Abstract
Individuals with kidney failure have multiple different options for further treatment. In addition to conservative care, discussed elsewhere, major kidney replacement therapy options include hemodialysis, peritoneal dialysis and kidney transplant. Within each of these modalities, there are subsets. For example, hemodialysis in the United States is most often performed thrice weekly in-center, but frequent home hemodialysis, incremental dialysis initially with twice weekly treatments, and nocturnal hemodialysis are additional options, and beyond the United States, hemodiafiltration may be an option. This chapter explores the epidemiology of kidney failure therapies, summarizing the data comparing the available kidney replacement therapy modalities and the patient-centered factors that may result in modality preference and concluding that patient preference and individual factors likely dictate the optimal therapy for people with kidney failure.
Keywords
Hemodialysis, Peritoneal dialysis, Kidney transplant, Epidemiology, End-of-life, Incremental dialysis
Patients with chronic kidney disease (CKD) stage 5 often require either kidney transplantation or maintenance dialysis to sustain life. In general, the paucity of living and deceased donors and the medical ineligibility of many end-stage renal disease (ESRD) patients for transplantation substantially limit the number of patients who can receive a kidney transplant. In contrast, many countries and states around the world have made a social compact to make maintenance dialysis therapy available to every kidney failure patient in whom this treatment is indicated. Thus maintenance dialysis is the dominant method of treatment for chronic kidney failure.
It is estimated that there are some 2 million patients treated with maintenance dialysis worldwide including almost half a million alone in the United States. The global census of patients undergoing dialysis continues to increase, in part reflecting exponential growth in patient populations undergoing dialysis in such emerging economies as China and India. It is believed that currently over 80% of patients undergoing dialysis worldwide are treated with in-center hemodialysis (HD), generally delivered thrice weekly in most industrialized nations, while once- to twice-weekly HD may be common in some other countries either as incremental transition to thrice weekly dialysis or as palliative dialysis, although resource constraints may also lead to less frequent treatment; most of the rest are treated with home peritoneal dialysis (PD).
In the last 10 years, variations in the conventional method of delivery of different therapies, but particularly HD, are increasingly being used, creating a veritable menu of kidney replacement therapies from which patients can choose ( Table 59.1 ). In recent years, there also has been heightened interest in better understanding the implications of transition from advanced CKD (estimated glomerular filtration rate [eGFR] <30 mL/min/1.73 m 2 ) to kidney replacement therapy as well as transitions within kidney replacement modalities ( Fig. 59.1 ).
| In-Center Hemodialysis |
| Thrice Weekly |
| Standard duration (3–4 hours) |
| Diurnal, long duration |
| Nocturnal, long duration |
| Frequent (four times weekly to daily hemodialysis) |
| Infrequent (once to twice weekly) |
| Incremental (upon transition to dialysis) |
| Palliative |
| Hemodiafiltration |
| On-line hemodiafiltration |
| Home Hemodialysis |
| Diurnal, twice to thrice weekly with conventional machines |
| Frequent with conventional or low-flow machines |
| Frequent long-duration nocturnal |
| Peritoneal Dialysis |
| Continuous ambulatory peritoneal dialysis |
| Automated peritoneal dialysis |
| Kidney Transplantation |
| Living related or unrelated donor |
| Deceased donor |
Notwithstanding the myriad options and different transitions, the median life expectancy of ESRD patients starting dialysis therapy still remains short, ranging from 3 to 5 years, prompting increased interest in determining if the method of delivery of kidney replacement therapy affects the survival of ESRD patients. More important, for reasons that remain uncertain, mortality of patients on dialysis is the highest in the first several months following dialysis initiation ( Fig. 59.2 ). Although it has been speculated that the excessively high mortality associated with transition to dialysis may be related to the rapid loss of residual kidney function in the first several months following dialysis initiation, other factors may also contribute, including loss of functionality and worsening quality of life in the setting of a dialysis-dependent lifestyle. Finally, acute kidney injury (AKI) as the cause of dialysis initiation, potentially mislabeled as transition to ESRD, may contribute, as AKI is associated with high death rates.
