Interobserver and Intraobserver Variability of Four Suprascapular Notch Classification Systems.

Abstract

Knowledge of the morphology of the suprascapular notch is clinically beneficial in patients withsuspected suprascapular nerve compression or palsy. Several classification systems have been proposed for the morphological classification of the suprascapular notch and its severalanatomical variations. The purpose of this study was to evaluate the inter- and intraobserver reliability of four different classification systems for suprascapular notch typing analysing shoulder computed tomography (CT) scans. Shoulder CT scans from 109 subjects (71.5% males) were examined by three raters of various experience levels, one senior, one experienced, and one junior orthopaedic surgeon. The CT scans were evaluated quantitatively and qualitatively and the suprascapular notch was classified according to four classification systems at two separate timepoints, four weeks apart. To determine consistency among the same or different raters, the Kappa statistic was performed and intrarater reliability for each rater between the first and the second evaluation was assessed using Cohen's kappa. Reliability across all raters at each timepoint was assessed using the Fleiss kappa. Agreement was almost perfect for all the classification systems and amongst all raters, regardless of their experience level. There were no significant differences between the raters on any of the evaluations. The overall interobserver agreement for all classifications was almost perfect. The four suprascapular notch classification systems are reliable, and the rater's experience level has no impact on the evaluation.