1 Introduction

Telemedicine is known to increase access to specialty care for patients in rural and underserved areas and lower patients physical and financial burden of traveling to appointments [1, 2]. For physicians, it offers scheduling flexibility and the potential to improve work-life balance. In neurology, telemedicine is shown to be none inferior to an in- person evaluation [3]. As a result, it has been adopted by several neurological sub-specialties such as stroke, epilepsy, neuro-oncology and headache [3,4,5]. However, data about the potential challenges encountered in telemedicine in neurology and its impact on patient care is scarce. Neurology is a unique specialty because of the heavy reliance on physical examination and the wide range of physical and cognitive abilities of patients (e.g., dementia, movement disorders or neuromuscular disorders) which may affect the feasibility and efficacy of telemedicine in those subsets of patients. Recognizing challenges in telemedicine in neurology may help in identifying opportunities to optimize and facilitate the integration of telemedicine across a wide range of subspecialties in neurology. Our objective is to explore neurologists’ experiences, challenges and barriers in telemedicine.

2 Materials and methods

We conducted a cross-sectional study from March 23rd to August 5th, 2021. Currently, there is no validated tool to examine barriers and facilitators to telemedicine. Therefore, we developed a survey to explore neurologists’ experiences (facilitators and barriers) in the following domains: (1) neurologists’ experiences in conducting telemedicine visits (e.g., technological difficulties, need for technological support to conduct telemedicine visit), (2) Availability of administrative staff support (e.g., faxing documents, preparation for virtual visits, medication prior authorization, and obtaining medical records), (3) patient cognitive and physical disability, use of interpreter services, and need for technology support from a family member, (4) impact of inability to perform a physical neurological examination on the diagnosis and management plan, (5) areas of potential improvement in telemedicine to facilitate its use (e.g., time allotted for patient visits, administrative staff support, integration of electronic medical records with video/audio software).

The survey was disseminated among three faculty members at our institution (Boston University School of Medicine) for readability and clarity. The survey was revised and edited according to their feedback. Data were collected and managed using REDCap electronic data capture tools hosted at Boston University and Boston Medical Center, CTSI 1UL1TR001430.11 [6]. The study was approved by the Institutional Review Board at Boston University School of Medicine IRB # H-40679.

2.1 Study subjects

The survey was randomly distributed to approximately 200 neurologists from academic neurology institutions in the United States. The institutions were selected randomly from the list of top 100 hospitals in the United States. Neurologists were recruited from only academic institutions for recruitment feasibility. The first author disseminated the survey via email to the chairperson who disseminated the survey among their faculty members. Three reminder emails were sent to the chairperson. Because of the exploratory nature of the survey, a power calculation was not performed.

2.2 Statistical methods

Descriptive statistics were reported as frequencies and percentages for each survey question. SAS 9.4 was used to perform all statistical analyses.

3 Results

A total of 159 neurologists participated in the survey, but only 110 completed it. Most participants were attending neurologists (85%), followed by clinical fellows (11%), neurology residents (2%), advanced practice providers (1%), and others (1%). The most common subspecialties were vascular neurology (22%), epilepsy (19%), headache medicine (15%) and movement disorders (15%). Most participants indicated that they use telemedicine in most of their visits since the start of the COVID-19 Pandemic. Details of participants characteristics are in Table 1.

Table 1 Participants’ characteristics
Full size table

  • Technological challenges in telemedicine

    Nearly half (51%) of the neurologists stated that they experienced technological issues (e.g., interruptions by poor connection, inability to connect to video, inability to connect to audio, dropped calls, etc.) in 1%–20% of telemedicine visits. Thirty-one percent experienced technological issues in 21%–40%, 12% experienced technological issues in 41%–60% of visits, and only 6% experienced technological issues in 61%–80% of telemedicine visits. Furthermore, 57% of neurologists required assistance from informational technology (IT) support in 1%–20% of telemedicine visits, 14% required it in 21%–40% of telemedicine visits, and 3% required IT support in both 41%–60% and in 61%–80% of telemedicine visits respectively.

  • Patient-related challenges in telemedicine

    Neurologists were queried about the most common patient-related challenges in telemedicine (e.g., mental cognitive impairment, physical disability, technological support, and interpreter services for non-English speaking patients). Ninety-five percent of neurologists rated patient’s difficulty in setting up telemedicine visit (technological support) as the most common encountered difficulty. Other  encountered difficulties were patient's cognitive disability (37%), limited interpreter services (37%) and patient's physical disability (18%). A smaller percentage, 7%, rated issues such as limited internet bandwidth, lack of patient cooperation with the exam, patient not taking the visit seriously, instability of virtual platform, patients forgetting their appointments and inappropriate telemedicine setup as additional patient-related challenges. Details of patient-related challenges in telemedicine are in Table 2.

