Surgical versus conservative interventions for treating acromioclavicular dislocation of the shoulder in adults

Dislocation of the acromioclavicular joint is one of the most common shoulder injuries in a sport-active population. The question of whether surgery should be used remains controversial. This is an update of a Cochrane Review first published in 2010.

To assess the effects (benefits and harms) of surgical versus conservative (non-surgical) interventions for treating acromioclavicular dislocations in adults.

We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register (to June 2019), the Cochrane Central Register of Controlled Trials (CENTRAL) (the Cochrane Library 2019, Issue 6), MEDLINE (1946 to June 2019), Embase (1980 to June 2019), and LILACS (1982 to June 2019), trial registries, and reference lists of articles. There were no restrictions based on language or publication status.

We included all randomised and quasi-randomised trials that compared surgical with conservative treatment of acromioclavicular dislocation in adults.

At least two review authors independently performed study screening and selection, 'Risk of bias' assessment, and data extraction. We pooled data where appropriate and used GRADE to assess the quality of evidence for each outcome.

We included five randomised trials and one quasi-randomised trial. The included trials involved 357 mainly young adults, the majority of whom were male, with acute acromioclavicular dislocation. The strength of the findings in all studies was limited due to design features, invariably lack of blinding, that carry a high risk of bias. Fixation of the acromioclavicular joint using hook plates, tunnelled suspension devices, coracoclavicular screws, acromioclavicular pins, or (usually threaded) wires was compared with supporting the arm in a sling or similar device. After surgery, the arm was also supported in a sling or similar device in all trials. Where described in the trials, both groups had exercise-based rehabilitation. We downgraded the evidence for all outcomes at least two levels, invariably for serious risk of bias and serious imprecision.Low-quality evidence from two studies showed no evidence of a difference between groups in shoulder function at one year, assessed using the Disability of the Arm, Shoulder, and Hand questionnaire (DASH) (0 (best function) to 100 (worst function)): mean difference (MD) 0.73 points, 95% confidence interval (CI) -2.70 to 4.16; 112 participants. These results were consistent with other measures of function at one-year or longer follow-up, including non-validated outcome scores reported by three studies. There is low-quality evidence that function at six weeks may be better after conservative treatment, indicating an earlier recovery. Very low-quality evidence from one trial found no difference between groups in participants reporting pain at one year: risk ratio (RR) 1.32, 95% CI 0.54 to 3.19; 79 participants. There is very low-quality evidence that surgery may not reduce the risk of treatment failure, usually resulting in non-routine secondary surgery: 14/168 versus 15/174; RR 0.99, 95% CI 0.51 to 1.94; 342 participants, 6 studies. The main source of treatment failure was complications related to surgical implants in the surgery group and persistent symptoms, mainly discomfort, due to the acromioclavicular dislocation in the conservatively treated group.There is low-quality evidence from two studies that there may be little or no difference between groups in the return to former activities (sports or work) at one year: 57/67 versus 62/70; RR 0.96, 95% CI 0.85 to 1.10; 137 participants, 2 studies. Low-quality but consistent evidence from four studies indicated an earlier recovery in conservatively treated participants compared with those treated with surgery. There is low-quality evidence of no clinically important difference between groups at one year in quality of life scores, measured using the 36-item or 12-item Short Form Health Survey (SF-36 or SF-12) (0-to-100 scale, where 100 is best score), in either the physical component (MD -0.63, 95% CI -2.63 to 1.37; 122 participants, 2 studies) or mental component (MD 0.47 points, 95% CI -1.51 to 2.44; 122 participants). There is very low-quality and clinically heterogenous evidence of a greater risk of an adverse event after surgery: 45/168 versus 16/174; RR 2.82, 95% CI 1.65 to 4.82; 342 participants, 6 studies; I² = 48%. Common adverse outcomes were hardware complications or discomfort (18.5%) and infection (8.7%) in the surgery group and persistent symptoms (7.1%), mainly discomfort, in the conservatively treated group. The majority of surgical complications occurred in older studies testing now-outdated devices known for their high risk of complications. The very low-quality evidence from one study (70 participants) means that we are uncertain whether there is a between-group difference in patient dissatisfaction with cosmetic results.It is notable that the evidence for function, return to former activities, and quality of life came from the two most recently conducted studies, which tested currently used devices and interventions in clearly defined participant populations that represented the commonly perceived population for which there is uncertainty over the use of surgery. There were insufficient data to conduct subgroup analysis relating to type of injury and whether surgery involved ligament reconstruction or not.

There is low-quality evidence that surgical treatment has no additional benefits in terms of function, return to former activities, and quality of life at one year compared with conservative treatment. There is, however, low-quality evidence that people treated conservatively had improved function at six weeks compared with surgical management. There is very low-quality evidence of little difference between the two treatments in pain at one year, treatment failure usually resulting in secondary surgery, or patient satisfaction with cosmetic result. Although surgery may result in more people sustaining adverse events, this varied between the trials, being more common in techniques such as K-wire fixation that are rarely used today. There remains a need to consider the balance of risks between the individual outcomes: for example, surgical adverse events, including wound infection or dehiscence and hardware complication, against risk of adverse events that may be more commonly associated with conservative treatment such as persistent symptoms or discomfort, or both.There is a need for sufficiently powered, good-quality, well-reported randomised trials of currently used surgical interventions versus conservative treatment for well-defined injuries.