• ABSTRACT
    • Variable outcomes in the prosthetic reconstruction of 4-part humerus fractures often can be attributed to inconsistent and nonanatomic tuberosity placement. To compare the effects of anatomic (anterior fin) versus nonanatomic (lateral fin) tuberosity placement, we developed a dynamic cadaver model for shoulder motion. With the use of a robotically driven, computer-controlled articulator, we tested external rotation torque in 5 fresh human shoulders. After evaluation of the intact shoulders, we experimentally induced 4-part humerus fractures in the specimens. These were then repaired by hemiarthroplasty, with the use of standard techniques to secure the greater and lesser tuberosities in either anatomic or nonanatomic positions; order was randomized. Nonanatomic tuberosity reconstruction led to significant impairment in external rotation kinematics and an 8-fold increase in torque requirements (P =.001). In contrast, anatomic reconstruction produced results indistinguishable from normal shoulder controls. This study underscores the importance of rotational alignment of tuberosities during reconstruction. Failure to properly position tuberosity fragments in the horizontal plane may result in insurmountable postoperative motion restriction.