Return to sport after surgical treatment for high-grade (Rockwood III–VI) acromioclavicular dislocation
Acromioclavicular (AC) joint dislocations are common in a young and active population, especially in people performing contact sports. Full recovery with a fast and high rate of return to sport is desirable. This systematic review aims to combine patient outcomes in order to help surgeons in addressing patient expectations regarding the return to sport after surgical intervention for AC dislocations.
To conduct this systematic review, the PRISMA guidelines were followed. Articles were included if written in English or Dutch and evaluated return to sport after any type of surgical intervention for Rockwood types III to VI AC dislocations in patients practicing sports. Outcome parameters were return to sport, time to return to sport, level of sport, functional outcome scores and complications.
Twelve studies involving 498 patients were included, of which 462 patients practiced sports. 432 (94%) patients returned to sport. The weighted mean time to return to sport was 4.0 months. 338 out of 401 patients (84%) returned to the same level of pre-injury sport and 35 patients (9%) lowered their level of sport. The weighted mean Constant score was 92 out of 100.
The rate of return to sport after surgical intervention for Rockwood (RW) III–VI AC dislocations is high. However, the level of evidence was low and due to the methodological heterogeneity between studies, subgroup analyses of return to sport outcomes were not feasible.
Level of evidence
Systematic review of level I–IV studies, level IV.
KeywordsAcromioclavicular dislocation AC joint Rockwood Return to sport Functional outcome Systematic review
Athletic shoulder outcome scoring system
Methodological Index for Nonrandomized Studies
Randomized controlled trial
Range of motion (aROM: active, pROM: passive, fROM: free)
Shoulder sport activity score
Visual analog score
Acromioclavicular (AC) joint injuries are common in young athletes with an incidence of 9.2 per 1000 person-years . The incidence of AC injuries is higher in contact sports and is highest in adults in their 20s [21, 31]. Full recovery with return to pre-injury level of sports and minimal time lost to injury is essential in the treatment of these injuries.
The Rockwood classification is most commonly used for classifying AC joint injuries according to the severity . Most type I and II dislocations are treated conservatively. For type IV to VI AC joint dislocations, surgical treatment generally is the intervention of choice [21, 29, 31]. The optimal management of type III AC joint dislocations remains controversial. A trend is set toward initial non-surgical treatment, unless non-surgical treatment fails in patients with persistent pain or the inability to return to sport or work [21, 29, 31].
Many surgical techniques have been described for the repair of AC joint dislocations [2, 21]. When evaluating surgical treatment, existing literature mainly measures functionality with outcome scores or self-reported outcome measures. Limited literature exists about the rate of return to sport in this population. However, return to sport has been found to be a good indicator of success after treatment for injuries to the musculoskeletal body [10, 14, 17, 22]. Analyzing the rate of return to sport could provide surgeons and patients with relevant outcome information to optimize shared decision making. The purpose of this systematic review is to assess the rate of return to sport after surgical intervention in high-grade AC dislocation (Rockwood III–VI), to help surgeons address patient expectations. In contrast to the existing literature on this topic, the present review aims to combine outcomes of sports-practicing patients of studies with larger sample sizes and more recent studies with higher levels of evidence.
Materials and methods
To conduct this systematic review, the PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) were followed .
Using the search terms “acromioclavicular joint” and “return to sport/to play”, an electronic search was performed in MEDLINE (PubMed), EMBASE and Cochrane with the assistance of a clinical librarian. There was no restriction on the date of publication. The final search was conducted in September 2018. Based on the titles and abstracts, two reviewers (D.V. and C.W.) independently identified potentially relevant articles for review of the full text. For each identified article, the reference list was screened and a manual search was conducted for articles that could potentially be relevant.
