Older Adults With Functional Limitations and Their Use of Telehealth During COVID-19


Yijung K. Kim1  & Shannon Ang2 

1Texas Aging & Longevity Center, The University of Texas at Austin

2School of Social Sciences, Nanyang Technological University


Published in Research on Aging

Highlights

Overview

The COVID-19 context changed the circumstances of many older adults who may not have needed to go online before the pandemic struck. Before 2020, pre-pandemic restrictions (e.g., geographic, administrative) around Medicare-eligible telehealth services had made it difficult to assess this relationship consistently. With access to telehealth services relatively equalized across the country due to COVID-19, we take this opportunity to assess how prior Internet use experience is associated with digital health equity among older adults with functional limitations. Data analysis using the National Health and Aging Trends COVID-19 Supplementary Survey on Medicare beneficiaries 70 years and older show that every additional functional limitation was associated with a 3-percentage point decrease in the probability of using the Internet, and 2-percentage point increase in the probabiltiy of using video-based telehealth services. Past internet use and receiving informal help with technology from family and friends also predicted Medicare beneficiaries' telehealth usage during the early phase of the COVID-19 pandemic. This project shows healthcare access may have been more difficult for older adults with functional limitations during the COVID-19 pandemic, especially for those with little prior experience with the Internet, or those without friends/family to provide technological support.

Challenge

The novel coronavirus (COVID-19) outbreak in 2020 put the promises and pitfalls of digital health care to the test, as social distancing measures forced healthcare providers to postpone or shift most routine medical visits online1. To facilitate older adults’ transitions to telehealth during the public health emergency, the Centers for Medicare & Medicaid Services (CMS) expanded the scope of remote medical visits covered by Medicare and lifted existing geographic or location restrictions, leading to many welcome changes in attitudes toward and structures surrounding telehealth2. As a result, early reports estimate that more than one in four community-dwelling Medicare beneficiaries overall had a telehealth visit between the summer and fall of 20203. The rate is similarly high for the Medicare beneficiaries 70 years and older, with studies showing that the adoption of video-based telehealth grew from ~4% pre-pandemic to ~21% in 20204. However, there are growing concerns that a rapid expansion of technology could further marginalize older adults with functional limitations, who often contend with other forms of social exclusion5-7. Older adults who are homebound, institutionalized, or have greater limitations in physical capacity are less likely to own Internet-enabled devices or go online8, less likely to use e-mail/text messaging for communication9, and more likely to discontinue using everyday technology over time10, compared to those older adults without comparable health limitations. If harnessed well, telehealth or other forms of online services have the potential to improve health care access and quality for individuals experiencing mobility restrictions or transportation challenges4.

This project examined whether the social distancing measures have changed the digital participation among older Medicare beneficiaries with functional limitations and whether prior Internet use and informal social support served as essential resources.


Method

The National Health and Aging Trends Study (NHATS) is a nationally representative longitudinal cohort study of Medicare beneficiaries 65 years and older, designed to assess late-life disability and functioning trends through annual in-home interviews. This study employed data from the NHATS COVID-19 survey, combined with some information (e.g., Internet use, income) from the three most recent rounds of the NHATS main surveys (2018-2020), sensitive demographic files, and COVID-19 Family and Friends survey. Because NHATS last recruited a cohort of Medicare beneficiaries aged 65 years and older in 2015, the data for this study comes from Medicare beneficiaries who were 70 years and older in 2020. Of 3,257 NHATS respondents who completed the COVID-19 survey, 106 respondents (3.2%) were missing information on key demographic, health, and technology use variables. For the first three research questions, we handled missing data by assigning zero weights to those with missing information. For example, those participants with missing information (or the response is marked as refused or don’t know) in their 2018 or 2019 Internet use variables were dropped from the analytic sample. Therefore, the final analytic sample size for the first three research questions was 3,151 NHATS respondents with complete data.To examine the associations between the extent of functional limitations and the use of technology around the times of COVID-19, we estimated a series of logistic regression models using NHATS COVID-19 survey weights. The four outcomes were: a) Access to a computer; b) Use of the Internet; c) Use of video calls to communicate with a healthcare provider; and d) Use of emails or texts or portal messages to communicate with a healthcare provider. We then assessed the moderating role of recent Internet use on the association between the extent of functional limitations and use of telehealth (i.e., the use of video calls and emails/texts/portal messages to communicate with a healthcare provider).


Of 3,257 NHATS COVID-19 respondents, 46% (n = 1,458) had a family member or friend complete the COVID-19 Family and Friends survey on their role during the pandemic. We conducted unweighted analyses on this subsample of 1,458 NHATS respondents to assess the moderating role of technology support. We examined the moderating role of informal technology support on the association between the extent of functional limitations and use of telehealth. This analysis was unweighted but included the same set of covariates as the previous models. STATA 16.1 was used for all data analysis.

