Metaversion can reliably predict humeral head version: a computed tomography-based validation study.

Abstract

Circumstances occur when the determination of anatomic humeral head version is difficult. In the setting of irreparable proximal humeral fracture, there are few reliable intraoperative landmarks to determine anatomic humeral head version. This study tested our hypothesis that the metaphyseal version (metaversion) is a landmark that can assist with correct head version and used computed tomography (CT) to evaluate its reliability as a predictor of anatomic version. CT scans from 50 consecutive patients (20 women, 30 men) were examined using commercial software. Patients were a mean age of 46 years (range, 17-85 years). Exclusion criteria included previous fracture, arthritis, or humeral deformity. The metaversion and humeral head version were measured. Measurements were conducted independently by 2 surgeons blinded to the results of the other. Interobserver and intraobserver reliability was calculated using intraclass correlation. The mean difference between the metaversion and the humeral head version was 2.5° (95% confidence interval [CI], 0.9°-3.9°). The mean difference between metaversion and humeral head version was 1.8° (95% CI, 0.0°-3.6°) in women, 2.9° (95% CI, 0.6°-5.1°) in men, 2.4° (95% CI: 0.6°-4.1°) in right shoulders, and 2.5° (95% CI, -0.1° to 5.1°) in left shoulders. Interrater and intrarater reliability was excellent, 0.97 and 0.98, respectively. Proximal humeral metaphyseal version (metaversion) is an accurate predictor of ipsilateral humeral head version.