Quality Indicators for Palliative and End-of-Life Care in Vulnerable Elders

Only a century ago, dying often quickly followed the onset of disease or injury, but today many patients and families struggle with the consequences of chronic illness and failing function for years before death. Leading causes of death include cardiovascular disease, malignancies, and infections, with nearly half dying with three or more conditions. Of the 15% of U.S. adults living with such conditions, one-twelfth have severe cognitive impairment, one-fifth have impaired vision, and one-third have difficulty walking. Because the exact timing of death is typically uncertain, medical care should routinely address palliative goals for the broad group of patients living with serious illness. Prognostic tools usefully characterize some populations (e.g., heart failure), but the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments (SUPPORT) demonstrated that many patients with serious, even fatal, conditions have substantial probability of 2- or 6-month survival, even in life's last week. A higher risk of death (i.e., an expected death or one associated with repeat hospitalizations, reduced function, and severe symptoms), rather than certainty about its timing, should trigger consideration of palliative principles. Despite increasing attention paid to end-of-life care in recent years, many studies have described difficulties in the final phase of life, including problems with access to hospice, inadequate symptom management, caregiving burdens, care mismatched with patient preferences, and inappropriate resource use. This set of quality indicators (QIs) broadly addresses palliative and end-of-life care as it applies to vulnerable elders (VEs). Several of the QIs in this article address pain and other aspects of suffering and the interrelationship between these QIs and those in other articles, such as pain management, and their applicability according to clinical condition and venue. METHODS: A total of 299 articles were considered in this review: 74 identified using a Web search, 120 through reference mining, 50 through the Assessing Care of Vulnerable Elders (ACOVE)-3 literature searches, and 55 from reference mining the ACOVE-1 monograph. RESULTS: Of the 24 potential QIs, the expert panel process judged 21 to be valid; three indicators were rejected. The literature summaries that support each indicator judged valid by the expert panel are described.