There are ongoing discussions as to whether very old persons with advanced CKD or those with more severe comorbid states and shorter life expectancy would have similar or even better survival and quality of life if they receive conservative management without dialysis or transplantation. This chapter reviews the contemporary studies comparing the survival of patients treated with different kidney replacement therapy modalities; more focused discussion of comprehensive conservative care occurs in Chapter 50 .
Maintenance Dialysis Versus Kidney Transplantation
For many patients, a functioning kidney transplant is less intrusive in their daily lives than maintenance dialysis and, hence, offers a substantial lifestyle advantage. However, these advantages are partially counterbalanced by the short-term surgical risks and longer term medical risks from lifelong immunosuppression. Studies suggest that, despite these short- and longer term risks, ESRD patients who receive a kidney transplant have a longer life expectancy when compared with individuals with equivalent health status who remain on the waiting list for a deceased donor organ while undergoing maintenance dialysis. The survival advantage with a successful transplant extends even to transplants where the kidney is harvested from marginal or expanded criteria deceased donors and in individuals whose age or coexisting medical conditions would have precluded them from being considered as donors over a decade ago. Furthermore, the earlier in the course of kidney failure that a patient receives a kidney transplant, the longer the allograft functions.
These observational reports are limited given systematic differences among individuals who receive a kidney transplant versus those who remain on the waiting list or those who receive an organ transplant earlier during the disease that cannot be fully accounted for in statistical models. While it is important to recognize these limitations, it is equally important to acknowledge that a clinical trial comparing kidney transplantation with maintenance dialysis in general, or the various dialysis modalities separately, is unlikely to be undertaken. Accordingly, the substantial lifestyle advantages along with the possibility of increased longevity make a compelling case for ensuring that every eligible kidney failure patient can undergo a kidney transplant as early as feasible within the constraints imposed by the availability of living donors and waiting time for a deceased donor transplant.
Hemodialysis or Peritoneal Dialysis
Both HD and PD require significant but different adaptations to patients’ lifestyles. First, HD requires high flow access to the bloodstream. Second, except for some recent trends, virtually all patients treated with HD are treated in a dialysis facility (referred to as in-center HD), and, throughout the developed world, HD is most commonly provided thrice weekly for 3 to 4 hours per session. Given transportation needs and time to achieve vascular access hemostasis posttreatment, HD requires a minimum total commitment of 4 to 5 hours thrice weekly. Variations from this general approach are increasingly being used and include differences in length and/or frequency of each treatment session (e.g., short daily HD of 2.5 to 4.0 hours, long thrice-weekly HD sessions of 5 to 7 hours each) as well as the site where dialysis is performed (home vs. in-center including conventional vs. nocturnal).
PD, in contrast, is almost exclusively performed at home. While traditionally patients performed four exchanges every day with continuous ambulatory peritoneal dialysis (CAPD), PD is increasingly performed with the use of a cycler at night, referred to as automated peritoneal dialysis (APD) or continuous cycling peritoneal dialysis (CCPD). APD usually requires patients to be connected to the machine for 8 to 10 hours with or without a daytime manual exchange. Thus the flexibility of dialyzing at home with PD is counterbalanced by the need to perform dialysis daily.
The necessary lifestyle adjustments for either dialysis modality are extremely important when considering the studies that have compared the outcomes of HD and PD patients. It is widely accepted that randomized, controlled clinical trials are the gold standard when comparing the outcomes of patients treated with two different therapies. However, the disparate effects of each treatment on patients’ lifestyles have stymied efforts to undertake randomized, controlled comparisons of HD and PD. For example, one clinical trial comparing these two modalities in the Netherlands was abandoned for futility as over 90% of eligible patients, when explained about the two treatment modalities, had a preference for one modality over the other and refused to be randomized ( Table 59.2 ). In this clinical trial, patients randomized to HD had a slightly better health-related quality of life, but these findings have substantially limited external validity. A more recent effort to conduct a randomized, controlled comparison of HD and PD in China ( Clinicaltrials.gov identifier: NCT01413074) that aimed to randomize 1370 patients with all-cause mortality as the primary outcome measure was terminated in December 2013 after 416 patients were enrolled. As of February 2017, results from this study remain unreported. Until then, one has to depend upon observational studies to compare the survival of HD and PD patients.
