A reliable method for classifying acromial shape

Abstract

The method commonly in use today to categorize the shape of the acromion is qualitative and not reliable, but often used to make clinical judgments. Other classifications are quantitative but have not been compared with each other. The objective of this study was to compare the intra- and inter-observer reliability of three methods for classifying acromial shape. Three clinicians evaluated standardized lateral-view radiographs of twenty-four human scapulas three times with intervals of at least two weeks for the intra-observer reliability. Thirteen clinicians evaluated 11 of these radiographs once for inter-observer reliability. We evaluated the classification described by Bigliani as well as two other classifications (acromial angle and modified Epstein classifications) that assign acromions to a type according to specific geometric measurements obtained from three anatomic landmarks. Reliability was assessed with Kappa statistics. The three classifications showed moderate to good intra-observer reliability (Kappa values between 0.50 and 0.72) without significant differences (p > 0.05), respectively. The classification described by Bigliani showed slight inter-observer reliability (Kappa value 0.25), the acromial angle classification revealed moderate inter-observer reliability (Kappa value 0.44), and the modified Epstein classification showed good inter-observer reliability with the highest inter-observer Kappa value of 0.62. These differences between the classifications were significant (p < 0.05), respectively. This study compared the intra- and inter-observer reliability of three classifications of acromial shape. The modified Epstein classification was the most reliable and should be used for classifying acromial shape to make clinical judgments and in future research studies.