Key words
telemedicine - video consultation - hip - pelvis - COVID-19
Introduction
In December of 2019, a previously unknown pathogen was identified in the city of Wuhan,
Hubei Province, China, that causes a severe viral pneumonia. This novel disease pathogen,
designated Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2), has since
spread globally at a rate never before seen [1], [2].
On 11 March 2020, this disease was declared a pandemic by the World Health Organization
(WHO) [3], [4]. To keep the national healthcare system from being overwhelmed as in Italy and Spain,
the German government enacted multiple infection prevention measures in March. These
measures included both postponement of elective surgeries and a drastic reduction
in numbers of face-to-face patient consultations [5].
A rapid response is now necessary to ensure further comprehensive care of orthopaedic
and trauma surgery patients. Telemedical consultations, also known as video appointments
or video consultations, are one way of maintaining direct patient contact. Telemedical
approaches have since seen greatly increased use in both Germany and the U. S. [6].
Use of telemedical consultations in orthopaedic and trauma surgery cases has been
fairly limited in Germany to date due to the inherent restrictions of palpation and
dynamic testing. As a result, experiential data and literature-based validation are
limited. To our knowledge, no detailed investigations have been published to date
with sufficient assessments of the feasibility of clinical hip joint and pelvic examination
using this approach. Since urgent action is now required of orthopaedic and trauma
surgery clinicians, validation and structuring of clinical telemedical examinations
have become an inescapable necessity.
The COVID-19 pandemic is not the first “natural catastrophe” that has led to increased
use of telemedical methods. Increased use of these methods was also observed in the
aftermath of Hurricane Maria in Puerto Rico in 2017 [7].
Until the COVID-19 pandemic struck, the development of telemedical methods focused
mainly on rural areas with limited medical resources [8]. Another area that saw a significant expansion of use of telemedicine was medical
care of military troops in hard-to-reach zones and at the front lines of conflicts
[9].
The metaphoric front line is now closer to home due to the organizational restrictions
presented by the COVID-19 pandemic. Even areas with ubiquitous medical care, such
as large German cities, are now in a situation characterized by limits to patient
care. Creativity in implementing and further expansion of the knowledge already gained
in this field is now needed if we are to ensure patient care beyond an emergency framework.
Relevant experience has been gained in other medical fields. Maia et al. published
an investigation of the paediatric telecardiological program, which has been used
to examine a total of 32,685 patients over the past 20 years. They reported improved
healthcare in Portugal and Africa based on a large volume of experiential data [10]. In a large-scale meta-analysis of telemedical diabetes mellitus therapy, the telemedical
follow-up regimen even proved superior to the conventional approach [11].
In Norway, initial investigations in the field of orthopaedics showed that telemedical
consultation within a defined patient cohort is safe and reliable and can operate
cost-effectively in both social services and healthcare [8], [12].
If we are to be prepared to meet the challenges facing us, we must now investigate
the use of telemedical methods in orthopaedics and trauma surgery.
This paper aims to investigate the feasibility of a clinical orthopaedic examination
of the hip and pelvis using telemedical methods in video consultations. To this end,
the reliability of the (virtual) clinical examination methods are assessed in comparison
with a conventional physical presence examination by a physician. Further, the impact
of individual patient factors are assessed for agreement between the examination methods.
Material and Methods
Patient collective and data collection
For this assessment, hospitalized patients in a German university clinic without prior
known hip joint pathologies were recruited and, after informed consent was obtained,
examined in a video consultation.
Exclusion criteria included significant cognitive impairments (e.g. dementia or mental
retardation), recent hip joint and/or pelvic surgeries, prescribed partial loading/bedrest,
refusal of consent, patient age under 18 years, language barriers and administration/intake
of analgesics > WHO II. 29 patients were included in the study following the selection
process. Age, sex, height, weight, BMI and ASA classification were documented in addition
to the examination results.