| First Author (Publication Year) a | Control Group | Intervention Group | Sample Size | Primary Outcome Measures | Key Results |
|---|---|---|---|---|---|
| Thrice-Weekly In-Center Hemodialysis Versus Peritoneal Dialysis | |||||
| Thrice-weekly hemodialysis | Peritoneal dialysis | 38 | Quality-adjusted life year score | 95% of eligible subjects refused to be randomized, limiting external validity; in first 2 years, hemodialysis associated with slightly better quality-adjusted life year score | |
| Thrice-Weekly In-Center Hemodialysis Versus Frequent In-Center Hemodialysis | |||||
| 2.9 treatments/week, mean duration, 213 min | 5.2 treatments/week, mean duration, 154 min | 245 | Death or 12-month change in left ventricular mass; death or 12-month change in physical health score | Frequent hemodialysis associated with significant benefits with respect to both coprimary outcomes and improved control of blood pressure and serum phosphorus but higher incidence of vascular access procedures | |
| Thrice-Weekly Hemodialysis Versus Frequent Nocturnal Home Hemodialysis | |||||
| 3 treatments/week, in-center | 5–6 treatments/week, at least 6 hours per treatment | 51 | 6-month change in left ventricular mass | Nocturnal hemodialysis associated with significant improvement in primary outcome, as well as kidney-specific domains of quality of life, blood pressure, and mineral metabolism | |
| 2.9 treatments/week, mean duration, 256 min, home | 5.1 treatments/week, mean duration, 379 min | 87 | Death or 12-month change in left ventricular mass; death or 12-month change in physical health score | No significant effect of nocturnal home hemodialysis or either of the two coprimary outcomes, but associated with improved control of hypertension and hyperphosphatemia; trend toward increased vascular access events | |
| Thrice-Weekly Low Flux Hemodialysis Versus Post-Dilution On-Line Hemodiafiltration | |||||
| Low flux in-center hemodialysis | Post-dilution online hemodiafiltration with 6 L/hour convective clearances with high-flux dialyzers | 714 | All-cause mortality | No significant difference in all-cause mortality or cardiovascular events; post-hoc analyses showed that patients with higher convective clearances had lower all-cause mortality | |
| Continuous Ambulatory Versus Automated Peritoneal Dialysis | |||||
| Continuous ambulatory peritoneal dialysis | Automated peritoneal dialysis | 82 | Patient and technique survival, time to first peritonitis, and hospital admission and catheter removal rates | Significantly fewer hospitalizations, and peritonitis rate with automated peritoneal dialysis; no difference in patient, or technique survival, or time to first peritonitis | |
| Continuous ambulatory peritoneal dialysis | Automated peritoneal dialysis | 34 | Health-related quality of life | No difference in quality of life; too few events to assess the effect on clinical outcomes; automated peritoneal dialysis patients reported more time for work, family, and social activities | |
a Please see the full bibliography on www.expertconsult.com for a list of these studies.
Over the 35 years since the advent of widespread use of PD for treatment of kidney failure, numerous single- and multicenter observational studies have compared the outcomes of patients treated with PD to those treated with HD. Despite differences among individual studies, analyses of survival data from national registries of patients who started treatment through the 1990s reveal a few common themes: overall, patients who started treatment with PD had a lower death risk for the first 1 to 2 years but a higher long-term risk. The apparent “early survival advantage” with PD was greater and of longer duration for younger and healthier patients, particularly among nondiabetics with no additional comorbidity. In contrast, there was little if any apparent “early survival advantage” with PD among older and sicker patients, and there appeared to be a higher long-term death risk.
