Association of Social Isolation and Loneliness with Telehealth Use Among Older Adults During COVID-19

Context: The COVID pandemic has renewed attention on social isolation and loneliness among older adults, while also rapidly accelerating the use of telehealth. It is unclear how social isolation and loneliness, which have previously been linked with poor health outcomes, may be associated with telehealth use.

Objective: To assess the rates of telehealth use in a Medicare population and explore possible determinants of telehealth use, including social isolation and loneliness.

Study Design and Analysis: Cross-sectional analysis; logistic regression to test association

Dataset: National Health and Aging Trends Study (NHATS), 2020, including COVID supplement. NHATS is an annual survey of a nationally representative sample of Medicare enrollees. The COVID supplement was collected as a mail survey of NHATS participants from June 2020 to January 2021, with data released in July 2021.

Population Studied: Medicare beneficiaries aged 70 years and older who completed both core NHATS questions and the COVID supplement.

Main and Secondary Outcome Measures: Main outcome (dependent variable) is video visit usage (ever or never used). Key predictors (independent variables) are social isolation (based on Berkman and Syme’s Social Network Index) and loneliness (direct measure). Covariates are self-reported health status, mobility difficulty, and sociodemographics.

Results: Among the 2309 participants, video visit usage increased from 6% prior to COVID to 24% during COVID. The majority of participants were somewhat isolated (63%), with fewer either least isolated (18%) or most isolated (19%). Similarly, 65% of participants felt lonely on some days (65%), compared to rarely/never (26%) or most/every day (9%). Participants who are most isolated are significantly less likely to use video visits (aOR 0.49, CI 0.29-0.83). Loneliness was not significantly associated with video visit usage (most lonely aOR 0.90, CI 0.60-1.35). Covariates significantly associated with video visit usage included age 80+ (aOR 0.61, CI 0.45-0.84) and metropolitan status (aOR 1.81, CI 1.25-2.63). Gender, race, Medicaid status, health status, self-reported anxiety or depression, mobility difficulty, and census division were not significantly associated with video visit usage.

Conclusions: Socially isolated older adults and those not living in metropolitan areas are less likely to use telehealth. More work is needed to identify these adults and tailor interventions to improve access and health outcomes.