Arthroscopic Bankart repair versus conservative treatment for first-time traumatic anterior shoulder dislocation: a systematic review and meta-analysis

Almost every orthopedic surgeon has to deal with anterior shoulder dislocation in their professional practices. Owing to its anatomical characteristics and increasing popularity of contact sports, shoulder is the most commonly dislocated joint and anterior shoulder dislocation accounts for nearly half of all joint dislocations [1]. A well functioned glenohumeral joint needs intactness and coordination of static component and dynamic force. Those structures that ensure normal function of shoulder are at high risks of injury at the moment of dislocation [2], including the glenoid labrum, bony glenoid rim, glenohumeral ligament, capsule, and humeral head.

Robinson et al. [3] reported an 87% to 100% incidence of Bankart lesion in first-time anterior dislocations. Eight nine percent incidence of Hill–Sachs lesions during first-time anterior shoulder dislocations was reported by Taylor et al. [4]. Nearly half of the dislocations occur in individuals aged 15 to 29, and the majority are males [5]. The recurrence of anterior shoulder instability could be as high as 87% in high-risk patients that managed conservatively after first episode of dislocation [6].

In a long-term prospective study, about half of the patients aged 20 to 25 who had encountered a primary anterior shoulder dislocation experienced recurrence of instability and about 25% needed surgical stabilization [7]. Follow-up data from the same group of patients revealed moderate to severe osteoarthritis in 18% of patients without recurrence. The corresponding figures were 26% for those undergoing surgical stabilization and 39% for patients with more than one recurrence who were managed conservatively [8]. The investigators had concluded that age older than 25 years at primary dislocation, recurrent instability, high-energy sports as the trigger of dislocation, and alcohol abuse were risk factors for developing osteoarthritis [8].

Patient expectation, age, gender, lifestyle, and sports level should all been taken into consideration during decision making process. There is a tendency towards arthroscopic stabilization at first episode of dislocation in recent years, especially for young patients with high-risk of recurrent instability. On the other hand, there was moderate-quality evidence that half of the patients did well after conservative management [9]. Previous systematic reviews have compared results of surgical and conservative treatment, with varied evidence levels and intervention methods [9‐12]. New high-level evidence studies focusing on the comparison of arthroscopic Bankart repair (ABR) and the conservative treatment had emerged since then. In this context, we tried to pool available evidence and conduct an updated systematic review and meta-analysis on this topic.

Pooled results of prospective studies showed that arthroscopic stabilization significantly lowered the rate of recurrent instability after primary anterior shoulder dislocation in young active population. Evidence regarding functional scores failed to reveal a distinct difference between the two treatment arms due to variation and incompleteness of outcome measures among studies. As the study population were mostly young men and large part of the injury were sport related, the strategy of early surgical stabilization cannot be easily generalized to other groups of patients. For those with high expectations of returning to contact sports, ABR could be a reliable choice after the first episode of anterior dislocation. However, as long-term prognosis of instability and prevalence of dislocation arthropathy is comparable among ABR, nonoperative treatment, and open procedure [26, 27], a wait and see strategy or delaying surgery depending on recurrence may be a reasonable choice for the non-athletes.

Traumatic anterior shoulder dislocation is accompanied with high incidence of pathologic changes including osseous and soft tissue lesion. Bankart lesion (soft-tissue or bony avulsion) and Hill-Sachs lesion are the most common pathological findings in arthroscopic and radiological examinations. Glenoid labrum deepens the glenoid socket and increases its surface area, serving as a type of “chock block” to the humeral head [2]. And it’s an essential supplement to the glenohumeral stability besides the surrounding ligaments. Biomechanical study has revealed that both Bankart lesion and anterior glenohumeral ligament complex elongation contribute substantially to the occurrence of anterior glenohumeral instability [28]. Long term observation by Kavaja et al. indicated that mild glenohumeral arthropathy was common following conservative management [26]. However, results of ABR were comparable to conservative ones, which rarely caused more than minor subjective symptoms or objectively perceived disadvantages during long-term follow-up. Another 13-year radiological study also indicated that osteoarthritic changes were common findings after ABR, which were comparable to that after open repairs reported in the literature [27]. The energy of trauma and age of patient were deemed to be more relevant to recurrence and long-term arthropathy than the kind of treatment following shoulder dislocation [27, 29].

In terms of recurrent instability, the results revealed that ABR was superior as compared with conservative management in young active population, which is consistent with previous studies [10, 12, 29]. However, glenohumeral instability is encountered in a wide range of activities and age groups, making extrapolation to other groups of population like women, older patients, and non-athletes not necessarily feasible [9]. Results of a network meta-analysis [9] were less favorable towards surgery than other systematic reviews and meta-analysis [12, 30]. The authors attributed this difference to a strict inclusion of RCTs only, the use of network analysis, the handling of labrum repair, and arthroscopic lavage and non-surgical treatment as separate entities. According to a previous long-term follow-up study, about half of the first-time dislocation patients younger than 25 would experience recurrent instability and need stabilizing surgery, while another half had not recurred or become stable over time [7, 8, 31]. Also, observation of another group of patients indicated that delaying surgery depending on the development of chronic instability after first-time anterior traumatic shoulder dislocation did not necessarily lead to less favorable quality of life, inferior prognosis of instability or glenohumeral joint arthropathy [26]. Kavaja et al. held the opinion that a wait and see strategy would direct resources more efficiently than routine surgery after first-time anterior traumatic shoulder dislocation and might save half of patients from unnecessary surgery. In this context, the superiority of immediate arthroscopic stabilization in precaution of recurrent instability can be questioned. Even for young athletes who desired to return to athletic activity after first episode of dislocation, nonoperative management could be feasible and effective during the playing season [32]. And for those who experienced a dislocation and managed to complete the season, surgery can be delayed but not essential for everyone [32, 33].

