Use telehealth as needed: telehealth substitutes in-person primary care and associates with the changes in unplanned events and follow-up visits

Abstract

Background: Telehealth rapidly expanded since the outbreak of the COVID-19 pandemic. This study aims to understand how telehealth can substitute in-person services by 1) estimating the changes in non-COVID emergency department (ED) visits, hospitalizations, and care costs among US Medicare beneficiaries by visit modality (telehealth vs. in-person) during the COVID-19 pandemic relative to the previous year; 2) comparing the follow-up time and patterns between telehealth and in-person care. Methods: A retrospective and longitudinal study design using US Medicare patients 65 years or older from an Accountable Care Organization (ACO). The study period was April-December 2020, and the baseline period was March 2019 – February 2020. The sample included 16,222 patients, 338,872 patient-month records and 134,375 outpatient encounters. Patients were categorized as non-users, telehealth only, in-person care only and users of both types. Outcomes included the number of unplanned events and costs per month at the patient level; number of days until the next visit and whether the next visit happened within 3-, 7-, 14- and 30-days at the encounter level. All analyses were adjusted for patient characteristics and seasonal trends. Results: Beneficiaries who used only telehealth or in-person care had comparable baseline health conditions but were healthier than those who used both types of services. During the study period, the telehealth only group had significantly fewer ED visits/hospitalizations and lower Medicare payments than the baseline (ED 13.2, 95% CI [11.6, 14.7] vs. 24.6 per 1,000 patients per month and hospitalization 8.1 [6.7, 9.4] vs. 12.7); the in-person only group had significantly fewer ED visits (21.9 [20.3, 23.5] vs. 26.1) and lower Medicare payments, but not hospitalizations; the both-types group had significantly more hospitalizations (23.0 [21.4, 24.6] vs. 17.8). Telehealth was not significantly different from in-person encounters in number of days until the next visit (33.4 vs. 31.2 days) or the probabilities of 3- and 7-day follow-up visits (9.2 vs. 9.3% and 21.8 vs.23.5%). Conclusions: Patients and providers treated telehealth and in-person visits as substitutes and used either depending on medical needs and availability. Telehealth did not lead to sooner or more follow-up visits than in-person services.