Advance Directives and End-of-Life Care among Nursing Home Residents Receiving Maintenance Dialysis


Background and objectives
Little is known about the relation between the content of advance directives and downstream treatment decisions among patients receiving maintenance dialysis. In this study, we determined the prevalence of advance directives specifying treatment limitations and/or surrogate decision-makers in the last year of life and their association with end-of-life care among nursing home residents.

Design, setting, participants, & measurements
Using national data from 2006 to 2007, we compared the content of advance directives among 30,716 nursing home residents receiving dialysis to 30,825 nursing home residents with other serious illnesses during the year before death. Among patients receiving dialysis, we linked the content of advance directives to Medicare claims to ascertain site of death and treatment intensity in the last month of life.

In the last year of life, 36% of nursing home residents receiving dialysis had a treatment-limiting directive, 22% had a surrogate decision-maker, and 13% had both in adjusted analyses. These estimates were 13%–27%, 5%–11%, and 6%–13% lower, respectively, than for decedents with other serious illnesses. For patients receiving dialysis who had both a treatment-limiting directive and surrogate decision-maker, the adjusted frequency of hospitalization, intensive care unit admission, intensive procedures, and inpatient death were lower by 13%, 17%, 13%, and 14%, respectively, and hospice use and dialysis discontinuation were 5% and 7% higher compared with patients receiving dialysis lacking both components.

Among nursing home residents receiving dialysis, treatment-limiting directives and surrogates were associated with fewer intensive interventions and inpatient deaths, but were in place much less often than for nursing home residents with other serious illnesses.


More than 80,000 Americans die each year while receiving maintenance dialysis therapy for ESRD (1). Although dialysis can sustain life, it rarely restores health or independence. Surveys suggest that a majority of patients receiving dialysis would prefer care focused on maintaining comfort rather than prolonging life if they were to become seriously ill (2–4). Yet many patients with ESRD receive treatments near the end of life that are aimed at prolonging life rather than maximizing comfort, and family members rate the quality of death for patients with ESRD lower than for other serious illnesses (5,6).

Advance directives are often promoted for patients with serious illness such as ESRD as a means to avoid interventions that are unwanted or of limited benefit. Although recommended in practice guidelines (7), available data suggest that only one in three patients with ESRD has completed an advance directive (8–10). It is unclear whether the low prevalence of advance directives in this population reflects lack of engagement in advance care planning or patient preferences for aggressive care (2). Because of uncertainty about the reasons for low use of advance directives and a lack of compelling data regarding their effectiveness in this population, efforts to expand advance care planning among patients with ESRD have not gained traction.

We sought to address these knowledge gaps by studying nursing home residents who are receiving maintenance dialysis. More than one third of Medicare beneficiaries with ESRD reside in a nursing home near the end of life. Among nursing home residents without ESRD, two thirds have an advance directive and more than half have a treatment-limiting advance directive (11). Although it is recognized that nursing home residents with ESRD have limited life expectancy (12), it is not known how frequently they complete advance directives requesting treatment limitations or naming a surrogate.

We used a national registry of nursing home residents linked to Medicare claims to determine (1) the content of advance directives that were in place near the end of life among nursing home residents receiving dialysis versus patients with other serious illnesses; (2) among patients with ESRD, whether having a treatment-limiting directive and surrogate decision-maker were associated with less intensive end-of-life care; and (3) how often patients with ESRD who had a treatment-limiting directive received care that was consistent with their advance directive.