Inter-observer reliability in determining the presence of hazardous disc disruption in flexion distraction injuries of the C-spine- an MRI study

Purpose

To perform an Interrater reliability study determining the agreement between an Orthopaedic team and a Radiology team on the MRI features of prolapsed uncontained cervical discs in Flexion-Distraction injuries of the cervical spine. This leads us to determine how many patients demonstrated evidence of a ‘dangerous’ disc: an uncontained disc herniating posteriorly that may be drawn into the spinal canal during closed reduction.

Methods

One hundred and ten patients who had pre-reduction MRI scans managed during the last 10 years were included. Variables were chosen and defined by the senior Author and explained to both teams prior to reviewing the scans. The review was performed by each team independently and without any access to clinical information. Data collection and interpretation was designed by a statistician to reduce risk of data entry errors. Interrater reliability/agreement was determined using the Cohen Kappa value.

Results

Disc Protrusion across Posterior Body Line, Interrater agreement Cohen-Kappa=0.46 (moderate agreement) Disc protrusion across Inferior Body Line Interrater agreement Cohen-Kappa=0.4 (moderate agreement) Disc protrusion across Corner to Corner Line, Interrater agreement Cohen-Kappa=0.15 (poor agreement) Disc disruption Line, Interrater agreement Cohen-Kappa=0.33 (moderate agreement) Disc Contained, Interrater agreement Cohen-Kappa=0.41 (moderate agreement) PLL, Interrater agreement Cohen Kappa=0.22 (Fair agreement). Prevalence of posterior disc herniation 42% (orthopaedic team), 51%(Radiology team). Prevalence of uncontained discs 70% (orthopaedic team), 76%(Radiology team). The most agreed upon variables were posterior herniation and the impression of whether the disc was contained.

Conclusion

Interrater agreement on the presence of an uncontained herniated disc between Orthopaedic surgeons and Radiologists in this study was never excellent. This leads us to believe it imperative for the surgeon to review MRI's in addition to reading the radiologists report.

NO DISCLOSURES