Key words
telemedicine - video consultation - hand - examination - COVID-19
Introduction
As part of the COVID 19 pandemic with its imposed lockdown, scheduled clinics in all
departments were cancelled and patients rescheduled to later dates. Telemedicine offers
patients the opportunity to interact and consult with experts. Examination of the
hands is an obvious choice, since both extremities can be “offered to the camera”
quite readily. It has already been demonstrated in recent years that telemedicine
may be used to obtain expert opinions [1]. In rural and sparsely populated areas in particular, this allows for sophisticated
medicine [2]; its utilisation there can also help reduce the number of patients transferred with
hand injuries [3]. Virtual telemedicine visits can also contribute to postoperative follow-up [4]. Until now, however, telemedicine has focused on transmitting and assessing findings.
Electronically transmitted radiographs [5], as well as the acquisition and transmission of smartphone images of the hand, facilitate
reliable treatment and follow-up [6]. Technical progress, with its ubiquitous access to computers, tablets and fast internet,
makes it possible to conduct virtual hand surgery clinics without any problems [7]. The quality of real-time images and videos offers outstanding possibilities for
examination and treatment in hand surgery in the era of COVID-19 and beyond.
Virtual clinics permit direct interaction between physician and patient even without
physical contact. Medical questions and instructions during the interview and examination
can be directly realised and observed. Despite the increasing use of telemedicine
procedures, to our knowledge there is no scientific data available to date comparing
the full examination of the hand in virtual clinics with the traditional doctor-patient
interaction.
This study aims to compare the virtual hand surgery clinic with a conventional hand
surgery clinic.
Material and Methods
Patient population and data collection
30 examinations of 60 hands were performed twice in succession: first, as part of
a virtual clinic with physical separation of physician and patient (VC), and then
by in-person contact (IPC) with the same physician. 5 unilateral pathologies of the
hand (Dupuytren contracture II° D V, trigger finger D IV, subcapital metacarpal fracture
D V, scaphoid pseudarthrosis, and de Quervain disease) were selected and assessed
by 6 senior physicians in the Department of Orthopaedics and Trauma Surgery of a German
university medical centre.
These virtual clinics used the same setup as for the regular virtual clinic sessions
in the department itself. For both the patient and physician, this included a computer
with a 23-inch screen, HD webcam, built-in microphone and audio output (HP EliteDesk
800 G1, HP EliteDisplay E232, CA, USA; Logitech C920 CA, USA). Audio and video transmission
was in real time, with both physician and patient facing a split-screen allowing them
to see themselves and the person they were talking to at all times. The examiner also
had the option of taking screenshots (Snipping Tool, Microsoft, WA, USA) and measuring
angles by drawing auxiliary lines.
The examination took place in a room in the department without the support of relatives
or medical staff. In order to be able to assess the feasibility without interfering
examiner dependency, both examinations were performed by the same physician.
All examinations were undertaken after obtaining detailed informed consent. The trial
was approved in advance by the institutional review board responsible for our department
(IRB #163/20).
Examination structure
The initial examination was first conducted by virtual clinic (VC), followed by a
traditional in-person contact with the same examiner (IPC). In order to structure
the examination and documentation of the findings, a systematic questionnaire and
examination form was developed and used in all examinations (Appendix 1). It included
selected questions on medical history, sketches of the identified findings, and documentation
of the range of motion of the joints. In addition, selected questions were asked for
the examination of strength, for motor function and nerve tests, as well as for examination
findings in common disorders. For the quick documentation of normal findings, preset
reference values could be check-marked.
The medical history of the hand was obtained directly during the virtual examination.
For the inspection, documentation of the range of motion and the subsequent examination
process, the examiner performed the required movements and manipulations in front
of the camera with direct control and adjustment of patient behaviour. In the virtual
clinic a thin-walled 1.5 l PET bottle, ⅔-full, was used to check the strength as well
as motion under load. Akin to intrinsic muscle testing, a sheet of paper was used
to examine the pincer, key pinch and power grips as well as the finger function. The
neurologic status was checked by assessing motion sequences; pressure points were
assessed to rule out nerve compression syndromes, and the correct technique was checked
in the virtual image. Common disorders were assessed separately through typical pain
locations, function and stress tests. At the end, the diagnosis and recommended course
of action were documented.
In the following traditional examination (IPC), the examination form was followed
in the same fashion. The examiner also had a goniometer to measure the range of motion.
