Purpose: Facet joint arthrosis is an entity that can have a key role in the etiology of low back pain, especially with hyperextension, and is a key component of surgical planning, especially when considering disc arthroplasty. Plain films and MRI are most commonly utilized as the initial imaging of choice for low back pain, but these methods may not truly allow an accurate assessment of facet arthosis. Our purpose was to observe the inter- and intraobserver reliability of utilizing CT and MRI to evaluate facet arthrosis, the inter- and intraobserver reliability of the facet grading system, and the agreement of surgeons as to when to perform disc arthroplasty after the lumbar facets are evaluated.

Method: A power analysis was performed which showed we would need 6 reviewers and 43 images to have 80% power to show excellent reliability. 102 CT and the corresponding MRI images of lumbar facets were obtained from patients who were to undergo lumbar spine surgery of any type. 10 spine surgeons and 3 spine fellows reviewed the randomized images at 2 time points, 3 months apart, graded the facet arthosis as well as indicated whether they would chose to perform a disc arthroplasty based on the amount of facet arthrosis. Both interobserver and intraobserver kappa values were calculated by result comparison between observers at the two time points and between CT and MRI images from the same patient.

Results: interobserver reliability for MRI was 0.21 and 0.07(fair to slight agreement), and for CT was 0.33 and 0.27(fair agreement), for the spine surgeons and spine fellows respectively. The mean intraobserver reliability for MRI was 0.36 and 0.26 (fair agreement) and for CT was 0.52 and 0.51 (moderate agreement). The kappa value for agreement of whether to perform a disc arthroplasty after grading the facet arthrosis utilizing MRI was 0.22 (fair agreement) and utilizing CT was 0.33 (fair agreement) among the senior spine surgeons.

Conclusion: The existing grading system for facet arthrosis and of whether to perform a disc arthroplasty utilizing the grading system has at best only fair agreement. CT is more reliable for grading facet arthrosis.