A scoping review of the unassisted physical exam conducted over synchronous audio-video telemedicine

Abstract

Background

This scoping review aims to provide a broad overview of the research on the unassisted virtual physical exam performed over synchronous audio-video telemedicine to identify gaps in knowledge and guide future research.

Methods

Searches for studies on the unassisted virtual physical exam were conducted in 3 databases. We included primary research studies in English on the virtual physical exam conducted via patient-to-provider synchronous, audio-video telemedicine in the absence of assistive technology or personnel. Screening and data extraction were performed by 2 independent reviewers.

Results

Seventy-four studies met inclusion criteria. The most common components of the physical exam performed over telemedicine were neurologic (38/74, 51%), musculoskeletal (10/74, 14%), multi-system (6/74, 8%), neuropsychologic (5/74, 7%), and skin (5/74, 7%). The majority of the literature focuses on the telemedicine physical exam in the adult population, with only 5% of studies conducted specifically in a pediatric population. During the telemedicine exam, the patients were most commonly located in outpatient offices (28/74, 38%) and homes and other non-clinical settings (25/74, 34%). Both patients and providers in the included studies most frequently used computers for the telemedicine encounter.

Conclusions

Research evaluating the unassisted virtual physical exam is at an early stage of maturity and is skewed toward the neurologic, musculoskeletal, neuropsychologic, and skin exam components. Future research should focus on expanding the range of telemedicine exam maneuvers studied and evaluating the exam in the most relevant settings, which for telemedicine is trending toward exams conducted through mobile devices and in patients’ homes.

Background

Although telemedicine has existed for decades, the pandemic inspired a new need to avoid face-to-face contact in both providers and patients. This, coupled with changes in payment and regulatory frameworks designed to encourage the use of telemedicine, resulted in a meteoric increase in the use of real-time video interactions for medical care [1, 2]. It appears that many of these regulatory changes may be long-lasting. The US Centers for Medicare and Medicaid Services has extended the payment changes through the end of 2023. Even if some of the payment and regulatory alterations put in place to facilitate telemedicine revert, it seems unlikely that the environment will return to the way it was before the pandemic. Many are convinced that the public, now familiar with the convenience of telemedicine, will continue to choose providers that offer telemedicine for at least a portion of their care.

Although many observers view the expansion of telemedicine capability favorably, the rapid substitution of telemedicine for face-to-face encounters has led to concerns regarding quality, access equity, and safety [3]. Academic discussion about what constitutes (and how to measure) high-quality video telemedicine has only just begun [4,5,6].

One criticism leveled at telemedicine is that a video encounter precludes the performance of a physical exam. The physical exam is one of the core pillars of information gathering in medicine; critics correctly assert, however, that it is a mistake to assume that video interaction precludes a physical exam [7]. Experience with telemedicine has shown that it is indeed possible to perform a physical examination over video, albeit a different sort of exam than that which occurs in person. National organizations have even begun to teach how to do an examination over video at the graduate medical education level [8]. Since it is a relatively new technique, the specific strengths and limitations of the video exam have not been explored. Hence, it is appropriate to investigate differences in examination over video as compared to face-to-face medical evaluation.

Like the face-to-face encounter, telemedicine examinations may use tools or technology to facilitate the exam [9], but even in the absence of devices such as web-enabled stethoscopes and otoscopes, it is possible to perform examinations and obtain actionable information. Visual inspection alone can yield crucial data, and many elements of a traditional exam that are identified with the examiner’s hands appear to be well approximated by providers instructing patients to assist them with the exam. For example, a tele-pediatrician can ask their young patient to jump up and down when evaluating them for appendicitis. Although providers of the real-time video evaluation are developing practice patterns and sharing their experiences about the patient exam, there has been no systematic assessment of what is and is not known on this topic.

