Abstract
Home hemodialysis allows for individualized therapy. It is an effective way to provide increased frequency, increased duration dialysis, or both. We explore the benefits and risk of the therapy and then describe the clinical implications of home hemodialysis.
Keywords
ESRD, home hemodialysis, increased frequency dialysis, left ventricular hypertrophy, nocturnal dialysis, ultrafiltration rate
Outline
Introduction, 437
Burden of Conventional Hemodialysis, 437
Home Hemodialysis: Benefits and Potential Risks, 437
Left Ventricular Hypertrophy and Cardiovascular Complications, 438
Blood Pressure and Antihypertensive Medication Use, 439
Mineral and Bone Disorder and Phosphate Binder Use, 440
Health-Related Quality of Life, 442
Treatment Complications and Tolerability, 443
Potential Risks of Intensive Hemodialysis, 445
Providing Home Dialysis: Clinical Implications, 447
Patient Selection, 447
Training and Clinics, 448
Dialysis Prescription, 448
Barrier to Home Dialysis, 449
Introduction
Therapy for end-stage renal disease (ESRD) in the United States is dominated by in-center hemodialysis, with less than 2% of the population using home hemodialysis for renal replacement therapy (RRT). However, since 2004 there has been a resurgence of interest in home hemodialysis, with an increase in utilization. This increase can be related to newer devices with less burden and more convenience, as well as the growing interest in increased frequency of hemodialysis. What percentage of patients in the United States could use home hemodialysis? At its peak in 1973, 40% of US patients were treated with home hemodialysis. Patient characteristics have changed over the past 40 years, with an older and sicker population receiving RRT. Certainly, home hemodialysis is not for every patient. However, most would agree that the therapy is underutilized. Moran and Kraus postulated a reasonable estimate of 25%, realizing this growth would take many years to achieve. The benefits of home hemodialysis include flexibility and convenience of home therapies, emphasis on patient-centered care with increased independence, and the simplicity and relatively low cost of providing increased frequency of dialysis, if desired.
Burden of Conventional Hemodialysis
Conventional hemodialysis (HD), as it is typically prescribed in the United States, is truly burdensome to the daily lives of patients. Overwhelmingly, conventional HD involves three sessions per week, each 3 to 4.5 hours in duration. Cumulatively, such time “on the machine” consumes 9 to 14 hours per week, typically during the day. Moreover, because patients typically undergo HD in a healthcare facility, there is requisite travel before and after each session. In the DOPPS (Dialysis Outcomes and Practice Patterns Study), more than 2100 US patients responded to the survey question, “How long does it take you to get to your dialysis unit or center (one way)?” Among respondents, 47% reported 15 minutes or less, 33% reported 16 to 30 minutes, 17% reported 31 to 60 minutes, and 3% reported longer than 60 minutes. By this distribution, each dialysis session reasonably necessitates 45 minutes of travel, thus increasing the time that is devoted directly to dialysis therapy to 11 to 16 hours per week. Not surprisingly, the employment rate among dialysis patients between the ages of 18 and 54 years remains low.
Arguably as important as time on the machine is time that is consumed by recovering from a dialysis session. During recovery time, which is highly variable but may tend to range from 6 to 8 hours, physical and mental function may be impaired. With 18 hours per week devoted to recovery, cumulative time either directly or indirectly devoted to dialysis therapy is 29 to 34 hours per week. In tandem with receipt of any additional health care, including hospital admissions, clinic appointments, and pharmacy visits, conventional HD therapy is essentially a (permanent) full-time occupation. Plausibly, patients would appreciate interventions that can ease treatment itself or return some normalcy to the interdialytic interval.
Home Hemodialysis: Benefits and Potential Risks
Home hemodialysis (HHD) represents a strategy to relieve the burden of conventional HD. HHD obviously shifts treatment from a healthcare facility to the home, thereby obviating peri-session travel. More importantly, HHD removes the constraint that shift-based scheduling on Monday–Wednesday–Friday and Tuesday–Thursday–Saturday cycles imposes on conventional HD. In other words, the cardinal feature of HHD is arguably not the setting of therapy, but rather the intensity of therapy. HHD facilitates intensive therapy (i.e., relative to conventional HD, any schedule that increases the number of treatment sessions per week, the number of hours per session, or both). HHD also readily permits treatment at any time of day or night, thus allowing the patient to dialyze during sleeping hours. The application of intensive HD in the home setting presents several benefits and potential risks, which are described in the following sections.
