Objectives:Recurrent shoulder instability can significantly increase in the presence of bony Bankart and Hill-Sachs lesions. Therefore, it is important to understand the changes in shoulder biomechanics due to bony defects. Limitations of using cadaveric model to investigate the effects of combined bony defects on shoulder instability is inability to test all combination in a single specimen. Utilizing the flexibility of computational methodology like finite element (FE) model provides the advantage of testing all combinations at multiple arm positions. The aim of this study was to develop a simple FE model of combined bony lesions and its effect on anterior shoulder instability. In addition, we wanted to determine the need for patient (specimen) specific modeling. We hypothesized that the shoulder instability would be similar for all three models (population-based model, specimen-specific model, and cadaveric model).Methods:Three specimens were randomly selected from specimens tested in our previous study and Computed Tomography (CT) arthrogram images were taken before and after experimentation to develop FE models. We also developed a simple population-based model representing a spherical humeral head, which was developed using the radii values for cartilage and bone from literature. The sizes of humeral head lesions chosen were: 6%, 19%, 31%, and 44% of humeral head diameter and glenoid defect sizes were 10%, 20% and 30% of the glenoid width. All simulations were performed at glenohumeral abduction angles (ABD) of 20°, 40°, and 60° and external rotation of 0°, 40°, and 80°. Each simulation comprised of translating the humeral head leading to an anterior dislocation under a constant 50 N medial load. This compressive load simulated the static load of soft tissue. The percent intact translation (%IT) was computed by normalizing the distance to dislocation value for each defect condition w.r.t intact condition of each specimen. Stability Ratio (SR) was computed as a ratio of horizontal reaction force to the compressive load.Results:The individual specimen-specific model results comparison to the experimental data for %IT had a good agreement as the values were similar for defect created. However, results for SR were over predicted by the FE model, but they had similar linear decreasing trends for both specimen-specific and cadaveric model. In addition, the humeral head defect size of 44% reduced the %IT from 100% to nearly 0% for all three models. The results for the comparison of all three models with increasing size of humeral defect with a 20% glenoid defect are shown in Figure 1 at three arm position.Conclusion:This study proposed a simple population-based model that can be used to estimate the loss in stability due to combined defects to determine a threshold for defect augmentation in clinical practice. It was demonstrated that a smaller glenoid defect size of 10% combined with a 19% humeral head defect can cause significant instability. Similar to past studies, it was also shown that a glenoid defect would lead to loss of translation and a humeral head defect would lead to instability at a functional arm position of increased abduction and external rotation [5-6]. All three models predicted similar results during validation, which shows that the population based model can be utilized to estimate the stability, instead of needing patient-specific FE models. The limitation of the study is the absence of soft tissue restraints.