Examination workflow (VC)
Each examination started with the patientʼs medical history and the description of
his/her symptoms. Active questions concerned the presence and timing of symptoms,
previous accidents or increased strain on the hands, pain intensity (VAS) and other
complaints, e.g. related to elbows or cervical spine. The symptomatic areas of the
hands, demonstrated upon request, were marked on hand-drawn sketches. By presenting
both hands to the camera, they were inspected for swelling; erythema; haematoma; atrophy/hypertrophy;
scars; and deformities. Findings were completed by active questioning.
The patients palpated their hands themselves as instructed and demonstrated by the
examiner. The examiner inquired about pain, hyperthermia and sensitivity with differentiation
from dysaesthesia and anaesthesia. These findings were also incorporated into the
manual sketches of the hands.
Further examination of the function was performed by a motion sequence demonstrated
by the examiner. Starting with the demonstration of pronation and supination while
standing with the elbows tucked in ([Fig. 1 a]), and continuing in a sitting position with dorsal extension and palmar flexion
as well as ulnar/radial abduction in the wrist, the range of motion of the thumb and
finger joints were also documented for both hands.
Fig. 1 a Supination while standing with the elbows tucked in. b Dorsal extension of the pronated hand holding a filled bottle. c, d Holding and stretching a sheet of paper with the thumb and index finger in pincer
grip.
In addition to the depicted measurement points, the examination form offered space
for documenting the range of motion as well as for check-marking preset reference
values.
Strength was tested with the help of a partially filled plastic bottle. Dorsal extension
with pronated wrist ([Fig. 1 b]), palmar flexion with supinated wrist, rapid pronation-supination, and strong squeezing
of the bottle had to be performed while holding the filled bottle and with the elbow
flexed. In the latter test under full power, the thin-walled bottle should exhibit
deformation and/or deformation noise. The degree of strength was classified as strong
(3 – performance without any problems), difficult (2 – performance with some problems),
weak (1 – motion only without bottle) and not possible (0).
Motor function tests were performed with the thumb-fingertip contact and the following
grip types: pincer; key pinch; power; and spherical. Differences in strength between
the pincer and key pinch grips of both hands were checked by jerking a sheet of paper
between both thumbs and index fingers ([Fig. 1 c] and [d]). The power grip was checked by opening or closing the bottle, the spherical grip
by holding the bottle cap in the hollow of the hand.
The examination process also included global neurologic tests. The radial nerve was
assessed by extending the wrist and fingers with the elbow flexed, while the deep
branch of the radial nerve was evaluated by powerful, rapid supination with the filled
bottle. The median nerve was assessed by fist closure, palmar flection of hand and
fingers, and by opposing the thumb with the ring finger and little finger, while the
ulnar nerve was evaluated by fist closure. In addition, the median and ulnar nerves
were examined by flexion of the supinated hand loaded with the filled bottle and checked
with the Ochsner test (median nerve) and intrinsic muscle test (ulnar nerve), the
latter by holding a sheet of paper between the ring and little fingers.
Palpation was performed according to the symptoms as a selective pain provocation
test with imitation of the motions and pressure points demonstrated by the examiner.
Triggered pain and abnormal motion patterns were regarded as pathological. The pronator
teres syndrome (a) was assessed by powerful pronation with strong compression of the
forearm by the other hand, the ulnar canal (b) by strong pressure with the thumb of
one hand on the ulnar wrist flexor crease of the other hand. The carpal tunnel (c)
was assessed by a (short) Phalen manoeuvre and the Tinel sign, the cubital tunnel
(d) by local pressure. Other examinations included pain on vigorous fist closure,
tenderness when putting pressure on the flat hand resting on the table and palmar
tenderness above the wrist medially and radially for scapholunate dissociation (e)
and scaphoid-centred pathologies (f). Other guided tests included tenderness above
the thumb saddle joint and when the thumb was compressed and
rotated as a guided grind test for carpometacarpal osteoarthritis (g); tenderness
over the first extensor compartment and a positive guided Finkelstein test for de
Quervain disease (h); tenderness over the A1 pulley with provoked pain on focal pressure
and extension of the affected finger in trigger finger/A1 pulley stenosis (i); limited
extension despite support with the unaffected hand and palmar scarring in Dupuytren
contracture (j); pain, swelling and laxity in the metacarpophalangeal (MCP) joint
of the thumb in ulnar collateral ligament rupture/skierʼs thumb (k); pain and swelling
with plump joint contour in osteoarthritis (l); synovitis of the MCP joints with ulnar
deviation of the fingers; extension deficit and swan neck or buttonhole deformity;
as well as pain and possible malalignment in fractures of fingers (m); metacarpals
(n); carpals (o); or distal radius (p).