Inclusion and exclusion criteria
The included articles should be written in English or Dutch and evaluate the return to sport after any type of surgical intervention for high-grade (Rockwood III–VI) AC dislocation in patients practicing sports. There was no restriction on the type of sport, level of sport and age of the patient. Exclusion criteria were reviews of the literature, expert opinions, non-clinical studies, case reports and studies with a sample size of less than 20 athletes. Studies were also excluded when there were insufficient data on the number of patients practicing sports pre- and post-injury. Studies that also included non-athletes in the analysis were only included in this review when separate and sufficient data on the sports-practicing patients were presented.
Data from eligible studies were extracted based on a predefined data extraction form. The following data and baseline parameters were recorded when available: author, publication year, study design, level of evidence, numbers of patients, sex, age, laterality, Rockwood classification (or Tossy III when the Tossy classification was used), acute or chronic injury (with definition of the interval according to the author), operation technique, type of sports, and follow-up time.
The primary objective was to determine the rate of return to sport. Secondary outcomes were the time to return to sport, level of sport pre- and post-injury, clinical outcome scores and complications. Self-reported outcome measures or clinical outcome scores were only considered for analysis when three or more studies reported on this outcome.
The methodological quality of included studies was assessed by assigning levels of evidence as previously defined by the Centre for Evidence-Based Medicine (http://www.cebm.net). Additionally, the quality of the included studies was evaluated according to the Methodological Index for Nonrandomized Studies (MINORS) checklist . This instrument is designed to assess the methodological quality of non-randomized surgical studies. To assess risk of bias for randomized trials, the Cochrane Collaboration tool will be used. Bias is assessed as a judgment (high, low or unclear) for individual elements from five domains (selection, performance, attrition, reporting and other) . A level of evidence was assigned by two authors (D.V. and C.W.). Disagreement was resolved by consensus.
For the rate of return to sport, a forest plot was created (Graphpad Prism; Graphpad Software, San Diego, CA, USA) with confidence intervals calculated by a binominal exact calculation for proportions. I2 tests (MedCalc; Medcalc Software, Ostend, Belgium) were used to determine heterogeneity between studies. Values of I2 between 25 and 49% were considered low, 50–74% was considered moderate, and values greater than 75% were considered to have high statistical heterogeneity . Further analysis on the outcomes was presented narratively. Means (weighted), medians, ranges and percentages were calculated using Microsoft Excel 2016 (Microsoft, Seattle, WA, USA).
Level of evidence
Mean age; y (range)
Dominant side; n
Mean follow-up; months (range)
Acute/chronic (cut-off, weeks)
Type of surgery
Modified Kirschner wire with CC repair using PDS-cord augmentation
Acute + chronic (3)
Modified Weaver–Dunn with CC repair using non-absorbable sutures
Modified Weaver–Dunn with AC repair using absorbable braided suture
Kirschner wire with CC repair using absorbable cord
De Carli 
CC repair using cortical buttons
CC repair using cortical buttons
Marcheggiani Muccioli 
CC repair using LARS
CC repair using double cortical buttons (arthroscopic)
28 (22–51) M
AC + CC repair using semitendinosus hamstring autograft
Non-overhead 46 M, overhead 33 (18–65)
CC repair using cortical buttons
G1: 39.3 ± 22.7
G2: 30.8 ± 8.4
G1: clavicular hookplate
G2: double cortical buttons (arthroscopic)
Single vs double cortical buttons (arthroscopic)
Return to sport outcomes
Return to sport; n
Return to sport protocol
Mean time to return to sport, w (range)
Level of Sports
Mean functional outcome measures
6 w immobilization → PT
Same level as pre-injury (90%)
4–6 w immobilization → PT
No contact sports < 36 w
Same level as pre-injury (93%)
4 w immobilization → after 6 w aROM
Same level as pre-injury (86%), Lower level (14%)
