Key words
spine examination - COVID-19 - telemedicine - video consultation - back pain
Background and Objective
The back is the commonest location of musculoskeletal pain [1]. According to the most recent Global Burden of Disease report, diseases of the musculoskeletal system contribute most to global invalidity, while low back pain is the most common cause of incapacity for work [2]. The lifetime prevalence of back pain in the German adult population is 85.5% [3]. Back pain is therefore extremely important medically and economically and often requires outpatient or inpatient treatment.
Because of the COVID-19 pandemic, contact restrictions (“social distancing”) and quarantine measures have become the new normal. Patients increasingly avoid visiting doctors in their practices or hospitals for fear of becoming infected [4]. Telemedicine has the potential to enable specialist medical consultation, at the same time minimising the risk of exposure to SARS-CoV2 [5]. There is therefore an increasing focus on telemedicine instruments such as video consultations. Previous studies have already described the general potential of telemedicine [6]. This certainly applies especially in disaster and infection prevention scenarios [7]. With the widespread use of smartphones, tablets, computers and commercial video consultation providers, the necessary technical equipment is available almost everywhere [8]. Previous studies have shown that
video consultation can be used successfully for postoperative wound review or to discuss radiological results [9].
Nevertheless, to date there are no recommendations or guidelines regarding video-assisted examination of patients with back pain using telemedicine. The aim of this study is to address the feasibility and practicability of clinical examination of the spine in a video consultation during the ongoing COVID-19 pandemic.
Study Design and Investigation Methods
Study Design and Investigation Methods
Patients with spinal complaints were examined once in a video consultation and then during a direct face-to-face medical consultation. For the study, the first video-assisted examination took place in an examination room in a university hospital outpatient clinic without assistance by relatives or medical staff. The examiner was in the next room at the same time.
Directly following the first video-assisted examination, the patient was examined again face-to-face. To limit the study to examining the feasibility and to avoid intra-observer variability, both examinations were performed by the same doctor. Two specialists in orthopaedics and trauma surgery carried out the examinations.
Technical equipment
The examiner used the usual standard setup for a video consultation (HP Elitedesk desktop computer, Logitech C270 webcam, Logitech H390 head-set).
The video consultation provider used in the study was arztkonsultation ak GmbH (Schwerin, Germany). We used the “Angle Meter 360” application (developed by Alexey Kozlov) to measure the active range of motion (AROM) ([Fig. 1]). The subjects used a tablet with integrated camera and microphone (Apple, iPad Air 2) for transmission to the examiner in real-time.
Fig. 1 Measurement of the active range of motion via the “Angle Meter 360” application.
Collected data
A systematic question and answer form was used at all examinations to structure the results ([Fig. 2]). This contains multiple-choice questions on inspection, pain location, pain severity, range of motion, provocation tests and a basic neurophysiological examination. In the video consultation the provocation tests were modified so that they could be performed by the patient on his own. The Lasegue, reverse Lasegue, Lhermitte, Adams and intervertebral cervical spine compression tests were performed.
Fig. 2 Systematic anamnestic questionaire and clinical findings. a Page 1. b Page 2.
Fig. 2 c Page 3. d Page 4.
Grading of the power of the key muscle in the basic neurophysiological examination was in only three levels (full power, against gravity and paresis) instead of the usual Janda five-point scale. For a rough assessment of power in the upper limb, the patient was provided with an ordinary plastic bottle (1.5 l PET) filled with water.
Study population
After giving informed consent, 43 patients were examined in our university hospital outpatient clinic.
The inclusion criteria were:
-
Patient age over 18 years
-
Referred by the specialist in orthopaedics and trauma surgery with spinal complaints
-
Knowledge of German
-
No cognitive deficits
-
Written consent to take part in the study
Statistical analysis
Statistical analysis was performed with “R” version 4.0.0. Frequency distributions and the Cohen kappa test (in the corrected Brennan-Prediger version) to calculate the intra-rater reliability (IRR). The IRR was interpreted as follows, after Altman: 0 to 0.20 poor; 0.20 – 0.40 low; 0.40 – 0.60 moderate; 0.60 – 0.80 good; 0.80 – 1.00 very good [10].
Informed consent and ethics committee approval
Verbal and written consent was obtained from all study participants. The mental and physical integrity of the participants was respected and protected in accordance with the Declaration of Helsinki [11]. The study was examined and approved by the universityʼs ethic committee (ethics application no. 163/20).
Results
43 patients in total (24 women and 19 men) were recruited for the study. The average age was about 60 years and the average body mass index was 28.6 ± 6.2 kg/m2 (18.6 – 38.5).
