Analysis of efficacy and failure in proximal humerus fractures treated with locking plates.

摘要

OBJECTIVE: The purpose of this study was to determine the efficacy of proximal humerus locking plates (PHLP) and to clarify predictors of loss of fixation. DESIGN: Retrospective review of patients with proximal humerus fractures fixed with a PHLP. SETTING: Five Level 1 trauma centers. PATIENTS: One hundred fifty-three patients (111 female, 42 male) 18 years or older with a displaced fracture or fracture-dislocation of the proximal humerus treated with a PHLP between January 1, 2001 and July 31, 2005. INTERVENTION: Demographic data, trauma mechanism, surgical approach, and perioperative complications were collected from the medical records. Fracture classification according to the AO/OTA, radiographic head-shaft angle, and screw tip-articular surface distance in true anteroposterior (AP) and axillary lateral radiographs of the shoulder were measured postoperatively. Varus malreduction was defined as a head-shaft angle of <120 degrees. MAIN OUTCOME MEASUREMENTS: Statistical analysis was done to establishcorrelations between loss of fixation and postoperative head-shaft angle in the true AP radiograph, patient age, fracture type, trauma mechanism, number of locking head screws, and type of plate. RESULTS: The mean age was 62.3 +/- 15.4 years (22-92) and the mean injury severity score (ISS) was 9.5 +/- 10.16 (4-57; n = 73). The surgical approach was deltopectoral (90.2%) or transdeltoid (9.8%). No intraoperative complications were reported. The mean postoperative head-shaft angle was 130 degrees (95 degrees to 160 degrees; SD = 13). The overall incidence of loss of fixation was 13.7%. There was a statistically significant association between varus reduction (<120 degrees) and loss of fixation (30.4% when the head-shaft angle was <120 degrees versus 11% when the head-shaft angle was > or =120 degrees; P = 0.02). CONCLUSION: This series presents the experience using PHLP in 5 Level 1 trauma centers. There were no intraoperative complications related to the locking plate systems. Despite the use of fixed-angle devices, loss of fixation occurred, primarily in the presence of varus malreduction. Our findings suggest that avoiding varus should substantially decrease the risk of postoperative failures.