The Telemedicine Takeover: Lessons Learned During an Emerging Pandemic

Introduction

The SARS-COV-2 pandemic stretched the capacity of the U.S. health care system in unexpected ways—while hospitals were filled with severely ill patients, traditional ambulatory care practices shut down. Evolving “stay at home” orders and social distancing rules required the rapid deployment of novel approaches to clinical care provision, including remote triage of patients with COVID concerns and delivery of essential primary care services. Health care systems rose to this challenge, creating or expanding existing telehealth programs (through a combination of video and telephone visits) to ensure that patients could access care without an in-person appointment. This expansion of telehealth services was further enabled by emergency declarations in March 2020 at both the Federal and State level, which loosened regulations that had previously curtailed access.

New York City’s (NYC) public health care system, Health + Hospitals (H+H), rapidly converted face-to-face encounters to telephone visits. This transition pushed the system from an average of 500 televisits in the month before COVID-19 to 57,000 in the first 3 weeks of the state of emergency (and ultimately to >1,250,000 as of June 12, 2020).¹ In the private health care system, a similar shift occurred, with New York University Langone Health (NYU Langone) experiencing an 80% decrease in the number of in-person visits between March and April 2020, whereas telemedicine urgent care visits increased from 82 to 1,336 visits in the first 3 weeks of March alone.^2,3

Historically, physicians have received limited formal telemedicine training. In light of COVID-19’s telehealth expansion, the American Medical Association (AMA) and American College of Physicians (ACP) published guides and resources to support physicians and practices in expediting the implementation of telemedicine.^4,5 The Association of American Medical Colleges (AAMC) is developing measurable telehealth competencies for future physicians, but clinicians who were already practicing were thrust into a new frontier with little or no additional training or practice. Although practicing physicians faced a rapid transition to a “new normal,” research suggests that many clinicians have expressed hesitancy toward telemedicine as the wave of the future.⁶

While a large body of literature supports the effectiveness of telehealth, much of this evidence surrounds specific targeted interventions and mobile health (m-health) tools.^7,8 Traditionally, these m-health tools have been used for remote monitoring or asynchronous assessment of vitals and symptoms. Few studies explore the ongoing routine use of telehealth to deliver comprehensive primary care.⁷ Effective primary care requires accurate information gathering, the establishment of trusting and constructive clinician–patient relationships, and utilization of patient education and activation skills. Understanding how best to educate, support, and prepare clinicians for practice in dynamic health care systems will facilitate quality telemedicine care.

To explore physician experience with telemedicine in the context of the COVID-19 pandemic, we developed and distributed a needs assessment survey to physicians who were practicing in our expansive system, which includes public, private, and veteran’s affairs (VA) hospitals. Ultimately, we sought to describe attitudes toward and challenges surrounding provision of quality virtual patient care.

Materials and Methods

For the purposes of this needs assessment, “telemedicine” was defined as a virtual visit or a real-time remote appointment between a patient and their physician over telephone or video, to an outpatient physician in our hospital network. Surveys were distributed via e-mail to all NYUs outpatient general internal medicine (GIM) faculty physicians working at NYU Langone, NYC Health + Hospitals/Bellevue and Gouverneur, or VA NY Harbor Health System clinics (n = 378). Adjunct physicians were excluded. The survey was distributed five times between May 20, 2020, and July 20, 2020, in an effort to maximize the response rate. Surveys were completed online and responses were completed anonymously. There were no incentives for participation, other than the opportunity to provide feedback on experiences.

Survey items were designed through review of the literature, with particular emphasis on content drawn from recent systematic reviews, primary care provider leadership consensus, and pilot testing.^9–11 Items included Likert and open-ended questions regarding demographics, experience with televisits (including average duration and number conducted), and attitudes toward care (Supplementary Appendix SA1). Specific questions covered experiences while conducting televisits, barriers to successful deployment, communication while working with a remote modality, and attitudes toward future utilization.

