Prescribing Peritoneal Dialysis for Elderly Patients Starting Peritoneal Dialysis

Abstract

Increased availability of dialysis services has led to both an increase in the number of elderly, frail, co-morbid patients with advanced chronic kidney disease now being offered dialysis and starting dialysis with residual kidney function. Traditionally, these patients would have been offered in-centre haemodialysis. However, the introduction of an assisted peritoneal dialysis service has allowed more of these elderly patients to be considered for peritoneal dialysis, a home-based treatment, with the exchanges performed by family members or visiting health care staff. It is now realised that the amount of dialytic clearance any individual requires varies, and as such, treatment targets have changed over time from achieving minimum clearance targets to a more holistic approach, considering patient lifestyles, and adapting dialysis prescriptions and schedules to the needs of the individual patient. As dietary intake is often lower in the elderly, coupled with the physiological loss of muscle mass, this results in a reduced generation of waste products of metabolism and consequently requires less dialytic clearance. Thus, this allows many elderly patients to benefit from an incremental approach to starting peritoneal dialysis, potentially beginning with only one or two continuous ambulatory peritoneal dialysis exchanges, or an overnight cycler for only a few nights/week.

1. Introduction

The number of patients with chronic kidney disease (CKD) continues to rise worldwide, and it is anticipated that during the next decade, CKD will become one of the most common non-communicable chronic illnesses on the planet [1]. Although there have been some recent advances in the management of CKD designed to slow progression [2,3], these will be outweighed by the increasing prevalence of obesity, diabetes, and cardiovascular disease in the ageing population. More patients who now progress to end-stage kidney disease (ESKD) are now being offered kidney dialysis treatments, and the number of countries now offering patients some or total financial support for dialysis treatments continues to increase. Although many more patients are currently treated by haemodialysis worldwide, continued expansion of haemodialysis centres may become limited by the ability to recruit and retain additional trained dialysis staff. Peritoneal dialysis (PD), being a home-based treatment, requires fewer trained nursing or allied health care staff to support larger numbers of patients compared to haemodialysis [4].

In addition, depending on the availability of peritoneal dialysis fluids, PD may be a substantially less costly form of dialysis than haemodialysis, and some countries, such as the UK, recommend PD as the first-line dialysis treatment [5,6].

In Japan, Western Europe, and North America, the widespread availability of dialysis services has led to patients now being offered to initiate dialysis with greater levels of residual kidney function. In addition, patient demographics have changed with increasing numbers of elderly, frail patients with additional co-morbidities now treated with dialysis. Historically, PD was advocated as a treatment for healthy elderly patients aged 60 years, who could perform their own dialysis at home [7]. As such, the age of PD patients in many centres was around a decade younger than their corresponding haemodialysis patients, as older, more frail patients who could not manage to perform PD exchanges were initiated on haemodialysis. Following the advent of an assisted PD service [8], whereby a trained family member or allied health care worker visits the home and performs the exchanges, has now allowed many more elderly, frail, co-morbid patients who were unable to perform their own PD exchanges to have a home-based dialysis therapy [9]. However, assisted programmes vary; particularly when assistance requires home visits from a nurse or allied health care worker visiting the home, compared to when a family member assists the patient [4]. Thus, depending on the amount of support, assisted PD patients may only have one or two visits to help with exchanges, or more may be available depending on family member support.

2. Peritoneal Dialysis Treatment Options for the Older Person

For elderly patients who can perform their own exchanges independently, continuous ambulatory peritoneal dialysis (CAPD), automated peritoneal dialysis cycler with either a dry day (APD), or a day dwell (continuous cycling peritoneal dialysis -CCPD) are options. Even so, some elderly patients may be able to connect and disconnect themselves but require assistance setting up a cycler machine, and so require one visit a day for support. If, however, patients cannot manage their own exchanges, then if treated by CAPD, treatment may be limited to one nocturnal exchange or, at most, two CAPD exchanges if relying on home visits from nurses or allied health care workers, although more exchanges may be possible depending on family or other support [10,11]. Whereas those treated with APD or CCPD may only require one visit in the evening if the patient can disconnect themselves in the morning, or two visits if totally unable to perform their treatment.