We sought to summarize the state of the published literature related to physical exam when conducted over live video. To this end, we conducted a scoping review to identify, appraise, and synthesize relevant data on the physical exam conducted via patient-to-provider synchronous, audio-video telemedicine in the absence of assistive technology or trained medical personnel on the patient side.

Methods

The scoping review methods are reported using the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) Checklist (see Additional file 1: PRISMA-ScR Checklist) [10].

Search strategy

A comprehensive literature search on the telemedicine physical exam was collaboratively developed by a research team member [PY] and medical librarian [KM], in consultation with senior investigators [PG, NN]. The initial search was performed on March 17, 2020, via OVID MEDLINE® ALL. It was then translated and rerun on OVID EMBASE and The Cochrane Library (Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials (CENTRAL), Cochrane Methodology Register). Search terms in each database included all MeSH terms and/or keywords associated with our research question, clustered around the key areas: (1) telemedicine and (2) physical examination. Search terms were joined using Boolean operators “OR” and “AND,” as appropriate. Search dates were limited to articles published after 1990 for relevance to current telemedicine practices. The full Ovid MEDLINE search strategy is available in Supplementary Appendix 1.

Study selection

Covidence (Veritas Health Innovation, Melbourne, Australia), a systematic review management software, was used for data management. After excluding duplicates using the automated Covidence workflow, reviewers [PY, HH, NN, MA, PG, MS] independently screened the titles and abstracts of the retrieved. Each citation was reviewed by two members of the research team, with consensus needed to move the citation forward to the next phase of screening. Discrepancies were resolved through group consensus. Inclusion and exclusion criteria were determined a priori. We included English primary research studies in which: (1) patients received any subset of a telemedicine physical exam by a licensed medical provider, (2) the exam was conducted over real-time audiovisual telemedicine, and (3) the exam was described in detail or validated by appropriate comparison to an in-person exam or another reference standard. We excluded studies in which (1) the physical exam was performed with the assistance of another medical provider or tele-presenter on the patient side and (2) the exam was assisted by equipment or technology obtained specifically for the purpose of the virtual physical exam. Excluded assistive equipment included but was not limited to equipment provided by the study to the patient for use in the virtual exam or equipment purchased by the patient specifically for the virtual exam. For example, a retinal camera or a dermatoscope provided to the patient for the purpose of the virtual exam would be excluded. Common medical devices sometimes found in the home, such as blood pressure cuffs, thermometers, glucometers, peak flow devices, and pulse oximeters, available to the patient during the virtual exam but not specifically intended for use only in the virtual exam, were not excluded.

Following the initial title and abstract screening phase, reviewers [PY, HH, NN, MA, PG, MS] independently screened the full text of citations using the same process outlined above.

Data extraction

A data extraction sheet was developed and pilot-tested iteratively to ensure relevant and consistent information capture. Definitions for each data extraction field were agreed upon by all authors (Supplementary Appendix 2). The following information was collected: study and patient characteristics, telemedicine encounter setting, physical exam component studied, diagnostic context, and validation measure of physical exam component if such a measure was reported. Data extraction was performed on included articles using Covidence and then exported for analysis in Microsoft Excel. Data extraction was performed by a team member [PY], and a second team member [HH, NN, MA, PG, MS, or SC] independently verified the extracted data.

Results

The literature search yielded 7920 unique results after de-duplication. Following the title and abstract screen, 7615 studies were excluded. Three hundred five studies underwent full-text review. Seventy-four studies ultimately met all inclusion criteria. Citations for all included studies are available in Supplementary Appendix 3. The PRISMA flow diagram outlining the study selection process is available in Fig. 1. A summary of the included studies with details on virtual exam maneuvers and study results is available in Table 1.