Constant: 97 (66–100)
2 w immobilization and 6 w of p/aROM
Same level as pre-injury (100%)
Not stratified for athletes
De Carli 
4 w immobilization, after 2 w pROM → after 8 w aROM
Contact sports and heavy work allowed > 12 w
Same level as pre-injury (83%), Lower level (17%)
6 w immobilization and pROM → aROM
Avoidance of lifting, carrying, pushing and pulling < 12w
Same level as pre-injury (86%), Lower level (12%)
Not stratified for athletes
Marcheggiani Muccioli 
III, IV, V
3 w immobilization and pROM → aROM
Contact sport allowed > 8–12 w
Professionals: 16 (12–20) M
Non-professionals: 20 (16–24)
Same level as pre-injury (100%)
Constant: prof: 97; non-prof: 91
6 w immobilization and PT → fROM
Sport allowed > 12 w, overhead or contact sports allowed > 24 w
Same level as pre-injury (62%), Lower level (38%)
Constant: 94 (86–100)
6 w immobilization and pROM after 3 w → after 7 w aROM
Full activity including contact sports allowed after 16 w
Same level as pre-injury (94%)
ASES: 85 (82–98)
4 w immobilization and pROM → after 6 w aROM
Non-contact sport allowed after 3 mo, no restrictions > 24 w
38 (12–72) M
Same level as pre-injury (79%), Lower level (16%)
Constant: non-overhead: 94 (49–100), overhead: 89 (63–100) M
III, IV, V
6 w immobilization → aROM
Fitness sport allowed > 16 w, contact sports allowed > 40 w
G1: SSAS 6.2 ± 1.5, ASOSS 77.8 ± 20.8
G2: SSAS 7 ± 1.4, ASOSS 91.4 ± 10.3
6 w immobilization and pROM → aROM
Limited rehabilitation sports allowed > 12 w
Single button: 21 (12–32)
Double button: 13 (12–16)
Same level as pre-injury (76%)
Constant: Single: 83 ± 4, Double: 92 ± 3
Return to sport
Acute or chronic
Among the 12 included studies, a variety of cut-off intervals for acute versus chronic surgery were used, ranging from 2 to 6 weeks (see Table 1). In four studies, the interval was not reported. Due to the heterogeneity of the definition for acute and chronic surgery, it was not possible to perform a subgroup analysis on return to sport outcomes for acute versus chronic surgery.
For subgroup analysis of the rate of return to sport for the Rockwood subtypes, studies could only be included when stratified data for each subgroup were presented. Two studies provided sufficient data on the rate of return to sport for Rockwood type III injuries [3, 20]. The rate of return to sport was 100% among 53 patients. In two studies [20, 39], for Rockwood type IV, 64 out of 83 patients (77%) returned to sport. For Rockwood type V, five studies reported the rate of return to sport [5, 20, 30, 32, 35]. Among 153 patients, 149 (97%) returned to sport, ranging from 94 to 100% (I2 = 7%, 0–66%). No data on return to sport were provided for patients diagnosed with Rockwood type VI.
Functional outcome scores
Many different functional outcome scores were used in the studies. Only the Constant score was used in more than three studies, in which the data on patients practicing sports were stratified. In six studies, the mean Constant score was reported [3, 20, 30, 32, 35, 39] (see Table 2). The weighted mean Constant score at follow-up for these studies was 92 out of 100 (range 84–98), indicating an excellent outcome. The mean time to follow-up in these studies ranged from 2 to 3.5 years.
Complications and re-surgery
The data on complications could not be stratified for athletes alone. The following complication rates are combined for all included studies. Among 498 operated patients, the rate of complications was 5.2% at final follow-up. There were 6 cases (1.2%) of failure or migration of osteosynthesis material, 5 cases (1.0%) of complete loss of reduction, 10 cases (2.0%) of wound infection, 2 cases of tunnel misplacement, 1 case of coracoid fracture due to non-compliance and 2 subacromial bursitis. The rate of revision surgery due to complications in these studies was 3.0%.
The most important finding of the present study was a very high (94%) overall rate of return to sport in patients surgically treated for Rockwood III–VI AC dislocation. Additionally, many patients (84%) returned to their pre-injury level of sports. The rate of return to sport in patients surgically treated for high-grade AC dislocation is superior to that of patients surgically treated for rotator cuff repair and arthroscopic Bankart repair (81–85%) [14, 22].