Analysis
The examination was analysed and the graphs were created with Graphpad Prism (v8.4.2,
Graphpad Software LLC, CA, USA). Range of motion statistics for the measurements by
virtual clinic (VC), in-person contact (IPC) and for the difference were descriptive
with mean value; median; standard deviation and error; and confidence level. The degree
of agreement was determined by Spearman correlation and the significance level defined
as p < 0.05. Descriptive and qualitative analysis was chosen for the sketches and
the strength, function and provocation tests.
Results
A total of 4560 individual range of motion measurements were obtained in 30 examinations
of both hands, each via virtual and in-person clinic.
The examination techniques demonstrated a significant (p < 0.0001) and high correlation
(R = 0.995, confidence interval 0.9946 to 0.9954) ([Fig. 2]).
Fig. 2 Correlation between hand range of motion measurements obtained by VC and IPC.
In the virtual examination 84.6% of the measurements deviated by less than 5° compared
to the in-person examination, with 92.8% deviating by less than 10° ([Fig. 3]).
Fig. 3 Distribution of the deviation of documented ranges of motion between virtual and
in-person measurements.
When documenting the medical history, pain intensity and location, the findings obtained
during the virtual examination were identical to those in the in-person examination
(30 of 30 examinations). Differences in the inspection of the hands were noted when
looking at scars. While scars were noted in 18 IPC hand examinations, scars were identified
in 8 (44.5%) VC hand examinations, with only 2 of those (11.1%) identifying all scars.
There were no unilateral sensitivity disorders and muscle changes (atrophy/hypertrophy)
in the hands examined, and thus none were documented.
The demonstrated motor function tests as well as the provocation tests specified by
the examination form also showed agreement with corresponding documentation in both
examination modalities. More detailed examinations, such as for strength with classification
according to Janda, and also more differentiated provocation tests were not feasible
in the video visit with the tools provided.
The unilateral pathologies with their disease-specific findings of limited finger
extension and scarring of the skin in Dupuytren contracture (j); snapping extension
of the ring finger with tenderness over the A1 pulley (i); pain, swelling and restricted
range of motion in subcapital metacarpal fracture (n); pain when leaning on the hand
in previous scaphoid fracture (f); as well as tenderness over the extensor tendon
sheath and a positive Finkelstein test in de Quervain disease (h); were identified
in all cases and diagnosed in both the virtual and in-person examination.
Discussion
The introduction of virtual clinics in hand surgery during the COVID 19 pandemic lockdown
was imperative for non-contact patient care. The benefits of obtaining expert or second
opinion via telemedicine consultations have already been confirmed [8] and the feasibility of conducting these as part of orthopaedic clinics has been
demonstrated [9]. Smartphone photographs can measure finger position and mobility with the same accuracy
as goniometers [7]. Up to now, it had been unclear whether hand surgery patients could also be adequately
assessed, diagnosed and managed via virtual clinics. This feasibility trial addressed
the first part of this challenge – the examination, assessment and clinical diagnosis.
The goal was to clarify whether the findings obtained during both the virtual clinic
and in-person examinations were identical. In this respect, an approach with identical
examiners for each constellation of findings was chosen. In order to prevent changes
in the findings over time, the virtual and in-person examinations were conducted without
any time in between. One source of bias is the fact that findings from the first examination
are still clearly present. Assuming that the virtual clinic would be inferior to in-person
examination, the virtual clinic was therefore conducted first, followed by the in-person
examination. The missing change of examiner between virtual clinic and in-person examination
also represents a source of error in terms of interrater reliability, which was only
mitigated by having 6 different examiners for each clinical entity. According to the
authors, the listed limitations do not significantly reduce the impact of this study,
which was designed
to identify and assess opportunities and problem areas.
In order to compare both types of examination in an orderly manner, a structured protocol
with hand surgery examinations was created. Besides the medical history, these included
in particular the inspection of both hands; measurement of the range of motion of
all joints in the hand; assessment of motor and sensory function; as well as the manual
examination or, in the case of VC, the guided examination by the patients themselves
Transmission of the medical history and description of the complaints was not limited
by any technical factors. It should be noted, however, that only findings of the hand
were documented. Although patient habitus and behaviour are also automatically assessed
on-screen, it cannot be ruled out that significant findings of the other extremities
may be overlooked.
