RESEARCHRacial and Socioeconomic Disparities in Access to Telehealth
Rapid implementation of telehealth has become a hallmark of our healthcare system's response to the COVID-19 pandemic. Telehealth has demonstrated feasibility and efficacy in the care of older adults including vulnerable homebound populations.1, 2
Implementation of telehealth was facilitated by both relaxation of the Health Insurance Portability and Accountability Act of 1996 (HIPAA) allowing the use of non–HIPAA-compliant platforms for video visits, as well as a temporary increase in reimbursement rates for telehealth for Medicare and Medicaid patients.3, 4 However, emerging evidence suggests that minority populations face substantial barriers to telehealth including lack of familiarity with technology, lower health and digital literacy, sensory impairments, lack of broadband access, and lack of assistance.5-10 Given the widespread implementation of telehealth as a tool to provide ongoing care to older adults during a pandemic, it is crucial to understand whether access to and use of telehealth services differed for disadvantaged groups including racial/ethnic minorities.
METHODS
We conducted a cross-sectional analysis of the electronic medical records of a large urban geriatric outpatient practice during the COVID-19 surge in New York City (NYC). The outpatient practice provides primary and consultative care to high-risk, high-need vulnerable, complex older adults, many with advanced illness. Our sample consisted of all active patients, defined as those who had a visit within the 19 months before March 1, 2020. For our analyses of patient visit data, we included all video and telephone visits completed during the peak of COVID-19 cases in NYC (March 18, 2020, to May 15, 2020).
To examine access to telehealth services, we measured whether or not a patient or proxy had activated the patient online portal (“MyChart”) within the electronic medical record required for telehealth use as of the start of the study period. Activation requires access to a mobile device or computer with Internet access and may occur either onsite with healthcare providers or at home. We estimated a multivariable logistic regression model to examine associations between patient characteristics and activation status. For the subset of patients who had activated the patient online portal before the study period and who had a visit during the study period, we estimated a multivariable logistic regression model to examine associations between patient characteristics and having a video versus telephone telehealth visit. This study was approved by the Mount Sinai institutional review board.
RESULTS
Of 4,571 total patients, one-third (35.1%) were aged 80 and younger; 21.0% were aged 90 and older. More than 70% were female (73.8%), 55.7% were White, and 18.8% were eligible for Medicaid. A total of 2,438 (53.3%) had activated their online portal and thus had the potential to have a telehealth video visit. Overall, 1,651 (36.1%) had not activated the portal, and 482 (19.8%) had declined activating the portal. Patients who had activated the portal were more likely to be younger, White, male, English speaking, married or partnered, and did not have dementia (Table 1).
| Total | Activated | Not activated | Declined | aOR (95%CI) | ||
|---|---|---|---|---|---|---|
| N = 4,571 | N = 2,438, | N = 1,651, | N = 482, | |||
| N (%) | % | % | % | |||
| Age, y | * | |||||
| <80 | 1,606 (35.1) | 61.2 | 30.2 | 8.7 | Reference | |
| 80–89 | 2006 (43.9) | 51.8 | 37.3 | 10.8 | .68 (.59–.78) * | |
| ≥90 | 959 (21.0) | 43.4 | 43.5 | 13.1 | .50 (.42–.60) * | |
| Sex | * | |||||
| Female | 3,371 (73.8) | 51.8 | 37.7 | 10.6 | Reference | |
| Male | 1,200 (26.3) | 57.8 | 31.8 | 10.5 | 1.00 (.86–1.16) | |
| Race | ||||||
| White | 2,547 (55.7) | 60.7 | 30.6 | 8.7 | * | Reference |
| African American | 563 (12.3) | 40.1 | 46.4 | 13.5 | .51 (.42–.62) * | |
| Asian | 128 (2.8) | 68.0 | 22.7 | 9.4 | 1.35 (.90–2.02) | |
| Other | 1,100 (24.1) | 44.6 | 41.9 | 13.5 | .61 (.52–.73) * | |
| Unknown | 233 (5.1) | 37.8 | 51.5 | 10.7 | .43 (.32–.58)* | |
| English language | * | |||||
| Yes | 4,145 (91.4) | 54.5 | 35.5 | 10.0 | Reference | |
| No | 388 (8.6) | 40.5 | 43.0 | 16.5 | .79 (.61–1.01) | |
| Medicaid coverage | * | |||||
| Yes | 860 (18.8) | 40.8 | 44.8 | 14.4 | .74 (.62–.90)* | |
| No | 3,711 (81.2) | 56.2 | 34.1 | 9.7 | Reference | |
| Marital status | * | |||||
| Married/Life partner | 1,433 (31.3) | 65.2 | 26.6 | 8.2 | 1.60 (1.38–1.86)* | |
| Widowed/Divorced/Single | 3,138 (68.6) | 48.5 | 39.8 | 11.7 | Reference | |
| Dementia | * | |||||
| Yes | 728 (15.9) | 46.3 | 42.6 | 11.1 | .90 (.76–1.08) | |
| No | 3,843 (84.1) | 54.7 | 34.9 | 10.4 | Reference |
- Note: *P value <.001; race category “Other” includes American Indian/Alaskan Native (.09%) and Native Hawaiian/Pacific Islander (.07%).
- Abbreviations: aOR, adjusted odds ratio (compares activated with not activated/declined); CI, confidence interval.
In adjusted analyses, patients were less likely to have activated their portal if they were older (adjusted odds ratio [aOR] = .68; 95% confidence interval [CI] = .59–.78 for aged 80–89 vs <80 and aOR = .50; 95% CI = .42–.60 for aged 90 and older vs <80), African American compared with White (aOR = .51; 95% CI = .42–.62), or eligible for Medicaid (aOR = .74; 95% CI = .62–.90). Patients who were married/partnered had a higher likelihood of activation than those without a spouse/partner (aOR = 1.60; 95% CI = 1.38–1.86). Among patients who activated their portal (N = 2,438) and sought care during the study period (N = 651), there were no statistically significant differences in the likelihood of having a telehealth video versus telephone visit in both unadjusted and adjusted models.
DISCUSSION
We found substantial differences by race and socioeconomic status in access to telehealth video services during the COVID-19 surge in NYC. African-American patients and those of other races were significantly less likely than White patients to have activated the online portal and more likely to have declined activation outright. Independent of race, those eligible for Medicaid were also less likely to have activated the portal. However, once activation of the online portal was considered, there were no differences by demographic or socioeconomic characteristics in the likelihood of receiving a telehealth video visit relative to a telephone visit during the COVID-19 surge in NYC, underscoring the key role of access to telehealth technology.
A patient's ability to access telehealth should be assessed and facilitated as part of their intake appointment. As telehealth becomes embedded in clinical practice for both emergency and standard clinical care, it is critical to ensure that its use does not exacerbate healthcare disparities among vulnerable older adults.
ACKNOWLEDGMENTS
Conflict of Interest
Melissa D. Aldridge was funded by Grant No. K07AG0660270. Katherine Ornstein was funded by Grant No. R01AG060967.
Author Contributions
Study concept and design: Rivera, Aldridge, Ornstein, and Chun. Analysis and interpretation of data: Aldridge, Moody, and Chun. Preparation of manuscript: All authors.
Sponsor's Role
Funders played no role in the study's design, methods, data collection, analysis, and preparation of the article.
