent hemodialysis on a higher dialysate calcium concentration, compared with conventional hemodialysis, 4 the major factor in hemodynamic stability may be lower ultrafiltration rates in patients treated with more frequent hemodialysis, as reported by Mahadevan et al. 56 There were 3 studies that provided information on the important outcome of parathyroidectomy. In the Bugeja et al 2 study, 2 patients on long hemodialysis underwent parathyroidectomy after being treated with a dialysate calcium concentration of 1.25 mmol/l. Two other patients, one treated with long hemodialysis and one treated with long-frequent hemodialysis, required a parathyroidectomy despite dialysate calcium concentrations 194 Am J Kidney Dis. 2013;62(1):

9 CSN Guidelines of 1.6 and 1.5 mmol/l, respectively. 53,57 Last, extraosseous calcification was improved in a patient with tumoral calcinosis despite a dialysate calcium concentration of mmol/l. 57 Benefits Increased dialysate calcium concentration appears to simplify the management of bone mineral metabolism (alkaline phosphatase and PTH) for patients on long and long-frequent hemodialysis. Similarly, an increase in dialysate calcium concentration was required (1.5 to 1.75 mmol/l) to prevent a decrease in BMD. Harms It is possible that the addition of calcium to the dialysate might result in a positive calcium balance, thereby theoretically promoting coronary artery calcification, peripheral vascular disease, and ectopic calcification, particularly in patients who continue to require calcium-based phosphate binders and vitamin D and have low weekly ultrafiltration volumes. The current literature, which consists primarily of case reports and small cohort studies, does not support the worsening of vascular or ectopic calcification. Conclusions and Research Needs For patients who are able to discontinue their calcium-based phosphate binders, a negative calcium balance is likely with long and long-frequent hemodialysis, particularly when a 1.25-mmol/L dialysate bath is used and weekly ultrafiltration rates are increased. This ultimately worsens secondary hyperparathyroidism and decreases BMD. Ultimately, the benefits and risks of dialysate calcium supplementation are related to the assessment of calcium balance. Currently, a simple method for determining a neutral calcium balance at the individual patient level does not exist, but would be expected to enhance patient care. An appropriately designed randomized controlled trial assessing the potential impact of different dialysate calcium and vitamin D management strategies is warranted. Potential outcomes of interest include fracture risk, maintenance of PTH levels, and bone histomorphometry, in addition to assessing potential harms such as coronary calcification by electronbeam computed tomography. Recommendation 6 Recommendation 6 is provided in Box 8. Background Hyperphosphatemia has been a consistent challenge for patients treated with conventional hemodialysis due to imbalance between phosphate intake and Box 8. Recommendation 6 For adult ESRD patients treated with long or long-frequent hemodialysis, we suggest using a phosphate dialysate additive to maintain the predialysis phosphate in the normal range if hypophosphatemia persists after stopping phosphate binders and liberalizing the diet. (Conditional recommendation, very low-quality evidence QŒŒŒ). Underlying Values and Preferences: This recommendation places a high value on preventing the potential adverse consequences of hypophosphatemia and a relatively low value on the potential inconvenience of adding phosphate to the dialysate. Remarks: Due to factors such as postdialysis rebound, the clinical relevance of a low postdialysis phosphate level is uncertain. However, given the uncertainty, and theoretical risk of harm with severe hypophosphatemia, members of the guideline panel currently consider supplementation when a postdialysis phosphate level drops below 0.4 to 0.7 mmol/l. We caution against phosphate supplementation for low postdialysis phosphate levels when the predialysis phosphate level is above normal. Note: See Box 1 for explanation of the grading of the quality of the body of evidence and the strength of the recommendations. removal. Increasing dialysis frequency has been shown to reduce serum phosphate levels, but increasing duration has a much greater impact on phosphate removal; for this reason the guidelines panel elected to focus on long and long-frequent hemodialysis for this recommendation. Summary of Findings No study directly addressed the benefits or harms associated with the addition of phosphate to the dialysate (compared to no phosphate additive) for a low pre- or postdialysis serum phosphate level on long or long-frequent hemodialysis. An action serum phosphate level that would prompt the use of a dialysate phosphate additive was also not defined in the included studies. It was also impossible to separate the relative effects of calcium versus phosphate manipulation on many of the reported mineral metabolism related outcomes. The prescription of higher dialysate calcium concentration was applied in a more universal fashion (across study populations) compared to phosphate additives, which were applied to a minority of patients. For this reason, the reporting of outcomes that might be related to calcium and phosphate such as vascular disease are cited for Recommendation 5. Studies that specifically addressed phosphate-related measures are summarized next. No direct evidence was available for critical and important outcomes related to phosphate supplementation. These outcomes included fragility fractures (which could result from osteomalacia due to phosphate depletion), coronary artery disease, peripheral vascular disease, calcific uremic arteriolopathy, mor- Am J Kidney Dis. 2013;62(1):

