After acute mid-shaft clavicular fracture, the shortening degree is one of the relative indication for surgery. When shortening distance is over 15 to 20 mm, the surgery is necessary. It is important to accurately measure the shortening degree of clavicular fracture in the absence of open fractures with absolute indications such as vascular and nerve injury. It is difficult to measure the shortening degree of clavicle accurately on the X-ray film because of its S-shape, and different angles on the X-ray films make the test results variable. Which is the optimal measurement method is still controversial, and thus it brings difficulty to make appropriate treatment plan.

To investigate and compare the accuracy and application value of bilateral length comparison method in the assessment of shortening degree after mid-shaft clavicular fracture.

Based on the premise of anatomic reduction of postoperative clavicle fracture, the accuracy of bilateral clavicle length comparison method (bilateral method) was analyzed with the calibration method as the standard. The bilateral method was used to take different films for the same group of patients, including posteroanterior (PA) view, anteroposterior (AP) 15° cephalad view, and posteroanterior (PA) 15° caudal view. Bland-altman consistency analysis was performed for the measured and calculated collarbone shrinkage (bilateral) and the calculated collarbone shrinkage (calibration) .

With the 3 methods, the 95% consistency range of error of bilateral length comparison method compared with calibration method is -10.98 ~ 10.00 mm of PA view, -18.98 ~ 10.95 mm of AP 15° cephalad view and -18.93 ~ 13.88 mm of PA 15° caudal view respectively. Among these methods, the PA view had the smallest error relatively. The d_{absolute value} ranged from small to large were 4.22 mm on PA view, 6.53 mm on <PA 15° caudal view and 7.12 mm on AP 15° cephalad view (P=0.044) . A total of 5 cases of clavicle shortening value over 15 mm were measured with bilateral method. Among these 5 cases, only 1 case of over 15 mm was measured on the same radiographs with calibration method, with a coincidence rate of 20%. A total of 10 cases in which the clavicle shortening value over 15 mm was measured with the calibration method. Among these 10 cases, only 1 case had the clavicle shortening value of over 15 mm measured with the bilateral method on the same radiographs, with a coincidence rate of 10%.

Bilateral clavicle length comparison method was applied to measure the degree of clavicular shortening (PA view, AP 15° cephalad view and PA 15° caudal view) had large error and insufficient accuracy, which was not applicable to the judgment of surgical indications for mid-shaft clavicular fractures.