How reliable is the Deltoid Tuberosity Index? Assessing intra- and interobserver variability and the influence of humeral rotation.

BACKGROUND

The Deltoid Tuberosity Index (DTI) is a radiographic parameter proposed as an indirect measure of bone mineral density in the humerus, potentially useful in assessing proximal humerus fractures. Despite suggestions that lower DTI values may predict osteosynthesis failure linked to osteopenia, the reliability of DTI measurements-particularly regarding intra- and interobserver variability and the influence of humeral rotation- has not yet been comprehensively validated.

METHODS

This study evaluated the reliability of DTI measurements in four parts, involving a total of 1448 measurements performed by seven observers. Interobserver variability was assessed by having seven trauma surgeons measure the DTI on 100 anteroposterior (AP) radiographs of proximal humerus fractures. Intraobserver variability was evaluated by having four observers measure the DTI on three selected radiographs ten times each on different occasions. The impact of humeral internal rotation on DTI measurements was examined using a synthetic humerus model at varying degrees of rotation (0°-90° in 10° increments), with repeated measurements to assess the impact of rotation. Cortical variation in the proximal humerus was assessed using computed tomography (CT) imaging in 10 patients, with DTI measurements conducted at different simulated internal rotation angles to evaluate consistency and compare with radiographic measurements. Finally, the clinical relevance of DTI was evaluated by correlating DTI values with the incidence of screw cutout or secondary displacement in 100 patients treated with PHILOS plates.

RESULTS

Interobserver variability was moderate (ICC = 0.55), with significant differences in DTI measurements across observers (p = 0.0079). Experience level had no significant impact (p = 0.71). Intraobserver variability was low, with a mean coefficient of variation of 3.45%, indicating high consistency within individual observers. Humeral rotation significantly influenced DTI measurements in the synthetic bone model (p < 0.01); however, in patient CT scans, DTI measurements showed minimal variation across rotation angles and correlated strongly with radiographic measurements (r = 0.82-0.89, p < 0.001). No significant correlation was found between DTI values and screw cutout incidence in the overall population. In the subgroup aged 65 to 80 years, only two of the seven observers found a significant correlation between lower DTI values and screw cutout.

CONCLUSION

This comprehensive study, involving 1448 DTI measurements, revealed not only considerable interobserver variability but also a significant sensitivity to humeral rotation, raising concerns about the DTI's reliability in clinical practice. The inconsistent correlation between DTI and clinical outcomes like screw cutout suggests that the DTI may have limited utility as a predictive tool in managing proximal humerus fractures.

LEVEL OF EVIDENCE: Level III.