Dialysis modality choice in elderly patients with end-stage renal disease: a narrative review of the available evidence Free

The number of elderly patients on maintenance dialysis has rapidly increased in the past few decades, particularly in developed countries, imposing a growing burden on dialysis centres. Hence, many nephrologists and healthcare authorities feel that greater emphasis should be placed on the promotion of home dialysis therapies such as peritoneal dialysis (PD) and home haemodialysis (HD). There is currently no general consensus as to the best dialysis modality for elderly patients with end-stage renal disease. In-centre HD is predominant in most countries, although it is widely recognized that PD has several advantages over HD, including the lack of need for vascular access, continuous slow ultrafiltration, less interference with patients' lifestyle and lower costs. Comparisons of outcomes between elderly patients on PD and HD rely on observational studies, as randomized controlled trials are lacking. The results of these studies are variable. However, most of them suggest that survival rates are largely similar between the two modalities, except for elderly patients with diabetes and/or beyond 1–3 years from dialysis initiation, in which cases HD appears to be superior. An equally important aspect to consider when choosing dialysis modality, particularly in this age group, is the quality of life, and in this regard most studies found no significant differences between PD and HD. In these circumstances, we believe that dialysis modality selection should be guided by patient's preference, based on comprehensive and unbiased information. A multidisciplinary team should review elderly patients starting on dialysis, aiming to identify possible barriers to PD and home HD, including physical, visual, cognitive, psychological and social problems, and to overcome such barriers by adequate care, education, psychological counselling and dialysis assistance.

TIMING AND STRATEGY OF DIALYSIS INITIATION IN ELDERLY PATIENTS WITH END-STAGE RENAL DISEASE

Elderly patients with end-stage renal disease (ESRD) have to deal with a series of choices. Since pre-emptive renal transplantation is rarely available, most of these patients are confronted with (i) the option between dialytic and non-dialytic (conservative) management, and, if they choose to undergo dialysis, (ii) the selection of dialysis modality and (iii) the decision upon the preferred timing of dialysis initiation. Given the reduced life expectancy of many elderly patients and the paucity of high-quality supporting evidence, it is very important for these decisions to be individual, with patients and their families playing a central role in the process [1]. Obviously, all management planning is best made in a non-emergency situation, after adequate patient medical and social assessment and education. This implies early referral of chronic kidney disease (CKD) patients to nephrologists by family physicians and other specialists. For patients who opt for renal replacement therapy (RRT), timely preparation and education for dialysis are crucial, as these are associated with a number of benefits, including elective dialysis start with access in place, reduction in hospitalizations, higher prevalence of patients choosing a home-based dialysis modality and, in those starting with haemodialysis (HD), a reduced prevalence of intravenous catheters [2]. Later on, the well-being and satisfaction of patients with their chosen therapy must be regularly reassessed and, if necessary, the initial choice should be reconsidered, discussed and adjusted or changed, as appropriate [1].

Conservative treatment may be a reasonable option for many elderly ESRD patients, provided that a realistic prognosis of life expectancy and disease course is considered, and is explained to and understood by patients. In several studies, older patients with a high burden of comorbidity experienced similar survival with and without dialysis [3, 4]. In addition, those managed conservatively spent significantly less time in hospital and were more likely to die at home [5]. However, there is little evidence on the comparative effects of conservative versus dialytic treatment on symptoms and other outcomes [1]. Patients who choose conservative treatment should continue to receive adequate follow-up, care and support.

Many patients who choose dialysis would nevertheless prefer to start it as late as possible. Current data suggest that, if dialysis is adequately prepared for in advance, it is safe to delay its initiation until the development of signs and symptoms of uraemia [6, 7]. However, many of these signs and symptoms are relatively non-specific, especially in older adults with other comorbid conditions. Therefore, it is often unclear whether and to what extent certain symptoms in a given patient would improve with dialysis initiation rather than with other methods [1].

