The key to accurate diagnosis, evaluation, and appropriate treatment begins with a detailed history to gain insight into the patient’s mechanism of injury and treatment goals. Important factors include circumstance of injury (during sport activity or not), type of sport, level of competition, position of play, timing during the season, and patient expectations or goals regarding return to play. Additionally, it should be determined if the patient has experienced a similar event in the past.

On physical exam, in addition to inspection, palpation, range of motion (ROM), and strength testing, several special tests can be utilized to evaluate shoulder instability. The apprehension/relocation and anterior load and shift tests are often recommended, and the mid-range apprehension test may indicate the presence of clinically significant glenoid bone loss [16]. Posterior and multidirectional instability can occur concomitantly with anterior instability. Therefore, these pathologies should be evaluated with Kim and Jerk, posterior load and shift, and sulcus tests. The Beighton score test, a clinical tool to screen for joint hypermobility [17], should be considered if there is concern for hyperlaxity or connective tissue disorders. Although a score less than 5/9 does not preclude an individual from hypermobility, additional investigation may be warranted if the score is 5/9 or greater.

Imaging evaluation should begin with radiographs. Obtain anteroposterior, axillary, scapular Y, and Grashey views to assess the acromion and glenohumeral joint. The use of advanced imaging helps to visualize soft-tissue pathology and assist with surgical planning. Anterior instability can cause damage to multiple structures, including the anterior-inferior glenoid labrum (Bankart lesion), anterior glenoid (bony Bankart lesion), inferior glenohumeral ligament (humeral avulsion of the glenohumeral ligament), and posterosuperior humerus (Hill-Sachs lesion) [1]. When focusing on patients aged 12–21 years with recurrent instability, Hill-Sachs lesions were present in 89% of patients who underwent revision stabilization and in 71% of patients who did not undergo further surgical stabilization [13].

Differences in injury patterns within the pediatric population exist and can be appreciated on imaging. In cases of pre-pubertal instability, cartilage injury can be observed or structural damage to the labrum, capsule, or ligaments may be absent. For example, when evaluating adolescents with traumatic anterior dislocations, Cannamela et al. [6] found that bony injuries such as bony Bankart and Hill-Sachs lesions were more common in older adolescents (≥15 years). Specifically, bony Bankart lesions were present in 34.2% vs. 18.2% and Hill-Sachs lesions were present in 89.5% vs. 74.6% of individuals 15 years or older compared to those younger than 15 years [6]. Conversely, younger (<15 years) adolescents were more commonly females diagnosed with an anterior labral periosteal sleeve avulsion (23.6% of <15 year olds vs. 10.5% of ≥15 year olds) or other atypical lesions [6].

Additionally, relevant measurements can be performed on advanced imaging. Glenoid index, the ratio of glenoid height-to-width, can be measured on MRI and may be a useful tool to evaluate dislocations, including in adolescent patients [10]. Specifically, glenoid height is measured in the sagittal oblique plane from the superior glenoid tubercle to the inferior glenoid, and the width is measured at the widest point perpendicular to the height axis [10]. In one cohort comparing known dislocations to normal controls, glenoid indices ≥1.45 were 83% sensitive and 79% specific for predicting subsequent dislocation [10]. While that retrospective study excluded patients with bony injury or glenoid bone loss, MRI was obtained after the first dislocation event for most patients. Therefore, it is unclear if an increased glenoid index is a predisposing factor or a result of traumatic dislocation. MR arthrogram is another modality used to assess soft tissue lesions, although it is typically reserved for the revision setting or when 3T (3 Tesla) MRI is unavailable. This MRI method is preferred due to its reliability for assessing soft tissue lesions.