Results

Sample Characteristics

Respondents on average were about 78 years old, 56% female, 54% married, 81% non-Hispanic White, and 62% had some college education or more. About 36% reported a presence of functional limitations in 2020, and individuals with functional limitations were at a socioeconomic disadvantage (e.g., less income, fewer years of education) relative to those without. Most respondents (94%) had access to a computer and used the Internet.


Functional Limitations and Technology Use

Results of logistic regression analyses for the extent of functional limitations and technology use are presented in Figure 1 as average marginal effects. Functional limitations were negatively associated with having access to a computer and using the Internet in later life. We found that on average, every additional functional limitation was associated with a 1 percentage point decrease in the probability of having access to a computer. Every additional functional limitation was associated with a 3 percentage point decrease in the probability of using the Internet. Functional limitations were positively associated with the use of video calls to communicate with a healthcare provider usage during COVID-19, such that every additional functional limitation was associated with a 2 percentage point increase in the probability of using video calls. 

Figure 1. Average marginal effects for logistic regression models predicting computer access, use of the Internet, and use of telehealth services (i.e., video calls, text/email/portal) to communicate with a healthcare provider during COVID-19; N=3,151 

The Role of Recent Internet Use 

There were no statistically significant associations between the extent of functional limitations and using emails/texts/portal messages for those with any Internet use (consistent or inconsistent) in the past two years, but the probability of using emails/texts/portal messages was higher for individuals without Internet use in the past two years (Figure 2). For those without Internet use in the past two years, every additional functional limitation was associated with an increase of 1 percentage point in the probability of using emails/texts/portal messages. Net of functional limitations, those with consistent Internet use were more likely to use emails/texts/portal messages to communicate with a healthcare provider.

Figure 2. Predicted probabilities of using emails/texts/portal messages to communicate with a healthcare provider during COVID-19, varying by the number of functional limitations and recent Internet use. Shaded area indicates confidence interval. AME=average marginal effects. **p <.01. N=3,151 

The Role of Technology Support from Family and Friends

On average, respondents whose family and friends completed the COVID-19 survey (n=1,458) were more likely to be older, married, non-Hispanic White, receive college education, and have more functional limitations, health conditions, and cognitive impairments than their counterparts. For this subgroup of respondents, we examined technology support from family and friends as a moderator. Informal help with technology moderated the association between functional limitations and the use of emails/texts/portal messages; every additional functional limitation was associated with a 4 percentage point increase in the probability of using emails/texts/portal messages among those who were receiving technology help from family and friends (Figure 3).

Figure 3. Predicted probabilities of using emails/texts/portal messages to communicate with a healthcare provider during COVID-19, varying by the number of functional limitations and technology support from family and friends. Shaded area indicates confidence interval. AME=average marginal effects. **p <.01. N=1,458 

Conclusion

Our study shows that older adults with more functional limitations were less likely to have access to computers and use the Internet less than their counterparts with fewer or no functional limitations. More functional limitations were associated with a higher probability of using video calls to communicate with a healthcare provider during COVID-19, but in light of the results on Internet access and utilization, it is unclear whether upticks in video-based telehealth usage will remain when the national health emergency is over. Our findings further show that the general divide between Internet users and Internet non-users extended to the use of emails/texts/portal messages to communicate with a health care provider during COVID-19, and older adults with functional limitations were more likely to use emails/texts/portal messages to communicate with a healthcare provider if they received technical support from family and friends. Overall, these results underscore the importance of addressing digital exclusion within the older adult population and suggest that ensuring social support is imperative to digital participation among older adults with functional limitations.

Limitations

Our study is foremost limited by the use of cross-sectional data. Although the findings are mostly in line with prior research that explored trends in digital access among older adults with serious health problems or mobility restrictions8, we do not claim any causal effects. Second, the current study is limited by the NHATS survey design, which assessed a cohort of Medicare beneficiaries aged 65 years and older in 2015 and has not added any new cohort since then. As such, our weighted results are generalizable to an even more specific subset of Medicare beneficiaries who were 70 years and older in 2020 (b.1950 or earlier), who were physically and cognitively able to complete a comprehensive survey. Lifetime exposure to and adoption of technology differ greatly by birth cohorts, and whether the association between the extent of functional limitations and the use of telehealth services remain consistent for the younger cohorts needs to be explored in future research. We further note that our study likely overestimated the proportion of older adults who have access to a computer because we included cellphone ownership as one of the indicators for computer access. Our data shows that at least 57 individuals who owned a cellphone but did not own a computer or a tablet were using the Internet in 2020, but it is unclear what fraction of working cellphones were in fact Internet-enabled devices. It is also possible that receipt of informal support is underreported, as respondents may have received support from those who did not complete the Friends and Family supplementary survey.  Lastly, while we explored four different indicators of digital access and utilization, we do not know how often or to what extent our sample used different functions of the Internet in more detail. Future research that uses multiple time points and includes more information on the type of telehealth services provided will be better suited to parse out mechanisms underlying the digital divide in later life.

References