Extra-articular all-endoscopic anchor fixation versus clavicular hook plate for Neer type II distal clavicle fractures: a retrospective comparative cohort study

Neer type II distal clavicle fractures are unstable injuries. Clavicular hook plates (CHP) provide fixation but can cause subacromial complications and require implant removal. We compared extra-articular all-endoscopic anchor fixation (BEA) with CHP. This single-center retrospective non-randomized cohort included 120 patients treated from July 2022 to June 2025 (BEA, 51; CHP, 69). After discussion of both procedures, patients selected treatment; both techniques were performed by the same general orthopedic team. The primary outcome was the 12-month Constant–Murley Score (CMS). Radiographic union by 6 months was the key secondary radiographic outcome; other secondary outcomes included healing status at scheduled visits, perioperative measures, pain, shoulder function, complications, cortical resorption, and secondary procedures. Because radiographs were obtained at scheduled visits, time to union was not estimated. At 12 months, before CHP removal, CMS was higher in the BEA group than in the CHP group (92.4 ± 5.6 vs. 84.8 ± 6.9; mean difference, 7.6 points; 95% CI, 5.3 to 9.9; P < 0.001). All patients had evaluable 6-month radiographs; union was achieved in 51/51 BEA patients (100%) and 65/69 CHP patients (94.2%; Fisher’s exact P = 0.136; absolute risk difference, 5.8 percentage points; 95% CI, -2.0 to 14.0). BEA was also associated with a shorter incision (2.8 ± 0.4 vs. 6.5 ± 1.2 cm, P < 0.001), less blood loss (35.6 ± 12.3 vs. 78.4 ± 25.6 mL, P < 0.001), and lower early postoperative pain scores. Four CHP patients were ununited at 6 months and met the study definition of nonunion at 9 months. Cortical resorption was observed in 9 BEA patients. At the primary endpoint, the BEA group had a higher 12-month Constant–Murley Score than the CHP group. For the key secondary outcome of radiographic fracture union at 6 months, no statistically significant between-group difference was observed. These findings warrant further validation in prospective studies. Not applicable.