A total of 29 test subjects were evaluated for the present study, 14 female and 15
male. Average age was 61.9 ± 17.1 years. Average height was 1.72 ± 0.10 m, average
body weight was 79.55 ± 18.36 kg. The resulting average BMI was 26.9 ± 6.6. The average
ASA score was 2 ± 0.7.
The Ethics Committee of our institution assessed and approved the study prior to the
start of data collection (Ethics Board Registration No. 163/2020).
Organization of the examinations
The test subject examinations were done in two segments, separated as to time and
place. In the first step, a complete telemedical examination was performed by a single
examiner. This procedure was based on the technical and organizational structure of
the telemedical consultation as practised in our own clinic. The examiner used a standard
clinical desktop computer (model Elitedesk, Hewlett-Packard, Wilmington, Delaware,
USA) equipped with a webcam (model C270, Logitech, Newark, California, USA) and a
headset (model H390, Logitech, Newark, California, USA). To simulate a mobile terminal,
the patient was provided with a tablet computer (model iPad Pro, Apple, Cupertino,
California, USA) with an integrated camera and microphone. Both the test subject and
the examiner were able to follow the audio and video transmission, and to communicate,
in real time on screen. The examiner could also pause the video transmission to draw
in lines for angular and length measurements
using a drawing tool. Audio transmission was not interrupted during this image
evaluation procedure.
Following completion of the video examination, a conventional physical presence clinical
examination by a physician was performed to validate the telemedical findings. This
presence examination was also performed by the initial examiner to avoid any potential
examiner-related interfering factors.
Clinical examination
A standardized examination form (see Annex) was used to structure the examination
procedure. The examination form was developed within the framework of this study and
has not yet been validated. It was used in both examination procedures to ensure uniformity
the examination procedure.
The examination form intentionally omits patient health history since the aim of this
study was to evaluate the feasibility of a clinical examination using telemedical
methods.
The examination was structured based on 5 main categories (Inspection, Palpation,
Function, Range of Motion and Provocation). The individual examination techniques
involved were explained in detail to the patient before each procedure, in some cases
with demonstration by the examiner. The patient was not supported by third persons
(e.g. family members or medical staff) for the telemedical examination. To ensure
the efficiency of the telemedical examination, if the explanation of a specific examination
procedure was not understood the explanation was interrupted after 3 minutes if it
proved non-feasible. The corresponding part of the examination was subsequently evaluated
as non-assessable.
The examination form in the Annex lists the clinical examinations performed. The individual
examination techniques were performed analogously to everyday clinical practice as
far as possible. However, further adaptations were necessary for the following examination
methods: For determination of the range of motion, active range of motion was measured
for both examination procedures as per feasibility in the telemedical examination.
The Drehmann sign procedure was performed actively by the patient, in a deviation
from standard practice, in both examination procedures. Muscle strength evaluation
was restricted to the categories: full range, against gravity, paralysis and non-assessable.
Axial compression pain was demonstrated by having the patient exert compression on
the affected leg while standing. The Thomas test was also actively performed by the
patient, who was instructed to maintain contact between lower back and treatment table
surface.
Statistical analysis
Statistical analysis and graphics were done using SPSS Statistics Version 23 (IBM,
Armonk, New York, USA). Significance was defined at p < 0.05. Checking of the collected
data for normal distribution was done using the Kolmogorov-Smirnov test. The Pearson
coefficient of correlation was then calculated to check for a potential correlation
of normally distributed metric data. The Spearman coefficient of correlation was used
as a non-parametric correlation test. Nominally-scaled data were evaluated using the
chi-squared test. In a cohort n < 30, a bootstrap sample was also derived to ensure
correct determination of significance. To assess agreement of the two examination
procedures beyond chance correlation, Cohenʼs kappa was determined. The Landis and
Koch scale was used to evaluate agreement [13]. Therefore, kappa values above 0.80 were designated as excellent agreement, 0.61 – 0.80
as good agreement, 0.41 – 0.60 as moderate agreement,
0.21 – 0.40 as adequate agreement and below 0.20 as poor agreement.