Recurrent instability episodes can result in increased glenoid bone loss. Three dimensional (3D) computed tomography is regarded as the gold standard for evaluating bone loss, which can change operative management. While it is important to assess for bone loss because of its effects on management, it must be weighed against the risks of radiation exposure. In a cohort of patients with single or recurrent anterior shoulder dislocation and a mean age of 31 years, glenoid bone loss was present in 41% of single and 86% of recurrent dislocations [18]. Of those with glenoid bone loss, 51% had <10% loss, 37% had 10%–20% loss, 6% had 20%–25% loss, and 6% had >25% loss [18]. In a retrospective review, Ellis et al. [16] found glenoid bone loss in 48.2% (55/114) of adolescent patients with recurrent traumatic glenohumeral instability evaluated with a combination of imaging modalities. Of 55 patients with glenoid bone loss, 27% (15/55) had critical bone loss, which was defined as “>20% of the glenoid during arthroscopy, or in cases where the treating provider performed an open procedure for glenoid bone loss [16].” Older, taller, and male patients who sustained primary dislocation while playing sports had increased odds of glenoid bone loss [16].

In the adolescent population, normal age-related ossification, fusion, or variants can mimic instability-related pathology. Sidharthan et al. [7] utilized 3D, frequency-selective, fat-suppressed spoiled gradient recalled echo MRI sequences to define the pattern of adolescent shoulder ossification and fusion centers with relation to age. This MRI technique was chosen due to the ability to assess physeal closure both quantitatively and qualitatively in skeletally immature individuals. Their study suggests that these growth centers follow a predictable pattern, often occurring earlier or in narrower age ranges for females and with the anterior glenoid lagging behind the coracoid and superior glenoid [7]. For example, all coracoids and 60% of the superior glenoids were ossified in the 9-year-old individuals [7]. However, ossification of the anterior glenoid rim occurred between 11 to 17 years of age for males versus 11 to 12 years for females [7]. Similarly, fusion of the anterior glenoid rim by age 17 years in males and age 13 years in females followed that of the superior glenoid at 14 years in males and 11 years in females [7]. Other authors caution against relying solely on imaging findings because secondary ossification may not be fully completed and not all facilities use highly specialized imaging techniques [19]. Therefore, it is imperative to consider imaging findings in conjunction with clinical date when deciding on management options, especially if there is potential for the findings to reflect normal development.

Similarly, attention should be paid to the timing and pattern of humeral head ossification. The MRI-based study of Kelly et al. [20] focused on patterns of development of the humeral head and reported that maturation occurred predictably. The authors noted younger patients to be at the highest risk of a false Hill-Sachs diagnosis, specifically girls aged 4–7 years and boys 5–14 years in earlier stages of ossification [20]. This again suggests that females have a younger and narrower age range for normal ossification.

Due to the increased risk of recurrence in young, highly active populations and individuals with glenoid bone loss, surgical stabilization should be discussed as an initial treatment option. Counseling should include the potential for repeat instability events. Some studies of early adolescents suggest that repeat dislocations are uncommon in the absence of structural abnormalities [23,24,26]. However, Leroux et al. [5] found that 19.6% of the 10–16-year-old patients from their study pursued surgical stabilization after primary dislocation. They suggested that this may be due to persistent apprehension causing an inability to return to sport, ongoing subluxation or self-reduction events, or a bias toward surgery in patients identified as high risk [5]. These factors should also be considered when counseling a patient and family.

In the survey mentioned above, consensus for operative treatment was reached for patients with closed or closing physes who sustained a first-time, contact-related dislocation [22]. Regardless of physeal closure or mechanism of injury, consensus favored arthroscopic stabilization in recurrent instability cases [22]. These survey results reflect a trend in management toward early operative intervention for some patients due to the risks of glenoid bone loss with recurrent instability [9,12,27-29].

There are multiple surgical options for stabilization, either to reduce the risk of recurrent instability and associated sequelae or after failing nonoperative management. These include arthroscopic or open labral repair with or without capsulorrhaphy and bone block augmentation procedures such as Latarjet, distal tibial allograft, and various other autograft transfers.

Arthroscopic Bankart repair is frequently utilized and studied as a less invasive procedure that can provide adequate stability. In a retrospective review of 15- to 20-year-old patients with recurrent anterior instability, Monk et al. [30] reported clinical outcomes following open stabilization for collision athletes, mostly rugby players. The authors found that open Bankart repair resulted in lower re-dislocation rates and higher return to play rates compared to published results for arthroscopic stabilization procedures. They also concluded that athletes typically have inferior surgical outcomes compared to nonathletes, and collision athletes return to play at lower rates than non-collision athletes [30]. However, other studies suggest that the higher rates for recurrence with arthroscopic stabilization were due to early technique flaws and implant designs, which have since improved [31,32]. A comparison of arthroscopic and open Bankart repairs in 99 patients found no significant difference in recurrence, secondary surgery rates, or patient reported outcomes [32].