A COMPARISON OF RELATIVE ADVANTAGES AND DISADVANTAGES OF PD AND HD IN THE ELDERLY

The prevalence of CKD and ESRD increases with age. Nowadays, advanced age is no longer an impediment for dialysis initiation, like it was some 30 years ago. Thus, the number of elderly patients on maintenance dialysis has rapidly increased in the past few decades, particularly in developed countries [8]. In the European Renal Association – European Dialysis and Transplant Association (ERA-EDTA) Registry Annual Report of 2012, patients aged 65–74 years represent 22% of the total prevalent RRT population, and those aged ≥75 years represent 20% [9]. In France, as much as 39% of prevalent dialysis patients are over the age of 75 years [10].

This trend poses an increasing burden on dialysis centres. Hence, many nephrologists and healthcare authorities feel that greater emphasis should be placed on the promotion of home-based dialysis therapies, such as peritoneal dialysis (PD) and home HD, which nowadays can be offered to patients living either in their own houses or in residential care. However, there is currently no general consensus as to the optimum modality of RRT in the geriatric population. From a medical perspective, both PD and HD have specific benefits, drawbacks and contraindications for those who opt to start RRT.

In-centre HD can be an effective and attractive dialysis modality for older patients, as it provides ongoing medical follow-up, as well as the opportunity to socialize during in-centre treatments [8, 11]. However, vascular access may be problematic, because the maturation of the arterio-venous (AV) fistula usually takes longer and has a lower success rate in the elderly, compared with younger patients [11, 12]. Central venous catheters should be avoided whenever possible, as they are associated with higher mortality rates than AV fistulas and grafts [13, 14]. AV grafts have superior patency rates to fistulas during the first 12 months after creation [14]. The risk of AV fistula failure in the elderly can be reduced by using proximal instead of distal sites [12, 15]. Timely and judicious choice of dialysis modality is very important for maximizing the chances to start HD with an AV fistula. In a study including 508 patients ≥57 years old, Hanko et al. [16] showed that among those who chose HD and started HD, 79% had AV fistula creation prior to dialysis start, whereas only 41% of those who chose PD but started HD instead and 50% of those who were undecided had predialysis AV fistula creation.

Geriatric patients on HD are also more prone to other HD-related complications, including intradialytic hypotension, malnutrition, infection and gastrointestinal bleeding, compared with younger patients [11, 17]. Due to cardiovascular disease and autonomic dysfunction, the elderly are particularly sensitive to fluid shifts and may tolerate ultrafiltration poorly [18, 19]. Last but not least, travel time to and from the HD centre may have a negative impact on patient's quality of life (QOL) and adherence to treatment [20].

PD does not require vascular access and provides a slower, continuous ultrafiltration, with better cardiovascular stability [11]. As a home-based therapy, PD is likely to be less disturbing for a patient's everyday life, avoids the discomfort of thrice weekly transportation to and from HD facilities, and may offer increased autonomy and independence, when compared with in-centre HD [21]. Several studies suggest PD to be associated with a slower decline in residual renal function (RRF) compared with HD, but this is still debated [22]. RRF is a significant factor contributing to adequacy of dialysis, QOL and mortality in dialysis patients, particularly in PD [23]. At the societal level, a meta-analysis of studies comparing the cost of PD versus HD in 46 countries found that most developed countries provide PD at a lesser expense to the healthcare system than HD. The evidence in developing countries is more mixed, but in most cases PD can be provided at a similar cost [24]. A ‘PD first’ approach has been suggested for ESRD patients, as it appears to be beneficial not only for the patients themselves, but also for physicians, healthcare systems and society, as well [25].