Survey collection and data maintenance were completed using Qualtrics, a survey distribution and data collection tool. Descriptives were examined for all respondents, but analyses focused on comparing type of site (public versus private). In our system, physicians working in the public outpatient practices predominantly provide telephonic care due to a limited capacity for video integration among the safety-net patient population. Respondents at NYU Langone provide predominantly video visits. Video visits at the private sites were conducted via a platform embedded in the electronic medical record. Audio visits were conducted over cell phone or landline. Results were reported both for the total group (including those who self-identified as VA and “other”) as well as based on the public versus private sites distinction (Tables 1 and 2). The 4- and 5-point Likert scale item response categories (Supplementary Appendix SA1) were dichotomized (Disagree vs. Agree, More Difficult vs. Equal/Less, and Sometimes/Often vs. Never) for analyses. Chi-square analyses were computed to assess differences by site. An alpha level of 0.05 was set for significant p-values. Open-ended question data were independently examined by authors J.A.W., L.A., and H.F. using a qualitative content analysis-based approach to identify emerging themes. Following initial independent assessment of physician responses, group consensus was reached through iterative discussion and recoding to identify final themes. This project was certified by NYU Langone’s Institutional Review Board as a quality improvement project, and data were collected without identifiers, so individual level consent was not required.

Results

Surveys were sent to 378 primary care physicians, 195 of which responded to the survey (a 52% response rate). The average number of years in practice for all participants was 18 years (range: 1–50 years), which did not differ from nonrespondents. 180/195 (92%) respondents had completed a virtual visit with a patient between March and July 2020. The average self-reported number of televisits conducted during this period was 177 (range: 5–1,000), and the average length of reported time spent with the patient per call was 17 min (range: 8–45 min).

Of those with complete site information (n = 152), 58 practiced in public outpatient settings (i.e., a Federally Qualified Health Center, or similar), 64 practiced in private outpatient settings (faculty practices), 11 practiced at the VA, and 19 were “other” or missing data. Physicians self-reported utilizing a mix of virtual technologies (42% audio, 33% video, 25% both), although with a preponderance of one or the other. Full demographic information is presented in Table 1. Respondents who identified as practicing at the VA and those with unclear or missing site information were included in the overall analyses but excluded from site comparisons due to small sample sizes (Table 2). Three significant themes emerged from the analyses of both the quantitative and qualitative data: technological challenges, new systems issues, and new patient/physician communication experience (Tables 2 and 3). Those surveyed also identified benefits of telemedicine.

TECHNOLOGY-RELATED BARRIERS

Ninety-eight percent of respondents in the total sample (n = 152) reported having problems establishing a connection from the patient’s end on occasion. Eighty-four percent reported difficulty in establishing connection on the physician’s end, with reported challenges including “patients didn’t know how to get on the video call” (Table 3). Issues with establishing connection and sound quality occurred frequently in both public and private settings (Table 2). Using interpreter services was also problematic for physicians, with 78% reporting occasional to frequent troubles. Public physicians reported challenges with interpreters more often than private physicians (p = 0.005). Issues with interpreters included visit participants speaking over each other and having more difficulty understanding patient’s concerns (Table 3).

NEW SYSTEMS-RELATED BARRIERS

Thirty-four percent of physicians found it more difficult to share information with other health care team members when conducting a virtual visit, and 39% found it more difficult to work collaboratively with health care team members. When subdivided, 23% of private and 38% of public physicians found information sharing more difficult (p = 0.082). Furthermore, 30% of the private physicians and 43% of public found team collaboration more difficult (p = 0.123) (Table 2). Qualitative comments expressed potential changes necessary, including the need to train clinical staff on protocols for each type of visit and to assess patient preferences (Table 3). Physicians also emphasized that they were now responsible for work that, traditionally, other support staff within the microsystem had owned: “the large administrative load that includes having to do much of the scheduling myself, having to learn the new system and help train my patients to do video calls.”

PATIENT/PHYSICIAN COMMUNICATION BARRIERS

Physicians identified challenges in several domains, including physical examinations (95%), establishing relationships with new patients (70%), taking a good history (47%), and educating patients (35%) (Table 2). In review of qualitative data (Table 3), physicians noted barriers to easy communication, by stating, for example, that “communication is much harder over the telephone without visual cues to evaluate emotions. Many of my patients are older and sometimes have difficulty communicating over the telephone where they would be able to do more in person.” Barriers existed in video visits as well, including “Need for extra pause to allow for lag [and to] look at camera, not at image for ‘eye contact’” and “Pts [patients] don’t always have privacy, so [they] don’t talk about sensitive issues (domestic violence, sexual sx [history]).” Noted difficulties were further solidified by the sentiment that “the real barrier is inability to examine the patient personally” and “Communication is fine but something essential is lost without face to face encounter.”

