The value of lateral glenohumeral offset in predicting construct failure in proximal humerus fractures following internal fixation.

BACKGROUND

Proximal humerus fractures are the third most common osteoporosis defining injury in the United States, yet operative fixation of these injuries remains technically challenging. Although several modifiable and nonmodifiable risk factors are correlated with failure of proximal humerus fixation, no study has investigated whether failure to restore glenohumeral offset plays a part in fixation failure. The goals of this study are: (1) to determine if lateral glenohumeral offset (LGHO) and humeral head diameter (HHD) can be measured radiographically with accuracy between observers, (2) to observe whether there is a correlation between failure to operatively restore an anatomic LGHO:HHD ratio and failure of fixation, and (3) if there is a correlation, can any recommendations be made in regard to the ideal LGHO:HHD ratio.

METHODS

Retrospective review found 183 patients meeting inclusion criteria who underwent operative fixation for proximal humerus fractures between 2005 and 2018. Patients suffering construct failure requiring reoperation were compared with clinically successful surgeries on the basis of age, sex, fracture morphology, head-shaft angle, smoking history, presence or absence of a calcar screw, and LGHO:HHD ratio. The groups were compared using a combination of Student t-tests, χ, and bivariate and multivariate logistic regression analyses where appropriate. The Student t-test and intraclass correlation coefficient were both used to assess interobserver reliability.

RESULTS

We found that LGHO and HHD can be measured by independent observers accurately (intraclass correlation coefficient = 0.80, 95% confidence interval: 0.65-0.89). Patients suffering implant failure had a significantly lower LGHO:HHD ratios compared with those who did not (0.94 vs. 1.03, P ≤ .001). The LGHO:HHD ratio was an independent predictor of implant failure even after controlling for other potential risk factors. Patients with an LGHO:HHD of 1.0 or above have a <10% chance of failure compared with a 20% risk with a ratio of 0.9 and a 40% risk at 0.8.

CONCLUSION

We found the LGHO:HHD ratio to be an independent predictor for construct failure after plate and screw fixation of proximal humerus fractures. Efforts should be made to restore an anatomic ratio of at least 1.0 to minimize the risk of failure.