Patients of family physicians in Ontario, Canada, who provided a high percentage of telehealth visits in the first years of the COVID-19 pandemic didn't have more emergency department (ED) visits than those of doctors who provided the lowest proportion of virtual care, finds a study published late last week in JAMA Network Open.
A University of Toronto–led research team analyzed administrative data on phone or video primary-care visits and ED visits among 12,951,063 patients and 13,820 family physicians with at least one primary-care claim from February to October 2021. The average age was 50.0 years among the family physicians and 42.6 years among the patients. The study period spanned January 2019 to October 2021.
Patients receiving only in-person care had most ED visits
The total number of primary care visits fell at the onset of the pandemic but rebounded to prepandemic levels by fall 2020. The proportion of virtual visits peaked in the first 2 weeks of the pandemic (82%), declining to 49% by October 2021.
ED visits decreased at the start of the pandemic and remained below 2019 volumes throughout the study period. Declines in ED visits corresponded to increases in COVID-19 cases, while a higher percentage of virtual primary care visits corresponded to more COVID-19 cases and fewer ED visits.
Most doctors provided 40% to 80% of care through telehealth visits; 336 physicians (2.4%) delivered 100% of care remotely, while 2,606 (18.9%) provided 81% to 99% of care virtually. Compared with doctors who provided 40% to 60% of care virtually, those who provided more than 80% of care virtually had a greater average number of primary-care visits per patient and less care continuity.
Physicians who provided more than 80% of care virtually tended to be older (65 years and up), women, and those who practiced in large cities. Most physicians who provided all care virtually worked in a traditional fee-for-service practice and had fewer than 100 patients, who tended to live in low-income, ethnically diverse neighborhoods.
The average number of ED visits was highest among patients whose doctors offered only in-person care (470.3 per 1,000 patients) and lowest among those of physicians who provided 81% to 99% of their care remotely (242.0 per 1,000).
An analysis adjusted for patient characteristics showed that patients of doctors who delivered more than 20% of care episodes virtually had lower rates of ED visits than those whose physicians provided less telehealthcare (eg, relative rate of 81% to 99% vs 0% to 20% virtual visits in large cities, 0.77%).
Mechanisms remain unclear
"The findings refute the hypothesis that family physicians providing more care virtually during the pandemic resulted in higher ED use," the study authors wrote.
Virtual care, they said, needs further study. "Mixed-methods research is needed to elucidate patients' experience with accessing their family physicians, reasons for seeking care in the ED, and views on virtual care as well as reasons behind physician-level variation in virtual care provision," they wrote. "Researchers and policy makers should be mindful of different patient subgroups wherein virtual care can either facilitate access (eg, for people in rural areas) or serve as a barrier (eg, for people with language or sensory barriers)."
The findings refute the hypothesis that family physicians providing more care virtually during the pandemic resulted in higher ED use.
In a related commentary, Jesse Pines, MD, MBA, of US Acute Care Solutions in Canton, Ohio, said that the study results, like many others, suggests that telehealth can be deployed without causing harm or dramatically affecting healthcare use.
The mechanisms behind the lower rate of ED usage among patients of physicians providing higher proportions of virtual care remain unclear, however.
"Randomized trials could test different doses of telehealth, or mixed-methods studies could help elucidate how telehealth may be changing clinical management or care seeking behavior," he wrote. "As telehealth continues to be a major fixture in care throughout North America, undoubtedly more studies will emerge providing physicians, payers, and policy makers with guidance on how best to design clinical models, reimbursement, and policies."