Data of return to play also demonstrated a statistically significant difference between arthroscopic and conservative treatments after first-time traumatic anterior shoulder dislocations. As the majority of study population are young and active, returning to play seems to be a priority. But return to play not necessarily means no recurrence or successful treatment, as we noticed that varied standards were adopted among included studies. Some of them provided precise number of patients who returned to previous level of sports [21, 23], while some have no clear statement regarding whether the patients returned to activities of previous level or not [3, 22]. The same is true of the composition of participants, as only a small part of the studies [15‐17, 22] enrolled pure athletes as research objects. The consistency among enrolled studies was not satisfactory due to above mentioned issues, and this led to a certain decrease of effectiveness of the statistically significant difference in the result of return to play. A uniform definition of return to play would have lend greater credence to the significance of our finding, and this need to be ameliorated in future studies.

We pooled available data regarding the function assessment. Final results concerning DASH score, (Western Ontario Shoulder Instability Index) WOSI score, Rowe score, ASES score, Constant-Murley score did not reach statically significant differences between groups. We also noticed that inconsistency (I²) of the results of functional scores among enrolled studies were too big to make an accurate comparison and solid conclusion. Additionally, these measurements include subjective items which might make it difficult to distinguish the superior treatment, especially when number of enrolled patients was relatively small.

Pooled result of subsequent instability surgery after primary anterior shoulder dislocation favored ABR without any suspense. The rate of recurrent instability was much higher after conservative treatment. It is reasonable that more additional surgery would be performed after primary treatment in that group. One major cause is that instability would cause inconvenient symptom in a quite portion of the patients and prevent them from participating in sports.

Functional evaluation failed to determine the better treatment. Number of studies that adopted functional evaluation as major measurement was relatively small. The reason we supposed was that dislocation caused varied trauma to the integrity of bony and soft structure, it might be difficult to apply a functional evaluation under various precondition.

Surgical timing is a crucial issue in handling shoulder instability. Duchman et al. found that glenohumeral bone and cartilage lesions were common findings at the time of both primary and revision shoulder stabilization surgery, and he appealed for further prospective study to compare the clinical results in these two groups [34]. Palth et al. pointed out that the age and extent of trauma sustained during preoperative dislocations were more relevant to long-term dislocation arthropathy than the kind of treatment [27]. Their study found that osteoarthritic changes at an average 13 years after ABR were comparable with that conservative treatment. Another radiologic evaluation and self-assessment study from Finland by Kavaja et al. indicated that incidence of glenohumeral arthrosis after ABR was quite common while the symptoms were generally mild and comparable to conservative treatment [26]. Though ABR can yield a more stable shoulder, these evidences indicated that long-term dislocation pathology was comparable between ABR and conservative treatment. Additionally, study by Barlow et al. failed to find a significant difference regarding recurrence rate in patients who had primary surgical stabilization after a single episode of dislocation compared with those experienced recurrent instability events [35]. And this further confirmed Plath’s view that the extent of trauma energy at primary dislocation was more relevant to recurrence [27]. It will be very interesting if more evidence on the relationship between traumatic energy and prognosis of shoulder dislocation could be disclosed in the future. More studies on long-term pathologic changes following both treatments should be present to provide more guidance for selecting the treatment.

The study has several limitations. Arthroscopic lavage was adopted in a conservative group in one study [3], whereas we pooled its result with other conservative ones. Previous studies [36, 37] revealed that arthroscopic lavage speeded reduction in effusion in the glenohumeral joint and lowered recurrence after primary anterior shoulder dislocation, however, its efficiency was deemed to be limited when compared with ABR. Sensitivity analysis also verified this. For post-reduction management, the immobilization position and duration were varied across the studies, and this is worthy of further unification in future studies. So is the definition of return to play, the unification of which would lend greater credence to the significance of pooled results. What’s more, adequate randomization was not achieved by most studies and this could reduce the reliability of the result. But so far, the pooled trials in this study presents the highest level of evidence available.

When handling first-time traumatic anterior shoulder dislocation in young active population, ABR demonstrated superiority over conservative management in terms of recurrence, return to play, and subsequent instability surgery. However, it is necessary to distinguish non-athletes from athletes as long-term prognosis is comparable. An immediate surgical stabilization might not be a routine approach for everyone. Specific concern on patient anticipations, pathologic changes, future sport demands and gender should be addressed when making the decision of surgery.