Variability in Primary Care Telehealth Delivery Methods Across Chronic Conditions.

Importance: Sociodemographic characteristics, such as age, gender, and urbanicity, are associated individual's engagement with telehealth. However, it is unknown whether an individual's specific clinical conditions (eg, hypertension, low back pain, or heart failure) are associated with the likelihood of receiving in person, telephone, or video-based care. Objective: To identify chronic conditions that are more or less likely to be managed in person, via telephone, or via video. Design, Setting, and Participants: This cross-sectional study used data from nationwide Veterans Health Administration (VHA) patient electronic health records between April 1, 2022, and March 31, 2023. Exposure: Thirty-nine chronic conditions managed at each encounter were identified using International Statistical Classification of Diseases and Related Health Problems, Tenth Revision diagnosis codes. Main Outcome and Measures: The primary outcome was the mean probability of an encounter being completed in person, via telephone, or via video, based on the diagnosis codes for that encounter. Analyses were adjusted for sociodemographic characteristics associated with care delivery methods and other chronic conditions managed during the encounter. Results: Of the included 3 975 328 US veterans with 7 144 371 outpatient primary care encounters, 1 203 436 were aged 45 to 64 years (30.3%), 3 582 876 were male (90.1%), 382 885 were female (9.6%), 736 960 were Black or African American veterans (18.5%), and 2 566 175 were White veterans (64.6%). Among 7 144 371 outpatient primary care encounters, 83.2% (95% CI, 83.1%-84.3%) occurred in person, 11.1% (95% CI, 11.0%-11.1%) occurred via telephone, and 5.8% (95% CI, 5.77%-5.82%) occurred via video. Conditions with the highest adjusted percentage of video-based care included dementia (10.7%; 95% CI, 10.39%-10.93%), HIV or AIDS (8.8%; 95% CI, 8.41%-9.18%), pressure and chronic ulcers (8.0%; 95% CI, 7.58%-8.56%), and Parkinson disease (7.5%; 95% CI, 7.20%-7.75%). A total of 21 conditions were less likely to be managed via telephone and video when the condition was documented compared with encounters without the condition documented, including ischemic heart disease, chronic obstructive pulmonary disease, atrial fibrillation, cancer, asthma, and deafness and hearing impairment. Only 3 conditions, heart failure, HIV or AIDS, and mobility impairment, were more likely to occur via telehealth (both telephone and video) when documented compared with not documented, suggesting increased use of telehealth for these conditions. Conclusions and Relevance: The findings of this cross-sectional study suggest an individual's chronic condition may be associated with whether they received care in person or through telehealth. The need for a physical examination, laboratory, or other diagnostic information, and the acuity of the condition may be potential drivers of this differential use. The percentage of care delivered via telephone or video may be a valuable measure to identify the optimal mix of telehealth and in-person care for various clinical scenarios.