In terms of inspection, on the other hand, the virtual clinic and in-person examinations
differed markedly. Without palpation, adequate assessment of many parameters, such
as skin moisture, turgor and heat distribution, was lacking due to the nature of this
modality. As part of the queried and tangible parameters, too, small skin changes
such as older scars were hard to identify. While the retractions in Dupuytrenʼ contracture
were certainly be seen on-screen, not all older scars and skin efflorescences were
identified.
Higher levels of accuracy can be achieved through active enquiry. It is possible that
higher camera resolution (better than the HD resolution used by us) may improve results.
Good lighting and a neutral background, typically the patientʼs outer clothing, is
also helpful.
For the clinical conditions studied in this trial, the incomplete identification of
the scars did not impact on the diagnosis. Nonetheless, this limitation should not
be underestimated, as existing scars should be taken into account, e.g., when choosing
the site of incision and thus also when obtaining informed consent for surgical measures.
As a result of this trial, in addition to sole observation patients should be asked
for information on scars and skin changes.
Range of motion measurements of the hand and fingers, which are important in hand
surgery, demonstrated a high degree of agreement between virtual and in-person examinations.
While Scott et al. (2019) [10] found valid results for visual range of motion assessment in the hand, which, however,
depended on the examiner and were therefore prone to errors, and Smergliuolo et al.
(2016) [11] also reported difficulties, particularly when measuring pronation and supination
on-screen, Zhao et al. (2019) [7] demonstrated that measuring finger position and mobility with a smartphone can be
on par with goniometer measurements.
Our results revealed a high correlation between the range of motion measurements obtained
in the virtual and in-person examinations. In particular, structured examination of
both hands revealed even small differences between them. When examining pronation
and supination, the results presented here were quite comparable only after standardisation
of the measurement protocol in the preliminary examinations. To measure pronation
and supination, the participants had to stand up, tuck in the elbow and keep it flexed
at 90°. With the camera on top of the monitor screen, the hands are well within the
field of vision and the position of the elbow joint can be checked. When measuring
extension and flexion in the wrist, valid measurement results were obtained only after
having gained experience in the preliminary examination. For the Phalen and reverse
Phalen manoeuvre the examiner demonstrated the correct motions and asked the patient
to keep both forearms in the correct position
without rotation in the visual axis. Here, too, this will reveal even small differences
on contra-lateral comparison. In general it can be said that patients almost “automatically”
adjust their hand position to that demonstrated by the examiner. This allows correct
and rapid examinations in a smooth sequence and problem-free “correction” of the demonstrated
joint positions. In terms of rapid assessment, the option of quickly check-marking
the reference values for the joint in question has proved particularly useful. However,
the considerable time saved could entail the risk of misinterpretation through implied
reference values. There was no apparent error on comparison in this trial; however,
since the examiner used the same form for the in-person examination as well, a “two-sided”
error cannot be ruled out. The high correlation may also be affected by the fact that
in both types of examination joint mobility was usually noted in steps of 5° and 10°.
Documentation of motor skills and strength also revealed good agreement between the
examination forms. Since both examinations were undertaken with the same tools, no
significant difference was expected. However, the result is marred by a lack of Janda
strength classification. While other authors [12] advocate the use of different size household items, such as milk cartons and wine
bottles, our trial did not find these to be practical. However, we do favour the use
of a single PET bottle, as this tends to provide a general assessment of the presence
of vigorous motions, restrictions and pareses. But the case number in the examination
selected here was too small for adequate statistical analysis.
The virtual examination also revealed good feasibility when conducting the function
and provocation tests. The findings characteristic of the respective clinical disorders
presented here were thus identified regularly and analogously in both examination
modalities. The fact that contra-lateral comparison with the unaffected hand was available
proved beneficial. Here too, however, further research is needed to provide statistical
confirmation of the results.
Conclusions
In the virtual examination as part of telemedicine, medical history, location of the
complaints, and range of motion of all joints of the hand can be documented with high
correlation with the traditional examination in person. To date, small skin lesions
and scars in particular cannot be adequately diagnosed by inspection. Strength testing
with simple tools, as well as function and provocation tests, can be reproduced well
as an initial examination. Virtual hand surgery clinics require further study. They
will not replace the traditional in-person expert clinic, but offer the potential
for future applications beyond the era of COVID-19.
Remark
K. Welle and S. Täger contributed equally to this project and should be considered
as co-lead authors. T. Jansen and K. Kabir contributed equally to this project and
should be considered as co-last authors. K. Welle and K. Kabir are the corresponding
authors.