Abdel Khalik et al. [33] separately conducted a systematic review and meta-analysis of cohort or comparative studies of first-time anterior shoulder dislocations, including 34 studies and 2,222 shoulder dislocations, with a mean patient age of 25.4±5.4 years. Their age-focused subgroup analysis of arthroscopic Bankart repairs after a single dislocation event revealed higher rates of cumulative instability (20.9%) and subsequent surgery (9%) in patients younger than 20 years compared to those older than 30 years (11.5% and 4.5%, respectively) [33]. In a systematic review, Shanmugaraj et al. [15] identified 24 studies with 696 adolescent shoulders that offered a variety of treatments and timing of surgical intervention. The majority of shoulders (n=525) underwent arthroscopic Bankart repairs, with the remainder experiencing open Bankart (n=75), modified Bristow (n=34), and Latarjet (n=26) repair. The authors ultimately concluded a benefit of surgical stabilization for first-time dislocations with known risk factors for recurrence. In addition, the authors reported no significant differences in outcomes of one study that compared arthroscopic and open Bankart repair [15]. Although they did not expand directly on factors that contributed to this lack of a difference, limitations of the systematic review included statistical and methodological heterogeneity and poor data documentation, which prevented performance of a meta-analysis [15]. On the other hand, Rosello et al. [27] reported on primary arthroscopic Bankart vs, arthroscopic Bristow-Latarjet with Bankart repair procedures in a retrospective review. The recurrence rate was 8% in the Bristow-Latarjet with Bankart repair group compared to 22% in the arthroscopic Bankart group. The four repeat dislocations in the primary arthroscopic Bankart group were repaired using the Bristow-Latarjet with Bankart procedure. The Bristow-Latarjet with Bankart repair group experienced failure earlier than the primary arthroscopic Bankart repair group, at about 2 years postoperatively. Following multivariate analysis, they concluded that patients with more than three preoperative episodes of instability and hyperlaxity had higher failure rates following primary arthroscopic Bankart repair and should be indicated for a bone block augmentation instead.

The glenoid track concept for Hill-Sachs lesions has been expanded in the literature, with some authors shifting toward terms such as “near track” to describe on-track lesions that have short distance to dislocation (DTD). In Li et al.’s retrospective analysis of 173 patients with a mean age of 20.0±6.3 years who underwent arthroscopic Bankart repair [34], 28 (16%) experienced recurrent instability postoperatively. Greater glenoid bone loss, larger Hill-Sachs lesions, smaller DTD values, and >1 instability episode preoperatively were associated with failure. Additionally, younger age was associated with increased DTD values. Subgroup analysis for individuals <20 and ≥20 years old revealed ideal DTDs to predict failure of 10 mm and 4 mm, respectively. However, the model was not as accurate in the younger age group and might not be as clinically useful until validated in larger cohorts. The “near track” and DTD concepts could be useful for treating adolescents, as their younger age and higher activity is thought to place individuals with shorter DTD at higher risk of treatment failure.

There have also been investigations on the utility of adding primary remplissage to an arthroscopic labral repair for individuals with Hill-Sachs lesions. Hughes et al. [35] performed a retrospective cohort study on adolescents with up to 20% glenoid bone loss who underwent arthroscopic Bankart repair with remplissage for recurrent anterior shoulder instability and compared them to a matched cohort undergoing only arthroscopic Bankart repair. The remplissage group included a mixture of primary and revision procedures following a failed Bankart repair. While there was no difference in patient-reported outcome score or ROM, the Bankart-only patients experienced higher rates of recurrence compared to the remplissage group (47% vs. 13%). The authors concluded that remplissage can provide additional stability in adolescents with a Hill-Sachs lesion and <20% glenoid bone loss [35].