Pain intensity and location
The average pain intensity measured by numerical rating scale (NRS) [12] was 4.7 (SD ± 2.3) in both examination with high IRR (kappa = 0.974). The majority of the patients suffered from pain of the lumbar spine (76.7%). There was complete agreement regarding pain location between the results obtained by video consultation and those in the face-to-face examination (kappa = 1.00).
Inspection
There was very high agreement in the examination results between the video-based (VB) and direct face-to-face examination (FTF) in gait assessment (kappa = 0.944; CI=(0.866; 1.000); p < 0.0001), wound inspection (kappa = 0.973; CI=(0.919; 1.000); p = 0.000 000 001) and lateral inspection (kappa = 0.814; CI=(0.672; 0.956); p < 0.0001) ([Table 1]). Posterior inspection showed somewhat poorer agreement of the examination results between VB and FTF (kappa = 0.752; CI=(0.592; 0.912);p < 0.0001).
Table 1 Agreement of the test results.
|
Category
|
Examination
|
Kappa
|
CI
|
SE
|
p
|
|
NRS: Numerical Rating Scale; CI: Confidence Interval; SE: Standard Error
|
|
Pain location
|
NRS pain
|
0.974
|
(0.923; 1.000)
|
0.026
|
< 0.0001
|
|
Pain location
|
1.000
|
(1.000; 1.000)
|
0.000
|
< 0.0001
|
|
Inspection
|
Dorsal inspection
|
0.752
|
(0.592; 0.912)
|
0.079
|
< 0.0001
|
|
Gait
|
0.944
|
(0.866; 1.000)
|
0.039
|
< 0.0001
|
|
Signs of infection
|
0.930
|
(0.833; 1.000)
|
0.048
|
< 0.0001
|
|
Lateral inspection
|
0.814
|
(0.672; 0.956)
|
0.070
|
< 0.0001
|
|
Wound inspection
|
0.973
|
(0.919; 1.000)
|
0.027
|
< 0.0001
|
|
Neurophysiological examination
|
C I and C II
|
0.938
|
(0.852; 1.000)
|
0.043
|
< 0.0001
|
|
C I – C IV
|
0.969
|
(0.907; 1.000)
|
0.031
|
< 0.0001
|
|
C III and C IV
|
0.969
|
(0.907; 1.000)
|
0.031
|
< 0.0001
|
|
C V
|
0.876
|
(0.757; 0.995)
|
0.059
|
< 0.0001
|
|
C VI
|
0.907
|
(0.802; 1.000)
|
0.052
|
< 0.0001
|
|
C VII
|
0.907
|
(0.802; 1.000)
|
0.052
|
< 0.0001
|
|
C VIII
|
0.935
|
(0.844; 1.000)
|
0.045
|
< 0.0001
|
|
Th I
|
0.907
|
(0.802; 1.000)
|
0.052
|
< 0.0001
|
|
Hip flexion
|
0.659
|
(0.48; 0.838)
|
0.089
|
< 0.0001
|
|
Knee extension
|
0.721
|
(0.554; 0.888)
|
0.083
|
< 0.0001
|
|
Dorsiflexion
|
0.814
|
(0.672; 0.956)
|
0.070
|
< 0.0001
|
|
Great toe extension
|
0.721
|
(0.554; 0.888)
|
0.083
|
< 0.0001
|
|
Plantarflexion
|
0.721
|
(0.554; 0.888)
|
0.083
|
< 0.0001
|
|
Sensory deficits
|
1.000
|
(1.000; 1.000)
|
0.000
|
< 0.0001
|
|
Urination frequency
|
1.000
|
(1.000; 1.000)
|
0.000
|
< 0.0001
|
|
Defecation quality
|
1.000
|
(1.000; 1.000)
|
0.000
|
< 0.0001
|
|
Provocation tests
|
Adams test
|
0.407
|
(0.181; 0.633)
|
0.112
|
0.001
|
|
Pain on heel strike
|
0.860
|
(0.726; 0.995)
|
0.066
|
< 0,0001
|
|
Cervical spine neuroforaminal compression test
|
0.938
|
(0.852; 1.000)
|
0.043
|
< 0.0001
|
|
Lasègue
|
0.512
|
(0.295; 0.728)
|
0.107
|
< 0.0001
|
|
Reverse Lasègue
|
0.407
|
(0.181; 0.633)
|
0.112
|
0.001
|
|
Lhermitte
|
0.686
|
(0.498; 0.874)
|
0.093
|
< 0.0001
|
Range of motion
In the measurement of the AROM differences of ± 5° were interpreted as method-related inaccuracies and accepted as equivalent [13]. There was good correlation between the two examinations.