Left Ventricular Hypertrophy and Cardiovascular Complications
The conventional HD schedule includes a nearly 72-hour interval between consecutive sessions on Friday and Monday or Saturday and Tuesday. Numerous studies have suggested that this interval is associated with higher risks for mortality and morbidity. In a retrospective study of over 32,000 patients, Foley et al. found that the death rate on the day after the long interval was 23% higher than on other days and that the cardiovascular (CV)-related hospital admission rate was 124% higher ( Fig. 28.1 ). These findings have been corroborated by CV-related death patterns in the DOPPS, the United Kingdom, and Australia and New Zealand. The long interval at the end of each dialytic week simply allows “unphysiological” fluid and solute accumulation, which puts substantial stress on the heart and the peripheral vasculature. Indeed, CV disease (CVD) is the leading cause of death among dialysis patients.
Left ventricular hypertrophy (LVH) is defined by an increase in LV mass (LVM), due to increased wall thickness. In dialysis patients, LVH itself and serial increases in LVM index are both associated with poor clinical outcomes. Two randomized clinical trials (RCTs) have assessed changes in LVM on nocturnal HHD. In the Frequent Hemodialysis Network (FHN) Nocturnal Trial, 87 patients were randomly assigned to receive 12 months of HHD for either six nocturnal sessions per week or three conventional sessions per week. HD was delivered with conventional equipment. Mean LVM decreased significantly with intensive HD, from 141 to 132 g, but remained unchanged with conventional HD, increasing from 132 to only 133 g. The difference in change between groups was almost 11 g, in favor of intensive HD, but lacked statistical significance. The earlier and smaller Alberta trial of nocturnal versus conventional HHD, which also used conventional equipment, generated qualitatively similar results: The effect of intensive versus conventional HD on LVM was over 15 g and statistically significant ( Fig. 28.2 ).
In post hoc analysis of the FHN trials, Raimann et al. examined concurrent influences of interdialytic weight gain and time-integrated estimate of fluid load (TIFL) on change in LVM. TIFL is a metric that estimates the total burden of extracellular fluid volume, with accounting for interdialytic weight gain and urine output. In both the FHN Daily and Nocturnal trials, 12-month changes in interdialytic weight gain and TIFL were positively correlated with 12-month change in LVM, albeit with a weaker association between TIFL and LVM in the Nocturnal Trial, possibly due to the lower prevalence of anuria at baseline in that trial.
Whether reduction of LVH portends improvement in clinical outcomes is not entirely certain. Data from the HOPE (Heart Outcomes Prevention Evaluation) trial of ramipril versus placebo strongly suggested such a relationship in the general population. The RCTs of intensive HD offer little guidance, as sample sizes were uniformly too low to evaluate risks for death and hospitalization. However, observational studies of data from the United States Renal Data System (USRDS) support the hypothesis of beneficial effects of intensive HD in the home setting on CV outcomes. Weinhandl et al. compared risks of hospitalization with daily HHD and conventional in-center HD in Medicare beneficiaries. With 3480 daily HHD patients on the NxStage System One, a device that delivers a low volume of dialysate per session, and 17,400 conventional in-center HD patients who were matched according to demographics and comorbid conditions, the hazard ratio (HR) of CV-related admission for daily HHD versus conventional in-center HD was 0.89 (95% confidence interval [CI], 0.86 to 0.93) in intention-to-treat analysis. Regarding types of CV-related admission, daily HHD patients had 25% lower risk for cerebrovascular disease; 41% lower risk for heart failure, fluid overload, and cardiomyopathy; and 16% lower risk for hypertensive disease in on-treatment analysis ( Fig. 28.3 ). In an earlier study, Weinhandl et al. compared risks for CV death with daily HHD on the NxStage System One and conventional in-center HD. In intention-to-treat and on-treatment analyses, the HRs of CV death for daily HHD versus conventional in-center HD were 0.92 (0.78 to 1.09) and 0.83 (0.67 to 1.01), respectively.
Considering the totality of evidence, including the compatibility of randomized studies that have evaluated LVM metrics and observational studies that have examined CV hospitalization, patients with either LVH after an initial trial of conventional HD and patients with overt heart failure, particularly due to diastolic dysfunction, may benefit from intensive HD in the home setting.