However, there are also numerous potential contraindications and barriers to PD, which may be classified as follows [17, 25–27]:

• (i) PD-related: infections, inadequate dialysis, ultrafiltration failure and catheter problems;

• (ii) Patient-related: loss of RRF, malnutrition/excess protein loss, multiple abdominal surgeries, hernias and ostomies, severe obesity, respiratory problems, stroke or severe illness limiting manual dexterity, visual troubles, cognitive deterioration, psychological reasons (fear of lack of supervision and fear of isolation) and social reasons;

• (iii) Healthcare system-related: the lack of patient modality education or training, insufficient expertise in PD among nephrologists, physician reimbursement policies, ownership of dialysis facilities and increasing availability of HD units.

PREVALENCE OF PD AND HD IN ELDERLY ESRD PATIENTS

The prevalence of PD is variable across the world; however, in most countries, it is used in <25% of patients on RRT. Although recent increases have been seen in states such as Argentina, Hungary, Portugal, Thailand and (due to a revised reimbursement) the USA, ∼40% of countries have reported an overall decline in the percentage of PD use from 2006 to 2012 [28].

Older ESRD patients, in comparison with younger ones, have a higher number of comorbid conditions, more commonly suffer from nutritional, functional, cognitive, sensory and gait impairments, and are more prone to depression and to social and financial difficulties. Therefore, such patients, as well as their doctors and caregivers, often tend to consider in-centre HD as a safer option than PD [8, 11]. Surveys have shown that nephrologists do not perceive age per se as a contraindication for PD, but they rather consider that it is the physical, mental and social problems of the elderly that could make home dialysis more difficult [29]. Late referral of ESRD patients to nephrologists also increases the probability of receiving HD rather than PD [17]. Xue et al. [30] observed that the time between the moment the patient was referred to a nephrologist and the start of dialysis was 3.5 weeks for individuals aged ≥75 years versus 20.5 weeks for those <75 years.

Therefore, in many countries, elderly ESRD patients are less prone than younger ones to start RRT on PD [21]. An analysis of the ERA-EDTA Registry showed that patients ≥70 years were 56% less likely to receive PD compared with those 20–44 years [31]. In 2009, in the UK, only 14.2% of incident dialysis patients ≥65 years were started on PD, when compared with 26.9% of those <65 years of age [32]. In the USA, the percentages of PD among incident RRT patients in 2012 were 6.1% for those ≥65 years versus 10.4% for those <65 years [28]. In fact, in a prospective cohort study conducted in seven North American nephrology units, age was the leading cause of non-eligibility for PD as RRT in patients with CKD stages III–V [33]. In contrast, in France, PD is almost equally distributed in all age groups; this modality is used in 20, 11.6, 10.8 and 11.8% of RRT patients aged 20–44, 45–64, 65–74 and ≥75 years, respectively [10]. This is probably due to the government funding of assisted PD in this country.

The Netherlands Cooperative Study on the Adequacy of Dialysis (NECOSAD) showed that patients of 70 years or older were six times more likely to choose HD over PD than those aged between 18 and 40 years, in the absence of contraindications to any modality. However, this study also showed that, with appropriate education and support from the multidisciplinary team, over 50% of older patients may be eligible for and choose PD [34]. Mehrotra et al. [35] reported that 36% of incident ESRD patients surveyed in the USA had not seen a nephrologist in the 4 months prior to the start of dialysis and that, of those who were seen, 66% claimed that PD had not been presented to them as a treatment option. Manns et al. randomized predialysis patients to receive a two-step process of written materials followed by small group, case-based sessions or standard education delivered in a multidisciplinary clinic. He found that the percentage of patients intending to start self-care dialysis (home PD, home HD or self-care HD unit) increased from 50 to 82% after the small group sessions [36]. The USA National Pre-ESRD Education Initiative involved 15 000 patients from 932 referring nephrologists between 1997 and 2001. Upon completion of the programme, 55% chose HD, whereas 45% chose PD, suggesting important influence of predialysis education on selection of modality of treatment [37].

Therefore, elderly patients starting dialysis should be reviewed by a multidisciplinary team to identify barriers to self-care PD, including physical, visual, cognitive, psychological and social problems. Such barriers should not be interpreted as contraindications to PD, as most of them can be overcome with appropriate care, education and support, including social aid, psychological counselling and assisted PD [17].

