Diagnostic validity of space available for the spinal cord at C1 level for cervical myelopathy in patients with rheumatoid arthritis

STUDY DESIGN. Retrospective cohort study. OBJECTIVE. To evaluate diagnostic validity of space available for the spinal cord (SAC) at C1 level for myelopathy in patients with rheumatoid arthritis (RA). SUMMARY OF BACKGROUND DATA. The relationship of SAC at C1 level with myelopathy has been evaluated by relatively small number of the patients, and 2 criteria have been proposed. METHODS. Two cohorts of the patients with RA were established. Group A consisted of 140 patients with myeopathy due to upper cervical involvement selected from the database. Group B consisted of 99 patients with upper cervical subluxation, but not associated with myelopathy selected from the consecutive series of the hospitalized patients. Distributions of SAC at C1 level in both groups were evaluated. Efficacy indexes for screening (sensitivity, specificity, etc.) were calculated for these patients' population by previously demonstrated 2 criteria. In addition, analysis according to receiver operating characteristic (ROC) curve was performed. RESULTS. The average values of SAC were 11.1 mm in Group A and 16.5 mm in Group B. When cut-off point for myelopathy was defined as 13 mm or less, sensitivity and specificity were 82% and 85%, respectively. When it was defined as 14 mm or less, sensitivity increased (88%) while specificity decreased (74%). Accuracies by these 2 criteria were almost the same (83%, 82%). The left upper corner point of ROC curve was located between these 2 cut-off points. CONCLUSION. Distributions of SAC showed that SAC was a reliable parameter for relating myelopathy in patients with upper cervical subluxation in RA. The plots according to ROC curve showed adequacy of previously demonstrated 2 cut-off points. For the purpose to screen the patients with high risk for myelopathy, 14 mm or less was recommended as a cut-off point of SAC.