Anterior glenohumeral instability secondary to a traumatic shoulder dislocation is a common problem that plagues both athletes and non-athletes of all age groups. Dislocations, which are significantly more common in men, represent about 4% of all high school sports injuries, with shoulder dislocations accounting for over 50% of all dislocations. The recurrence rate for a shoulder dislocation in the young athlete is between 50-80%, with approximately 5-30% of acute anterior dislocations showing evidence of a glenoid rim fracture (bony Bankart lesion) (Figure 1). Patients under the age of 20 who sustain a shoulder dislocation often have an associated Bankart lesion (Figure 2), while patients over 60 will likely have a concomitant rotator cuff tear.
Acute glenohumeral instability typically has a classic presentation. Patients will present after having sustained an acute anterior dislocation event in which they either spontaneously reduced, reduced themselves, or went to the emergency department and were reduced there. They will complain of some pain and weakness in the arm, but should not have any neurological deficits (although axillary nerve palsy, especially the sensory branch which innervates the skin over the lateral deltoid, is possible). Older patients (greater than 60 years old) may complain of weakness in the arm secondary to an associated rotator cuff injury. Following the event, the patient’s main complaint will be a feeling of instability when their shoulder is brought into the “at-risk” position of 90 degrees of both abduction and external rotation. Often times they will not be able to sleep on the affected shoulder. They may also complain that the shoulder has almost popped out a few times since the incident, leading the examiner to believe they have experienced several subluxations.
The physical exam begins with inspection of the shoulder, looking for muscle wasting or gross malalignment. Palpate the acromioclavicular (AC) joint, bicipital groove, and anterior and posterior joint lines. Next, focus on range of motion (ROM). Test ROM in forward elevation, extension, abduction, and external/internal rotation both with the patient’s hands at their sides and with their shoulder abducted to 90 degrees. Then, evaluate strength; isolate the subscapularis, supraspinatus, infraspinatus, teres minor, and deltoid. Compare all ROM and strength exams to the contralateral side. Finish with shoulder stability testing. (There are many tests for shoulder instability including the hyperabduction test, anterior and posterior drawer, sulcus sign, etc., but for brevity purposes we will review the following two tests.)
The patient lies supine with their shoulder off the edge of the examination table and the scapula stabilized; the shoulder is brought into abduction and external rotation, both at 90 degrees. In this position the patient becomes apprehensive – reporting that the shoulder feels like it is going to dislocate. The same maneuver is then performed except now with a posteriorly-directed force applied over the anterior glenoid/humeral head. If the patient reports less apprehension this is considered a positive test.
Load and shift test
The patient is supine with their shoulder off the edge of the examination table and the scapula stabilized. The arm is in 90 degrees of abduction. With one hand the examiner provides an axial load to the shoulder while the other hand translates the humeral head anteriorly and posteriorly. The amount of humeral head translated is graded (grade 0: minimal movement/firm endpoint; grade 1: head is translated onto labrum; grade 2: head is translated over glenoid but spontaneously reduces; grade 3: head is translated over glenoid and remains dislocated once force is removed). Load and shift testing is often useful in the operating room while evaluating the patient under anesthesia.
Imaging studies should include standard anteroposterior (AP) of both internal and external rotation, axillary, and scapular Y views. Stryker Notch and West Point axillary views should also be performed to best evaluate for Hill-Sachs (Figure 3) and bony Bankart lesions respectively.
A shoulder computed tomography (CT) exam (Figure 4) including three dimensional reconstructions (Figure 5) is useful in evaluating the shape of the glenoid and to quantify glenoid bone loss. This can also aid in evaluating for a Hill-Sachs lesion. A normal glenoid will appear as an upright pear on CT scan. If the patient has sustained a bony Bankart lesion (anteroinferior glenoid bone loss) after their dislocation, the glenoid shape will appear as an inverted pear (wider superiorly than inferiorly; correlates with bone loss, often greater than 25%) (Figure 1). The amount of bone loss can be quantified as a percentage of the total surface area of the glenoid, with more than 6 mm of bone loss (equating to 25% bone loss) leading to an increased likelihood of recurrent dislocations (normal width of the glenoid is approximately 24 mm, making the radius 12 mm, and ½ the radius 6 mm). We often will perform a CT scan of the contralateral shoulder if there is a question about the patient’s native glenoid shape.
