Purpose: The purpose of this study is to evaluate the biomechanical properties and the stability between locking clavicle plate, dynamic compression plate and external fixation systems on an unstable displaced fracture model under torsional and 3 point bending loading.

Materials and Methods: Forty eight human adult formalin fixed clavicles were paired according to their BMD (DEXA) homogeneously into three groups; Group 1: Locking clavicle plate, Group 2: Dynamic compression plate and Group 3: External fixator. Each specimen was then osteotomized in the midshaft; and a 5mm bone segment was removed in order to stimulate a displaced fracture model. Biomechanical tests were applied in a cyclic loading model in MTS, Bionix 2. Torsional and three point bending forces were performed for 1000 cycles in all subgroups, stiffness was recorded at 10 cycles (initial) and periodic every 100 cyclic intervals. Failure load and moment were obtained after 1000 cycles. Initial stiffness, failure loads and the percentage of initial stiffness for each subgroup were compared across each group. One-way ANOVA and Bonferoni post- hoc tests were utilized to determine which were significantly different from one another with the significance level set as p< 0.05.

Results: The mean initial stiffness(Nmm/deg) - mean failure moments(Nmm) for torsional tests were 703.2 – 7671.7 (locking plate), 448.1 – 4370.3 (compression plate), 365.2 – 2999.7 (ex-fix) and the mean initial stiffness(Nmm) – mean failure loads(N) for bending tests were 32.6 – 213.2 (locking plate), 23.4 – 131.1 (compression plate), 20.6 – 102.7 (ex-fix) respectively. ANOVA test confirmed an overall significant difference between the three constructs in terms of both failure loads and a significant difference only between locking plate and others in terms of initial stiffness. At all cyclic intervals after 100 cycles there was significant difference of percentage of initial stiffness between locking plate and others in bending and torsion. There was a significant difference between compression plate and ex-fix after 700 cycles in torsional group and no difference found in bending group between (any of) them at any cyclic interval.

Conclusions: Locking anatomic clavicle plate is significantly more stable than unlocked dynamic compression plate and external fixator under torsional and bending cyclic loading in an unstable displaced fracture or non-union clavicle model.