Interobserver agreement on MRI evaluation of patients with cervical radiculopathy.

Aim To evaluate the interobserver agreement on magnetic resonance imaging (MRI) evaluation of herniated discs, spondylotic neuroforaminal stenosis, and root compression in patients with recent onset cervical radiculopathy and in addition, to assess the added value of disclosure of clinical information to interobserver agreement. Materials and methods The MRI images of 82 patients with less than 1 month of symptoms and signs of cervical radiculopathy were evaluated independently by two neuroradiologists who were unaware of clinical findings. MRI analysis was repeated after disclosure of clinical information. Interobserver agreement was calculated using kappa statistics. Results The kappa score for evaluation of herniated discs and of spondylotic foramen stenosis was 0.59 and 0.63, respectively. A kappa score of 0.67 was found for the presence of root compression. After disclosure of clinical information kappa scores increased slightly: from 0.59 to 0.62 for the detection of herniated discs, from 0.63 to 0.66 for spondylotic foramen stenosis, and from 0.67 to 0.76 for root compression. Conclusion Interobserver reliability of MRI evaluation in patients with cervical radiculopathy was substantial for root compression, with or without clinical information. Agreement on the cause of the compression, i.e., herniated disc or spondylotic foraminal stenosis, was lower.