Provider Perspectives Regarding Ambulatory Telemedicine at a Large Academic Health System During the COVID-19 Pandemic

Introduction

While we have known for at least the past decade that telemedicine has the potential to transform care, before COVID-19, the evidence for telehealth at large scale was limited. During the COVID-19 pandemic, telemedicine has experienced exponential growth and has been viewed as a promising and positive practice by providers. ¹

Previous work in understanding provider perspectives have focused on satisfaction with using telemedicine services ^{2 –7} outlining perceived benefits including improved patient access, reduction in no-show rates, and decreased patient travel time. Perceived negative aspects have included the inability to perform physical examinations, limited training, and facilitation of providing bad news at a distance. ^2,8

In addition, there has been appropriate focus on patient satisfaction with telemedicine. ^6,7,9,10 However, as has been discussed by Nguyen et al., the practice will only succeed if both patients and providers consistently consider telemedicine at least as positively as traditional visits, ⁹ or in other words, if there continues to be both demand from patients and supply from providers.

Work has been done to show why there may be continued supply and demand in specific disciplines, such as in psychiatry. ^11,12 To date, there has not been a broad cross-specialty comparison on provider perceptions of clinical appropriateness and quality for ambulatory telehealth, or provider preferences for the amount of telemedicine providers prefer to offer beyond the COVID-19 pandemic. Understanding these preferences will be essential to sustainably operationalize ambulatory telemedicine in the coming years. We sought to survey all providers who had conducted a telemedicine visit across Johns Hopkins Medicine to understand their perspectives on the clinical appropriateness of telemedicine in their fields and their desires for future telemedicine volume.

Methods

Johns Hopkins Medicine, in Baltimore, Maryland, comprises 6 hospitals and 40 ambulatory clinics in Maryland, the Washington D.C. Capital Region, and Florida. Ambulatory sites across Johns Hopkins Medicine include hospital-based academic clinics, nonhospital-based academic clinics, and community sites. All providers who conducted at least one ambulatory telemedicine visit were invited to complete a de novo survey administered by the Johns Hopkins Office of Telemedicine in April 2021. The survey was distributed by e-mail on the Verint platform by the Johns Hopkins Market Research team. Demographic variables (provider sex, provider type, primary clinical department, primary entity/context) were collected from the electronic health record (EHR) provider record and matched with responses. Providers were asked to indicate their percent clinical effort and years in practice on the survey instrument.

The de novo survey instrument included questions on the clinical appropriateness of telemedicine with both new and existing patients and preferences for practicing telemedicine in the future. For new and existing patients, providers were asked, “For common conditions treated in your specialty/practice area, how often do you feel you can provide clinically appropriate, high-quality care via telemedicine in comparison with in-person visits? Assume use of the most appropriate modality—video or phone.” For future telemedicine scheduling, providers were asked, “Assuming there is continued reimbursement at an equivalent rate for telemedicine visits after the COVID-19 pandemic, what percentage of your schedule template would you like to have open for telemedicine visits (video and/or phone, all platforms)?” Providers could enter a whole percentage number between 0% and 100%.

Providers were also asked regarding their preferences for expanding after-hours care through telemedicine. “If your practice/clinic planned to offer more after-hours access for patients, what would be your preferred modality for providing after-hours care? (Assume your total workweek hours and compensation were held constant.)” For the options of “Weekday early morning (before 8am),” “Weekday evening (After 5pm),” “Saturday hours,” and “Sunday hours,” participants could select “In-person,” “Telemedicine,” “No preference,” or “Unavailable to see patients.”

Descriptive analyses were conducted to provide a profile of perceptions about telemedicine clinical appropriateness and preferences for telemedicine scheduling. Wilcoxon rank sum tests were run to compare provider responses based on demographics and departmental affiliation. Departments were categorized into clinical department groups via content relevance and sample distribution (e.g., “surgery” included neurosurgery, orthopedic surgery, plastic surgery, general surgery). Descriptive statistics were calculated using STATA Version 15 MP. ¹³ R Studio version 4.1.2 was used to create boxplot visualizations of sample response distributions across clinical departments. ¹⁴

Analysis of this institutional survey was approved by the Johns Hopkins University School of Medicine Institutional Review Board. A waiver of consent approved to analyze the results of this survey was collected for operational purposes.