10 Nesrallah et al tality, or extraosseous calcification. Theoretically, one might expect vascular disease or ectopic calcification to stabilize or improve because hyperphosphatemia is improved by more intensive hemodialysis. This might occur as a direct consequence of the reduced phosphate and secondarily due to a reduction in calciumbased phosphate binders. However, the impact of maintaining phosphate levels in the normal range on these outcomes is unclear. Despite the lack of information about the critical and many important outcome measures, there is consistent reporting of a reduction in serum phosphate levels and the discontinuation of phosphate binders with more intensive hemodialysis despite an increase in dietary phosphate. 8,61-63 The observed decrease in serum phosphate levels on long-frequent hemodialysis was 0.36 mmol/l (2 randomized controlled trials) 4,64 to 0.5 mmol/l (6 observational studies) 8,52,53,60,61,63 compared to 0.45 mmol/l on long hemodialysis (6 observational studies). 2,58,65-68 In the observational studies of long hemodialysis, 40% of patients still require phosphate binders. 2,56,58,67 These findings are consistent with a study by Kjellstrand et al 69 in which the amount of phosphate removed per week was related to hours and frequency of dialysis. Benefits Based on our current understanding of vascular smooth muscle cell biology, correction of hyperphosphatemia is expected to impact on progressive vascular calcification. 70 However, hypophosphatemia and phosphate depletion have negative health consequences in the general population that might also be problematic for people with ESRD, such as osteomalacia and proximal myopathy. The addition of phosphate to the dialysate is intended to prevent these potential complications. However, we were unable to find any studies to support a reduction in fragility fractures that might be expected with osteomalacia or an improvement in muscle mass. Van Eps et al 71 reported an improvement in the 6-minute walk test in patients on long dialysis, but the role of phosphate balance on this outcome was not directly studied. Harms Without appropriate monitoring of serum phosphate levels, hyperphosphatemia may result from dialysate phosphate supplementation, with all of the negative health consequences that have been documented in the conventional hemodialysis literature. Additionally, there is a theoretical risk that the addition of phosphate to the dialysate may precipitate other dialysate constituents; however, this is not supported by a technical study of dialysate composition after the addition of phosphate by Su et al. 72 Furthermore, phosphate is added as sodium phosphate, which may increase the dialysate sodium concentration, resulting in a positive sodium balance, thirst, and hypertension. However, there was no information in the literature to support or refute this. Conclusions and Research Needs Long and long-frequent hemodialysis are associated with a reduction in serum phosphate levels despite an increase in dietary phosphate intake and a reduction or cessation of phosphate binders. A reduction in serum phosphate levels and phosphate depletion may ensue, with associated negative health consequences. An optimal pre- and postdialysis serum phosphate level likely exists, but is not currently known. A first attempt to define these levels will need to be based on large observational studies. ACKNOWLEDGEMENTS Support: A grant was provided by the CSN to support this guideline development, and this work was funded in part by a Canadian Institutes of Health Research Meeting Planning and Dissemination Grant FRN # Financial Disclosure: The authors declare that they have no relevant financial interests. SUPPLEMENTARY MATERIAL Item S1. Questions to be addressed. Item S2. Outcomes. Note: The supplementary material accompanying this article ( is available at REFERENCES 1. Chertow GM, Levin NW, Beck GJ, et al. In-center hemodialysis six times per week versus three times per week. N Engl J Med. 2010;363(24): Bugeja A, Dacouris N, Thomas A, et al. In-center nocturnal hemodialysis: another option in the management of chronic kidney disease. Clin J Am Soc Nephrol. 2009;4(4): Nesrallah GE, Lindsay RM, Cuerden MS, et al. Intensive hemodialysis associates with improved survival compared with conventional hemodialysis. J Am Soc Nephrol. 2012;23(4): Rocco MV, Lockridge RS Jr, Beck GJ, et al. The effects of frequent nocturnal home hemodialysis: the Frequent Hemodialysis Network Nocturnal Trial. Kidney Int. 2011;80(10): Nesrallah G, Suri R, Moist L, Kortas C, Lindsay RM. Volume control and blood pressure management in patients undergoing quotidian hemodialysis. Am J Kidney Dis. 2003;42(suppl 1): Traeger J, Galland G, Delawari E, Arkouche W, Hadden R. Six years experience with short daily hemodialysis: Do the early improvements persist in the mid and long term? Hemodialysis Int. 2004;8: Raj DS, Ouwendyk M, Francoeur R, Pierratos A. Beta(2)- microglobulin kinetics in nocturnal haemodialysis. Nephrol Dial Transplant. 2000;15(1): Mucsi I, Hercz G, Uldall R, Ouwendyk M, Francoeur R, Pierratos A. Control of serum phosphate without any phosphate 196 Am J Kidney Dis. 2013;62(1):