The volume of outpatient office visits did not increase for specialties that were more likely to adopt telehealth

Le résumé fourni par la source

Telehealth use expanded rapidly during the COVID-19 pandemic and has become an integral part of healthcare delivery. However, concerns persist that increased telehealth availability may contribute to higher healthcare utilization without significantly improving quality. As federal policymakers deliberate permanent telehealth coverage policies, it is critical to evaluate whether expanded telehealth use has increased overall utilization. Several studies have examined the association between expanded telehealth use and healthcare utilization in Medicare fee-for-service (FFS) populations, with mixed findings. Analyses by the Medicare Payment Advisory Commission and the American Institutes for Research found that hospital service areas with high telehealth use saw higher total clinician encounters compared to those with low telehealth use.1 Another study comparing health systems with varying levels of telehealth use found a small increase in total outpatient office visits among high telehealth-use systems.2 However, disentangling the impact of telehealth adoption from underlying health system or geographic factors remains a key challenge. Our objective was to assess whether increased telehealth adoption was associated with changes in overall outpatient office visits among Medicare beneficiaries by comparing specialties with low, medium, and high telehealth adoption. We analysed 100% of Medicare FFS claims from Carrier and Outpatient files from January 2019 to June 2024. We included FFS beneficiaries aged 65 and older with Part A and Part B coverage. We excluded counties with fewer than 10 FFS beneficiaries or fewer than 10 outpatient office visits per month per specialty. We identified all outpatient evaluation and management visits using Berenson-Eggers Type of Service codes. Telehealth visits were classified based on Medicare's list of eligible telehealth services and the corresponding modifier or place of service codes for each study year (Appendix Table S1). Our main exposure was how likely specialties were to adopt telehealth, categorized into low, medium, and high telehealth groups (Appendix Table S2). The low telehealth group included otolaryngology, ophthalmology, orthopaedic surgery, sports medicine, plastic and reconstructive surgery, hand surgery, podiatry, and maxillofacial surgery. The medium telehealth group included primary care specialties and additional medical and surgical subspecialties. The high telehealth group comprised behavioural health specialties. The primary outcome was total outpatient office visits, including telehealth and in-person visits. We conducted a difference-in-differences (DID) analysis comparing total outpatient office visit volumes pre- (January 2019–February 2020) and post-pandemic (January 2021–June 2024) periods across telehealth groups, using the low telehealth group as the control. The analysis included county fixed effects and adjusted for county-level demographics and comorbidities. We visually assessed that pre-pandemic trends in our outcome were parallel across the specialty groups. A formal test indicated that while trends were similar, small differences in slopes existed between groups (Appendix Figure S1). A total of 60 531 036 patients were included (Appendix Table S3), with 538 801 822 outpatient office visits. In the post-pandemic period, telehealth comprised 5.3%, 9.1%, and 43.8% of total outpatient office visits for the low, medium, and high telehealth groups, respectively (Appendix Figure S2). Despite variation in telehealth adoption post-pandemic, all 3 groups experienced declines in predicted number of total outpatient office visits: a 14% decrease for the low telehealth group (from 12 588 to 10 831 predicted visits per month per county), a 17% decrease for the medium telehealth group (from 18 873 to 15 733), and an 18% decrease for the high telehealth group (from 3264 to 2682) (Table 1). Compared to the low telehealth group, both medium (incidence rate ratio: 0.96, 95% CI: 0.96-0.97) and high (incidence rate ratio: 0.91, 95% CI: 0.90-0.92) telehealth groups saw significantly larger declines post-pandemic (Table 1). Difference-in-differences model estimates and predicted total outpatient office visits per county-month by telehealth group. Authors' analysis of a 100% sample of Medicare fee-for-service beneficiaries from January 1, 2019, through June 30, 2024. Low telehealth group includes otolaryngology, ophthalmology, orthopaedic surgery, sports medicine, plastic and reconstructive surgery, hand surgery, podiatry, and maxillofacial surgery. Medium telehealth group includes general practice, family practice, internal medicine, osteopathic manipulative therapy, hospice and palliative care, geriatric medicine, preventive medicine, gastroenterology, pulmonary disease, nephrology, infectious disease, endocrinology, medical oncology, neurosurgery, thoracic surgery, and surgical oncology. High telehealth group includes psychiatry, general psychiatry, psychology, clinical psychology, and neuropsychiatry. Pre-pandemic period was classified as January 2019–February 2020, and post-pandemic period was classified as January 2021–June 2024. Model adjusted for county-level average patient age, sex distribution, racial composition, average Hierarchical Chronic Condition score, proportion of dual eligible beneficiaries, and proportion of patients in fee-for-service (FFS). The model includes county fixed effects. Counties with fewer than 10 fee-for-service beneficiaries or fewer than 10 outpatient office visits per month per specialty group were excluded. aP < 0.0001. Difference-in-differences model estimates and predicted total outpatient office visits per county-month by telehealth group. Authors' analysis of a 100% sample of Medicare fee-for-service beneficiaries from January 1, 2019, through June 30, 2024. Low telehealth group includes otolaryngology, ophthalmology, orthopaedic surgery, sports medicine, plastic and reconstructive surgery, hand surgery, podiatry, and maxillofacial surgery. Medium telehealth group includes general practice, family practice, internal medicine, osteopathic manipulative therapy, hospice and palliative care, geriatric medicine, preventive medicine, gastroenterology, pulmonary disease, nephrology, infectious disease, endocrinology, medical oncology, neurosurgery, thoracic surgery, and surgical oncology. High telehealth group includes psychiatry, general psychiatry, psychology, clinical psychology, and neuropsychiatry. Pre-pandemic period was classified as January 2019–February 2020, and post-pandemic period was classified as January 2021–June 2024. Model adjusted for county-level average patient age, sex distribution, racial composition, average Hierarchical Chronic Condition score, proportion of dual eligible beneficiaries, and proportion of patients in fee-for-service (FFS). The model includes county fixed effects. Counties with fewer than 10 fee-for-service beneficiaries or fewer than 10 outpatient office visits per month per specialty group were excluded. aP < 0.0001. Despite variation in telehealth adoption, total outpatient office visits remained stable or declined slightly in groups with higher telehealth use compared to the low telehealth group. Our findings align with some studies reporting minimal changes in overall outpatient volume after telehealth expansion. An analysis of primary care encounters across 3 large healthcare systems found little difference in total outpatient visits despite increased telehealth utilization.3 Similarly, a study of commercially insured patients found that the overall number of outpatient office visits in the 30 days following hospital discharge did not change significantly, despite a rapid increase in telehealth use.4 Several factors may explain why total outpatient office visit volume did not increase despite expanded use of telehealth. First, provider capacity likely remained a limiting factor, as most clinicians operate on fixed schedul

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.