Magnetic resonance imaging (MRI) can be useful to evaluate the capsule and labrum. If this is done acutely, whatever blood is present from the initial trauma will act almost as contrast and can aid in visualization. If done in the sub-acute or chronic setting, an arthrogram may be considered to better evaluate the labrum, although this is often unnecessary. Frequently the capsule will be stretched and the labrum may be torn or displaced.
In most instances, patients who present as first time dislocators should initially be treated non-operatively. The exception is young athletes (less than 20 years old) who are playing at-risk sports (basketball, football) and are at the end of their season or in the off-season, and would like surgery so they are ready to play the following season. Also, patients with a significant amount of glenoid bone loss should be counseled that they have a significantly higher risk of recurrent instability.
Immobilization after the initial dislocation, either in internal or external rotation, has not been proven to decrease the number of subsequent dislocations. Hence, our protocol is a sling initially for comfort followed by shoulder movements at or below the horizontal. Patients should avoid excessive abduction and external rotation or any other position that makes their shoulder feel unstable. They should avoid push-ups and bench presses initially until the capsule has time to scar down.
Indications for Surgery
Indicating a patient for a stabilization procedure must involve a discussion between the patient and treating physician as there is no set algorithm for who should and should not have surgery. Patients who have failed non-operative management should be considered for surgery, and based on the amount of bone loss (either humeral or glenoid), should be consented for an arthroscopic stabilization versus open bony stabilization (i.e. Latarjet). Typically if there is more than 20% bone loss on the glenoid as evidenced by the inverted pear shape on CT scan, then a bony procedure (Latarjet) is recommended as opposed to a soft tissue procedure secondary to the increased rate of failure with the soft tissue procedure.
Fortunately the Instability Severity Index Score was developed by Balg in 2007 to help surgeons determine if patients would benefit more from a soft tissue or open stabilization procedure. The score consists of 6 categories with a maximum score of 10 (See Table I).
A patient with a score >6 will have a 70% failure rate with a soft tissue procedure alone and should therefore undergo an open bony stabilization (Latarjet). However, patients with a score of <6 have a 10% rate of recurrent instability with an arthroscopic stabilization. Arthroscopic Bankart repairs are more common in essentially every country except France, where open procedures are the common treatment of choice; in the United States between 2004-2009, arthroscopic procedures made up 84% of shoulder stabilization cases while open procedures made up 16%. Rates of recurrence for the open technique have been quoted at a mean of 7% while that of the arthroscopic technique is 9%.
High-level athletes who are nearing the end of their season or are in the off-season are good candidates for surgery as they can rehab and be ready to play before the start of their next season. If they participate in a sport that is high contact and are competing at an elite level, consideration must be given to an open procedure compared to an only arthroscopic procedure.
Patients who are 20 years of age or under should strongly consider surgery as the likelihood for recurrent instability is significantly higher than patient in their 30s-40s (80-90% versus <50%).
There are numerous surgical techniques for anterior stabilization procedures. These can be broken down into soft tissue and bony stabilizations. For the purposes of this review we will focus on the arthroscopic anterior stabilization and Latarjet procedures, although open anterior stabilization, distal tibia/iliac crest allograft, and the Bristow procedure are all viable alternatives.
We prefer to perform our arthroscopic Bankart repairs in a lateral decubitus position, with the unaffected shoulder down on the bed and the patient under general anesthesia with or without a supplemental nerve block.
The patient is placed on a beanbag, which is then contoured and hardened to ensure proper stability.
Care must be taken not to cover the posterior portal site with the beanbag.
