A first-in-the-nation research project launched by Duke rheumatologists will evaluate the effectiveness of telehealth encounters by asking physicians to score remote visits.

Launched in December 2020, the endeavor brought rheumatologists together with digital and bioinformatics specialists to create an assessment tool within the EMR. Supported by a grant from Pfizer, the year-long project is designed to ensure that remote visits serve patients effectively and to identify patients most likely to benefit from telehealth visits.

The results will help contribute to new approaches to telehealth across Duke and perhaps nationally, says David Leverenz, MD, the rheumatologist leading the research. Telehealth was launched by the Duke Division of Rheumatology and Immunology in March 2020 as a response to the COVID-19 pandemic.

“We were seeing patients with every sort of presentation—vasculitis, lupus, rheumatoid arthritis. There was a wide range of reactions among our faculty about the value of these visits,” Leverenz says.

Within three weeks of implementing telehealth care in March 2020, approximately 87% of rheumatology and immunology patient encounters were conducted using telehealth. The number of telehealth visits declined during the rest of the year, but Leverenz estimates that telehealth continues to comprise about a third of Duke rheumatology visits.

“At the end of 2020, we decided to take a step back,” he says. “Telehealth has become such a substantial part of our practice that we recognize the need to assess this experience and make sure we are providing optimal care for patients.”

Several Duke organizations have come together to contribute to this work, including Duke Forge, the Duke University Digital Health Learning Health Unit led by Megan E.B. Clowse, MD, MPH, and the artificial intelligence health team.

Each patient encounter will be scored using the “Encounter Appropriateness Score for You” (EASY), a ranking for each appointment. Providers also choose the most effective appointment type for each patient:

• In-person preferred
• Telehealth preferred
• In-person or telehealth acceptable

“By asking the physicians to assess each visit, we hope to learn more about the optimal care for our patients,” Leverenz says. “Did the telehealth consult work effectively in this instance? Should this have been in-person? We want to know how to achieve optimal care for each patient.”

In addition to EASY scores, the data analysis team will analyze disease activity scores and other electronic health record data to try to predict which patients are appropriate for telehealth or in-person care.

The project represents a large-scale collaboration, Leverenz says. “This has been a big ask for all faculty members to complete this with every patient, but they know it’s a priority and they want to help,” he adds.

“It’s been amazing to work with the informatics and statistical teams within Duke Health. As I am a relatively new faculty member, it’s been wonderful that we could come together and collaborate.”