Temporal Trends and Sociodemographic Differences in Telemedicine Utilization, 2019-2024

Clinical Snapshot

Key Findings

• P Value: Not explicitly reported in abstract; statistical significance of all reported aORs is implied given sample size

• Effect Size: Telemedicine surged from ~1% of visits pre-pandemic to 17% in April 2020, stabilised at 8–13% through late 2020, and settled at 4–6% from 2022–2024. Key adjusted associations: age ≥65 vs. <40 (aOR 0.47); age 40–64 vs. <40 (aOR 0.67); male vs. female (aOR 0.90); new vs. established visit (aOR 0.46); patient portal user (aOR 1.44); distance ≥15 miles vs. <5 miles (aOR 1.42); income <$50,000 vs. $50,000–$100,000 (aOR 1.06); primary vs. specialty care (aOR 1.23); Non-Hispanic Black vs. White (aOR 0.88); Hispanic vs. White (aOR 0.94); Asian vs. White (aOR 0.82); unmarried vs. married (aOR 1.10)

• Primary Outcome: Whether each outpatient encounter was conducted via telemedicine vs. in-person across 46,149,734 visits among 2,248,341 patients at Penn Medicine, 2019–2024

• Nnt Or Sensitivity: Not applicable for this utilisation study; no NNT, sensitivity, or specificity reported. Adjusted odds ratios serve as the primary association metric. The aOR of 0.47 for age ≥65 represents the strongest single predictor of lower telemedicine use.

• Confidence Interval: Not reported in abstract; assumed narrow given sample size >46 million visits

Clinical Application

Findings are directly actionable for health system administrators and policy makers. Targeted interventions suggested by the data include: digital literacy and portal onboarding programmes for older adults and racial/ethnic minority patients; telemedicine infrastructure support for patients in rural or distant locations; and differentiated telemedicine pathways for new vs. established patient encounters. Implementation requires investment in digital equity infrastructure and clinician training. Australia's telehealth landscape is directly relevant to these findings. Medicare Benefits Schedule (MBS) telehealth items were introduced in March 2020 and have been progressively refined, with ongoing debates about permanent vs. temporary status for various item numbers. The RACGP has published telehealth guidelines emphasising appropriate patient selection and continuity of care. The Australian Digital Health Agency's My Health Record and patient portal ecosystem parallels the Penn Medicine portal findings — digital engagement predicts telehealth uptake. Aboriginal and Torres Strait Islander communities, rural and remote patients, and elderly Australians face analogous access barriers to those identified in this study. The TGA does not directly regulate telehealth delivery, but prescribing via telehealth is subject to specific MBS and state-based regulatory requirements. Australian health services should note the finding that new patient encounters have markedly lower telemedicine uptake (aOR 0.46), consistent with RACGP guidance recommending in-person assessment for new or complex presentations. The racial/ethnic disparity findings have direct parallels to documented digital divide issues in CALD (Culturally and Linguistically Diverse) communities in Australia. Adult outpatients in large academic or metropolitan health systems with established EHR and patient portal infrastructure. Most directly applicable to established patients with prior healthcare engagement. Findings regarding older adults, racial/ethnic minorities, and new patients are broadly relevant to any health system planning telemedicine service delivery.

Abstract

BACKGROUND: Telemedicine usage surged during the COVID-19 pandemic, shaping how patients access healthcare services. Its sustained role in post-pandemic healthcare may uncover long-term trends and variations in utilization. OBJECTIVE: To characterize telemedicine utilization from 2019 to 2024 and identify patient characteristics associated with telemedicine use. DESIGN AND PARTICIPANTS: This retrospective cohort study analyzed outpatient visits across five hospitals within the University of Pennsylvania Health System (Penn Medicine) from January 1, 2019, to September 30, 2024. MAIN MEASURES: The primary outcome was whether each outpatient encounter was conducted via telemedicine (vs in-person). We used multivariable logistic regression clustering on patients to assess associations between telemedicine use and patient- and encounter-level characteristics, including demographics, insurance, patient portal use, income, clinical comorbidity, distance from care, provider specialty, encounter type, hospital index, and visit year. KEY RESULTS: The study included 46,149,734 visits among 2,248,341 patients. Telemedicine surged from 1% of visits pre-pandemic to 17% in April 2020, stabilized at 8-13% through late 2020, and remained 4-6% from 2022 to 2024. Telemedicine use was lower among older adults (aOR 0.67 for ages 40-64; 0.47 for ≥ 65 vs. < 40 years), males (aOR 0.90), and new visits (aOR 0.46). Higher use was observed among unmarried (aOR 1.10), patient portal users (aOR 1.44), patients with fewer comorbidities, those living ≥ 15 miles from care (aOR 1.42 vs. < 5 miles), lower-income (< $50,000 aOR 1.06 vs. $50,000-$100,000), and primary care (aOR 1.23 vs. specialty care). Telemedicine use was lower among Non-Hispanic Black (aOR 0.88), Hispanic (aOR 0.94), and Asian (aOR 0.82) patients compared to Non-Hispanic White patients. Patterns differed by clinical condition, with disproportionately higher use among White patients with mental health disorders. CONCLUSIONS: Telemedicine use persists post-pandemic but reflects differences in access by age, race/ethnicity, socioeconomic status, and prior engagement with the patient portal. Targeted policies are needed to ensure equitable telemedicine adoption and accessibility for all patients.