End-of-Life Health Care Use Among Socially Isolated Older Adults with Cognitive Impairment

TitleEnd-of-Life Health Care Use Among Socially Isolated Older Adults with Cognitive Impairment
Publication TypeJournal Article
Year of Publication2022
AuthorsKotwal, A, Cenzer, I, Smith, A
JournalJournal of Pain and Symptom Management
ISSN Number0885-3924
Keywordscognitive impairment, End-of-life care, Medicare, socially isolated

Outcomes 1. Understand how social isolation is defined and should be considered among older adults with cognitive impairment. 2. Understand how social isolation and cognitive impairment can lead to reduced healthcare use at the end of life and the role of clinicians in addressing social needs Original Research Background At the end of life, older adults may rely on others to access healthcare, particularly if they have impaired cognition. Research Objectives To determine whether socially isolated older adults with and without cognitive impairment have different patterns of end-of-life healthcare use. Methods We used nationally representative 2006-2016 Health and Retirement Study data linked to Medicare claims to examine adults age >50 interviewed once in the last 4 years of life (N = 2,073). We measured three self-report subscales of social relationships: household contacts, social network interaction, and community engagement. The three subscales were combined to create an overall social isolation measure. End-of-life health outcomes included emergency department (ED) visits, hospitalizations, ICU stays, and hospice use in the last months of life. Cognitive impairment (CI) included CIND or dementia. We used logistic regression to test the adjusted association of each social measure with each end-of-life outcome and tested for significant interactions with CI (p < 0.1). Results Our sample had a mean age of 82 (52% female, 9% Black, 5% Hispanic, 47% had CI). There were significant interactions between social isolation or network interaction and CI for all end-of-life outcomes; social isolation was associated with lower hospice use (aOR = 0.61, p = 0.05), ED use (aOR = 0.55, p = 0.03), ICU use (aOR = 0.62, p = 0.05), and hospitalizations (aOR = 0.63, p = 0.04) for patients with CI, whereas there was no association for those with no CI (hospice, aOR = 1.1; ED, aOR = 1.1; ICU, aOR = 1.0; hospitalizations, aOR = 1.1). Conclusion Cognitively impaired older adults who were socially isolated or had reduced social network interaction had fewer ED visits, hospitalizations, ICU stays, and hospice use at the end of life. Implications for Research, Policy, or Practice Cognitively impaired, isolated older adults may struggle to access end-of-life healthcare. Further research is needed to determine whether low rates of acute care and hospice are concordant with their wishes or are due to a lack of support in accessing health services.

Citation KeyKOTWAL2022867