Results

The survey was distributed to 3,576 providers and completed by 1,342 with a 37.5% response rate (Table 1). Respondents were more frequently female (n = 666; 62%), with more than 15 years in clinical practice (n = 555, 41.4%). Most respondents were physicians (n = 819, 61%). The most frequently represented clinical departments were medicine (n = 343, 27.6%), psychiatry/behavioral sciences (n = 177, 14.2%), and surgery (n = 147, 11.8%). Most respondents were based in the academic setting and employed by the Johns Hopkins University School of Medicine (n = 824, 62%). Most respondents had more than 70% effort dedicated to clinical care (n = 789, 58.8%).

Figure 1 shows boxplot distributions of responses to questions related to the clinical appropriateness of telemedicine across new and established patients by clinical department. For new patients, the median response to how often providers felt they could provide clinically appropriate care via telemedicine ranged from 20% (range 0–95%) in pediatrics to 80% (range 0–100%) in psychiatry/behavioral sciences. For established patients, the median response to how often providers felt they could provide clinically appropriate care via telemedicine ranged from 40% (range 0–100%) in other departments (e.g., dermatology, genetic medicine, nutrition) to 90% (range 0–100%) in psychiatry/behavioral sciences. In almost every department across both new and established patients, the perception of how often telemedicine was clinically appropriate ranged from 0% to 100%.

OPEN IN VIEWER

Fig. 1.

Provider perceptions of clinical appropriateness of telemedicine by specialty. Providers answered the question “For common conditions treated in your specialty/practice area, how often do you feel you can provide clinically appropriate, high-quality care via telemedicine in comparison with in-person visits?” and could respond with a number between 0% and 100%. Black dots indicate mean responses. (A) Shows the distribution of the percent of time clinicians feel telemedicine is clinically appropriate for new patients, by departmental affiliation. (B) Shows the distribution of the percent of time clinicians feel telemedicine is clinically appropriate for existing patients, by departmental affiliation.

Figure 2 shows a boxplot distribution of telemedicine scheduling volume preferences of providers by clinical department. Under the assumption of continued reimbursement, providers in family medicine and in psychiatry/behavioral sciences indicated that they would like a median of 20% (range 25–75%) and a median of 70% (range 0–100), respectively, of their schedule templates open for telemedicine visits after the COVID-19 pandemic. Responses in all departments except for family practice ranged from 0% to 100%.

OPEN IN VIEWER

Fig. 2.

Provider preferences for amount of telemedicine on schedule template. Providers were asked, “Assuming there is continued reimbursement at an equivalent rate for telemedicine visits after the COVID-19 pandemic, what percentage of your schedule template would you like to have open for telemedicine visits (video and/or phone, all platforms)?” They could enter a number from 0 to 100. Black dots indicate mean responses.

We found overarching departmental trends in perspectives on telemedicine, with department affiliation showing significant associations with provider rating of telemedicine appropriateness and preferences for its future use. This was particularly pronounced in the department of psychiatry and behavioral sciences: psychiatrists and psychologists tended to find telemedicine more appropriate for both new patients (median score: 80% psychiatry vs. 30% all other specialties, p < 0.0001) and existing patients (90% psychiatry vs. 60% all other specialties, p < 0.0001), and preferred a greater degree of telemedicine use after COVID (70% psychiatry vs. 30% all other specialties, p < 0.0001) when compared with other departments.

