Utilization of Asynchronous and Synchronous Teledermatology in a Large Health Care System During the COVID-19 Pandemic

The coronavirus disease 2019 (COVID-19) pandemic has infected millions globally, with spread largely attributed to close contact and droplet exposure. Telehealth, along with other forms of social distancing, has been vital in preventing further disease spread, while also continuing exemplary care for patients.^1,2 Several regulatory bodies and academies, including the American Academy of Dermatology, have encouraged dermatologists to substitute nonessential face-to-face patient visits with teledermatology.^3,4

Teledermatology has been used for over two decades to provide improved access and care, especially to underserved communities, while simultaneously reducing health care costs.⁵ Asynchronous visits use store-and-forward technologies such as secure text, e-mail, or direct-to-physician electronic medical records.^6–8 Synchronous visits are performed using live, real-time, and collaborative telecommunications technology (audio and/or video) between the provider and the patient.^7,8 Hybridization is the use of asynchronous and synchronous modalities simultaneously.⁵ Both technologies provide significant advantages: asynchronous visits have lower costs and flexible timing, whereas synchronous visits allow immediate interactions between patient and physician.⁷ Yet, both systems also have disadvantages such as privacy concerns, technological difficulties, and a lack of communication for asynchronous visits.^2,6 Limited data exist comparing the technologies and, as such, a consensus on patient and provider preference has not yet been reached.^9–11

In general, telehealth utilization by patients in a large commercially insured population in the United States averaged ∼6.6 telehealth visits per 1,000 patient-visits per year.^12,13 The use of teledermatology is slightly more popular with ∼8.6 teledermatology visits per 1,000 patients per year.¹⁴ However, during times of crisis, such as Hurricane Maria in 2017, there has been increased utilization of telehealth.^15–17 During Hurricane Maria, the daily percentage of telehealth visits compared with daily outpatient visits rose to 26% compared with a baseline of 14%.¹⁷ As such, during the COVID-19 pandemic crisis, access to direct health care was invariably reduced. In response, telehealth systems have increased in volume and scope during the COVID-19 pandemic; however, systematic quantification of teledermatology utilization has not yet been published.^3,18,19 In this study, we performed a single-center retrospective assessment of teledermatology utilization during the COVID-19 pandemic.

Teledermatology visits were defined as either asynchronous or synchronous visits. Asynchronous visits constituted direct patient-to-physician store-and-forward visits in which a patient, at any time, completed a questionnaire and supplied three photos of their skin lesion. This was performed through the University of Pittsburgh Medical Center (UPMC) Health System's EPIC EMR and secure patient web-based portal through the UPMC Health Plan. This system had been run by 3 practitioners at this academic center since 2013; however, during the COVID-19 pandemic, this was expanded to 26 practitioners.

Synchronous visits constituted real-time audio and/or video teleconferencing between the patient and the practitioner. These were scheduled live synchronous visits using our medical center's EPIC EMR with work station cameras, EPIC Haiku, or Canto functionalities for iPhones and iPads, respectively. When patients called to cancel in-clinic appointments or when our clinics reached out to cancel clinic appointments, they were given an unbiased choice to use either asynchronous or synchronous teledermatology based on their own preference by front desk personnel or through our systems central scheduling. All teledermatology visits, were reviewed and completed by a physician assistant or dermatology resident under the direct supervision of an attending board-certified dermatologist or by the attending individually, regardless of whether asynchronous or synchronous. Synchronous visits were officially instituted on March 20, 2020, at this academic center.

All teledermatology visits between March 16 and May 1, 2020, were collected, with a total of 35 days of visits. Twenty-nine records were removed as the visits were either erroneous or not completed by the practitioner. A total of 2,623 patient records (951 asynchronous, 1,672 synchronous) were analyzed and data for demographics and visit details (diagnoses, prescriptions, and treatment plans) were documented. Diagnoses and prescriptions ordered were grouped based on subjective criteria. For example, the category of “viral infection” constitutes herpes, whitlow, molluscum, and warts. Diagnoses and prescriptions that were <1% of the total for both teledermatology modalities were compiled into a group referred to as “other.”