Similarly, if patients are unable to measure their own weight or check blood pressure, then either family members or visiting nurses or allied health staff should provide monitoring information to the home PD centre. Dietary restrictions, with a reduction in roughage and cooking practices to reduce potassium, increase the prevalence of constipation, particularly in the elderly. As such, elderly patients may have increased low-drain volume alerts when using APD cyclers. This not only impacts sleep but also reduces treatment effectiveness. The introduction of external monitoring of APD cyclers has been a major advance, with information on total ultrafiltration, but also a review of individual cycles transmitted to the PD centre for review. Thus, advice can be fed back to patients, carers, and allied health staff to adjust treatments to reduce alarms and improve treatment efficiency [12,13]. In addition, stimulated by the COVID pandemic and advances in information technology, more centres are moving to using telemedicine to review patients at home and amend treatments [14,15,16].

3. Residual Kidney Function

Whereas haemodialysis treatments may be associated with episodes of intra-dialytic hypotension and temporary hypoperfusion to vital organs, including the heart, brain and kidney [17], blood pressure and cardiac perfusion are not affected by peritoneal dialysis exchanges [18]. Thus, there have been numerous observational reports that residual kidney function is better maintained with PD compared to haemodialysis [19].

It has been recognised for many years that residual kidney function is important in determining both PD technique and patient survival [20]. Traditionally, residual kidney function is calculated from 24 h urine collections. Urine volume and the amounts of both urea and sodium fall after a haemodialysis session and then start to recover during the inter-dialytic interval; therefore, measuring residual kidney function must be carefully timed to obtain reliable measurements [21]. Whereas, due to the cardiovascular stability associated with PD, 24 h urine collections are more dependable.

As haemodialysis dosing has traditionally been based on dialyser urea clearance, urinary urea excretion has been used to assess residual kidney function and added to dialyser urea clearance to generate a measure of total urea clearance. Urea clearance by the kidney depends primarily on glomerular filtration; however, kidney function is more than just glomerular clearance. As such, current assessments of residual kidney function, when applied to haemodialysis, exclude any contribution from kidney tubular function. Neither haemodialysis nor PD effectively clear protein-bound solutes, including indoxyl sulphate and p-cresyl sulphate. These protein-bound uraemic toxins have been demonstrated to cause inflammation and damage of the vascular endothelium in cell culture experiments and small animal models, and clinical observational studies have reported increased haemodialysis patient mortality with increasing serum concentrations [22]. However, as these protein-bound toxins are excreted by organic acid transporters in the distal tubule, then serum levels are much lower in dialysis patients with some residual kidney function. Creatinine is also secreted in the distal tubule, so as glomerular filtration declines, the proportion of urinary creatinine from secretion increases, and as such, measurements of urinary creatinine clearance overestimate the glomerular filtration rate. Studies demonstrated that blocking renal tubular secretion of creatinine with cimetidine improved the accuracy of creatinine-based measurement of glomerular filtration rate when compared to inulin-measured glomerular filtration rate in kidney dialysis patients [23]. Additional studies reported that 24 h urinary urea measurements underestimate and 24 h urinary creatinine measurements overestimate glomerular filtration rate; thus, by combining the relative proportions of both, the results were similar to the glomerular filtration rate measured by inulin infusion [21,23]. As such, residual kidney function for PD patients has been taken as the average of urinary urea and creatinine clearances and so differs from that reported for haemodialysis patients [21]. In addition, there were differences noted in the tubular secretion of anionic and cationic solutes between PD and haemodialysis patients, with greater clearances reported in PD patients [24].

However, many elderly, frail, co-morbid patients are unable to reliably collect 24 h urine collections due to cognitive decline, relative immobility, and incontinence. Thus, several plasma biomarkers have been investigated as potential alternatives to estimate residual kidney function, including β2-microglobulin, β trace protein, and cystatin-C, with β2-microglobulin being the most promising [25,26,27].

4. Estimating Delivered Dialysis Dose

Traditionally, the amount of dialysis delivered to PD patients has been based on weekly clearance, based on the combination of the 24 h peritoneal urea clearance (KtPerit urea), and the combined urinary urea and creatinine clearance (Kt Urinary urea and creatinine). However, it must be remembered that these clearances are not equivalent, as 1.0 mL of urinary clearance provides a greater total clearance than 1.0 mL of peritoneal urea clearance. This was demonstrated by the CANUSA study, which showed that both PD technique and patient survival depended more upon residual kidney function rather than the combination of urinary and peritoneal clearance [19]. As such, clinical guideline committees initially advocated target clearances based on both weekly combined peritoneal and urinary urea clearance and creatinine clearance [28,29,30]. However, subsequently, due to the variance between urea and creatinine clearances affected by residual kidney function, some clinical guideline committees dropped creatinine clearance targets [31].

