Patient-, provider-, and facility-level contributors to the use of cardiology telehealth care in the Veterans Health Administration: retrospective cohort study

Abstract: Background: Telehealth (care delivered by phone or video) comprises a substantial proportion of cardiology care delivered in the Veterans Health Administration (VHA). Little is known about how factors specific to patients, clinicians, and facilities contribute to variation in cardiology telehealth use. Objective: The aim of this study is to estimate the relative extent to which patient-, clinician-, and facility-level factors affect cardiology telehealth use in VHA. Methods: This was a retrospective, nation-wide cohort study of veterans' use of VHA cardiology telehealth care during the first 2 years of the COVID-19 pandemic (March 11, 2020, to March 10, 2022). We constructed multilevel, multivariable, logistic regression models of patient-level cardiology telehealth use (telephone or video-based care). Models included random effects for the patient, the patient's main cardiology provider, and the patient's primary facility (ie, VHA medical center) for specialty care and fixed effects for patient sociodemographic and clinical characteristics. Results: Our analytic cohort comprised 223,809 veterans with 989,271 encounters among 2235 unique clinicians. The veterans' average age was 70.2 years, and 3.4% (n=7616) were women. Of the 989,271 encounters, 4.2% (n=41,480) were video based and 34.3% (n=338,834) were phone based. Adjusted odds of telehealth use were slightly higher for women versus men (adjusted odds ratio [AOR] 1.08, 95% CI 1.05-1.10), individuals identifying as Hispanic or Latino versus not Hispanic or Latino (AOR 1.46, 95% CI 1.43-1.49), and those with medium and long drive times versus short drive time (AOR 1.11, 95% CI 1.10-1.12 and AOR 1.09, 95% CI 1.07-1.10, respectively). Further, 40.5% of the variation in a veteran's likelihood of using cardiology telehealth care was found at the patient level, 30.8% at the clinician level, and 7% at the facility level. Conclusions: The largest share of the attributable variability in VHA cardiology telehealth use in this cohort was explained by the patient, followed closely by the clinician. Little variability was attributed to the primary facility through which the veteran received their cardiology care. These results suggest that policy solutions intended to improve equity of cardiology telehealth care use in VHA may be most impactful when directed at patients and clinicians.