Increasing soft tissue damage with increasing grades of AC joint injury lead to greater functional deficits and pain. Particularly poor outcomes have been noted in higher grades that are associated with deltotrapezial fascia disruption and horizontal plane instability. Authors have reported concomitant intra‐articular glenohumeral pathology in up to 84% of patients with an acute AC injury, which impacts treatment and outcome in select patients.1

There are no prospective comparison studies specifically looking only at low grade (Rockwood type I–III) injuries and nonoperative treatment is considered standard of care for type I and II AC separations. However, Bergfeld et al. reported persistent pain and activity limitations in 9% of type I and 23% of type II AC injuries in an active military population.²

Type III injuries have been the center of debate. Gstettner et al. performed a retrospective cohort study comparing patients with type III injuries treated either with hook plates (n = 24) or nonoperatively (n = 12). Treatment choice was based on patient’s preference. They found improved Constant scores (90.4 vs 80.7; p <0.05) and AC joint reduction in the operative group but no difference in return to sport.³ Smith performed a meta‐analysis comparing operative (trans‐acromial Kirschner wires [K‐wires], coracoclavicular screws, or hook plates) and nonoperative treatment of type III separations. The operative group had improved cosmesis (risk difference [RD] = 0.64; 95% confidence interval [CI]: 1.09, 0.19; p <0.0001), slower return to work (RTW) (mean difference [MD]: 3.3 days’ sick leave; 95% CI: 2.10–4.50; p <0.001), similar function, but there was no difference in strength, throwing ability, and AC arthritis.⁴ A more recent systematic review of 22 low‐quality studies showed a 14% loss of reduction in type III injuries with fixation and no difference in Constant score (87.3 operative vs 88 nonoperative; p = 0.6832), or arthritic change (38.4% operative vs 40.5% nonoperative; p = 0.9413). There was a trend toward lower persistent pain with operative treatment (11% vs 25%; p = 0.07).⁵ These analyses should be interpreted with caution due to the inclusion of heterogeneous (and sometimes outdated) techniques, but in general they support the conservative management of most type III separations with acceptable outcomes. However, a portion of patients treated nonoperatively will have persistent symptoms.^4,5 In general, operative management is reserved for those who fail nonoperative management. This is most commonly due to persistent pain and weakness. Other causes of failure include intra‐articular glenohumeral pathology, dynamic posterior instability, scapular dyskinesia, thoracic outlet syndrome, deformity, or cosmetic concerns.

Operative indications have remained largely unchanged since the first randomized trial by Bannister in 1989.⁶ They randomized 54 patients with AC dislocations of various grades to nonoperative treatment versus fixation with a coracoclavicular screw. They noted significantly faster RTW (manual workers RTW average 4 vs 11 weeks; clerical workers RTW average 1 vs 4 weeks; p <0.01), return to sport (7 vs 16 weeks; p <0.05), and fewer unsatisfactory results in the conservative management group (0% vs 16% fair at 4 years). Better results were noted with early surgical fixation for severe displacement of ≥2 cm displacement (consistent with Rockwood type V).⁶ Bannister confirmed these results in another prospective study of 48 patients, showing the highest benefit for early surgery in those with displacement ≥2 cm.⁷

The Canadian Orthopaedic Trauma Society performed an RCT of 83 patients with grade III–V AC separations comparing hook plate versus nonoperative treatment.⁸ Hook plate offered better radiographic reduction (p <0.05) but more frequent complications. Nonoperative treatment yielded better early function – DASH up to three months 16.03 (standard deviation [SD]: 17.03) vs 28.76 (SD: 15.3); p = 0.005, Constant score up to six months – 91.53 (SD: 7.09) vs 80.22 (SD: 17.56); p = 0.0001 – earlier RTW (76% vs 43% at three months; p = 0.004) and lower re‐operation rate (38 planned and 2 unexpected in the operative group vs 2 conversion to surgical in the conservative group; p <0.05). This is biased by the standardized removal of hardware necessitated by this procedure which occurred in 38 (79%) of the patients during this study. There was no difference in perceived cosmesis at one year (p = 0.091) or function from six months through two years.⁸ A follow‐up study of this data analyzing health‐related quality of life showed these patients have pre‐injury function higher than population norms and they return to this physical function faster with nonoperative management (six months vs two years), concluding that hook plate fixation does not lead to improved general health status.⁹ Although there were patients in the nonoperative group who had a poor outcome and required later reconstruction, the authors were unable to predict factors associated with nonoperative failure. Despite the rigorous nature of this multicenter trial, they were unable to show a benefit in high‐grade injuries due to lower recruitment of these rare entities.