Conversely, providers in other departments such as pediatrics tended to prefer less telemedicine post-COVID than other departments (25% pediatrics vs. 33% all other specialties, p = 0.001). Pediatricians also tended to find telemedicine less appropriate for new patients than other providers (20% pediatrics vs. 40% all other specialties, p = 0.0001), and providers in the department of physical medicine and rehabilitation (PM&R) medicine tended to find telemedicine less appropriate for existing patients (50% PM&R vs. 70% all other specialties, p < 0.0001).

We also examined provider responses related to the clinical appropriateness of telemedicine and future desired telemedicine volume by provider demographics (Table 2). In general, providers who were female (40% vs. 25%, p = 0.003) and had fewer than 15 years in clinical practice (50% vs. 25%, p < 0.0001) tended to feel that telemedicine was more often clinically appropriate for new patients. Female (33% vs. 30%, p = 0.02) and fewer years in practice (40% vs. 30%, p < 0.0001) providers also tended to desire more telemedicine on their schedule templates in the future.

With respect to after-hours telemedicine care (Table 3), more than 40% of providers were willing to add evening hours via telemedicine, while fewer than 19% of providers were willing to add Sunday hours. This trend was consistent across provider demographic groups and specialties.

Discussion

Across all departments at a large academic medical center, almost all providers reported that telemedicine is a clinically appropriate modality to deliver high-quality care to a subset of patients in their respective fields. Furthermore, many providers wanted to have a meaningful portion of their scheduling templates available for telemedicine care after the end of the pandemic assuming continued reimbursement. These findings suggest that many providers across a broad array of specialties in this large academic health system see the clinical importance of this tool for the future. In our data set, there was wide variability within each specialty of provider perspectives regarding both the clinical appropriateness and the amount of telemedicine scheduling desired, suggesting the presence of both telemedicine champions and telemedicine hesitant providers within every specialty surveyed.

In addition, we found varied perceptions of telemedicine across demographic and departmental groups, indicating the importance of clinical context, provider experience, and provider personal circumstances when considering how to best operationalize telemedicine from the provider perspective.

Before COVID, key operational and logistical barriers, particularly related to reimbursement as well as patient and provider location limited the feasibility of incorporating telemedicine into daily ambulatory practice. ¹⁵ When telemedicine rapidly expanded at the beginning of the COVID-19 pandemic, provider use of telehealth visits was largely met with satisfaction. ^{2 –7} However, there were those who indicated that once the pandemic concluded, a return to face-to-face medical visits might be preferable. ¹⁶ Even providers who found the practice of telemedicine beneficial indicated concerns about whether its use could be sustained. A 2021 study evaluating radiation oncologists' perceptions of telemedicine found that >80% wished to continue using telemedicine post-COVID, but nearly 50% were concerned about continued reimbursement. ¹⁷

Telemedicine has many potential benefits for patients including increased access to care, promoting patient-centered care and patient engagement, providing the opportunity for multidisciplinary care, and providing more flexible hours of access. ^18,19 We believe our data support the argument that “telemedicine is here to stay,” as a large subset of providers across all specialty groups would like to continue offering high rates of telemedicine in their ambulatory schedules beyond the COVID-19 pandemic, providing a first glimpse at future provider supply to meet patient demand.

Clear differences exist in the perceptions about seeing new versus established patient visits across departments in this data set. Providers viewed telemedicine as more clinically appropriate and high quality for established patients over new patients. Furthermore, providers in psychiatry/behavioral sciences and neurology wanted to offer more telemedicine in the future. Previous research on telepsychiatry indicates that psychiatrists/psychologists not only find the use of telemedicine beneficial during the pandemic, but are in favor of exploring its post-COVID use on a systemic level. ^11,12 Psychiatrists have been satisfied with the practice and found their patients to be engaged in the process. ¹¹ Additional studies have indicated that the ability to perform physical examinations is a significant barrier to telemedicine; this may be less impactful in therapeutic mental health visits. ² Our data mirror these findings, with psychiatrists and psychologists consistently indicating telemedicine as both highly appropriate and preferable on average.

