Purpose of the study: The objective was to analyse shoulder motion, particularly abduction and anterior elevation, in patients with an reversed prosthesis. A radiocinematographic study enable an analysis of the movements of the prosthesis it self and movements due to scapulothoracic participation.

Material and methods: This study was based on the analysis of 33 patients with an reversed shoulder prosthesis. A videoscopic recording (25 images per second) of anterior elevation from a workstation used for abduction arteriography was used. The Constant score was noted and a standard x-ray work-up (four views) was obtained for all patients.

Results: The cohort was a homogeneous continuous series of 21 women and 12 men, mean age 72.5 years (range 39–84). Two modes of motion were observed. The first (group 12, n=17 shoulders) was «monoarticular»: shoulder motion was almost exclusively related to movement of the scapulothoracic junction. Abduction did not exceed 90°. The second mode (group 2, n=16 shoulders) was «bi-articular»: joint motion began with the prosthesis (50° on average) followed by scapulothoracic participation (50° on average). The implant then was involved in the final part of the motion (in six of the 16 shoulders in group 2) to complete the range of motion exceeding 120° abduction and anterior elevation.

Discussion: This study confirmed the presence of an initial intrinsic mobility of the prosthesis followed by scapulothoracic participation. It was noted however, that for the majority of cases, the intrinsic mobility of the prosthesis was limited. In all cases, the range of motion recorded by clinical examination was greater than the range of motion measured objectively by radiocinematography.

Discussion: This study demonstrated the in vivo mobility of the reversed prosthesis. However, a range of motion greater than 100° anterior elevation and abduction is exceptional. Clinical findings reflect imperfectly the real mobility of this type of prosthesis.