Student's t-test was used for comparison of continuous variables with a threshold for significance of 0.01. Pearson's chi-squared or Fisher's exact test was used for categorical variables, with a threshold for significance of 0.05. For any diagnosis or prescription category resulting in a Pearson statistic >5, post hoc analysis was performed. This resulted in post hoc analysis between 7 diagnosis categories (21 pairwise tests) and 4 prescription categories (7 pairwise tests). Statistical significance required Bonferroni correction and was set at p < 0.002 for diagnoses and p < 0.02 for prescriptions.

An intradepartmental survey was sent to all providers to evaluate provider preferences between asynchronous and synchronous formats. Billing codes and subsequent reimbursement data were collected 4 months after our study time frame.

From March 16 to May 1, 2020, there were 2,755 teledermatology visits; however, 132 visits were a “no-show” (4.7%) and were excluded from further analysis. All no-show visits were from synchronous visits. Two thousand six hundred and thirty-two teledermatology visits were completed (19.5 visits per 1,000 visits per year) across 2,530 unique patients, of which 1,780 (67.8%) were female and the average age was 39.4 ± 0.4 years (Table 1). A total of 951 asynchronous visits were completed (36.2%), with an average age of 35.3 ± 0.5 years. A total of 1,672 synchronous visits were completed (63.8%). with an average age of 41.8 ± 0.5 years, which was significantly higher than the age of those completing asynchronous visits (p < 0.001). However, there was no significant difference between the number of pediatric and senior patient cases completed (p = 0.3)

Initially, most completed visits were asynchronous, and at day 17, visits were approximately equal. Thereafter, there were more synchronous visits on a daily basis (Fig. 1). By the end of this period, synchronous teledermatology visits constituted more than half of all completed teledermatology visits. Of the 951 asynchronous visits, 286 were converted from previously scheduled clinic visits, which were cancelled due to COVID-19–related concerns (30.1%). In-person follow-up was requested for only 131 of the 2,623 encounters (4.9%), with 98 follow-up visits completed. Sixty of those visits resulted in biopsies (61.2%), whereas the remainder had minor procedures performed including intralesional steroid injection, cryotherapy, excision, or incision and drainage.

Over time, the average number of teledermatology visits increased significantly until week 3, then remained stable from weeks 4 to 6 (p < 0.01; Fig. 2). Interestingly, there was a significant decrease between weeks 6 and 7. Asynchronous visits did not show a significant increase over time however, synchronous visits significantly increased between weeks 2 and 3 (p < 0.01). Synchronous visits were significantly greater than asynchronous visits at weeks 6 and 7 (p < 0.01).

We found that both asynchronous and synchronous teledermatology were heavily utilized during the pandemic, with increased use over time. During nonpandemic times, dermatologists nationally conduct ∼8.6 teledermatology visits per 1,000 visits per year.^5,14 We showed that at our center, during the COVID-19 pandemic, this number more than doubled, as ∼20 teledermatology visits per 1,000 visits per year were completed. Furthermore, we also showed that teledermatology was effectively able to treat patients remotely, as <5% of visits resulted in an acute clinic follow-up. For comparison, our physician-to-physician asynchronous teledermatology service during nonpandemic times notes ∼37% clinic follow-ups (unpublished data). Over time, the total number of visits increased until week 3 and then stabilized before ultimately declining between weeks 6 and 7. This largely parallels progression of COVID-19, as well as shifts in governmental and institutional policies. During the time period of this study, confirmed regional COVID-19 cases increased from 22 on March 16, 2020, to 1405 on May 1, 2020.²⁰ Furthermore, during this period, institutional and governmental policies regarding social distancing were expanded, causing patients to explore virtual visitation options.²¹

Although we noted consistent increases in teledermatology visits from weeks 1 to 6, there was a drop in teledermatology visits at week 7. This pattern parallels what has been seen for telehealth use during acute crises, such as Hurricane Maria in 2017. Telehealth rose to a peak of 26% of visits during the peak of the Hurricane, and reduced progressively as the crisis abated.¹⁷ Interestingly, after Hurricane Maria, telehealth use remained higher than precrisis levels, suggesting that telehealth technologies implemented during a crisis can have long-term impacts.¹⁷ Indeed, across this study's teledermatology visits, <5% necessitated an in-person follow-up. Thus, we suspect that teledermatology not only provided care during the pandemic, but as a result, may also become more heavily utilized in the future, largely for its ability to triage and decrease in-live patient appointment waiting times once the pandemic resolves.

