This cadaveric biomechanical study used 10 fresh-frozen shoulders and a six-degrees-of-freedom testing machine to determine at what glenoid defect size Bankart repair fails. Four progressively larger defects (9%, 21%, 34%, and 46% of glenoid length) were tested across two arm positions, with and without Bankart repair.
For decades, the only guidance on glenoid bone loss was Rowe's vague "one-third rule" — a visual estimate with no biomechanical validation and no standardized measurement method.
This paper gave surgeons the first rigorously measured threshold: 21% of glenoid length (~6.8 mm). When you see this much bone loss on a CT or arthroscopically, Bankart repair alone is biomechanically insufficient. Consider a Latarjet or iliac crest graft.
Measure width on axial CT, not area. A defect that looks like one-third by eye may be only one-fifth by area. Linear measurement (percentage of glenoid length) is more accurate and reproducible than any area estimate.
For throwing athletes, the 25-degree-per-centimeter external rotation penalty matters even when stability thresholds are not crossed. A defect just below 21% can still cost a pitcher meaningful rotation. Bone restoration may be warranted for function even when pure stability would be preserved.
This cadaveric biomechanical study used 10 fresh-frozen shoulders and a six-degrees-of-freedom testing machine to determine at what glenoid defect size Bankart repair fails. Four progressively larger defects (9%, 21%, 34%, and 46% of glenoid length) were tested across two arm positions, with and without Bankart repair.
For decades, the only guidance on glenoid bone loss was Rowe's vague "one-third rule" — a visual estimate with no biomechanical validation and no standardized measurement method.
This paper gave surgeons the first rigorously measured threshold: 21% of glenoid length (~6.8 mm). When you see this much bone loss on a CT or arthroscopically, Bankart repair alone is biomechanically insufficient. Consider a Latarjet or iliac crest graft.
Measure width on axial CT, not area. A defect that looks like one-third by eye may be only one-fifth by area. Linear measurement (percentage of glenoid length) is more accurate and reproducible than any area estimate.
For throwing athletes, the 25-degree-per-centimeter external rotation penalty matters even when stability thresholds are not crossed. A defect just below 21% can still cost a pitcher meaningful rotation. Bone restoration may be warranted for function even when pure stability would be preserved.