Results
The individual examinations were separately evaluated, then combined to obtain an
average for the respective superordinate category (Inspection, Palpation, Function,
Range of Motion and Provocation). The inspections agreed closely with a mean Cohenʼs
kappa of 0.76 ± 0.37 (swelling κ = 0.818; redness κ = 1.0; atrophy κ = 0.220; scar/wound
anomalies κ = 1.0). Palpation showed adequate agreement with a mean Cohenʼs kappa
of 0.38 ± 0.19 (symphysis κ = 0.588; trochanter major κ = 0.223; groin κ = 0.209;
gluteal κ = 0.482).
Function showed good agreement with a mean Cohenʼs kappa of 0.61 ± 0.26 and thus good
agreement of the examinations (muscle strength hip flexion κ = 0.473; muscle strength
hip extension κ = 1.0; muscle strength hip abduction κ = 0.482; muscle strength hip
adduction κ = 0.482; gait κ = 1.0; feasibility of one-legged stand κ = 1.0; feasibility
of knee bend κ = 1.0). Evaluation of the clinical examination of range of motion showed
adequate agreement with a mean Cohenʼs kappa of 0.36 ± 0.19 (extension/flexion κ = 0.380;
outer rotation/inner rotation κ = 0.486; abduction/adduction κ = 0.265). Analysis
of the various provocation tests also revealed merely adequate agreement of the different
examination procedures with a Cohenʼs kappa of 0.33 ± 0.13 (Apley test κ = 0.181;
Drehmann sign κ = 0.386; Patrickʼs test κ = 0.291; Trendelenburg sign κ = 0.475; axial
compression pain κ = 0.1; Thomas test κ = 0.370; posterior impingement test κ = 0.146;
ventral impingement test κ = 0.147;
foveal impingement test κ = 0.045). [Fig. 1] shows an overview of the Cohenʼs kappa values grouped according to the examination
categories. The individual Cohenʼs kappa values are also listed in [Table 1].
Fig. 1 Agreement between modalities grouped according to examination categories.
Table 1 Cohenʼs kappa values grouped according to the different examination methods.
|
Examination
|
Cohenʼs κ
|
|
Inspection
|
0.76 ± 0.37
|
|
|
0.818
|
|
|
1.0
|
|
|
0.220
|
|
|
1.0
|
|
Palpation
|
0.38 ± 0.19
|
|
|
0.588
|
|
|
0.223
|
|
|
0.209
|
|
|
0.482
|
|
Function
|
0.61 ± 0.26
|
|
|
0.473
|
|
|
1.0
|
|
|
0.482
|
|
|
0.482
|
|
|
1.0
|
|
|
1.0
|
|
|
1.0
|
|
Range of motion
|
0.36 ± 0.19
|
|
|
0.380
|
|
|
0.486
|
|
|
0.265
|
|
Provocation tests
|
0.33 ± 0.13
|
|
|
0.181
|
|
|
0.386
|
|
|
0.281
|
|
|
0.475
|
|
|
0.370
|
|
|
0.146
|
|
|
0.147
|
|
|
0.045
|
The correlation analysis revealed a significant positive correlation between age and
number of deviations among the different examinations (r = 0.588 p < 0.01). Test subject
sex showed neither a significant effect on the number of deviations among the different
examinations (p = 0.55) nor on the number of examinations that proved telemedically
non-assessable (p = 0.52). Both BMI (r = 0.389 p < 0.05) and ASA (r = 0.396 p < 0.05)
correlate significantly with an increasing number of deviations between telemedical
and conventional examinations. A significant positive correlation of telemedically
non-assessable examination results was also observed with the ASA classification score
(r = 0.509 p < 0.01), BMI (r = 0.485 p < 0.01) and age (r = 0.579 p < 0.01). The correlations
of BMI and age with the rate of non-assessable examination methods are illustrated
by way of example in [Fig. 2].
