The use of teledermatology has significantly increased since the onset of the COVID-19 pandemic. Teledermatology provides several benefits, including improving access to care for underserved communities and saving patients time when in-person visits are not medically necessary [1]. This modality has been integrated into the care of medically complex patients (levels of service 4 and 5) and serves as a valuable tool for managing their conditions [2]. However, there remains a limited understanding of teledermatology use across varying levels of clinical complexity. Levels of service refer to medical billing categories that describe the complexity involved in a patient visit, ranging from 1 (brief) through 5 (long, with greater clinical judgment). In this study, we explore trends in synchronous teledermatology use across levels of service, providing insights that can inform the clinical integration of teledermatology.

Before 2020, level 2 and 3 visits accounted for the highest proportion of teledermatology visits among both patient groups (Figure 1). From 2020 to 2023, there was an increase in teledermatology use for level 4 and 5 visits among new patients at level 4 (2020: 570/22,571, 2.53%; 2021: 1468/6069, 24.19%; 2022: 838/3170, 26.44%; and 2023: 814/2468, 32.98%) and level 5 (2020: 115/22,571, 0.51%; 2021: 137/6069, 2.26%; 2022: 32/3170, 1.01%; and 2023: 52/2468, 2.11%), as well as for established patients at level 4 (2020: 21,594/184,044, 11.73%; 2021: 14,337/49,666, 28.87%; 2022: 10,417/28,643, 36.37%; and 2023: 9524/23,329, 40.82%) and level 5 (2020: 1352/184,044, 0.73%; 2021: 1422/49,666, 2.86%; 2022: 826/28,643, 2.88%; and 2023: 867/23,329, 3.72%). In contrast, level 2 teledermatology visits decreased from 2020 to 2023 for new (2020: 13,506/22,571, 59.84%; 2021: 1280/6069, 21.10%; 2022: 503/3170, 15.87%; and 2023: 330/2468, 13.37%) and established (2020: 39,881/184,044, 21.67%; 2021: 6632/49,666, 13.35%; 2022: 3009/28,643, 10.51%; and 2023: 2371/23,329, 10.16%) patients. Compared to teledermatology visits from the same period, in-person visits had a lower percentage of complex visits (levels 4 and 5) for both patient types (Figure 2). Although in-person level 4 and 5 new-patient visits increased, their growth remained lower than for teledermatology. Among established patients, in-person visits showed a level 4 decline and slight level 5 increase, contrasting the growth in teledermatology for these levels. Due to these trends, in 2023, the level 4 and 5 visits among total teledermatology visits numbered 814 of 2468 (32.3%) and 52 of 2468 (2.11%) among new patients and 9524 of 23,329 (40.82%) and 867 of 23,329(3.82%) among established patients, respectively. In comparison, for in-person visits, level 4 and 5 visits accounted for 308,577 of 1,342,022 (22.99%) and 3727 of 1,342,022 (0.28%) of visits among new patients, and 2,130,561 of 9,196,175 (23.17%) and 31,858 of 9,196,175 (0.35%) among established patients (Multimedia Appendix 2).

Our study identified that early on, teledermatology focused on lower-level visits (levels 2 and 3), but shifted toward higher-level encounters (levels 4 and 5) after 2020. This trend toward higher-level visits may be driven by multiple factors. One possible factor is the 2021 evaluation and management billing change, enabling health care providers to determine visit level by time or by medical decision [4]. This change may have simplified justification for higher-level virtual visits, leading to coding changes. However, the greater growth of higher-level encounters in teledermatology compared to in-person visits suggests factors beyond billing requirements contributed to this trend. The COVID-19 pandemic led to greater reliance on teledermatology, likely increasing dermatologist comfort with using this modality for managing complex conditions that do not require in-person visits [5]. However, as total teledermatology visits declined from 2020 to 2023, dermatologists still using virtual visits were likely those with greater comfort using telehealth or medical systems that easily incorporate virtual care. Our findings highlight important considerations for policymakers and health systems. For policymakers, coding trends can inform future telehealth billing changes to balance access, quality, and cost. For health systems, the shift highlights the importance of integrating teledermatology into practice for effective use.

Our study is limited in its capacity to explore the multiple factors that influenced the findings and by its sole examination of Medicare beneficiaries, which may not have captured the entirety of teledermatology use, reducing generalizability. Limitations in the dataset may have led to undercounting teledermatology visits, as only the first and second modifier fields were available. Furthermore, teledermatology is often used in settings where health care providers feel comfortable with virtual care, potentially skewing the patient population. Further research is needed to continue evaluating the scope of teledermatology in dermatologic care and improve our understanding of the factors that influence its increasing use for complex visits.