Blood Pressure and Antihypertensive Medication Use
Hypertension is very common in patients with ESRD. Hypertensive kidney disease is the primary cause of ESRD for nearly 30% of US patients and the prevalence of hypertension is 85% in incident ESRD patients. In prevalent hemodialysis patients in the United States, as of December 2016, mean predialysis systolic blood pressure (SBP) was 147 mmHg, with 75th and 90th percentiles of 163 and 178 mmHg, respectively. From a different viewpoint, 33%, 21%, and 9% had predialysis SBP of 140 to 159, 160 to 179, and ≥180 mmHg, respectively. The burden of hypertension cannot be readily attributed to sporadic application of antihypertensive medications, as use of β-blockers, calcium channel blockers, and renin–angiotensin system inhibitors in hemodialysis patients is approximately 70%, 50%, and 40%, respectively. Although observational studies are mixed and many suggest a U-shaped association between either predialysis or postdialysis blood pressure (BP) measurements and clinical outcomes, both ambulatory and home measurements of SBP are strongly associated with mortality risk, with increasing risk along a linear gradient beginning at 120 mmHg. Thus lowering BP is likely an effective strategy for improving CV outcomes.
Multiple RCTs have shown that intensive HHD significantly reduces BP. In the FHN Nocturnal Trial, predialysis SBP decreased with intensive HD, from 145 to 142 mmHg at month 2, and steadily declined to 137 mmHg during months 10 to 12. Concurrently, predialysis SBP decreased only slightly with conventional HD, from 153 mmHg at baseline to 151 mmHg at the end of the study. The treatment effect of intensive HD on predialysis SBP was 8 mmHg, in favor of intensive HD, and statistically significant. The Alberta trial of nocturnal versus conventional HD generated similar findings, but without statistical significance ( Fig. 28.4 ). Therefore in patients with uncontrolled hypertension, especially with use of multiple antihypertensive medications, intensive HD in the home setting may be useful. What remains uncertain is whether intensive HHD in the home setting lowers ambulatory SBP, which is a stronger predictor of CV risk than predialysis SBP.
In tandem with lowering BP, intensive HHD also appears to reduce the use of antihypertensive medications. In the FHN Nocturnal Trial, the mean number of antihypertensive medications per patient decreased from 1.9 to 1.1 with intensive HD, whereas the corresponding number of medications with conventional HD remained at 1.6 during follow-up. In the Alberta trial of nocturnal versus conventional HD, antihypertensive medication use was either reduced or discontinued in 62% of patients who underwent nocturnal HD and only 12% of patients who underwent conventional HD ( P < 0.001). Similar patterns have been seen in observational studies. In the prospective FREEDOM (Following Rehabilitation, Economics and Everyday-Dialysis Outcome Measurements) study, in which patients underwent daily HHD with the NxStage System One, the mean number of antihypertensive medications per patient decreased from 1.7 to 1.0 after 12 months of follow-up ( P < 0.0001).
Mineral and Bone Disorder and Phosphate Binder Use
Mineral and bone disease (MBD) is another common complication of ESRD. Only one in three hemodialysis patients have serum calcium, phosphorus, and parathyroid hormone concentrations in target ranges of 8.4 to 10.2 mg/dL, 3.5 to 5.5 mg/dL, and 150 to 600 pg/mL, respectively. There are many pharmacological interventions for MBD, including vitamin D sterols, calcium- and noncalcium-based phosphate binders, and calcimimetics. At least in the United States, these interventions are incredibly costly. In dialysis patients with Medicare Part D coverage during 2014, cumulative spending on phosphate binders exceeded $840 million, whereas cumulative spending on cinacalcet hydrochloride was nearly $480 million; in other words, over 58 cents of every dollar of Part D spending in dialysis patients was devoted to the treatment of MBD. This allocation of resources is problematic, in no small part because of poor adherence.