In addition, since the mid-1990s, there has been a differential change in outcome of patients treated with the two dialysis therapies such that improvements in survival of patients treated with PD appear to outpace those seen for patients treated with HD, whereas the outcome of any types of kidney transplantations remains superior to both HD and PD ( Fig. 59.3 ). This differential improvement in outcomes appears worldwide, with data emerging from the United States, France, Australia, New Zealand, Canada, Denmark, and Taiwan. The reasons for these differential changes over time are unclear but highlight the importance of considering an “era effect,” or secular trend, in addition to the complexities in comparing the HD and PD outcomes mentioned earlier.
The results of observational studies that have compared the outcomes of HD and PD patients who started treatment after 2000 are summarized in Table 59.3 . These contemporary studies, even though nonrandomized, have been more diligent in attempting to account for potential bias when comparing two therapies as different as HD and PD, with some applying advanced statistical tools like propensity scores and/or marginal structural models. Notwithstanding the sophistication of statistical models used, the risk for residual confounding persists. Stated differently, it remains uncertain if differences in outcomes of patients treated with the two therapies are a result of the dialysis modality or simply reflect the differences in outcomes of patients treated with HD and PD. With this caveat, these studies suggest that the “early survival advantage” with PD described in studies of earlier cohorts may be attributable to the high risk of death for patients who start HD with central venous catheters and disproportionate representation of late-referred patients in the HD cohort. In other words, the “early survival advantage” with PD may not be a direct benefit of PD but rather a result of differences in patients who were treated with PD compared with HD. More important, these studies demonstrate that the 4-, 5-, and 10-year survival of patients treated with HD and PD in different parts of the world with different PD utilization rates are similar. In addition, there are studies that demonstrate equivalent outcomes with HD and PD in subgroups of patients like those infected with hepatitis C, those with atheroembolic disease, or those returning to dialysis after a failed kidney transplant. Two studies seem to be the exception to the theme of equivalency of outcomes with HD and PD. One study, with data from the European registry, showed a robust survival advantage for patients who started treatment with PD, while a study from France showed a higher death risk among kidney failure patients with congestive heart failure treated with PD.
| First Author (Publication Year) a | Cohort Period/Country | Sample Size | Statistical Approach | Follow-up Duration | Key Results |
|---|---|---|---|---|---|
| 1987–2002 Netherlands | 16,643 (HD 10,841; PD 5802) | Cox proportional hazards model | Up to 16 years | In younger diabetic and nondiabetic patients, lower risk for PD patients for the first 15 months; no difference thereafter. In older nondiabetics, lower risk for PD patients in the first 6 months, but higher risk after the first 15 months. In older diabetics, no difference in early death but higher risk for PD patients after the first 15 months | |
| 1995–2002 Taiwan | 48,629 (HD 45,820; PD 2809) | Cox proportional hazards model | Up to 6 years | Overall similar 5-year (HD, 54%; PD, 56%) and 10-year survival (HD, 34%; PD, 35%); subgroup analysis showed higher risk for death among all diabetics, and older nondiabetics (>55 years age) | |
| 2001–2003 Colombia | 923 (HD 437; PD 486) | Cox proportional hazards model | Up to December 2005 | No difference in overall adjusted mortality rates between HD and PD; lower death risk for young, nondiabetic patients treated with PD but similar outcomes in all other groups | |
| 1991–2005 Australia and New Zealand | 25,287 (HD 14,733; PD 10,554) | Cox proportional hazards model, including analyses in propensity-score quartiles | Up to December 31, 2005 | Overall 11% lower risk for death for PD patients in the first year, but 33% higher risk after the first 12 months; early survival advantage with PD seen only in young patients without comorbidities; in the most recent cohort (2004), no difference in long-term mortality of HD and PD patients | |