To be able to compare clearances between patients, traditionally, urea clearance has been adjusted using an anthropometric estimate of total body water, assuming that urea is equally distributed within all body water [32], and creatinine clearance has been adjusted to body surface area of 1.73 m2. However, the original calculations of total body water were based on a healthy and now what would be considered a younger population, and therefore differ from the average dialysis patient. Dialysis patients are more likely to suffer from malnutrition due to dietary restrictions, changes in body composition typified by muscle wasting, termed sarcopenia, increased fat mass, and sarcopenic obesity [33]. These changes in body composition alter total body water, as the water content of fat is around 10% or lower, while that of muscle is around 90%. Not only do body composition changes occur with age and disease, but body composition also varies with ethnicity [34]. Thus, when calculating Kt/Vurea, the anthropometric “V” value may overestimate true V in the obese person [35] or equally underestimate “V” in smaller men or women [36]. As with adjusting creatinine clearance, studies have suggested adjusting Kturea for body surface area as a more accurate metric of the delivered dialysis dose of small solute clearance, as there is a closer association between body surface area and body composition than anthropometric estimates of body water [37].

As the majority of patients starting PD will have residual kidney function, and clearance targets are a composite of residual kidney function and peritoneal clearance, then, depending on the individual patient, an incremental approach can be considered. Patients could be given the option of dialysing less frequently than the daily standard, and additionally starting with smaller dwell volumes or fewer exchanges (

Table 1. Approach to starting peritoneal dialysis in an elderly, frail patient.

❖ Consider an incremental approach ○ reduced frequency of treatments ○ 5 days/week rather daily ○ reduced volume of exchange ○ 1.5 L dwell rather than 2.0 L ○ If CAPD ○ 1 or 2 daily exchanges rather than more ❖ Requirements ○ monitoring residual kidney function ○ (24 h urine collections/serum β2 Microglobulin) ○ monitoring volume status (N-terminal brain natriuretic peptide/Bioimpedance) ○ monitoring dietary intake Food recall/Nitrogen appearance rate ○ monitoring body composition Anthropometrics/Bioimpedance/Dual-energy X-ray absorptiometry ❖ Action ○ increase dialysis dose as required ○ frequency of treatment ○ dwell volume ○ number of cycles ○ hypertonic solutions ○ nutritional counselling

).

5. How Much Dialysis Clearance Do Patients Require?

The kidney plays a major role in the excretion of the waste products of metabolism. As such, metabolism and the daily generation of waste products depend on three main factors: dietary intake, basal metabolic rate, and active energy expenditure (

Table 2. Factors to consider when starting patients on peritoneal dialysis to aid in the choice of the starting peritoneal dialysis prescription. Continuous ambulatory peritoneal dialysis (CAPD), automated peritoneal dialysis with dry day (APD) or with long day dwell—continuous cycling peritoneal dialysis (CCPD).

Suitable for CAPD ○ Consider metabolic requirements ○ Dietary protein intake ○ Food recall ○ Estimated dietary protein intake ○ Basal metabolic rate ○ Sex/age/weight/height ○ Active energy expenditure ○ Sedentary/exercising ○ Muscle mass ○ Action on metabolic requirements ○ Frequency ○ Daily or less frequently ○ Dwell volume ○ 2.0 L or lower volume ○ Session time ○ Short or long ○ Exchanges ○ One or more ○ Volume requirements ○ Dietary Sodium intake ○ Food frequency questionnaire ○ 24 h urine sodium ○ Volume status ○ Clinical examination ○ N-terminal brain natriuretic peptide ○ Bioimpedance ○ Predicted peritoneal transport ○ Sex/serum sodium/albumin ○ Glucose concentration ○ Icodextrin ○ Action on Volume requirements ○ Dietary sodium ○ Nutrition counselling ○ Volume excess ○ Loop diuretics ○ 1 or 2 icodextrin exchanges ○ Hypertonic glucose dialysates

Suitable for APD ○ Consider metabolic requirements ○ Dietary protein intake ○ Food recall ○ Estimated dietary protein intake ○ Basal metabolic rate ○ Sex/age/weight/height ○ Active energy expenditure ○ Sedentary/exercising ○ Muscle mass ○ Action on metabolic requirements ○ Frequency ○ Daily or less frequently ○ Dwell volume ○ 2.0 L or lower volume ○ Session time ○ 8 h or less ○ APD or CCPD ○ Dry day or day dwell ○ Volume requirements ○ Dietary sodium ○ Food frequency questionnaire ○ 24 h urine sodium ○ Volume status ○ Clinical examination ○ N-terminal brain natriuretic peptide ○ Bioimpedance ○ Predicted peritoneal transport ○ Sex/serum sodium/albumin ○ Icodextrin day fill ○ Action on Volume requirements ○ Dietary sodium ○ Nutrition counselling ○ Volume excess ○ Loop diuretics ○ 6 or 7 APD cycles ○ Shorter cycler dwell time ○ Hypertonic glucose ○ Long icodextrin day dwell

).