Fig. 2 Scatter diagram of correlation between age in years and BMI with the number of non-feasible
examinations.
Discussion
This study evaluated the feasibility of telemedical examination of the hip joint and
pelvis. Good agreement of results with a conventional clinical examination was observed
for the categories Inspection and Function. The evaluation of Range of Motion revealed
only moderate agreement. Palpation and Provocation showed only adequate correlations
between the examination results. The correlation analysis with patient-specific factors
showed a significant positive correlation between age, BMI and ASA classification
score and the rate of deviations between the two examinations as well as the number
of non-assessable tests.
Inspection was shown in this study to be readily feasible in a telemedical consultation.
Good agreement was seen between telemedical and conventional findings. This finding
opens up possibilities for a number of diagnoses and follow-up examinations, e.g.
wound monitoring. Caution is advised in evaluation of atrophies and in highly adipose
patients. Evaluation of coarse pelvic and hip function based on telemedical examination
is valid.
However, specification of pelvic and hip joint function, for example by exact determination
of ranges of motion, showed only moderate agreement. This may have to do with the
active character of the motions. Patients experiencing pain tend to avoid positions
that elicit further pain. Further, even assuming proper execution by the patients,
there are many influencing factors that can reduce measurement accuracy. Investigations
already published in the literature differ, for instance claiming only moderate results
for the wrist and good agreement for elbow evaluations [14], [15].
Provocation testing was determined to be not reliably assessable in our study design.
Agreement with a conventional clinical examination were minimal at best. These results
are supported by a physiotherapeutic meta-analysis by Mani et al. [16]. Here as well, it may be that the underlying mechanism is conscious or unconscious
pain avoidance by the test subjects. To avoid influencing the examiner, this assessment
intentionally omitted structured health history reports, whereas recording of a thorough
health history is an absolute necessity in conventional clinical examination practice.
A structured health history assumes a special significance in telemedical evaluation
of the hip joint and pelvis. Particularly in view of the limited information value
of specific function and provocation tests, detailed information provided by the patient
can provide essential diagnostic information. It could be concluded that the limited
information value of
telemedical provocation test assessment could be improved by a specific health
history. However, since the present study did not evaluate this connection, further
studies would be required to reach a conclusive assessment.
Patient selection represents an essential aspect of setup and optimization of a telemedical
consultation. As far as the authors are aware, no current studies have been published
that cover this aspect adequately. Various recommendations have been published, but
validation has not been established [17], [18], [19].
This assessment revealed a statistically relevant connection between patient-specific
factors such as age, BMI and ASA classification score and the valid and successful
performance of a telemedical examination. However, since the number of test subjects
is small, a differentiated view of the value of the statistical evaluation and its
interpretation is necessary. Even though the limited collective in this study cannot
provide definitive conclusions regarding significance and causality, the results do
highlight the importance of adequate patient selection. This study cannot provide
definitive conclusions regarding inclusion and exclusion criteria, but results for
patients with the following characteristics must be interpreted cautiously: The present
assessment, supported by the subjective perception of the examiner, tends to the view
that a telemedical examination has limited validity for older, adipose and multimorbid
patients. In such cases, the examiner must be aware that
the significance of the results of a telemedical examination may be limited and
that an additional conventional face-to-face consultation should be arranged if the
doubts appear well-founded.
In the end, patient and physician acceptance of a telemedical examination must be
the goal. Here as well, pitfalls abound. In the literature, it appears that the patients
who willingly undergo a telemedical consultation are those without major health problems,
whereas patients with relevant symptoms tend to be sceptical of the new methods [20]. Generally speaking, patients appear to view the notion of telemedical consultation
critically [21].
Telemedicine, not a novel invention, is increasingly important in the COVID-19 crisis.