| 2003 USA | 6337 pairs (HD 6337; PD 6337) | Propensity-score matched cohort | Up to 4 years | Overall mortality risk was 8% lower for PD patients. Similar adjusted 4-year-survival (HD, 48%; PD, 47%) | |
| 1996–2004 USA | 684,426 (HD 620,020; PD 64,406) | Marginal structural model | Up to 5 years | In 2002–2004, no significant difference in the 5-year adjusted survival of HD and PD patients (35% and 33%, respectively); lower risk for death for younger, nondiabetic PD patients; higher death risk for older diabetics—particularly those with additional comorbidity—treated with PD | |
| 2001–2008 Canada | 38,512 (HD, 31,100; PD, 7412) | Proportional and nonproportional piecewise exponential survival model | Up to 5 years | First-year mortality of HD patients who started dialysis with a fistula or graft was similar to that with PD but significantly higher for those HD patients who started with central venous catheters; during entire follow-up period, HD patients with fistula/graft had lower and those with central venous catheters higher death risk than PD patients | |
| 1998–2006 Ontario, Canada | 6573 (HD, 4538; PD, 2035) | Cox proportional hazards model | — | No significant difference in early or late mortality of HD and PD patients who started dialysis electively as outpatients and had at least 4 months of predialysis nephrology care | |
| 1982–2006 Scotland | 3197 (HD, 2107; PD, 1090) | Cox proportional hazards model | Through December 31, 2006 | No significant difference in survival of nondiabetic transplant-listed patients treated with either hemodialysis or peritoneal dialysis | |
| 1998–2006 ERA-EDTA Registry | 15,828 (HD, 12,731; PD, 3097) | Cox proportional hazards model | Up to 3 years | 18% lower 3-year death risk in patients starting treatment with PD; survival benefit greater in those with no underlying comorbidity and no difference in death risk in those with comorbidity | |
| 1991–2004 Canada | 46,839 (HD, 32,531; PD, 14,308) | Cox proportional hazards model | Up to December 31, 2007 | For the 2001–2004 cohort, PD patients had a lower death risk than HD patients for the first 2 years, but there was no difference in death risk thereafter | |
| 2000–2009 Finland | 4463 (HD, 3246; PD, 1217) | Cox proportional hazards model | Up to December 31, 2009 | No significant difference in adjusted survival of patients treated with peritoneal dialysis or in-center hemodialysis | |
| 2008–2011 South Korea | 1060 (HD, 736; PD, 324) | Cox proportional hazards model | Up to December 30, 2011 | Overall, patients treated with PD had a 51% lower risk for death compared with those treated with in-center hemodialysis | |
| 1990–2010 Denmark | 12,095 (HD: 8273; PD, 3822) | Piecewise Cox regression | — | Significantly greater improvement in risk for death for patients treated with peritoneal dialysis than hemodialysis; starting 1995, significantly lower risk for death for patients treated with peritoneal dialysis, compared with hemodialysis; for the 2005–2010 cohort, 20% lower death risk | |
| 2001–2013 Kaiser Permanente, Southern California | 1003 propensity-score matched pairs | Stratified proportional hazards model | Up to June 30, 2013 | There was a significantly higher risk for death for patients treated with in-center hemodialysis compared with peritoneal dialysis; the cumulative hazard ratio for death for patients treated with hemodialysis was 2.38 (1.68–3.40) and 2.10 (1.50–2.94), compared with peritoneal dialysis, in the as-treated and intent-to-treat analyses, respectively | |
| 2005–2012 Norway | 692 propensity-matched pairs | Cox proportional hazards model | Up to December 3, 2012 | No significant difference in 2- or 5-year mortality of patients treated with hemodialysis or peritoneal dialysis or in subgroups by gender, diabetic status, or patients >65 years of age; significantly lower risk for death for patients ≤65 years age treated with PD compared with in-center hemodialysis | |
| 1993–2012 ERA-EDTA Registry | 140,527 (HD, 112,258; PD, 28,269) | Cox proportional hazards model | Up to 5 years | No significant difference in 5-year survival of patients treated with hemodialysis or peritoneal dialysis for 1993–2002; 9% lower 5-year mortality for patients treated with peritoneal dialysis in 2003–2007, compared with hemodialysis. Generally lower risk for death with peritoneal dialysis in patients <65 years and without diabetes and higher risk for patients >65 years and with diabetes |
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