Although the pathophysiology of MBD is complex, the mechanics of phosphorus intake and excretion merit special attention. Phosphorus is excreted through the kidneys; in the absence of residual renal function, the only exit route for phosphorus is dialysis itself. On the intake side, Noori et al. found that most patients consume between 500 and 1500 mg of phosphorus each day, according to data collected from a food frequency questionnaire. Between 60% and 70% of this intake is absorbed in the gastrointestinal tract. Thus the typical HD patient absorbs around 4500 mg of phosphorus each week. By the way of comparison, a 4-hour HD session clears approximately 1000 mg of phosphorus; the rate of clearance is initially high, but decreases later. Thus there is a significant gap between phosphorus absorption and clearance. With poor adherence to binders, phosphorus accumulates, thus contributing to vascular calcification.
Intensive HD usually increases the cumulative number of treatment hours per week, thus necessarily facilitating increased dialytic clearance of phosphorus. Multiple RCTs have shown that intensive HHD reduces serum phosphorus. In the FHN Nocturnal Trial, mean serum phosphorus declined with intensive HD, from 5.74 mg/dL at baseline to 4.38 and 4.72 mg/dL after 3 to 5 and 10 to 12 months of follow-up, respectively. The adjusted change in serum phosphorus from baseline to end of study was –1.11 mg/dL with intensive HD versus 0.12 mg/dL with conventional HD, thus resulting in a treatment effect of –1.24 (–1.79 to 0.68) mg/dL. In the Alberta trial of nocturnal versus conventional HD, the treatment effect of intensive versus conventional HD was –1.5 mg/dL ( Fig. 28.5 ). In patients with persistent hyperphosphatemia, especially in light of poor adherence to phosphate binders, intensive HD in the home setting, particularly at night, may be a highly effective strategy for controlling serum phosphorus.
Use of phosphate binders also appears to decrease with intensive HHD. In the FHN Nocturnal Trial, the percentage of patients who used any phosphate binders decreased from 97% at baseline to only 27% at months 10 to 12 on intensive HD, whereas no such decrease was observed on conventional HD. At the end of the trial, 42% of patients on intensive HD actually required phosphorus supplementation in the dialysate to avoid development of hypophosphatemia.
In contrast to these data, there is little evidence that intensive HHD meaningfully alters either serum calcium or parathyroid hormone, although results about the latter biochemical are limited by the brief follow-up of studies. Of course, serum calcium also reflects the choice of dialysate calcium concentration. Many patients who have performed nocturnal HD have required dialysate calcium ≥3.0 mEq/L.
Health-Related Quality of Life
There is increasing recognition that dialysis patients value quality of life (QoL) more highly than traditional clinical outcomes, including risks for death and hospitalization. In a survey of over 4500 German dialysis patients, safety of treatment, satisfaction with care, and health-related quality of life (HRQoL) were the only outcomes that were rated by greater than 90% of patients as “very important.”
HRQoL is typically separated into physical and mental components. In dialysis patients, the prevalence of low physical HRQoL is especially high. In a sample of prevalent HD patients in 2016, 38% of patients had a physical composite summary (PCS) score between 35 and 44 points and another 19% of patients had a PCS score less than 35, according to the 12-item Short-Form Survey, for which the mean and standard deviation are 50 and 10 points, respectively, among US adults. On the other hand, 28% of patients had a mental composite summary (MCS) score less than 45.
In RCTs, intensive HHD appears to exert a positive effect on physical HRQoL, although direct comparisons of effects on intensive and conventional HD suggest cautious interpretation. In the FHN Nocturnal Trial, mean PCS score increased with intensive HD, from 37 at baseline to 40.3 at 12 months, but also increased with conventional HD. Accordingly, the treatment effect of 0.6 points, in favor of intensive HD, lacked statistical significance. In the Alberta trial of conventional versus nocturnal HD, the effect of intensive HD on PCS score was 1.5 points, in favor of intensive HD, despite only 6 months of follow-up, but the effect lacked statistical significance ( Fig. 28.6 ). In an interim analysis of the FREEDOM study of daily HHD with the NxStage System One, but without a comparator group, mean PCS score increased significantly between baseline and at both 4 and 12 months. The percentage of participants with PCS score ≥50 increased between baseline and 12 months in the intention-to-treat (from 8% to 24%) and the per-protocol (from 9% to 21%) cohorts. Interestingly, in the FHN Daily Trial, which was conducted in a facility, the effect of intensive versus conventional HD on PCS score was larger and statistically significant. One possibility is that although intensive HD improves physical HRQoL, the home setting adds a physical burden to patients’ lives, thus resulting in an attenuated effect of intensive HHD on physical HRQoL.
