Reliability of radiological measurements of type 2 odontoid fracture

It is recognized that radiological parameters of type 2 dens fractures, including displacement and angulation, are predictive of treatment outcomes and are used to guide surgical decision-making. The reproducibility of such measurements, therefore, is of critical importance. Past literature has shown poor interobserver reliability for both displacement and angulation measurements of type 2 dens fractures. Since such studies however, various advancements of radiological review systems and measurement tools have evolved to potentially improve such measurements.

To re-examine the inter-rater reliability of measuring displacement and angulation of type 2 dens fractures using modern radiological review systems. Besides quantitative measurements, the reliability of raters in identifying diagnostic classifications based on translational and angulational displacement was also examined.

Radiographic measurement reliability and agreement study.

Thirty-seven patients seen at a single institution between 2002 and 2017 with primary diagnosis of acute type 2 dens fracture with complete computed tomography (CT) imaging.

Radiological measurements included displacement and angulation. Diagnostic classifications based on consensus-based clinical cutoff points were also recorded.

Measurements were performed by five surgeons with varying years of experience in spine surgery using the hospital's electronic medical record radiological measuring tools. The radiological measurements included displacement and angulation. Diagnostic classifications based on consensus-based clinical cutoff points were also recorded. Each rater received a graphic demonstration of the measurement methods, but had the autonomy to select a best cut from the sagittal CT to measure. All raters were blinded to patient information.

Measurements for displacement and angulation among the five raters demonstrated "excellent" reliability. Intra-rater reliability was also "excellent" in measuring displacement and angulation. The reliability of diagnostic classification of displacement (above vs. below 5 mm), was found to be "very good" among the raters. The reliability of diagnostic classification of angulation (above vs. below 11°) demonstrated "good" reliability.

Advancement of radiological review systems, including review tools and embedded image processing software, has facilitated more reliable measurements for type 2 odontoid fractures.