Fig. 1

Preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow diagram

Table 1 Summary of physical exam maneuvers and outcomes of the included studies organized by physical exam category. In the “Equivalence” column, “equivalent” means the telemedicine exam and the in-person exam were equivalent, “inferior” means the telemedicine exam was inferior to the in-person exam, and “NS” means the equivalence between the telemedicine exam and the in-person exam was not specified

Study descriptors

Publication dates of included studies spanned 1993 to 2020 with 2 citations from 1991 to 1995, 4 citations from 1996 to 2000, 6 citations from 2001 to 2005, 10 citations from 2006 to 2010, 23 citations from 2011 to 2015, and 29 citations from 2016 to 2020.

By country, the studies were most frequently conducted in the USA (44/74, 59%), Australia (10/74, 14%), the UK (4/74, 5%), Canada (3/74, 4%), Belgium (2/74, 3%), and Italy (2/74, 3%). By international region, the studies were most frequently conducted in the Americas (47/74, 64%), Europe (12/74, 16%), Oceania (11/74, 15%), and Asia (4/74, 5%).

Eighty-five percent of included studies were cross-sectional studies with the other 15% representing cohort studies, case-control studies, case series, qualitative studies, and case reports (Fig. 2).

Fig. 2

Pie chart showing the frequency of study designs

Sixty-seven (91%) of the studies were peer-reviewed articles and 7 (9%) were abstracts, presentations, or posters.

Study population descriptors

Reporting of demographic information, including age, sex, gender, race, ethnicity, and smoking status, was assessed. Fifty-seven (77%) studies reported at least one piece of demographic information as defined here, while 17 (23%) studies did not report any demographic information about the participants.

Fifty-four (73%) studies reported the mean or median age of the study population while 20 (27%) studies did not. Among the studies that reported mean or median age, the average mean or median age of the study population was 57.7 years (SD: 20.4 years). Four (5%) of studies were conducted in pediatric populations only. The median number of participants per study was 27 (IQR: 13-51).

Health conditions

The most studied medical conditions were Parkinson’s disease (14/74, 19%), Alzheimer’s disease (12/74, 16%), dementia (8/74, 11%), mild cognitive impairment (7/74, 9%), stroke (7/74, 9%), multiple sclerosis (4/74, 5%), and depression (3/74, 4%). Grouping the studied medical conditions more broadly, the most studied disorder classes were neurologic disorders (40/74, 54%), musculoskeletal disorders (8/74, 11%), psychiatric disorders (5/74, 7%), dermatologic disorders (4/74, 5%), urinary disorders (2/74, 3%), pulmonary disorders (2/74, 3%), cardiovascular disorders (2/74, 3%), otolaryngologic disorders (2/74, 3%), hematologic disorders (1/74, 1%), and digestive disorders (1/74, 1%).

Telemedicine exam

The telemedicine physical exam was performed, in descending order of frequency, by physicians (37/74, 50%), physical therapists (8/74, 11%), psychologists (6/74, 8%), nurses (4/74, 5%), psychometrists (3/74, 4%), advanced practice providers (3/74, 4%), speech and language pathologists (2/74, 3%), occupational therapists (2/74, 3%), and respiratory therapists (1/74, 1%) (Fig. 3). In 11 (15%) studies, the profession of the examiner was not specified.

Fig. 3

Bar graph showing the professions of the telemedicine providers

During the telemedicine exam, the patients were most commonly located in outpatient offices (28/74, 38%), homes and other non-clinical settings (25/74, 34%), hospitals (10/74, 14%), research facilities (5/74, 7%), emergency rooms (3/74, 4%), and ambulances (1/74, 1%) (Fig. 4). In 5 (9%) studies, the patient’s location was not specified.

Fig. 4

Bar graph showing the patients’ locations during the telemedicine exam

During the telemedicine exam, the providers were most commonly located in outpatient offices (37/74, 50%), hospitals (11/74, 15%), research facilities (4/74, 5%), emergency rooms (2/74, 3%), skilled nursing facilities (1/74, 1%), and at home (1/74, 1%) (Fig. 5). In 19 (26%) studies, the provider’s location was not specified.