It has been estimated that >60% of patients over the age of 80 years require assistance with dialysis exchanges, exit-site care and medications [38]. These issues may also arise in patients who previously performed PD independently and lost functional independence for several reasons. In such cases, assisted PD may be a viable option [39]. Oliver and Quinn [17] reported a prospective non-randomized study comparing a group of patients who were offered assisted home PD with a group who were not. Among patients who were not offered assistance, 65% were suitable for home PD, while the offer of assistance increased the percentage to 80%. Assistance may be provided by spouses, family members, private caregivers and/or visiting homecare nurses. For patients without family members or the means to pay a caregiver, homecare assistance is an effective strategy and is now funded in several countries [17]. In a study by Povlsen and Ivarsen [40] including 100 PD patients >65 years old from one centre, assisted PD patients (N = 58) had lower survival and higher peritonitis rates, but similar technique survival, compared with self-care PD patients, after adjustment for some baseline characteristics. Verger et al. [41] also found that patients on automated PD assisted at home by private nurses had a higher risk of developing peritonitis than autonomous PD patients and family-assisted patients, but they showed that this increased risk could be eliminated by regular follow-up home visits and retraining of the caregivers. In France, assisted PD is actually the preferred dialysis modality for older patients entering dialysis, when patients are informed on treatment options [42]. A recently published analysis by Couchoud et al. [43] predicted that in patients >70 years, assisted PD would be less costly than hospital-based HD, although more expensive than non-hospital facility-based HD. The relatively high cost of assisted PD probably explains why this modality is not reimbursed and therefore rarely used worldwide.

A low prevalence of PD leads to insufficient exposure of in-training fellows to this modality. This, in turn, leads to a vicious cycle, with nephrologists not offering PD to their patients and further reducing its use [44]. Deficiencies in infrastructure, incentives and resource allocation for PD may additionally explain why some dialysis centres are less prepared or willing to offer PD to their elderly patients [45].

SURVIVAL STUDIES OF ELDERLY PATIENTS ON PD VERSUS HD

Starting from the mid-1990s, the survival of dialysis patients has improved with both modalities. However, much greater progress has been made in the outcomes of PD patients than in those of HD patients, and this is particularly true for the first 12 months of treatment [46].

As demonstrated by NECOSAD [47], randomized controlled trials (RCTs) to compare outcomes between HD and PD are difficult to perform. Therefore, we have to rely on observational studies in this regard. The most significant prospective cohort studies comparing PD with HD patients—NECOSAD [47] and CHOICE [48]—enrolled patients who started dialysis in the 1990s; thus, the only available contemporary comparisons are derived from national registries of dialysis patients. However, observational studies have inherent limitations, arising mainly from substantial differences between patients non-randomly assigned to the two dialysis modalities. Inclusion of several demographic, clinical and laboratory data as covariates in statistical models is insufficient to account for such differences. Using statistical techniques, such as propensity scores and matching, may improve the value of the results. Prospective cohort studies have a limited statistical power, as they usually include a rather small number of subjects, whereas dialysis registries are limited by the paucity of medical information available [46]. With these considerations in mind, it is not surprising that observational studies have yielded conflicting results with respect to mortality comparisons between PD and HD.

With regard to elderly patients, most of the available information was obtained from subgroup analysis of general PD versus HD comparative survival studies. The outcomes are not uniform, since these studies differ by country, era, statistical approach, duration of follow-up, patients' comorbidity and definition of ‘elderly’. The key results of the most important studies published so far [27, 31, 48–64] are summarized in Table 1. Some of these studies show better outcomes with PD, others show worse or similar outcomes, in comparison with HD. However, most of them suggest that survival rates on PD may be inferior to HD, especially in patients with diabetes. The risk of death on PD, compared with HD, also seems to increase with dialysis vintage (after 1–3 years).