ATTITUDES TOWARD UTILIZATION AND PERCEIVED BENEFITS

While 69% (75% of private physicians, 69% of public physicians when subdivided) look forward to further integrating telemedicine into their regular practice, 46% of the physicians (39% private, 52% public) in the sample were concerned with idea of conducting too many televisits in the future (Table 2). Despite concern over future quantity of televisits, a majority (89%; 84% private, 93% public) believe that a mix of in-person and virtual visits could be beneficial for their patients in the future. Ninety-two percent of participants noted that gaining a view of a patient’s home life is easier during a televisit, a theme further reinforced by open-ended commentary. Qualitative feedback from respondents identified their awareness of the benefits of the rapid influx of telemedicine and remote care provision including easier connectivity to patients, an uptick in remote monitoring, and a more thorough understanding of patient’s home life (Table 3). Physicians recognize the power of using technology for clinical care provision. One reported, “even though I was apprehensive at first, I really see the value of telemedicine. I was able to care for sick patients from the confines of their homes regularly” and that they “… really enjoyed seeing my patients’ apartments. It gave me access to parts of their lives that had been invisible.” Other physician-reported telemedicine benefits included instant access to patient medications, the ability to connect with immobile/no-show patients, and general usefulness of remote care for uncomplicated cases.

Discussion

Respondents felt that telemedicine presents both challenges and opportunities for effective care. The challenges of telemedicine were specifically related to (1) technology, (2) the interdisciplinary microsystem, and (3) communication skills. Notably, there were differences between public safety-net sites when compared with private outpatient sites. We found that the most notable differences in physician experiences with telemedicine were rooted in the type of underlying technology employed. Public safety-net practices conducted mostly telephonic visits, whereas private outpatient sites largely conducted video visits, despite both systems using the same electronic medical record. Importantly, our results suggest that telephone visits might limit information gathering compared with video visits and that efforts must be made to increase the provision of video visits for all patients.

Although there are high rates of satisfaction among telemedicine utilizers,^12,13 some specific challenges exist for both physicians and their patients, which were noted in the current study. Eighty-four percent of respondents reported “sometimes or often” having problems establishing a connection with the patient. In particular, they noted troubles with joining the call and similar technological issues. They also commented on patients’ difficulty in showing specific body parts. Patients in safety-net systems often lack broadband internet or a computer at home and mostly rely on mobile phones for connectivity.¹⁴ In many cases, these same patients lack technological literacy necessary for successful utilization.¹⁴ Addressing technological concerns and challenges through increased investment in technology infrastructure and training for patients will be essential for the widespread success of telemedicine, particularly video visits.^12,13 Furthermore, all practices, including safety-net sites, need to pivot to higher percentages of video visits. In addition, other patient-level issues that must be acknowledged for sustained effective telemedicine utilization include a lack of private space/distractions in the home as well as potential language challenges during visits, both of which were noted in qualitative comments.

Many practices now emphasize that the clinical microsystem—the smallest unit at which care is delivered—should be organized around patient needs; however, the organization of the telemedicine microsystem remains unclear and minimally defined. Factors that operate at the level of the clinical microsystem have been shown to be important influencers of both processes and outcomes of health care.^15,16 Broadly, microsystem factors can include practice structure, staffing patterns, patient volume, resources and support systems, and organizational culture. Within primary care, all these factors are knit together through effective multidisciplinary teamwork. Infrastructure, care coordination, workflow, and internal processes have a major impact on patient experience. The traditional clinical microsystem cannot be transferred directly into the telemedicine domain, as telemedicine requires an adapted re-creation of these systems on a virtual level. In our needs assessment, physicians noted issues with the microsystem as a rate-limiting factor for a successful televisit experience. Study physicians noted having to take on additional work that had previously been done by other members of the care team. Ensuring that physicians are supported in virtual care provision will be pivotal to continued adoption of telemedicine and satisfaction with use.¹² Health care systems must develop effective protocols and workflows to ensure that telemedicine primary care practices involve an integrated team of nursing, medical assistants, and social service workers who can provide efficient, meaningful virtual patient care experiences.