    Table 2 Patient-related barriers in telemedicine
    Full size table

  • Limited neurological examination

    We asked neurologists about the effect of the limited ability to perform neurological examination using the statement: “the inability to examine my patients makes me worried that I am providing a suboptimal care to my patients”. The responses were graded on a Likert scale (agree, strongly agree, neutral, disagree, strongly disagree). Thirty-four percent of neurologists agreed that the inability to perform neurological examination makes them worried that they are providing a suboptimal care to their patients, 26% disagreed, 9% strongly agreed, 8% strongly disagreed and 24% were neutral to the statement. Furthermore, we asked participants about the percentage of telemedicine visits in which they advised a subsequent in-person evaluation. Fifty percent advised an in-person evaluation in 1%–20% of their telemedicine visits, 16% advised it in 21%–40% of the visits, and 11% recommended it in 41%–60% of telemedicine visits. Smaller percentages, 8% and 11% of neurologists, recommended subsequent in person evaluation in 61%–80% and 81–100% of telemedicine visits, respectively.

    Neurologists were asked ‘how many additional studies do you order on average due to the lack of in-person physical examination. The majority (72%) stated no change in the number of studies ordered. Nineteen percent stated they order somewhat more additional studies and 9% stated ordering somewhat fewer additional studies due to the limited physical examination. Furthermore, 45% of neurologists reported somewhat more additional telemedicine-related responsibilities (e.g., extra documentation, additional paperwork, obtaining medical records, etc.), 13% reported significantly more responsibilities and 35% reported no change in administrative responsibilities. Smaller percentages reported significantly fewer responsibilities (2%) and somewhat fewer responsibilities (5%).

  • Issues to be addressed to improve telemedicine in neurology

    Lastly, neurologists were asked to rank the most common issues to address to improve telemedicine in neurology. The most frequently rated issues were increasing administrative support (39%), integrating EMR with video and telephone calls (37%), and sufficient time allocation to complete telemedicine visits (27%). Other responses included providing telemedicine devices to patients, improving internet access in rural areas, determining patient technological capability to participate in telemedicine visit, adding allotted time for patients requiring interpreter services, ensuring equal reimbursement for telemedicine and in-person visits, virtual rooming for patients via medical assistants and other staff members, and improving patient preparation for the visit.

4 Discussion

Our study provides an insight into the challenges encountered in telemedicine in neurology. These challenges can be categorized into technology challenges, limited physical examination challenges, and patient-related challenges.

  1. 1.

    Technology challenges

Nearly half of the participants encountered technological difficulties that required IT support in 1%–20% of telemedicine visits. While the range of 1%–20% is wide, our findings are consistent with the existing literature that showed similar technology support challenges [1]. Common technological difficulties include virtual platform functionality, connection reliability, accessibility, and availability of IT support [2,3,4,5, 7]. Suboptimal technology support may cause technical inefficiencies that may increase workload on neurologists as well as neurologists and patients frustration and delays care delivery [2, 8]. Establishing an effective telemedicine practice requires developing a strong technological infrastructure and using adequate equipment (e.g., hardware, software) that meet the requirements of the Health Insurance Portability and Accountability Act (HIPAA) [8]. In addition, it requires effective technological implementation, training and maintenance [2].In addition, supporting telemedicine programs via IT requires trained personnel with special set of skills such as understanding how video conferencing works, networking, firewall, and can interface and communicate effectively with clinical staff members [4].

  1. 2.

    Limited neurological examination challenges

Nearly one-third of the participants agreed with the statement that the inability to examine their patients is concerning and that they may be providing a suboptimal care to their patients. And nearly half of the participants stated that they recommended a subsequent in-person visit in up to 20% of their telemedicine visits. Nonetheless, they did not report a significant change in the percentage of additional ancillary tests ordered (e.g., MRIs) due to the limited physical neurological examination. Concerns about the consequences of both the limited examination and or inability to perform hands on neurological examination have been reported in neurology as well as in other specialties that have heavy reliance on physical examination, such as neurosurgery (specifically spine surgery) and orthopedics [2, 4, 5, 7]. While research showed that neurological examination via telemedicine is non-inferior to face-to-face evaluation [9,10,11], our results showed that one third of the participants have negative perception about the impact of the limited neurological examination via telemedicine. And that it is influencing their decision in recommending a subsequent in-person evaluation. There have been creative efforts to overcome some of these barriers, for example using Telemedicine Examination Medical Devices Peripherals [12]. These medical devices (e.g., tele-ophthalmo-scope, video-otoscope) allow transfer of data stream through a shared digital interface to be reviewed by the physician [12]. In neurology, some disease-specific examination tools such as the Unified Parkinson’s Disease Rating Scale (UPDRS) have been modified to enable examination virtually by removing the rigidity and retropulsion tests (mUPDRS) which has been shown to have a high reliability and validity compared to the standard UPDRS [13]. Similar validated modified scoring systems can be developed for other neurological diseases to facilitate neurological examination, especially for conditions that have heavy reliance on motor examination.