In a recent meta-analysis, Han et al. [63] reviewed 15 comparative studies of PD versus HD, published after 2000, involving 631 421 elderly patients. Compared with HD, the pooled hazard ratio (HR) with PD was 1.10 [95% confidence interval (CI), 1.01–1.20]. The survival benefit from HD was particularly strong in subgroups with diabetes, dialysis duration >1 year and in those starting dialysis in the 1990s. The pooled HR for mortality was 1.26 (95% CI, 1.13–1.40) in the diabetes group and 1.10 (95% CI, 1.02–1.18) in the non-diabetes group. During the first year after initiating dialysis, there was no difference between PD and HD (HR, 0.95; 95% CI, 0.72–1.18), but after 1 year the PD group had a higher mortality rate than the HD group (HR, 1.41; 95% CI, 1.20–1.61).

Importantly, the survival advantage of HD over PD also seems to depend on the type of vascular access used in the former. Perl et al. [65] showed that mortality rates among elderly Canadian dialysis patients were similar on PD and on central venous catheter-based HD, but significantly lower on PD than on AV fistula- or AV graft-based HD.

Nevertheless, it is also important to understand that in all these studies absolute differences in survival duration between HD and PD elderly patients, even when statistically significant, are usually in the range of months, given the short survival on dialysis of these patients, in general. This aspect should be pointed out to elderly persons starting on dialysis, during presentation and discussion of these data.

QOL IN ELDERLY PATIENTS ON PD VERSUS HD

QOL is a very important outcome to consider, possibly as important as or even more important than survival in elderly (and especially in very elderly and frail) patients. When they are aware of a short life expectancy, such patients seem to focus on supportive care more than on prolonged survival [66]. Some studies even suggest that high comorbidity elderly patients on PD would choose conservative treatment (and, thus, a likely shorter survival) rather than switching to HD, when PD is no longer feasible [67, 68].

Most studies on QOL have examined the dialysis population as a whole and have not been restricted to specific age groups. These studies generally suggest at least equivalent or better QOL for patients treated with PD compared with in-centre HD. A systematic review by Boateng and East [69] showed that PD patients mostly rate their QOL higher than 2HD patients. Mental health components are comparable between both dialysis populations, yet HD patients may enjoy a relatively better QOL in the physical dimensions.

Harris et al. [70] sought to compare clinical outcomes and QOL in elderly patients on PD and HD from the North Thames Dialysis Study. This was a prospective cohort study, which included 174 incident and prevalent dialysis patients ≥70 years of age (78 PD and 96 HD) from four hospital-based renal units in London, UK. QOL was assessed at baseline and at 6 and 12 months, using the Short Form (36) Health Survey (SF-36) and the Symptoms/Problems scale of the Kidney Disease Quality of Life Questionnaire (KDQOL). PD and HD patients were comparable for sociodemographic and clinical characteristics, as well as for QOL at baseline. The authors found similar annual mortality and hospitalization rates and similar 6- and 12-month QOL in PD and HD patients.

Broadening Options for Long-term Dialysis in the Elderly (BOLDE) [21] was a cross-sectional, multicentre study in the UK, which assessed QOL, using the SF-12 Mental and Physical Component Summary scales, Hospital Anxiety and Depression Scale and Illness Intrusiveness Ratings Scale, in 140 patients ≥65 years on PD and HD. The groups did not differ in the unadjusted QOL and there was no difference in physical component scores; however, PD patients had marginally, but significantly, better mental component scores. PD patients also had a lower number of symptoms, less possible depression and lower illness intrusion.