While technological and microsystem issues posed significant challenges in providing care, physicians also noted a number of communication challenges, including noise, distractions, and language barriers. Given that physicians’ communication skills and their interactions with patients are consistently shown to be essential to quality care and outcomes, especially in primary care, it is critical to understand how telemedicine can impede or facilitate this communication.^17–19 Studies suggest that specific physician behaviors during telemedicine visits, including a lack of small talk, minimal empathic statements and praise, and limited chances for patients to speak up can diminish patient activation, when compared with in-person visits.^20,21 This, when coupled with the communication barriers noted by respondents, reinforces the need for medical education curricula to enhance communication skill for this burgeoning modality of care delivery. Clinicians will need to hone their rapport building (e.g., understanding the patient’s care setting) and patient education (e.g., understanding next steps and how the care team will assist) skills. Additional essential skills should include defining the scope of each visit, performing physical examinations virtually, and leveraging the virtual environment/setting to obtain additional information (e.g., asking a patient to go into the other room to obtain a medication bottle or demonstrate a medical technique). Furthermore, ensuring that physicians know how to effectively engage with a variety of patients remotely should be a priority. Establishing an understanding of how best to utilize and call upon resources (e.g., interpreters, social work) in a way that enhances a visit is also essential.

Recent data from the University of California at San Francisco’s GIM primary care practice note a decline in visits for any medical treatment (even when inclusive of televisits) post-COVID when compared with the pre-COVID era. Specifically, this was notable for older patients (65+ years), those with a preferred language other than English, those who had Medicare or Medicaid, and those who identified as Black/African American, Latinx, and Asian/Pacific Islander.²² Many of the challenges of primary care for vulnerable populations are not unique to telemedicine; however, the modality amplifies some of them.²³ Our survey’s findings confirm the particular challenges of telemedicine in the setting of primary care for the underserved or vulnerable: less access to video visits, challenges with the use of interpreters during a virtual visit, and the impact of a lack of access to the internet or technology in receiving care. These findings also emphasize the critical importance of placing equitable access at the forefront of the telemedicine expansion and are consistent with concerns noted in recent literature in this area.²⁴

Our descriptive study is not without limitations. This survey was administered to a group of physicians affiliated with a large, urban academic hospital system. We can only postulate that these findings are applicable to the larger NYC community. As our data relied on self-report for clinicians only, cross-validation with hospital administrative and an exploration of clinical data could be a logical future step. Furthermore, the data collected do not take into consideration the experiences of subspecialties or other types of clinical care providers. Developing complementary surveys to assess the experiences of other types of care providers could expand our findings and provide data for quality improvement projects. Our survey provides an overview of experiences in early 2020 during the initial surge of the pandemic in NYC. In addition, as both clinicians and patients become more comfortable with the modality, we expect priorities and needs to shift. Follow-up surveys will be necessary as telemedicine needs and practices continue to evolve postinitial surges.

Moving forward, clinical care systems must understand the essential role that telemedicine now plays in outpatient medicine as the United States undergoes widespread community transmission of COVID-19 infections. Systems need to invest in understanding the patient-level factors associated with telemedicine usage, and changes can be made to ensure widespread preparedness and reception of telehealth. These include (1) a critical review and enhancement of the telemedicine microsystem, (2) better integration of language services when necessary, (3) provision of internet services to the most vulnerable patients, and (4) clinician training on communication skills. Finally, systems that lack video visit capacity should work toward implementing video applications and tools to ensure the highest quality care for their patients.

Conclusions

Our needs assessment identified challenges that physicians experienced as televisits were implemented during the COVID-19 pandemic. Importantly, although, findings provided insights into attitudes toward the future of telemedicine, which will, undoubtedly, be integrated into the new outpatient care delivery model. Physicians, while struggling occasionally, saw the benefit of telemedicine and believe that it will be an important adjunct to delivering high-quality patient-centered care as we move forward. Our sample of physicians believe that virtual care can contribute to a stronger, more holistic view of a patient’s home environment, remote monitoring of biometric measures, and that telemedicine can help connect with hard-to-connect with patients. As we consider a “new normal” and a possible prolonged community transmission of COVID-19 infection, we believe that it is essential to establish telemedicine training, tools, and protocols that meet the needs of both patients and physicians.

Authors’ Contributions

J.A.W. and H.F. were primary data holders and collectors. J.A.W., H.F., C.G., L.A., and S.Z. designed the survey and contributed to analyses. All authors contributed substantially to the writing and revision of the article, and results were reviewed and approved by each author.

Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

Supplementary Material

Supplementary Appendix SA1