  1. 3.

    Patient-related challenges

Patient- related barriers in telemedicine such as patient’s age, education and ability to navigate technology have been described [14]. In our study, patients’ difficulties in connecting with telemedicine visits was the most common challenge encountered. Similar findings were reported in a study that examined patients’ barriers in telemedicine, which showed 65% of patients were interested in telemedicine visits but only 54% completed one successfully [15]. Challenges related to internet data access/connection stability (reported in at least 1 in 4 patients) and inability to download a video application on the call are common reasons for failure to complete telemedicine visits [15]. Factors such as patient’s age, socioeconomic status, and computer literacy are associated with unsuccessful telemedicine visits [14]. As a result, approximately 25% of patients who face challenges in connecting to a telemedicine visit, often need help from a family member for technological support [15]. Notably, the rate of successful completion of telemedicine visits was especially low among non-English and non-Spanish speaking patients and those who are > 55 years old [14,15,16]. This is concordant with our findings, where using interpreter services for non-English speaking patients was one of the barriers for successful telemedicine visits. Language dis-concordance between physicians and patients is associated with confusion, frustration, and medical errors [17]. As a results patients with low English language proficiency may face significant challenges navigating the healthcare system [17,18,19]. Telemedicine may exacerbate this issue because patients need to navigate several steps to login a virtual platform which often requires English proficiency. In addition, some virtual platforms lack integrated interpreter services which may cause delay in connecting with interpreter services resulting in overall workflow and clinic schedule delays [17, 19].

Lastly, patients cognitive disability was another commonly encountered challenge in telemedicine visits. Although studies examining efficacy of telemedicine in patients with dementia were promising, adequate supporting staff and availability of a caregiver that can navigate the virtual platform  are needed for using telemedicine successfully [20]. However, patients with suboptimal support or those with sensory (auditory/visual) or language impairments may not have the same success with telemedicine visits [20]. Unfortunately, most telemedicine platforms do not yet have features to facilitate communications with patients who are deaf, blind, or have language or mobility impairments, or cognitive disabilities [8, 20]. While patients with cognitive and or physical disabilities may face similar challenges with in-person visits, using telemedicine may amplify these challenges and increase the gap of inequitable healthcare access for them.

  • Areas of potential improvement to facilitate telemedicine in neurology:

Participants rated providing adequate administrative support and facilitating ordering ancillary testing (e.g., EEG, EMG, etc.) as the most important issues to address in order to optimize telemedicine in neurology. In our study, nearly half of the participants reported that their department did not develop a telemedicine protocol or provide virtual workflow training. Examples of common administrative issues that are reported in the literature include missing or incorrect patient’s contact information and lack of patients preparation for the visits [7]. Both, suboptimal virtual work-flow and administrative support can cause negative physicians’ and patients’ experiences, and generate more work for the physicians [5]. Therefore, establishing a virtual-workflow and staff training t  to support telemedicine visits are needed to optimize both physicians’ and patients’ experiences [5]. Lastly, integrating video and telephone calls into the EMR was another common issue to be addressed to improve telemedicine visits. One of the organizational challenges in adopting telemedicine is choosing the appropriate electronic platform. During the last three years, there has been an exponential increase in the different software and electronic platforms for telemedicine [21]. The decision of which platform to purchase is often met with hesitancy because of the variability in functionality and  cost. Having a video-conferencing function integrated into EMR has the advantage of being more functional and minimizes physicians and administrative staff burden to learn different software platforms.

4.1 Practice change implications

Opportunities to improve the quality of telemedicine in neurology include developing an efficient EMR system with integrated video-telephone functionality, creating an effective virtual- workflow, and ensuring the availability of adequate administrative staff support. To address the limitations in physical examination, develop a triaging system to determine which patients are likely to need in-person evaluation and to adopt new technological tools to augment the virtual neurologic examination. Finally, to overcome some of the disparities in accessing healthcare, investing in built-in interpreter services and providing support for patients with difficulty accessing technology and or require special assistance due to cognitive disabilities.

5 Study limitations

Our study limitations include its purely descriptive nature, small sample size that limited the ability to analyze based on subspeciality, inclusion of only neurologists practicing in academic institutions, lack of response rate calculation and the unvalidated survey. In addition, we did not include neurologists’ demographic data (e.g., gender, years of experience), and did not dichotomize inpatient versus outpatient experiences. Lastly, these experiences and challenges are reflections of participants special interest in this topic which may reduce the generalizability of our findings. Lastly, some of the survey questions included a wide percentage range (e.g., 1%–20%).