Although this outcome is promising, it must be noted that 10 of 12 studies are level IV, of which the majority are non-comparative studies. In this review, only one randomized study could be included. The RCT of Muller et al.  showed that, at 24 months, 100% of patients treated with arthroscopically assisted double-suture-buttons participated in sports compared to 93% of patients treated with a clavicular hook plate.
A subgroup analysis on the rate of return to sport for different types of sports was not feasible. The study of Porschke et al.  compared sports activity for overhead athletes versus non-overhead athletes. Overhead athletes needed to change sport activity significantly more often (54% vs 12%; p = 0.011) and 27% had to reduce their sports level to low-demanding sports, while none of the non-overhead athletes lowered their level (p = 0.029). Muller et al.  found that overhead athletes had more benefit from a double-suture-button technique than collision athletes and suggested non-anatomic clavicle hookplate as a good alternative for the collision athletes.
The rates of return to sport for Rockwood III and V dislocations were similar. However, there were no comparative studies determining the return to sport outcomes for the different Rockwood subtypes. In addition, not all patients could be included in the Rockwood subgroup analysis since some authors used the Tossy classification, in which Tossy type III corresponds to Rockwood types III to VI. Notable is that the rate of return to sport in Rockwood IV dislocations was found to be lower, although this is mainly based on one study  analyzing 78 out of 83 patients with Rockwood IV dislocation included in this analysis. This study found a lower rate of return to former sports, especially among patients treated by a single cortical button technique. However, no data was presented on the possibility that patients practiced other sports at follow-up.
The surgical techniques for AC dislocation varied greatly among the included studies. A subgroup analysis was not feasible. A recent review of Gowd et al.  showed, although there is a trend toward minimally invasive procedures, there are no differences in loss of reduction, complication rate and revision rate between open and arthroscopic AC joint reconstruction. Considering return to sport outcomes, the RCT by Muller et al.  showed that patients treated arthroscopically by double cortical buttons had a superior level of performance in shoulder sport, close to the level of performance in healthy athletes, compared to patients treated by a hookplate.
Recently, a review on the return to sport after surgical management for AC dislocation was published by Kay et al. . Five out of 12 studies included in the present review were also reviewed by Kay et al.; however, the present review included 7 additional studies reporting on return to sport outcomes after high-grade AC dislocation, one of which was an RCT. In contrast to Kay et al., we excluded three studies [4, 11, 37] due to lack of data on the practicing of sports pre-injury and four studies due to a sample size of less than 20 athletes [1, 6, 28, 36]. Another methodological difference was the approach to non-stratified data on mixed populations. The present review only combined patient outcomes from different studies when data on athletes were stratified, for instance in the analysis of functional outcome measures, while Kay et al. combined patient outcomes from all studies reporting on these outcomes, including studies with mixed populations. In addition, Kay et al. combined return to sport outcomes of acute versus chronic surgery although the cut-off intervals between studies were very different. No conclusion could be drawn on the rate of return to sport for acute versus chronic AC dislocation because of the many different definitions used for these injuries. For future research, it is important to work with one definition and one cut-off interval to be able to compare acute and chronic surgeries in different studies.
The present review was limited by the overall low level of evidence provided by the many level IV studies. As a consequence, no robust conclusions can be drawn on return to sport outcomes for different subgroups comparing surgical techniques, acute versus chronic surgery, types of sports and Rockwood subtypes. In addition, the rates of return to sport were mostly secondary outcomes in these studies, with insufficient attention paid to the many aspects of the return to sport. This may lead to publication bias when only high rates of return to sport are reported.
