Abstract 12277: Impact of Telemedicine New Patient Visits on Total Visit Utilization at an Academic Cardiovascular Center: An Instrumental Variable Analysis

Abstract

Introduction: Telemedicine use has increased, but little is known about the impact of telemedicine on downstream healthcare utilization. We sought to determine the effect of conducting new patient visits (NPVs) through telemedicine on the number of follow-up visits for common cardiovascular conditions. Methods: We extracted data for all visits at an academic cardiovascular center from Jan 2019 to Feb 2022. We evaluated the effect of telemedicine vs in-person NPVs on total follow-up visits within 6 months for patients seen for 8 common primary diagnoses between Jun 2020 and Aug 2021. We utilized variation in the rate of telemedicine use among clinicians as an instrumental variable, adjusting for patient characteristics, diagnosis, subspecialty, and site. We used two-stage least squares (2SLS) and instrumental causal forest (ICF) models. Results: There were 4,856 NPVs and 3,594 follow-up visits. NPVs were delivered by 120 clinicians with 58.8% provided by telemedicine (range 0-100% across clinicians). There was no association between the average number of follow-up visits before Mar 2020 and subsequent clinician telemedicine adoption. In 2SLS analysis, there was no significant overall effect of delivering NPVs through telemedicine on the number of follow-up visits within 6 months (-0.029 visits per patient, 95% CI -0.15, 0.089). The ICF analysis similarly showed no significant overall effect of telemedicine NPVs on the number of follow-up visits. In 2SLS subgroup analyses, the effect of telemedicine NPVs differed across diagnoses (Figure). Conclusions: At an academic cardiovascular center, use of telemedicine NPVs for several common diagnoses did not affect the total number of follow-up visits within 6 months. Telemedicine NPVs can successfully replace in-person visits though their impact on total visits may differ across cardiovascular conditions. Future research should evaluate the effect of visit modality on other domains of care utilization and outcomes.