Video-oculographic measures of eye movement control in the smooth pursuit neck torsion test can classify idiopathic neck pain patients from healthy individuals: A datamining based diagnostic accuracy study

Background: Idiopathic neck pain patients frequently experience oculomotor disfunctions with deficits in eye movement control between neutral and neck torsion position (SPNT test) being commonly investigated in clinical and research settings. Objectives: The aim of the study was to determine accuracy of SPNT test in classifying idiopathic neck pain patients. Design: a datamining based diagnostic accuracy study. Methods: The study was conducted on a referred sample of 38 chronic neck pain patients from orthopaedic outpatient clinic and 40 healthy controls. Video-oculography was used to study gain and SPNTdiff during SPNT test under three target movement velocities and amplitudes and two different angles of neck torsion. A Naive Bayesian predictive model was used to classify neck pain patients based on gain or SPNTdiff. Results: Gain during two target movement profiles at velocities of 30s(-1) and amplitudes of 30 and 40 under 45 of neck torsion presented with highest area under the curve (0.837), specificity (92%), sensitivity (94%), highest true positive and lowest false negative predicted value. Highest area under the curve (0.760), specificity (50%), sensitivity (71%), highest true positive and lowest false negative values were observed for SPNTdiff at velocities of 30 s- 1 and amplitude of 30 applying 45 of neck torsion. Conclusion: SPNT test provides useful diagnostic tool for classifying neck pain patients when using single or combination of two target movement profiles. Neck torsion of 45 as opposed to 30 should be used during SPNT test when investigating patients with neck pain disorders.