Fig. 5

Bar graph showing the providers’ locations during the telemedicine exam

On patient side, the telemedicine encounter was most frequently conducted using only computers (49/74, 66%), any devices including computers, tablets, and smartphones (10/74, 14%), only tablets (5/74, 7%), and only smartphones (4/74, 5%). In 7 (9%) studies, the device used by the patient for telecommunications was not specified.

On provider side, the telemedicine encounter was most frequently conducted using only computers (52/74, 70%), only tablets (3/74, 4%), any devices including computers, tablets, and smartphones (2/74, 3%), and only smartphones (1/74, 1%). In 16 (22%) studies, the device used by the provider for telecommunications was not specified.

Although our methodology excluded studies where the exam was facilitated by a physically present health care provider, 14% of studies reported that the telemedicine exam was assisted by an assistant with no medical training and 86% of studies reported that the telemedicine exam was not assisted.

The most common components of the physical exam performed over telemedicine were neurologic (38/74, 51%), musculoskeletal (10/74, 14%), multi-system (6/74, 8%), neuropsychologic (5/74, 7%), skin (5/74, 7%), psychiatric (4/74, 5%), cardiovascular (2/74, 3%), abdominal (1/74, 1%), and head and neck (1/74, 1%) (Fig. 6). In 2 (3%) studies, the physical exam component was not specified.

Fig. 6

Pie chart showing the components of the telemedicine exam performed in all studies

In 61 (82%) studies, the outcome of the study was a comparison of the telemedicine exam and the in-person exam. Of these studies, 54 (89%) concluded that the telemedicine exam performed was equivalent to the in-person physical exam whereas 7 (11%) studies [11,12,13,14,15,16,17] concluded that the telemedicine exam was inferior to the in-person exam.

Discussion

This scoping review aimed to synthesize prior research on the unassisted physical exam conducted via real-time audio-video telemedicine.

Study descriptors

Analysis of the publication dates of included articles demonstrates a steady, shallow increase in publications from 1991 to 2010 followed by a sharp sloping rise in publications from 2011 to 2020. We hypothesize that the change in slope is because the decade from 2010 to 2020 saw telemedicine cross the gap from use by a few visionaries to acceptance by an early majority of pragmatists [18]. Though our study does not capture the literature from the COVID era, we expect the rise of publications on the telemedicine physical exam to accelerate even more rapidly as COVID catapulted telemedicine to the forefront of care delivery.

The English literature demonstrated a skew toward the studies conducted in the USA, which accounted for 59% of publications. Notably, none of the included studies were conducted in Africa, though all the other major international regions—the Americans, Europe, Oceania, and Asia—were represented. Telemedicine is an important component in addressing healthcare needs and access in Africa [8, 19], and increased attention to the telemedicine physical exam in this region is needed to address this gap in the literature.

Evaluated by study design, the body of literature is skewed toward a low level of evidence, with cross-sectional studies representing 86% of included studies. Notably, there were no randomized controlled trials that compared the telemedicine exam and in-person physical exam.

Study population descriptors

Almost a quarter (23%) of studies did not report demographic information about the study participants. Improvements in demographic reporting in this field would contribute toward a collective responsibility to seek clarity and transparency in the representativeness and diversity of study participants [20] and is especially important in the context of recent research showing disparities in telehealth utilization [21].

The sample sizes of all included studies were relatively small with a median size of 27 participants among included studies and future studies may benefit from larger sample sizes. The bulk of the literature focuses on the telemedicine physical exam in the adult population, with only 5% of studies conducted specifically in the pediatric population. Given the increased adoption of telemedicine by pediatricians and pediatric medical and surgical specialists [22], evaluations focused on the telemedicine exam in children are an important direction for future research.

Telemedicine exam

Though most telemedicine exams in included studies were conducted by physicians (50%), our research demonstrates a wide variety of other healthcare providers including nurses, physical therapists, and speech language pathologists, which fits well with our perception that telemedicine will be a way of providing evaluations and delivering health care across the medical spectrum.