Many clinical guideline groups, including the Kidney Disease Outcomes Quality Initiative (KDOQI), recommend that patients with CKD stages 3 and higher restrict dietary protein intake to 0.55–0.60 g [38]. It is well established that patients naturally reduce their dietary protein intake as kidney function declines [39]. Once patients transition to dialysis, a greater dietary protein intake is then recommended to compensate for protein losses in waste dialysate [38,40]. However, protein losses, both in terms of urinary and peritoneal protein losses, will differ between patients, depending on the primary kidney disease, peritoneal dialysis prescription, and patient co-morbidities [40,41].

Despite advice, many dialysis patients fail to achieve this increased target dietary protein intake of 1.0–1.2 g/kg/day [42], particularly the elderly, as older patients generally have a reduced appetite compared to younger patients. Additionally, some patients simply continue with their reduced dietary protein intake following transition from pre-dialysis to dialysis [43]. Whereas others, particularly those treated by PD, do not appear to have the same hunger profile for food compared to healthy individuals or haemodialysis patients [44], and may have a reduced appetite due to increased abdominal pressure, reflux oesophagitis, glucose absorption from the dialysate, and increased amounts of circulating anorexic hormones, including des-acyl ghrelin from the stomach and peptide YY from the colon [45,46].

Traditionally, dietary protein intake has been estimated from combined peritoneal and urinary urea excretion with a varying compensatory factor to account for urinary, faecal, and other protein losses [47]. However, many of the equations currently used to estimate daily nitrogen appearance rates were derived from a different era, when PD patients were younger with less co-morbidity and treated with continuous ambulatory peritoneal dialysis (CAPD), using four daily glucose dialysate exchanges [48,49]. More recent studies have suggested that these equations tend to underestimate total peritoneal nitrogen losses [40]. As many patients starting PD now have residual kidney function, urinary protein losses may well exceed those proposed in these equations. As such, diet histories may provide a more accurate estimate of daily protein and calorie intake. However, when dealing with older, frail, co-morbid patients, a 24 h diet recall has been suggested to provide a more reliable assessment compared to the standard 3-day dietary history [49].

Resting energy expenditure can be formally measured by direct or indirect calorimetry [50,51], although in clinical practice, basal metabolic rate is often estimated using a series of standard equations based on sex, age, weight and height [52,53,54]. Investigational studies have suggested that the basal metabolic rate of dialysis patients does not significantly differ from that of healthy control subjects [51,55].

Active energy expenditure has often been estimated by using physical activity scores or activity questionnaires [56]. Studies in dialysis patients have demonstrated that, for most patients, particularly the elderly dialysis patients, they are relatively inactive with a low level of active energy expenditure [57]. For most patients, there is an association between dietary protein intake, muscle mass, and active energy expenditure [58]. Although dietary intake and energy expenditure may be lower in many PD patients, it is important that those who are eating well and are physically active receive an adequate amount of dialysis, as observational studies have demonstrated that patient survival is dependent on the amount of dialysis urea clearance when adjusted for total energy expenditure [59,60] (Table 2).

6. Volume Control in Peritoneal Dialysis Patients

Although maintaining residual kidney function is important for PD patients, maintaining patients in an over-hydrated state does not preserve residual kidney function [61]; instead, it increases the risk of developing left ventricular hypertrophy and heart failure. Patients should follow a dietary salt restriction to limit thirst and fluid gains and thus aid volume homeostasis, as patients who ingest excess salt will require greater use of icodextrin and hypertonic glucose dialysates to restore volume control, increasing their risk of losing residual kidney function [62]. Loop diuretics have been demonstrated to increase both urine volume and sodium removal in PD patients [63]. The observational Dutch NECOSAD study suggested that residual kidney function declined faster when PD patients were treated with nocturnal automated peritoneal dialysis cyclers (APD) using glucose dialysates and a dry day compared to continuous ambulatory peritoneal dialysis (CAPD) [64]. It was suggested that removing fluid overnight with APD could have led to periods of relative dehydration, which then resulted in a faster loss of residual kidney function compared to fluid removal with CAPD.