The suitability of telemedical consultations in specific fields of application has
already been shown. In telemedical consultations in Denmark, improved fast-track capacities
were demonstrated, resulting in earlier patient discharges without any relevant risk
to patient satisfaction [22]. Good results were also achieved in a telemedical rehabilitation programme with
patients with lower extremity injuries [23]. Time will tell whether these results can be extrapolated to larger patient collectives
and broader contexts. One possible task of a telemedical consultation service could
be targeted triage with the aim of determining whether a conventional consultation
is necessary and how urgent such an appointment is. Telemedical preselection and assignment
to conventional consultation appointments is also conceivable. These applications
would open a channel of communication that would demand relatively little patient
effort.
The relatively small test subject collective is one of the main limitations of this
study. In future, additional large-scale prospective studies will be required to achieve
adequate validation of telemedical examinations in orthopaedics and trauma surgery.
Given the many different hip pathologies one sees in a conventional consultation practice,
large case numbers with sufficient test subjects for each clinical picture will be
necessary if the assessment of diagnostic sensitivity and specificity for defined
pathologies is to be of sufficient quality. The present study cannot draw conclusions
of this scope due to the limited number of test subjects. It can therefore serve as
a pilot study considering the possibility, in principle, of examining the hip joint
telemedically. We therefore decided to limit the recruited patient collective to patients
with no hip joint pathologies. More extensive follow-up studies can now be designed
on this basis. The current state of scientific
knowledge is still insufficient to support a definitive assessment of the sufficiency
of telemedical examination. Statistical evaluations determine only correlations that
do not necessarily reflect causality. Here as well, future studies with much larger
collectives will provide important and interesting results. Further, the clinical
examinations in this study were performed without prior training of the test subjects.
Additional studies must determine whether a structured patient information sheet would
improve agreement. It must also be noted that little experiential data has been collected
on clinical telemedical examination. In time and with increasing establishment of
the method, data from experience, as well as increased structuring and optimization
of the examination procedure, could also impact future results. This study also did
not collect information on previous patient experience with clinical hip joint and
pelvis examinations. The patients in this case were
hospitalized orthopaedic patients classified as having healthy hip joints when
the data were collected. Nonetheless, the possibility of previous experience and clinical
examination of the hip joint and pelvis in an inpatient context cannot be excluded
and could result in bias accordingly.
The patients in our study were provided with a high-quality terminal device to prevent
any impact of technical equipment quality. Everyday clinical practice will however
of course involve contacts with patients whose equipment is older or whose internet
connection is poor. As far as the authors are aware, no definitions have been arrived
at regarding minimum quality standards in this respect, although this could potentially
have a considerable negative impact on an examination.
Conclusion
It was shown in this study that a telemedical examination is possible within limits.
Patient-specific factors such as age, BMI and levels of prior morbidities would appear
to have a relevant impact on validity and execution.
As the matter currently stands, targeted patient selection is a must, whereby patients
should meet the technical, physical and cognitive preconditions required to ensure
a reliable examination. Patients with severe prior morbidities, advanced age and/or
adiposity may predispose the telemedical examination to limited significance of the
results obtained. We recommend considering the findings in this cohort with caution,
with arrangement of an additional conventional examination appointment as indicated.
It is the assumption of the authors that the near future will see the existing results
significantly augmented by experience, telemedically adapted examination techniques
and technical innovations such as augmented reality and artificial intelligence. Whether
telemedical assessment of patients currently considered poorly suited for this approach
will then be acceptably valid remains to be seen. Possible applications of telemedicine
within the framework of an orthopaedic and trauma surgery hip and pelvic consultation
include triaging and assignment of patients to conventional appointments and performance
of clinical follow-up monitoring, e.g. of wounds and swellings. This study establishes
a starting point for scientific work on the theme of telemedical hip and pelvis examinations.
It can serve as a reference for future work and will hopefully provide motivation
for further investigations of this interesting and relevant thematic complex. Telemedical
applications are being
developed apace, with the current COVID-19 pandemic providing the impetus to anchor
them in standard clinical practice.