In this review, no comparison was made between conservative and surgical treatment for patients with high-grade AC dislocation. Non-operative treatment for high-grade AC dislocation is being studied more extensively since functional outcome may be non-inferior to surgery. A recent meta-analysis comparing surgical and conservative treatments of Rockwood III AC dislocation showed no significant differences in terms of functional outcome scores . Similar findings are presented for Rockwood IV and V AC dislocations [12, 16, 25, 26]. Murray et al. , comparing open reduction and tunneled suspensory device with non-operative treatment in Rockwood III and V, states that non-operatively managed patient generally recover faster, although a substantial part of remain dissatisfied and require delayed surgical reconstruction.
The outcomes of this review may help surgeons in addressing patient expectations in a sporting population facing surgery for AC dislocation. Patients can now be properly informed on the rate of return to sport, time to return to sport and the patient satisfaction with the results. Still, surgeons need to consider conservative therapy, especially in Rockwood III, since non-surgical treatment can result in similar functional outcome. For future research, high-quality comparative studies are needed to provide adequate data for subgroup analysis on functional and sport outcomes, for both conservative and surgical treatment. When assessing return to sport outcomes, it is imperative to incorporate aspects of sports activity, frequency and intensity.
Overall, the rate of return to sport following surgery for AC joint dislocation is high and most patients can expect to return to their pre-injury level of sport.
The authors are grateful to Bert Berenschot, medical librarian, for his assistance in performing the literature search.
DEV performed the screening, data extraction and analysis, and drafted the manuscript. CLW performed the screening and helped with the analysis and to draft the manuscript. AJS aided in the design and process of all aspects regarded with performing this review. MPJB aided in the design and process of all aspects regarded with performing this review. All authors read and approved the final manuscript.
No funds were received.
Compliance with ethical standards
Conflict of interest
The authors declare that they have no competing interests. No benefits have been received or will be received from a commercial party related (in)directly to the subject of this article.
This chapter does not contain any studies with animals performed by any of the authors.
For this type of study formal consent is not required.
- 5.Garofalo R, Ceccarelli E, Castagna A, Calvisi V, Flanagin B, Conti M et al (2017) Open capsular and ligament reconstruction with semitendinosus hamstring autograft successfully controls superior and posterior translation for type V acromioclavicular joint dislocation. Knee Surg Sports Traumatol Arthrosc 25:1989–1994CrossRefGoogle Scholar
- 7.Gowd AK, Liu JN, Cabarcas BC, Cvetanovich GL, Garcia GH, Manderle BJ et al (2018) Current concepts in the operative management of acromioclavicular dislocations: a systematic review and meta-analysis of operative techniques. Am J Sports Med. https://doi.org/10.1177/0363546518795147 CrossRefPubMedGoogle Scholar
- 12.Joukainen A, Kroger H, Niemitukia L, Makela EA, Vaatainen U (2014) Results of operative and nonoperative treatment of Rockwood types III and V acromioclavicular joint dislocation: a prospective, randomized trial with an 18- to 20-year follow-up. Orthop J Sports Med. https://doi.org/10.1177/2325967114560130 CrossRefPubMedPubMedCentralGoogle Scholar
- 24.Muller D, Reinig Y, Hoffmann R, Blank M, Welsch F, Schweigkofler U et al (2018) Return to sport after acute acromioclavicular stabilization: a randomized control of double-suture-button system versus clavicular hook plate compared to uninjured shoulder sport athletes. Knee Surg Sports Traumatol Arthrosc. https://doi.org/10.1007/s00167-018-5044-x CrossRefPubMedGoogle Scholar
- 31.Rockwood CJ, Williams G, Young D (1998) Disorders of the acromioclavicular joint. In: Rockwood CA, Matsen FA III (eds) The shoulder, 2nd edn. WB Saunders, Philadelphia, pp 483–553Google Scholar
- 36.Triantafyllopoulos IK, Lampropoulou-Adamidou K, Schizas NP, Karadimas EV (2017) Surgical treatment of acute type V acromioclavicular joint dislocations in professional athletes: an anatomic ligament reconstruction with synthetic implant augmentation. J Shoulder Elbow Surg. https://doi.org/10.1016/j.jse.2017.05.032 CrossRefPubMedGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.