RISK OF PERITONITIS AND TECHNIQUE FAILURE IN OLDER VERSUS YOUNGER PD PATIENTS

It has been hypothesized that older patients are at increased risk of peritonitis, due to immunodeficiency of aging and malnutrition [39]. However, most studies did not confirm this hypothesis. Based on data collected from 1996 to 2005 in the multicentre Baxter Peritonitis Organism Exit sites Tunnel infections database, including 1265 incident Canadian PD patients aged ≥70 years, Nessim et al. [71] found that older age was independently associated with a higher peritonitis rate only among those who initiated PD between 1996 and 2000, but not in those initiated after 2001. A study by Lim et al. [72] reported that, compared with younger patients (<50 years), elderly patients (≥65 years) had a similar peritonitis-free survival and a lower risk of death-censored technique failure, although they had higher peritonitis-related and all-cause mortality. In 328 Hong Kong PD patients, Li et al. [73] observed no significant difference between patients ≥65 years and those <65 years in cumulative patient survival, technique survival and peritonitis-free periods. In addition, no differences were observed in these regards between the elderly self-care group and the elderly assisted group. Among 235 PD patients, Taveras et al. [74] found that technique failure rates at 12 months and peritonitis rates were not significantly different between patients ≥75 and those <75 years of age. Other studies [75, 76] have also shown that, compared with younger patients, older patients on PD have similar rates of technique failure and transfer to HD.

CONCLUSIONS AND RECOMMENDATIONS

In most countries, facility HD is currently the preferred dialysis modality for ESRD patients, including the elderly. With few exceptions, a decline in the use of PD across the world has been seen over the past several years. In many countries, older ESRD patients are more rarely initiated on PD than younger patients. However, we believe that greater emphasis should be placed on the promotion of home dialysis therapies such as PD and home HD. In favour of this view, we suggest that several facts should be taken into consideration:

• Compared with HD, PD has a number of advantages, including the lack of need for vascular access (and, thus, lack of related complications), continuous slow ultrafiltration, less interference with patients' lifestyle and lower costs.

• During the past 15–20 years, an improvement of clinical outcomes was seen in PD patients, more significant than in HD patients.

• Most of the perceived barriers to PD in the elderly, such as dexterity, visual and cognitive impairments, can be overcome with appropriate care, education and support, including social aid, psychological counselling and assisted PD.

• The common view that in elderly patients, PD is associated with worse survival than HD is supported only by observational studies, since RCTs are lacking. However, observational studies have inherent limitations and their results must be interpreted with caution. Furthermore, these results are extremely variable, some showing better survival, others showing worse or similar survival on PD versus HD.

• QOL is a very important outcome to consider, possibly as important as or even more important than survival duration in elderly patients. Existing studies indicate that there are basically no differences in QOL between PD and HD in this population.

• In contrast with previous beliefs, there are also no differences in the risk of peritonitis and technique failure between older and younger patients on PD.

In conclusion, for elderly ESRD patients starting on dialysis, we recommend the following policies:

• Patients should be offered both home-based and facility-based modalities as options for RRT. Patients should receive balanced and unbiased information about HD and PD, including their relative benefits and drawbacks.

• Dialysis modality choice should be a patient-centred, individual decision. In the absence of compelling contraindications, this choice should be based on the preference of a well-informed and well-prepared patient.

• Planning of dialysis should be made in advance, whenever possible. A multidisciplinary team should review patients, aiming to identify potential barriers to PD and home HD, including physical, visual, cognitive, psychological and social problems, and to overcome such barriers by providing adequate care, education, psychological counselling and dialysis assistance.

ACKNOWLEDGEMENTS

European Renal Best Practice (ERBP), supported by ERA-EDTA, is currently in the process of developing a clinical practice guideline on the diagnosis and treatment of frail and elderly patients with advanced CKD. K.F. and A.C. are the Co-Chairs of ERBP Elderly Guideline Development Group. I.N. is a member of the Methods Support Team of ERBP. W.V.B. is the Chair of ERBP. I.N. is a fellow of ERBP and supported by a grant of the ERA-EDTA.

CONFLICT OF INTEREST STATEMENT

The results presented in this paper have not been published previously in whole or part.

REFERENCES