Basic neurophysiological examination
When the results of the basic neurophysiological examination are considered, a distinction should be made between the upper and lower extremity. There was very good agreement overall between VB and FTF in the examination of the upper limb (kappa between 0.876 and 0.969). Examination of the lower limb yielded good agreement levels (kappa between 0.659 and 0.814).
Provocation tests
The provocation tests showed very varied but overall poorer agreement between VB and FTF. On the one hand, agreement was only moderate in the Lasègue test, reverse Lasègue test and Adams forward bend test. On the other hand, there was very good agreement for the cervical spine neuroforaminal compression test.
General trends
Agreement between the different dimensions of physical examination diminished in the following order: pain location, inspection, neurophysiological examination and provocation tests ([Fig. 3]). Moreover, an age-dependent decrease in agreement was measured across all dimensions ([Fig. 4]).
Fig. 3 Match scores between different physical examination.
Fig. 4 Age-dependent cross-dimensional agreement values.
Discussion
The rapid introduction and integration of telemedicine into orthopaedic and trauma surgery care is possible today due to the rapid advances in communication technology [14]. The technical requirements for setting up a video-based consultation are easy to meet and present almost everywhere in medical practices and hospitals [15]. Patients are open to telemedicine solutions [16] and are often just as satisfied with a video consultation as with a conventional outpatient treatment [17], [18].
Factors that can make physical examination via video consultation more difficult are low internet bandwidth [19], low camera resolution on the patientʼs side, poor lighting, excessive complexity of the tests performed [20] and poor videoconferencing etiquette [21]. Since palpation of the patient is not possible, a body diagram can be sent before the video examination in order to narrow down the pain [19]. Physical examination with palpation of, for example, muscle tension, tender points, instability tests or manual therapy tests such as the standing flexion test is likewise not possible.
The patient population with spinal complaints is very varied as regards age, socioeconomic status [22] and technical equipment. Some patients are therefore very suitable for video-based consultation, are comfortable with the technology and value the time efficiency and lack of travel time associated with video consultations [23], [24]. Other patients have great difficulty in following and implementing the doctorʼs instructions by video. In our study, correct camera positioning by the patient in particular was a critical point in assessing the overall course of the examination. It is probably often difficult to evaluate gait disorders or clinical signs of myelopathy in a video consultation. With the majority of patients, however, it is possible to perform an inspection and examine AROM and power during a video consultation [25]. Examination of muscle power had to
be adapted to make it practically possible in a video consultation. Manual testing of muscle function according to Janda [26] with five levels is not possible in a video consultation without an examiner. We therefore simplified the measurement of power into three levels (“full power”, “against gravity” and “paresis”). In addition, examination of the key muscles of the lower limb by video was often more difficult to assess than that of the upper limb. The provocation tests were difficult to impossible for a few patients. Elderly patients in particular had problems in our study in carrying out specific tests in the cameraʼs field of view. The mortality rate due to COVID-19 is markedly increased precisely in the group of elderly and multimorbid patients and special infection prevention would be particularly important for this group of patients. Possible examination by video consultation would therefore be a useful addition to the conventional
medical consultation especially for elderly patients. Assistance in carrying out a video consultation with elderly patients by relatives, friends or home carers could solve this dilemma but was not investigated in our study.
Limitations
This study has a range of limitations. The spinal examinations by video consultation were performed in a simulated setting in a hospital outpatient clinic. The technical equipment was provided for the patient on site. The intra-rater reliability was measured in two successive examinations. Because of this method, the examiner can still recall the previous examination during the second examination, so examiner bias is possible. The question of whether an examiner who is not familiar with carrying out a video consultation obtains the same results was not addressed. Furthermore, the study was conducted in a relatively small group of patients.
Conclusion
Our study shows the feasibility and limits of video-based spinal examination. Video consultations are a form of technology accepted by patients and readily usable for diagnostic investigation of back pain. Examination with direct face-to-face doctor-patient contact is and remains the gold standard. In the current COVID-19 pandemic, specialist consultation and spinal examination are possible by this means, without the risk of possible virus exposure. However, the quality and safety of using telemedicine for patients with back pain should be examined in further larger studies.
Remarks
The authors T. R. Jansen and M. Gathen contributed equally to this project and should be regarded as joint first authors. The authors K. Welle and K. Kabir contributed equally to this project and should be regarded as joint last authors. T. R. Jansen and K. Kabir are corresponding authors.