Other studies have suggested that telemedicine may be less frequently utilized in subspecialties outside of primary care and mental health care. For example, a recent study from 2022 found that while over half of primary care providers serving Medicare beneficiaries provided telemedicine services, the same was true for only about 25% of specialists. ²⁰ Our results demonstrate a nuanced provider perception of telemedicine specialty-based care—the appropriateness of telemedicine for a given clinical encounter may relate to specialty, clinical condition, as well as provider characteristics. Based on the perspectives in providers in our sample, there are likely appropriate use cases for telemedicine in almost every specialty.

Telemedicine offers unique opportunities to fundamentally change medical appointment scheduling by incorporating appointment times during weekends or outside of traditional business hours. ²¹ Our results suggest that, regardless of medical specialty, more physicians and advanced practice providers may prefer adding evening—as opposed to early morning or weekend—telemedicine appointments. Given this preference and the desirability of such evening clinic hours for patients, expanding access to evening week-day telemedicine appointments may be a cost-effective way to operationalize improved access for patients without investing in additional overhead for in-person clinic space outside of business hours.

Our data also suggest that providers with fewer years of practice may be more likely to both view telemedicine as a clinically appropriate method of health care delivery and prefer the incorporation of telemedicine visits in their clinic schedules. Although it is difficult to distinguish the impact of provider years of experience from provider age, this result is similar to prior research's findings that physicians with fewer years of practice were initially more likely to view electronic medical records as clinically useful and prefer using electronic medical records in clinical practice in the late 1990s. ²²

In addition, we found that female providers tended to find telemedicine more clinically appropriate and desired a greater proportion in the future than their male counterparts. This has implications for both gender differences in the practice of medicine, and the potential benefits of telemedicine to individual providers. For example, a body of work has indicated that female doctors tend to engage in more patient-centered care such as behaviors that promote provider–patient partnerships, integration of psychosocial counseling, and positive or emotionally focused dialogue. ^23,24 Although some providers have indicated concerns about rapport during telemedicine visits, ² others perceive patient-centered care as a benefit in telemedicine visits. ^18,25 In addition, although the gender divide has diminished over the years, women physicians have been shown to have more domestic responsibilities than their male colleagues. ²⁶

Female physicians have also been shown to be more susceptible to burnout, which, in turn, has been linked to work environment, home–work demands and demanding or conflicting hours. ^{27 –29} Practicing telemedicine may provide more flexibility in caring for home and children, which may appeal to physicians with increased domestic responsibilities or those experiencing burnout.

Notably, our results include a wide range of responses within any given clinical department, as nearly every department grouping had responses ranging from 0% to 100% in each of the questions regarding the appropriateness and future desired amount of telemedicine offerings. This distribution is similar to the perceptions around early EHR adoption. ²² As the health care workforce evolves, our data suggest that we may see telemedicine “champions” in each clinical area, as each area had providers who report that telemedicine is a clinically appropriate and high-quality way to conduct care for 100% of new and established patient visits and who want to offer 100% telemedicine care in their ambulatory schedules. These findings suggest that operational leadership should leverage the existence of telemedicine champions across varied specialties to build new telemedicine offerings.

There are a number of limitations to consider when drawing conclusions from these data. This was a voluntary de novo survey from a single institution with a 37.5% response rate, and thus, further research will be needed to determine if these findings are replicable. We have presented a generally descriptive analysis as a landscape of our initial findings. Additional multivariable statistical analyses could be a next step to determine which provider factors (e.g., years in practice, gender) are independently associated with telemedicine practices. Finally, these data were taken during the COVID-19 pandemic, during which perspectives are rapidly changes. Provider perspectives should be continually reassessed as the clinical care delivery and regulatory landscapes evolve.

These survey data provide a window into understanding provider expectations regarding telemedicine use in a large academic setting beyond the COVID-19 pandemic. Future work could supplement these findings by the addition of qualitative data that would allow a more comprehensive understanding of the factors underlying the interdepartmental and intradepartmental differences in providers' preferences for telemedicine and the reasons why telemedicine is or is not considered clinically appropriate.