We found that there were differences in diagnoses and prescriptions between the two modalities. There were a greater number of diagnoses per visit for synchronous teledermatology, with the number of diagnoses ranging from 1 to 5 for synchronous teledermatology and from 1 to 3 for asynchronous teledermatology. From our provider survey, the main advantage for synchronous teledermatology related to better patient interaction. Improved interaction and more closely mimicking in-office visits allow for the discussion for multiple problems, often resulting in more diagnoses.²² Furthermore, synchronous visitation was preferred by providers for more complex diseases, which may result in multiple diagnoses. Indeed, we show that there was a greater percentage of complex medical dermatology cases performed through synchronous teledermatology. Lastly, immunomodulators and biologics were more commonly prescribed for synchronous visits and such treatment plans often require multiple diagnoses for management.

Another important difference was the greater number of prescriptions per visit for asynchronous visits. This is likely a function of the greater number of acne cases seen through asynchronous teledermatology. Polytherapy is the mainstay of acne treatment, as >90% of cases require more than one prescription, with the average patient requiring more than three medications.²³ Indeed, we show that certain acne therapies, such as nonretinoid acne treatment, were more often prescribed through asynchronous teledermatology. Most importantly, the enhanced provider–patient interactivity of synchronous visits allows for improved patient education, and thus may reduce the need for increased pharmacological interventions.²⁴

Our providers performed either form of teledermatology based on the patients selected mode. Providers did not choose the type of interaction with the patient and reimbursement data were not available during the time of the initial COVID-19 restrictions, thus no bias influenced our providers or health system.

This study is limited by being a single-center study. It lacks direct input from patients regarding preferred teledermatology modalities, thus further survey work may be done to assess this.

Our study concluded that teledermatology was effectively utilized during the pandemic, matching telehealth trends during times of crisis. It allowed for a marked reduction in live patient visits during a time of mandatory personal distancing. Asynchronous teledermatology visits were more commonly used for acne, whereas synchronous teledermatology can be an effective means for management of complex medical dermatology. Postanalysis of the data collected at our institution, after this designated period of evaluation, led to instituting a hybridization of asynchronous and synchronous teledermatology. All patients with scheduled synchronous appointments are now highly encouraged to also submit digital images (asynchronous) through patient web-based portal before live video interaction with our teledermatology team. In addition, patients who request initial asynchronous consultations have been instructed to follow up with our synchronous modalities, especially if more complex medical conditions are diagnosed. We support the need for training, implementation, and funding for both asynchronous and synchronous teledermatology technologies for wider availability.

We thank the following providers for their participation: Oleg Akilov, MD, PhD; Olga Bunimovich, MD; Yuri Bunimovich, MD, PhD; Sonal Choudhary, MD; Mary-Katherine Collins, MD; Robin Gehris, MD; Corey Georgesen, MD; Arthur Huen, MD, PhD; Alaina James, MD, PhD; Daniel Kaplan, MD, PhD; Viktoryia D Kazlouskaya, MD, PhD; Ellen Koch, MD; Grace Lee, MD; Melissa Pugliano-Mauro, MD; Suzan Obagi, MD; Timothy Patton, DO; Mary Sheehan, MD; Sarah Whitley, MD, PhD; Carolyn Willis, MD; Megan Carrigan, PA-C; Emily Clarke, PA-C; Hannah Glass, PA-C; Morgan Matisko, PA-C; Talia Scherer, PA-C; Erin Skaros, PA-C; Kristin Smith, PA-C; and Noel Prevost, PA-C. We thank Kylie Zamperini for aide in data collection and management.