The Acromion Index Reliability in radiographic and magnetic resonance imaging

The present study evaluates the inter-rater reliability (IRR) of the Acromion Index (AI) in radiographic (x-ray) and magnetic resonance imaging (MRI), the correlation between both types of imaging and the reliability of the AI as a predictor of rotator cuff tears (RCT). 116 patients who suffered a partial or total rotator cuff and/or long biceps tear, an impingement syndrome and/or calcific tendinitis and who were surgical treated in our clinic were included in this study. We were able to confirm the excellent inter-rater reliability of the AI in x-ray (p = 0.969) and could also show an excellent IRR in MRI (rho = 0.97). However, due to more single extreme AI in MRI, the Pearson's correlation showed only an average correlation between both types of imaging (p = 0.391). No significant correlation (p = 0.35) of the AI and patients without (AI 0.709, SD 0.07) and with (AI 0.695, SD 0.07) arthroscopic evidence of a RCT could be found. However, when grouping the patients according to their AI, we could show that an AI = 0.76 leads to a 76% presence of a RCT and an AI < 0.75 only leads to a 54.9% presence of a RCT in shoulder arthroscopy. The AI has an excellent IRR in x-ray and MRI. However, a clearer definition of the AI calculation in MRI is needed in order to improve the only moderate correlation between both types of imaging. By grouping patients according to their AI we discovered an increase in the probability of RCT with decreasing AI. This could lead to a more parable rather than linear correlation of the AI and RCT showing that there is a possible ideal force vector of the deltoid muscle and a change of the AI in either direction leads to harm of the rotator cuff tendons. Further studies with a greater number of patients are needed to prove this hypothesis.