Abstract
Reverse shoulder arthroplasty (RSA) is used to treat patients with cuff tear arthropathy. Loosening remains to be one of the principal modes of implant failure and the main complication leading to revision. Excess micromotion contributes to glenoid loosening. This study sought to determine the predictive accuracy of an experimental system designed to assess factors contributing to RSA glenoid baseplate micromotion. A half-fractional factorial experiment was designed to assess 4 factors: central element type (screw versus peg), central element length (13.5 versus 23.5 mm), anterior-posterior peripheral screw type (locking versus nonlocking) and cancellous bone density (10 versus 25 pounds per cubic foot (pcf)). Four linear variable differential transducers (LVDTs) recorded micromotion from a stainless-steel disk surrounding a modified glenosphere. The displacements were used to interpolate micromotion at each peripheral screw position. The mean absolute percentage error (MAPE) was used to determine the predictive accuracy and error range of the system. The MAPE for each condition ranged from 6.8% to 12.9% for an overall MAPE of (9.5 ± 0.9)%. The system had an error range of 2.7 μm to 20.1 μm, which was lower than those reported by prior studies using optical systems. One of the eight conditions had micromotion that exceeded 150 μm. These findings support the use of displacement transducers, specifically LVDTs, as an accurate system for determining RSA baseplate micromotion in rigid polyurethane foam bone surrogates.