Abstract
Introduction: (1) Determine whether initial MRI findings correlate with clinical outcome.(2) Study the reproducibility of MRI measurements of large disc prolapses.(3) Estimate the ability to predict CES based on MRI alone.(4) Does CES only occur in degenerate discs?
Method: 31 patients with CES were identified and invited to attend clinic. 19 patients who underwent discectomy were identified. Digital photographs of all 50 MRIs were obtained. Observers: 1 Radiologist, 2 Spinal Surgeons and 1 Trainee did not know the number of patients in each group. Observers estimated the percentage spinal canal compromise on each view (0–100%), indicated whether they thought the scan findings could produce CES and commented on disc degeneration. Measurements were repeated after 2 weeks.
Results: 26 patients attended clinic – mean follow up 51 months (range 25–97). 12 of the 26 patients with CES had, on average, > 75% canal compromise. No significant correlations were found between MRI canal compromise and clinical outcome. Kappa values for intra-observer reproducibility of measurements ranged from 0.4–0.85 and inter-observer 0.63–5. Based on MRI, the correct identification of CES has sensitivity 68%, specificity 78%, positive predictive value 84% and negative predictive value 58%. Over 80% of the CES causing discs were degenerate.
Discussion: Canal compromise does not appear to predict clinical outcome. MRI measurement reproducibility has substantial agreement. CES is a clinical diagnosis supported by an MRI scan. In less clear cases the presence of a large disc on an MRI scan supports a diagnosis of CES (PPV 84%). CES occurs in degenerate discs.
Please send correspondance to BASS at the Royal College of Surgeons, 35 - 43 Lincoln’s Inn Fields, London WC2A 3PE, UK.