Glenoid bone grafts and coracoid transfer procedures are typically reserved for individuals with >15% glenoid bone loss or who have failed a prior stabilization procedure. Coracoid transfer procedures include the classic Latarjet and modifications, such as the Bristow and congruent arc techniques. Bone graft options include iliac crest autograft, distal tibial allograft, and distal clavicle autograft [36]. While the most concerning complication following Latarjet is nerve injury, potential risks specific to the adolescent population include damage to the physis, transferred bone affecting future growth in skeletally immature patients, and the earlier development of arthritis. These risks ultimately can affect the long-term morbidity and functionality of the shoulder.

There is some debate pertaining to primary bone block augmentation procedures in adolescents. In a retrospective comparative cohort study of adolescents aged 13–18 years who underwent an arthroscopic Bankart repair or open Latarjet procedure, Waltenspül et al. [29] concluded that the Latarjet group had significantly longer time to treatment failure at final follow-up. Specifically, 20 of 35 (57%) failed Bankart repair after a mean of 4 years (range, 0.5–13 years), while 2/31 (6%) experienced failure of Latarjet at a mean of 6 years (range, 3.5–8.5 years). Failure in their study was defined as recurrent instability after a stabilization procedure or persistent apprehension resulting in revision surgery. Additionally, they found a reoperation rate of 43% in the Bankart group at a mean of 4.7 years (range, 0–12 years) compared to 13% in the Latarjet group at a mean of 2 years (range, 1–4 years). In the Bankart group, 37% of reoperations were due to recurrent instability; that in the Latarjet group was 3%. After undergoing open Latarjet as revision for a failed Bankart repair, two shoulders had persistent instability. In addition, the Latarjet group had a 13% rate of complications that did not require reoperation, compared to 6% in the Bankart group [29].

Clinically, Domos et al. [37] found that skeletally immature patients with recurrent anterior instability and soft tissue Bankart or bony lesions on computed tomography arthrogram with any size Hill-Sachs lesion had high rates of high functional outcome scores, return to play, and satisfaction after open Latarjet. Postoperatively, no patients required a second operation for or radiologic evidence of bone block complications, including those who had mild (grade I) postoperative OA at a median follow-up of 6.6 years. The authors reported a similar recurrence rate (4%) to adult populations (11%); however, there was lower overall recurrence of instability compared to arthroscopic Bankart repairs in young populations [13]. Domos et al. [37] also found lower rates of postoperative arthritis (9%) than previously reported Hyperlaxity, female sex, and large or deep Hill-Sachs lesions were frequently associated with persistent apprehension, but none achieved significance. However, when present, persistent apprehension was associated with inferior clinical results [37].

Other surgeons argue that the arthroscopic Bankart repair can be a reasonable first-choice option in non-athletes and patients without significant bone loss, and that revision Latarjet remains possible in the case of failed arthroscopic stabilization [38]. Coracoid anatomy may affect hardware placement, with smaller grafts yielding a higher potential for fracture or bone healing complications [38]. For high-level athletes, collision athletes, significant bone loss, and off-track lesions, there is greater consensus for a Latarjet procedure in a primary setting [38].

Recurrent instability remains a risk following surgical stabilization procedures. Cheng et al. [9] reported a 20.5% failure rate at a mean 1.2 years post-arthroscopic stabilization in patients younger than 18 years. In their age- and sex-matched cohort analysis, increased glenoid bone loss, decreased glenoid retroversion, and more than one instability event before surgery were significant risk factors for recurrent instability. Notably, risk of recurrence was three times greater in patients with two risk factors and four times greater with all three risk factors [9]. Additionally, open proximal humeral physis was a significant independent predictor of recurrent instability on multivariate analysis [9]. Labrum tear size, number of anchors used in stabilization, collision sport participation, and size of Hill-Sachs lesions were not significant factors for failure. In contrast, Egger et al. [39] conducted a retrospective study evaluating clinical and radiographic predictors of failure in individuals <19 years old who underwent arthroscopic Bankart repair. The authors found a revision rate of 17%, reporting larger anterior labral tears in patients who experienced surgical failure. Of the 59 patients, 38 had a measurable Hill-Sachs defect, seven of which underwent revision stabilization surgery and four remained subjectively unstable without additional procedures. Larger Hill-Sachs intervals were relevant in subgroup analysis of the seven revision cases compared to 31 who did not undergo revision (20 mm vs. 14 mm). The authors also concluded that off-track instability was not a helpful factor in their study as it affected only four patients. Last, patient-reported outcome measures did not statistically differ among those who did or did not undergo revision surgery; however, compared to the rest of the cohort, patients with persistent subjective instability without revision had significantly lower scores [39].

