A Patient-Centered Vision of Care for ESRD: Dialysis as a Bridging Treatment or as a Final Destination?

Do All Patients with ESRD Benefit from Best Practice Guidelines?

During the past decades, numerous clinical practice guidelines aiming to improve and harmonize dialysis procedures and overall medical care for dialysis patients were developed. These guidelines are mainly based on epidemiologic data, cohort studies, trials with secondary end points, and expert opinion, rather than on well powered trials with hard end points.³ Although these limitations are acknowledged in the guidelines,³ the guidelines are frequently used as a quality standard applicable to all patients with ESRD, irrespective of their age, prognosis, and individual preferences, and as such are widely endorsed by policy makers and regulatory agencies.

Patients with ESRD experience high levels of mortality, morbidity, and distressing symptoms (Table 1). One-year mortality rates after dialysis initiation range from 10% for patients aged 20–44 years up to 41% for patients aged≥75 years. Dialysis patients are frequently hospitalized, on average 11.8 days per year.¹ Chronic pain, fatigue, cognitive impairment, and depression are present in up to half of patients with ESRD.^4,5 There exists, however, considerable heterogeneity in outcome among different patient groups.^1,6,7 For example, 25% of 70-year-old patients starting dialysis in the United States survive <9 months, whereas 25% survive for >4 years. The survival benefit of dialysis treatment may be limited for patients with severe comorbidity or ischemic heart disease.^6,8

Clinical trials in dialysis highlight important limitations of existing treatment strategies. The conventional wisdom that dialysis treatment could compensate for native kidney function and thus reduce morbidity and mortality for all patients with ESRD has been tempered by findings from the Initiating Dialysis Early and Late (IDEAL) and Hemodialysis (HEMO) trials.^9,10 For example, in the IDEAL trial, patients randomized to early start of dialysis, compared with those randomized to late or uremic symptom–triggered start of dialysis, were dialyzed an average of 5.6 months longer at an average expense of $18,715 per patient, without improvement in mortality, morbidity, or quality of life.⁹ In the HEMO trial, increasing the dosage of thrice-weekly dialysis or the use of high-flux dialyzers had limited, if any, health benefits for the dialysis population as a whole.^10,11 On the other hand, observational data and two clinical trials suggest that more frequent daily or nocturnal dialysis may improve self-reported quality of life and left ventricular mass, whereas a third trial of frequent nocturnal dialysis failed to find similar results.^12–14 These potential favorable effects should be balanced against the burdens, risks, and as yet undetermined effects on mortality. Three randomized controlled trials of online hemodiafiltration revealed conflicting results.¹⁵

Similarly, several interventions targeting ESRD-related complications have produced mixed results. Compared with a strategy of lower hemoglobin targets with restricted use of erythropoietin-stimulating agents, higher hemoglobin targets with more liberal use of erythropoietin-stimulating agents reduce transfusions and modestly improve quality of life, but do not reduce mortality and may increase the risk for stroke. Dietary phosphorous restriction and phosphate binders are difficult for patients to adhere to, and the appropriate intensity of treatment has not been validated in a randomized clinical trial.

Despite evidence of substantial need, there has been less emphasis on palliative and other supportive interventions to address symptom burden, rehabilitation, and end-of-life care in patients with ESRD. Palliative care is defined as care that aims to relieve suffering and improve the quality of life for patients with advanced illnesses and their proxy.^4,16 This care intends to neither hasten nor postpone death, and should be offered simultaneously with all other medical treatment.^4,16 Guidelines from the Renal Physicians Association as well as the Kidney Disease Improving Global Outcomes initiative address some of these needs^3,17; however, these recommendations have not yet been incorporated into quality indicators.

A Patient-Centered Vision of Care for All Patients with ESRD

When a patient develops ESRD, he or she is confronted with a likely, but not invariably, life-threatening disease. Therapy is burdensome, and the best attainable goal is frequently care endeavoring a deferment of morbidity and mortality and an improvement or maintenance of quality of life.^17,18 Although the risks and burden of many health care interventions are immediate, the benefits may be delayed or may not even materialize for patients with a poor prognosis.^18–20 Consequently, life expectancy is an important element of the risk/benefit profile of most dialysis-related interventions.^18–21 Moreover, the most appropriate care for patients with complex multimorbidity, as is the case for many dialysis patients, may differ substantially from the sum of the most appropriate care strategies of each discrete condition.^17–19,21