In addition to clinical examination, monitoring with bioimpedance and measurement of natriuretic peptides may add additional information to guide patient management. Bioimpedance equipment is increasingly being used in dialysis centres, and devices vary not only in terms of using single electrical frequencies to multiple frequencies and bioimpedance spectroscopy, but also whether they report on whole body or individual body segments [65]. As body composition changes with age, manufacturers have developed proprietary software models to take age into account; however, these vary between devices regarding whether models are based on 2, 3, or 4 body compartments [66,67]. Changes in body composition result in a loss of intracellular water (ICW), so devices that report the ratio of extracellular water (ECW) to ICW show increases in the elderly, without necessarily implying volume overload [68]. Similarly, devices that aim to estimate normohydrated weight are confounded by changes in muscle composition in the elderly, with loss of myocytes and increasing fibrous tissue [69]. As such, single absolute values may have some errors, but serial trends are more reliable in tracking whether elderly patients are gaining or losing ECW [70]. Similarly, although natriuretic peptides increase as the plasma volume and intracardiac chambers increase in size, these peptides are cleared by the kidneys; thus, an increase may reflect the loss of residual kidney function and can also increase with inflammation and sepsis [71,72,73]. However, underlying cardiac diseases, particularly atrial fibrillation and cardiac valvular disease, which are commonly encountered in the elderly, also increase natriuretic peptides. So again, as with bioimpedance, single measurements may not accurately reflect plasma volume, but serial measurements and trends can provide additional guidance as to volume status [74]. Other potential volume biomarkers, such as CA125, are currently under investigation [75,76].

7. Starting Elderly Frail Patients with Residual Kidney Function on Peritoneal Dialysis

Ideally, patients should collect a 24 h urine sample, as this not only can be used to determine residual kidney function but can also be tested for urinary sodium to provide an estimate of dietary sodium intake, allowing patients with a raised urinary sodium to be given appropriate dietary advice. The amount of PD to be prescribed to patients has changed over recent years from fixed minimum targets [28,30,77] to a more holistic approach [78,79,80], considering not only residual kidney function but also the lifestyle of the patient, which includes nutritional intake, body composition, and physical activity. Thus, elderly patients with a healthy appetite, who are physically active, will require greater overall clearance than those eating less with lower muscle mass and less physical activity. The introduction of assisted PD has now allowed many more elderly, frail patients who previously could not have been treated by PD, as they were unable to perform their own exchanges, to be treated at home. However, most assisted PD programmes only offer two visits per day. So, in practice, this limits patient treatment to one or two long-dwell CAPD exchanges, or APD with either a dry day or a long daytime exchange (continuous cycling peritoneal dialysis (CCPD)) [81]. Thus, the more physically active older patients with a good appetite or having limited residual kidney function will more often require APD or CCPD, whereas those who are less active with residual kidney function can be considered for one or two CAPD exchanges/day. The PD prescription can then be further individualised by taking into consideration volume status, cardiac reserve and by predicting peritoneal solute transport rate from equations [82]. So that the duration of the PD treatment, number of cycles and glucose concentration for APD can be predicted as patients initiate dialysis, rather than waiting several weeks before performing a formal peritoneal equilibrium test. Similarly, the use of icodextrin exchanges for CAPD will depend on volume status and faster peritoneal transport. Although icodextrin exchanges result in greater volume control than standard glucose dialysates [83], the lymphatic absorption and metabolism of icodextrin vary between individuals [84]. As such, standard laboratory measurements of serum sodium using indirect sodium electrode analysis are characteristically lower in patients using icodextrin dialysates due to the interference of icodextrin metabolites. So, especially when prescribing two icodextrin exchanges to an elderly patient, measurement of serum osmolality is useful to determine whether a low laboratory serum sodium is due to the interference of icodextrin metabolites [85], a relative excess of water [86], or volume depletion [87]. If an elderly, frail patient becomes unwell and reduces their intake of fluids and food, but continues dialysing with icodextrin, then they can rapidly become volume-depleted. Therefore, it is important that, in addition to simply performing PD exchanges, the assisted PD team regularly records patient blood pressure and weight and feeds back to the supervising PD centre so that alterations can be made to PD prescriptions.

8. Conclusions

Many elderly, frail patients with progressive chronic kidney disease and additional co-morbidities can now be considered suitable candidates for PD; however, as they may not be able to manage their own exchanges, they require the support of an assisted service or family members to perform one or two CAPD exchanges/day or to connect and disconnect them to an APD cycler. As many of these patients starting PD have residual kidney function and often follow a restricted diet while being relatively inactive, the rate of generation of uraemic toxins can be low. Consequently, less peritoneal dialysis is required compared to younger, more active, healthier patients. As such, they may potentially not need treatment every day or, similarly, fewer daily PD exchanges, thus enjoying a better quality of life.