Bone block augmentation can also be used in the revision setting. Rosello et al. [27] used the Bristow-Latarjet technique for those who failed initial arthroscopic Bankart repair. The authors also used iliac crest bone autograft with Bankart repair and remplissage in their revision stabilizations for patients who experienced recurrent instability following Bristow-Latarjet with Bankart repair [27]. Waltenspül et al.’s comparative study of arthroscopic Bankart vs. open Latarjet procedures [29] found no significant difference in failure rates of primary vs. secondary open Latarjet procedures. They noted that four shoulders in the Latarjet group had failed Bankart repairs, two of which had undergone a prior stabilization procedure [29]. The lack of significant difference in failure rates of primary vs. secondary Latarjet is interesting, as revision cases typically do not demonstrate the same level of success as primary Latarjet procedures. While some surgeons have used a distal tibial allograft for reconstructions, no studies analyzing its use have been performed solely in adolescents.

Desire and timing to return to sport are critical considerations in the management of anterior instability patients and may vary depending on sport, position, individual postoperative course, and surgeon preference. Studies often report return to sport as a measure of treatment success. For example, Shanley et al. [40] compared return rates in high-school patients treated nonoperatively and operatively. Failure was defined as the inability to play the next full season or experiencing any time loss due to an upper extremity injury [40]. Following nonoperative intervention, patients with dislocations were more likely to fail to return to sport compared with those experiencing subluxations. However, in those treated operatively, instability severity demonstrated no statistical difference in the rate of return to sport [40]. Additionally, of those who failed nonoperative treatment, 82% were able to successfully return to sport after undergoing surgical stabilization [40]. As a result, the authors suggest that patients can expect reasonable success with surgery after failing nonoperative treatment.

Augmentations to nonoperative treatment have also been investigated. Kwapisz et al. [41] investigated the role of functional bracing in a cohort of high-school athletes, defining success as completion of the current season and one subsequent season without surgical intervention. They found no statistical difference in success for braced vs. unbraced athletes (mean age, 15.8 ±1.4 years), including for a subgroup analysis of football players [41].

Operative studies typically focus on recurrence of instability as the definition for failure, which often affects the athlete’s ability to return to play. Rosello et al. [27] investigated the return to sport rates between patients stabilized with arthroscopic Bristow-Latarjet with Bankart repair versus arthroscopic Bankart repair. The Bristow-Latarjet with Bankart repair group had a significantly higher return rate at 79%, compared to 47% in the primary arthroscopic Bankart repair group (P≤0.001) [27]. The arthroscopic Latarjet procedure is technically challenging, and success rates different by surgeon experience. On the other hand, Harada et al. [42] found that all of the competitive teenage athletes who underwent arthroscopic Bankart repair in their study were able to return to sport, with 76% returning to pre-injury level. Contact athletes had an 82% complete return to sport rate, whereas 59% of overhead athletes and 100% of noncontact-nonoverhead athletes returned [42]. Notably, the athletes who did not completely return cited instability, anxiety to play fully, slight pain, restriction of motion, or vague discomfort as factors [42]. Ozturk et al.’s prospective study of 53 patients (42 male, 11 female) who underwent arthroscopic capsulolabral repair [31] found an overall return rate of 86% for male patients and 89% for female patients. In total, 75% of patients returned to pre-injury level, including 15 of 22 contact athletes [31]. They found no difference in rates of recurrence or return to sport to pre-injury level in their age-focused analysis of individuals younger than 20 years vs. those 20–24 years old [31]. The athletes in their cohort cited caution, weakness on the affected side, and increased pain with overhead activity as reasons for not returning to sport despite being able to resume daily activity without limitations and denying further instability events. Anxiety, caution, and perceived differences in the operative shoulder were mentioned by Harada et al. [42] and Ozturk et al. [31] as potential psychological factors that can also contribute to patient ability to return to sport.