Informing Our Elders About Dialysis: Is an Age-Attuned Approach Warranted?

Impact of Age, Functional Status, Comorbid Medical Conditions, and Dialysis on Survival

The trajectory of illness is variable and unique to CKD, and neither renal failure nor its sequelae is predictably the terminal or even sentinel event. For patients on dialysis, the presence of comorbid conditions is the single most predictive factor of outcome and is additive with age in predicting survival (10–14). Average first year mortality for dialysis patients over the age of 65 years is 46%, with those individuals older than 85 years having 22% greater likelihood of death in the first year (2) and a median survival of 11.6 months compared with 24.9 months for individuals between 66 and 79 years. A high prevalence of frailty has been noted among older patients on dialysis (15), and it is associated with a threefold higher risk of death, supporting a primary role for frailty as a predictor of outcome. Marked functional impairment, a history of falls, and the inability to transfer are other indicators of poor prognosis (16,17) along with hypoalbuminemia, a strong predictor for mortality in patients on dialysis (13,18–20).

Several methods of predicting outcomes in CKD have recently been advanced. A clinical score using nine comorbid conditions to predict 6-month prognosis in patients over 75 years starting dialysis found that individuals with the highest scores had the highest mortality, with rates as high as 70% (21). The surprise question, “Would I be surprised if this patient died in the next 12 months?,” has been recognized as a simple and reliable method of identifying sicker incident dialysis patients with high risk for early death (22). Cohen et al. found five variables to be independently associated with early mortality: older age, dementia, hypoalbuminemia, diagnosis of peripheral vascular occlusive disease, and negative response to the surprise question were used to develop an integrated 6-month prognostic tool that was validated for the prevalent hemodialysis population (23). Among octogenarians offered dialysis, predictors of death within 1 year included poor nutritional status, functional dependency, and late referral (24).

Finally, a growing amount of literature suggests that, for patients with a poor prognosis, RRT may not confer a significant survival benefit or prevent significant functional deterioration during the first year on dialysis (10–14). For patients over 75 years with characteristics suggesting a poor prognosis (Table 3), studies comparing dialysis with active medical management without dialysis do not uniformly show benefit to survival with dialysis (7,10–14), and the survival advantage, when present, may come with substantial burdens that negatively affect quality of life.

Likelihood of Renal Disease Progression before Death

Older patients lose renal function at slower rates than their younger counterparts, have a low incidence of progression to ESRD, and have shorter survival (6,24–29). Many older CKD patients die long before reaching dialysis because of significant cardiovascular mortality with higher stages of CKD (6,24–29). Older adults were 13 times more likely to die of any cause than progress to ESRD and six times more likely to die of cardiovascular disease than develop ESRD (27). The decline in renal function was slowest in CKD stage 4 patients over 75 years of age, and the higher the age, the lower the likelihood of living long enough to require dialysis (26). Irrespective of baseline renal function, patients over 85 years of age were more likely to die before progressing to the need for dialysis (6), and 30% of octogenarians followed over 4 years died of nonrenal causes before RRT (28). More than one-third of patients over 80 years of age were found to have stable nonprogressive CKD, which was predicted by low levels of proteinuria, lack of hypertension, and low cardiovascular comorbidity (29). Thus, many older patients as well as those patients without deteriorating renal function or proteinuria may not progress to ESRD and the need for RRT.

Burdens of Dialysis and Risk to Quality of Life

For most patients considering RRT, the impact of dialysis on day to day life may be difficult to envision. The symptom burden in patients with ESRD is high, and along with increased comorbidity, it is linked to reduced quality of life and depression (30,31), also common in ESRD (21). In addition to the occurrence of adverse physical symptoms while on dialysis, many patients experience multiple interdialytic symptoms, the aggregate of which can cause difficulty with activities of daily living (31). Furthermore, patients living distant to a dialysis unit must add the time and expense of traveling to and from dialysis, the long hours spent on dialysis, and a reduction in the time available for other activities (11–14).

The burdens of dialysis also include surgery for vascular or peritoneal access placement and their potential surgical complications. Primary maturation failure of the arteriovenous fistula (AVF) is reported to be as high as 60%, with the highest rates observed in older patients, females, nonwhites, and individuals with cardiovascular disease (32,33). An important prerequisite for achievement of a successful AVF is the presence of suitable vascular anatomy, the likelihood of which decreases with age and presence of many conditions common to patients with ESRD (32–35). Requiring all patients with advanced CKD, irrespective of age, to undergo AVF placement may be for naught in those individuals who are unlikely to live long enough to need their AVF for RRT. Faced with the possibility for little to no survival advantage with dialysis, older patients may be less accepting of such required clinical practices as well as the inter- and intradialytic symptoms that commonly accompany dialysis.

Clinically Meaningful Outcomes Are Relative

Dialysis can sustain life, but for patients suffering many of its indignities, dialysis may be viewed as merely a forestalling of death. Such a “clinically sustained existence,” as characterized by Kaufman et al. (8), fosters a sense of prolongation without progress—a place somewhere between getting neither better nor worse. This dilemma highlights the need for emphasis on clinically meaningful outcomes, the focus of which may be quality of life rather than quantity of survival.

