#156 Visual Estimation of Cervical Spine Range of Motion Can Be Improved with Training
Cervical Therapies and Outcomes
Poster Presented by: M. Webb
M.L. Webb (1)
J.A. Gruskay (1)
B.P. Hirsch (1)
K.A. Jegede (1)
P.G. Whang (1)
J.N. Grauer (1)
(1) Yale Department of Orthopaedics and Rehabilitation, New Haven, CT, USA
Introduction: Cervical range of motion (ROM) is a common measure of cervical conditions, surgical outcomes, and functional impairment. In fact, there is growing emphasis being placed on such functional measures before and after treatments. However, even though ROM is routinely assessed and documented with visual estimation in clinical practice, these assessments have been shown to be unreliable and inaccurate. Although more reliable goniometer devices exist for such assessments, logistics and costs generally limit their clinical acceptance.
Methods: To investigate if training might improve visual estimation of cervical ROM, two attending spine surgeons, four orthopedic residents, and two medical students at our institution visually estimated cervical ROM of four healthy subjects fitted with a validated electrogoniometer device while an impartial investigator recorded electrogoniometric measurements. 840 independent visual estimations of 120 distinct movements were recorded.
The eight examiners then received verbal instruction on cervical ROM assessment and observed the investigator, fitted with the electrogoniometer, announcing electrogoniometric measurements while demonstrating partial and full cervical motions. Visual estimation and electrogoniometric measurement of cervical ROM of those same subjects were then repeated immediately following and one month following this training session for a total of 2520 observations of 360 distinct movements.
Results: The interexaminer ICCs for visual assessment of flexion-extension, lateral bending, and axial rotation were 0.82, 0.84, and 0.82, respectively before training and were not significantly different after.
Initial errors in visual estimation of flexion-extension were 23.9° ± 3.3° (mean absolute error ± 95% CI), in lateral bending 15.5° ± 2.1°, and in axial rotation 19.3° ± 3.3°.
Immediately after the training session, the average error of visual estimation decreased in flexion-extension to 12.0°, in lateral bending to 11.7°, and in axial rotation to 16.4°. These improvements were statistically significant in all three planes of cervical motion.
One month following the training session, the errors in visual assessment of flexion-extension, lateral bending, and axial rotation were 14.4°, 13.9°, and 16.2°, respectively. Only flexion-extension improvement from baseline remained statistically significant, a 39.7% decrease in error.
Discussion and Conclusion: Visual estimation of cervical ROM was noted to have very high ICC before and after targeted training. However, the accuracy of visual estimation of ROM (relative to a validated electrogoniometer) was significantly improved with a training session. Clearly, the flexion-extension plane was that most affected by training, as the initial errors were the greatest. In this plane, statistically significant improvement of approximately 10° remained, even a month after training. In a time when we are putting increased emphasis on functional assessments of outcomes, this degree of improvement seems to be of clinical, as well as statistical, significance. More formalized instruction of ROM assessment is supported by this study.