In the included studies, the two most common patient locations were outpatient offices (38%) and homes and other non-clinical settings (34%). In many studies, patients were in the outpatient offices for research convenience or access to the non-portable telecommunications equipment supplied by researchers. Also, the most studied device used for telemedicine encounters by both providers and patients was the computer. Portable devices such as tablets and smartphones were only used by patients in 12% of studies and by providers in 5% of studies. As modern-day telemedical examination increasingly moves to the home and to mobile devices, it will be crucial for future studies to further examine the efficacy of the physical exam in a home setting on mobile devices [18].

The studies of the virtual neurological exam, in conjunction with the musculoskeletal, neuropsychologic, and skin exams, comprise 79% of the literature and are highly overrepresented compared to other exam areas. The overrepresentation of these physical exam components may reflect the early adoption of telemedicine in the corresponding medical fields. For example, telestroke services, first described in 1999 [23], have been integrated in stroke systems of care for more than 10 years [24]. The success of the neurologic exam in telestroke may have subsequently spurred the application of telemedicine for a range of other neurologic conditions such as dementia, epilepsy, movement disorders, and multiple sclerosis [25]. In addition, the prevalence of studies on the neuropsychologic exam may reflect the ease of transitioning a verbal exam to the telemedicine setting. Nonetheless, common exam maneuvers performed over telemedicine are not limited to these areas. An analysis of the American Medical Association’s 2016 Physician Practice Benchmark Survey shows that the specialists using telemedicine the most to interact with patients are radiologists, psychiatrists, and cardiologists. The subspecialties most often using video-conferencing-based telemedicine to interact with patients or other health care professionals are emergency medicine, psychiatry, and pathology [26]. Given the range of medical specialists practicing telemedicine, future research should aim to expand the range of exam components studied with particular attention to less researched exam areas.

Limitations

There are several limitations to this scoping review. First, the literature search was conducted on March 17, 2020. As a result, our review does not include articles on the telemedicine exam from the COVID era. Given the meteoritic rise in telemedicine use spurred by region-wide lockdowns and concerns over COVID transmission, a follow-up review including articles on the telemedicine exam in the COVID era would be an important future research direction. Second, our review focuses on the most basic form of a telemedicine physical exam—one without the assistance of technology and trained medical personnel on the patient side. We chose to focus on an unassisted exam to investigate the exam components that could be performed under the widest variety of clinical practice settings, but we acknowledge that assistive technology will be important in facilitating an accurate and effective physical exam over telemedicine in the future. Third, though many publications in our review use comparison between the telemedicine physical exam and the in-person exam as a primary outcome, in practice, outcomes with direct impact on patient care may be more meaningful. After all, the goal of the telemedicine examination is not to directly translate the in-person exam, but to create a set of virtual exam maneuvers that provide the highest degree of actionable information to the clinician. Lastly, our restriction on English language articles may have excluded relevant studies in other languages, as telemedicine is a global phenomenon.

Conclusions

In conclusion, the current research on an unassisted physical exam performed over synchronous audiovisual telemedicine is focused on the neurologic, musculoskeletal, neuropsychologic, and skin exam components. The majority (89%) of studies that directly compared the virtual exam to the in-person exam concluded that the telemedicine exam was equivalent to the in-person exam. Research evaluating the virtual physical exam is at an early stage of maturity with current limitations including body of literature mostly comprised of cross-sectional studies representing a low level of evidence, incomplete demographic reporting, paucity of studies focusing on the pediatric exam, and a minority of studies looking at exams performed on mobile devices and in home settings. Future research in this area should focus on expanding the range of telemedicine exam maneuvers studied and evaluating the exam in the most relevant settings, which for telemedicine is trending toward exams through mobile devices and in patients’ homes. In the long term, this review aims to guide the development of a virtual physical exam that provides the highest degree of clinically relevant information to the clinician.