There is no consensus regarding the ideal surgical technique for AC joint dislocation . Different open and arthroscopic techniques have been proposed in the literature. Recently, clinical and biomechanical studies have shown that suture button fixations are effective for coracoclavicular reconstruction of the joint [5,12,13]. In this procedure, the CC ligaments are expected to heal since the AC joint is kept reduced and the AC and CC ligament remnants are brought into contact. In a systemic meta-analysis, loop fixation devices were found to have higher shoulder function scores and lower postoperative pain . Another study comparing the results of loop button fixation and Bosworth screw fixation showed that more treatment satisfaction is achieved with loop fixation . Our study showed that arthroscopic fixation of acute acromioclavicular joint disruptions (type III and V) by using a double button device achieves satisfactory outcomes. Our findings are consistent with other authors reporting satisfactory outcomes using similar techniques and device.
Biomechanical studies have shown that button systems provide excellent supero-inferior stability with load to failure higher than native ligament [16,17]. However, horizontal instability in AC dislocations is increasingly being treated with AC repair, with an additional acromioclavicular sling . AC repair was not performed in any of the patients in this study group since it requires wider dissection and longer surgery, has higher morbidity and the additional stability is questionable. In a cadaver study, Weiser et al. did not observe any additional stability with direct AC repair .
The patients in our study were operated by using a single tunnel to avoid the risk of coracoid fracture. In a biomechanical study by Beitzel et al. , there was no difference in stability between the single and double device; nevertheless the coracoid fracture incidence was much higher with the double device. Likewise, a recent study demonstrated that single tunnel reconstruction demonstrates similar biomechanical properties to the intact state and double tunnel reconstruction .
Coracoid fractures have been reported as a potential complication in the literature . We did not have any such complications. We believe that appropriate visualization of the inferior surface of coracoid process and drilling with a 4mm drill close to the base of the coracoid is essential. In a cadaveric study, Rylander et al.  showed that a 4-mm tunnel technique is significantly stronger than a 6-mm tunnel technique when using a transcoracoid reconstruction technique. In addition, tunnel placement is also a significant factor to avoid coracoid fractures and failure of fixation. In a recent study, higher peak load to failure was found when a centre-centre or medial-centre tunnel orientation was performed during drilling .
Recently, some authors have suggested CC ligament reconstruction by using tendon grafting . However, in another study, 47% reduction loss and 20% complications were seen after CC ligament reconstruction with an autologous tendon graft, which adversely affected the results . Many previous open procedures treating the AC joint dislocation without tendon reconstruction, such as hook plate ﬁxation, K-wire, and Bosworth type screw fixation have shown good results after the implants were removed. Therefore, we believe it is viable to do the CC ligament reconstruction without ligamentous augmentation.
The most commonly reported complication after AC joint reconstruction is loss of reduction. In this study, 9 of 36 patients had a reduction loss greater than 3 mm CC distance compared with the unaffected side. There are some possibilities for the reduction loss. In patients operated with the Single Flip Button Device Technique, reduction loss has been shown to be caused by a longer duration between injury and treatment of more than 5 days and poor quality of initial reduction (more than 2 mm CC distance difference in early postop radiological examination) . In our study, no statistically significant difference was observed between groups with and without reduction loss in terms of time from injury to treatment. In addition, early radiological examination revealed a CC difference of less than 2 mm in all patients. Our adjustable loop system only mimics one component of the CC ligament and the AC ligament was not also reconstructed. Undue forces on the AC joint may cause instability and damage the healing process of the CC ligaments. In addition, excessive force on the bone metal button interface may result in bone erosion, clavicular tunnel widening and cause loss of reduction . Another reason could be loosening of the adjustable loop system. Zhang et al.  showed 25% fixation failure rate within 3-6 months after CC fixation using a similar suspensory fixation device. In different series, reduction loss varies; however, good or excellent outcomes are reported regardless of fixation loss. Murena et al.  reported outcomes of 16 patients with AC dislocation treated by the double flip button technique. Although 25% of the patients had a reduction loss, the average constant score was 97.
De Carli et al. compared the functional results of patients with double button fixation and conservative treatment of 3rd degree AC joint dislocations. According to this study, although there was no difference in terms of non-AC joint-specific objective scores in the surgical group, a statistically significant difference was found in the AC joint-specific objective measurements, subjective evaluation of patients and aesthetic satisfaction . In our study, although there was no difference in non-AC joint-specific objective scores in the group without loss of reduction, a statistically significant difference was found in AC joint specific objective scores, subjective assessment and aesthetic satisfaction. Therefore, the authors believe that reduction maintaining is crucial for excellent functional and aesthetic results after fixation of the AC joint with a double button device.
There are some limitations of this study. First, the number of patients was small. Second, our follow up time was not sufficient to make conclusion about the incidence of post-traumatic arthritis; however, we were able to show early results and complications since most of the reduction loss occurs in a short time.