Although the effects of initiating dialysis on patient status will vary between patients, current research suggests that functional status in older individuals is likely to decline rather than improve (16,36). The anticipation that chronic comorbid conditions will require hospitalization and the desire to die at home (or fear of dying while hospitalized) may be important considerations (9). Patients choosing dialysis over maximum conservative management survived three times as long but had higher hospitalization rates and were less likely to die at home (11); 1- and 5-year survival rates of veterans >75 years of age on dialysis were 53% and 2.4%, respectively, although approximately 20% of time was spent hospitalized (37). Comorbidity did not independently predict mortality, and survival on dialysis was not significantly different from survival of patients choosing not to pursue dialysis; however, significantly more patients choosing dialysis died while in the hospital (65%) than individuals who chose the nondialytic course (10). Disability and functional loss requiring caregiver or nursing home support were reported in more than 30% of octogenarians within 6 months of initiation of dialysis (36). For older patients suffering the loss of independence and other indignities of advancing age, meaningful components of the informed consent process may include symptom management, dying at home, and avoidance of hospitalization rather than duration of survival.

Informing Prospective Dialysis Patients about the Contingencies of Their Consent

The initiation of dialysis presumes the appropriate provision of informed consent (Table 4). Ideally, the process of informed consent begins as part of ACP long before a patient is faced with a decision about RRT. The importance of ACP is underscored by the high rates of withdrawal from dialysis (7), which is second only to cardiovascular disease as a cause of death and accounts for one-quarter of dialysis patient deaths (3).

The complexities of ACP for patients with kidney disease are of formidable magnitude (9) and confounded by preferences about dialysis that may change over time (38). Elderly patients evaluated before treatment for depression were rarely interested in dialysis as a life-sustaining therapy when depressed; after treatment, patients showed increased desire to pursue a variety of life-sustaining therapies subsequent to improvement in their depression (39). Factors influencing an older patient’s decision to start dialysis may be heavily weighted on functional status, the cognitive ability to appreciate quality of life, or even the perception (right or wrong) of the dying experience should they choose to forego dialysis. Low levels of cognitive function are highly prevalent in older patients with CKD (40); thus, many older patients may be less able to comprehend the magnitude of their disease, let alone the trappings associated with the day to day experiences of dialysis. Among elderly hemodialysis patients interviewed to determine whether informed consent for performing dialysis had been obtained, most of the patients lacked sufficient understanding and knowledge of their circumstances (41). Patients often do not understand the options (or the options presented may not include the option to forego dialysis), and many feel as if they have no choice (8). In situations of medical emergency, the presumed certainty of death is likely to trump any risk to benefit considerations that, under elective circumstances, might result in a different decision. Ironically, patients rarely include dialysis in the category of life support (9), although some view the act of refusing a life-sustaining treatment (as opposed to foregoing a treatment that may or may not prolong life) as ethically or morally wrong. Most patients wish to receive detailed information about their condition to be prepared and plan ahead in case of death (9), and the vast majority desire information about their life expectancy and prognosis, whether positive or negative (42). A high percentage of patients, however, have not had these discussions with their nephrologists (43).

The increasing numbers of older patients starting dialysis, coupled with the persistently high withdrawal rates, might be thought to reflect a passive acceptance on the part of the patient, a lack of good judgment on the part of the physician, or both. Some patients with good cognitive function choose death over dialysis (44), citing satisfaction with the totality and finality of their life. An estimated 5–7% of patients with advanced CKD chose to forego dialysis or were not offered dialysis by their nephrologists (45). Significantly more patients with advanced CKD were not referred for dialysis at all by primary care physicians (22%), who cited age as the primary reason for withholding treatment (60%) (45). By contrast, principle factors influencing the decision to not offer dialysis to octogenarians included late referral, social isolation, and poor functional status (24). The lack of data on frequency of withholding dialysis notwithstanding, patients making the informed choice to forego RRT and opt for medical management deserve a clear treatment plan with timely arrangements for active medical treatment during the end of life.

Informed Consent for the Dialysis Procedure Is Not Enough

Nephrologists’ ability to predict outcomes has been significantly enhanced with the available predictive tools and evidence-based guidelines that can be used when counseling older patients about the burdens, risks, and benefits of dialysis (7) (Tables 1 and 2). Prognosis of the individual patient is ultimately uncertain, however, and the ability to predetermine which patients will most likely benefit from dialysis remains a challenge and an area of needed research (17,46). This prognostic ambiguity should not limit but rather, encourage open discussions of the implications of either choice and as Swidler further advocates, communication processes should address and update goals of care as the effect of dialysis on a patient’s status becomes apparent: the “trajectory of decline progresses and end of life issues become more prominent” (47). When contemplating the propriety of dialysis, patients and nephrologists alike must anticipate and adapt to changes in patient status to assure an informed consent process that is individualized and dynamic. Unique aspects of ACP for patients with CKD include, in addition to the availability of replacement therapy for the failing organ, the opportunity to accept or refuse its implementation, a choice that is increasingly invoked (3). The finding that patients with nephrology care before starting dialysis had 50% lower risk of death because of withdrawal and cardiovascular disease in the first 120 days of dialysis (48) suggests a benefit of proactive discussion between patient and nephrologist. Specific issues to address include estimates of survival, impact of dialysis on daily life, projections of change in functional status, and alternatives of nondialytic management of progressive CKD (Tables 5 and 6). When medical management is chosen in lieu of dialysis, ACP discussions that outline the activity of CKD treatment and the implementation of patients’ goals are recommended (7).

Disclosures

None.