In a patient-centered vision of care, three important elements are combined to determine an “attainable treatment goal” for each individual patient: namely, a thorough knowledge of the patient’s medical condition and treatment options; his or her preferences, expectations, and psychosocial context; and his or her prognosis. We propose to distinguish three attainable treatment goal groups: (1) dialysis as a bridging treatment or long-term maintenance treatment, (2) dialysis as a final destination treatment when a bridging or long-term maintenance strategy is no longer feasible and dialysis is intended as short-term support, and (3) active medical management without dialysis. This framework is intended as a means to help clinicians prioritize treatment decisions when the existing evidence base is weak and/or not generalizable to all patients with ESRD, analogous to a therapeutic framework used for left ventricular assist devices in patients with refractory heart failure. It is important to emphasize that this stratification is not intended to be self-fulfilling, but rather to provide clinicians and patients with therapeutic beacons. A single patient may evolve from one group to another during his or her treatment course as a function of changes in disease state and preferences, as will be the case in a time-limited trial of dialysis, or after a decision to stop dialysis treatment. Some patients may also be willing to trade longer life expectancy for less burdensome treatment that preserves their well-being and quality of life.²¹ Table 2 provides an example of a possible interpretation of this framework. Ideally, guidelines should suggest the optimal care for each group, on the basis of evidence with clear mortality, morbidity, and quality-of-life end points.¹⁷ In the absence of such data, guidelines might indicate the lag time to benefit for various interventions, so that clinicians can judge whether patients with a more limited prognosis still could expect a benefit. Rates of dialysis withdrawal may approximate a lower bound for the proportion of patients who are candidates for dialysis as a final destination. In the United States, the rate of dialysis withdrawal is between 2% and 9% per year, depending on age, and withdrawals precede approximately 20% of deaths.²²

Consider a 73-year-old woman with oxygen-dependent pulmonary fibrosis, pulmonary hypertension, and advanced heart failure who has been hospitalized multiple times in the past year, culminating in ESRD. Although many patients in these circumstances will choose dialysis in order to increase longevity, some patients may prefer active medical management without dialysis. If she chooses dialysis, the appropriate intensity of dialysis and other medical therapies will depend on the burden of the treatment and whether the intervention is for treatment versus prevention. Some patients may wish to receive interventions with modest or uncertain survival benefit because this may give them hope and the impression of control. Other patients in these circumstances may feel the gain in quality and quantity of life with the former treatment approach is small, and may desire minimizing treatment burden as an equally important goal. For example, pulmonary rehabilitation and more frequent dialysis for volume control could produce benefit in the short to intermediate term, but these strategies also carry significant burdens. Conversely, dietary phosphorus restriction and treatment of hyperphosphatemia and secondary hyperparathyroidism are mainly for prevention; these therapies might be relaxed or even discontinued. Conversion of a tunnelled hemodialysis catheter to an arteriovenous graft may reduce risk for infectious complications, but the patient’s preferences and limited prognosis should also be factors in the decision-making process. A colonoscopy to assess iron-deficiency anemia in this patient with heme-positive stool is unlikely to improve her longevity or quality of life, whereas higher hemoglobin targets with erythropoietin therapy may alleviate dyspnea.

Consider another 73-year-old active woman without severe comorbidity who develops ESRD after infarction of a solitary kidney. For this patient, dialysis as long-term maintenance treatment, or possibly as a bridge to transplantation, is an attainable goal. Given her favorable prognosis, existing guidelines for dialysis adequacy, vascular access, anemia and mineral bone disease management, and cardiovascular risk reduction are more likely to improve her short- and long-term health outcomes.

Challenges in Achieving Patient-Centered Care for Patients with ESRD

The proposed patient-centered approach offers a framework for all patients with ESRD to live their lives as fully as possible within the restrictions imposed by their illness and treatment. To achieve this vision of care, several real and perceived barriers must be addressed.²¹

First is the concern that existing prognostic tools are not sufficiently accurate to permit allocation of patients into the appropriate treatment goal groups. Validated tools to predict prognosis exist for patients receiving dialysis.²³ For patients approaching ESRD, geriatric syndromes such as frailty (defined by three of the following five features: unintentional weight loss, self-reported exhaustion, weakness, slow walking speed, and low physical activity) may help to refine prognostic estimates based on age and comorbidity.⁵ By incorporating patient preferences in addition to predicted prognosis, clinicians can ensure that goals are both attainable and desired by patients. Second is the concern that the goals of treatment are too patient specific to be measurable (and by implication, valuable).²¹ The National Institutes of Health recently identified universal, patient-centered outcome measures that could be objectively measured and routinely recorded in healthcare delivery for patients with multiple chronic conditions. Third, nephrologists need to receive education and training in communication skills to initiate discussions about goal-based care.¹⁷

Finally, well designed trials in patients with limited life expectancy are needed to validate this new model of care. These trials should include patient-centered end points such as functional status, symptom burden, and quality of and satisfaction with life, in addition to mortality. Rather than exploring the effect of single disease-centered interventions, clinically relevant treatment strategies should be compared. For example, a strategy of less stringent BP, glycemic, lipid, and phosphorous control could be compared with a strategy of usual care based on clinical practice guidelines, or usual care could be compared with usual care plus palliative care.

Since its introduction, maintenance dialysis treatment gradually evolved from a means to bridge relatively young and fit patients with ESRD to transplantation, into a universal treatment option for a much more heterogeneous patient population with ESRD, including a growing proportion of elderly patients or patients with multiple and severe comorbidities for whom dialysis is their final treatment destination. In these latter groups, the burden of dialysis and associated treatments as proposed by current clinical practice guidelines is frequently high, and the evidence supporting its beneficial effects on mortality, morbidity, and quality of life is variable, but sometimes rather weak. This effect may be partially mitigated by integrating the patients’ preferences and their overall prognosis into medical decision making. When caring for an individual patient, the question should no longer be “What interventions are available?”; rather, “What treatment goals are attainable and desired for this patient, and how can they be translated into medical care for this individual patient?” This small difference in approach could make a world of difference in the quality of life of many patients with ESRD.