Isolated fractures of the greater tuberosity represent up to 20% of all proximal humeral fractures. Conservative and surgical treatment strategies are reported. Fracture displacement on radiographs is a key factor in treatment decisions, but the accuracy of measurement methods can limit this process. This study aimed to (Aim I) assess the inter- and intraobserver reliability of established radiographic measurements of fracture displacement and (Aims II/III) determine if these measurements can predict superior, posterior, or combined superior + posterior fracture displacement quantified on CT scans in a patient cohort.
Forty-nine cases of isolated greater tuberosity fractures with complete radiographs and CT scans were analyzed. (Aim I) Displacement was measured on radiographs in millimeters as superior ('sup') and lateral ('lat') displacement, along with Mutch's superior ratio ('sGT'), anterior/posterior ratio ('apGT'), and Nyffeler's impingement index ('I-Ind'). Reliability was assessed by three independent observers. (Aim II) A humerus saw-bone model with predefined superior, posterior, and combined fracture displacements was used to validate a CT-based measurement technique. (Aim III) This CT-based method was applied to patient CT scans, and linear regression was used to test if radiographic measurements predicted the CT-measured displacements.
(Aim I) Inter- and intraobserver reliability was excellent for 'I-Ind' (inter 0.98/intra 0.94), good-excellent for 'lat' (0.82/0.76) and 'sGT' (0.75/0.94), and moderate for 'sup' (0.73/0.66) and 'apGT' (0.64/0.49). (Aim II) The CT-based technique accurately measured superior (R ² =0.99), posterior (R ² =0.99), and combined (R ² =0.99) displacement. (Aim III) Patient CT scans showed mean displacement of 3.3±2.5 mm superior, 8.4±5.6 mm posterior, and 11.6±7 mm combined. Superior displacement correlated with 'sup' (p<0.001), posterior with 'lat' (p<0.001), 'apGT' (p=0.004), 'I-Ind' (p=0.048), and combined displacement with 'lat' (p<0.001) and 'apGT' (p=0.006). In fractures displaced ≥5mm, 'sup' (p=0.002) correlated with superior, 'lat' (p=0.003) and 'I-Ind' (p=0.049) with posterior, and 'lat' (p<0.001) with combined displacement.
The CT-based technique accurately measured fragment displacement in the saw-bone model. For fractures with ≥5mm displacement, 'lat' was the best predictor of combined displacement (interobserver reliability: 0.82-0.94). However, in the reported cohort, 'lat' underestimated the "true" fragment displacement. Therefore, in 'lat' measurements ≥ 3mm, a CT scan for quantification of fragment displacement should be considered. However, these findings must be confirmed in bigger patient populations before clinical translation.