We use a shoulder distraction system (STaR Sleeve and 3-Point Shoulder Distraction System, Arthrex, Naples, FL.) to improve visualization with 5-10 lbs used for balanced suspension.
The arm is examined under anesthesia with the load and shift test.
After hanging the arm from an intravenous (IV) pole and then sterilely prepped, the arm is draped sterilely and placed into the distraction system sterilely as well.
The arm is in 20-30 degrees of abduction and 20 degrees of forward elevation in this system.
The arthroscope, along with suction, inflow, outflow, radiofrequency device, and power for a burr or shaver are readied.
The posterior portal is established by marking out the “soft spot” – 2 cm medial and 2 cm inferior to the posterolateral edge of the acromion.
0.5% marcaine is injected into the proposed path of the needle.
An 18 gauge spinal needle is passed into the joint, feeling the humeral head and glenoid to confirm placement.
Thirty cc of normal saline is injected into the joint.
The stylet is removed and fluid should begin to leak out of the needle to confirm intra-articular placement.
A skin incision is made, the needle is removed, and a trocar is advanced into the joint, using the same trajectory as the needle, aiming for the coracoid, twisting as one applies pressure.
Once intra-articular, the scope is inserted and the water is turned on to a pressure of 60 mmHg.
Under direct visualization the anterosuperior portal is established with a spinal needle approximately 2 cm lateral to the coracoid and 1 cm anterior to the acromion (ideally the needle should be entering the joint near the biceps attachment).
A 6.5 mm cannula is inserted into this portal.
The anteroinferior portal is established next in a similar manner with the needle placed just above to the superior aspect of the subscapularis.
An 8.25 mm cannula is placed in this portal.
Care must be taken to ensure that these two portals are far enough apart to prevent portal crowding.
A diagnostic arthroscopy is performed with specific attention paid to the superior and posterior labral attachments (which, if torn, are fixed prior to the anterior stabilization).
The anteroinferior labrum is examined.
If it is torn (Bankart lesion) an arthroscopic elevator is used to mobilize the capsulolabral tissue from the glenoid (the subscapularis tissue should be identified underneath the mobilized tissue) back to the glenoid rim (Figure 6).
A bone cutting shaver or burr is used to create a bleeding bone bed on the glenoid rim.
The anchors will be passed through the anteroinferior portal.
An arthroscopic drill creates the pilot hole and the suture anchor is placed into the hole.
A braided wire is passed through the capsulolabral tissue using a shuttling device with the bite taken inferior to the site of the anchor insertion to allow for an inferior capsular shift in addition to fixation of the torn anteroinferior labrum.
This wire is brought out the anterosuperior portal.
The first suture is retrieved through the anterosuperior portal and passed through a half-hitch thrown in the braided wire.
It is then shuttled across the capsulolabral tissue.
This is repeated for the other sutures and potentially more anchors based off the size of the defect.
Arthroscopic knots are tied with stacked simple half hitches and cut.
The labrum is now repaired back to the glenoid (Figure 7)
The patient is placed in a sling post-operatively.
While a large amount of attention in an arthroscopic Bankart repair is paid to the labrum, it is also important to evaluate the humeral head for a Hill-Sachs lesion which could lead to recurrent instability.
The humeral head lesion should be evaluated to determine if it “engages” the glenoid as this lesion would necessitate surgical intervention. Intra-operative testing with ROM of the shoulder in flexion, abduction, and external rotation will help elucidate the severity of the Hill-Sachs defect. Multiple options exist for addressing this humeral lesion. These include open capsular shift, humeral head bone augmentation, and remplissage. The theory behind a capsular shift is that by tightening down the capsule one limits external rotation and anterior translation of the humeral head, preventing the Hill-Sachs lesion from engaging and encouraging an instability event.
The remplissage procedure has undergone many modifications and is performed differently by different surgeons. The purpose is to fill in the humeral head defect with posterior capsule and the infraspinatus tendon, similarly limiting humeral head external rotation and anterior translation preventing Hill-Sachs engagement.
Our Latarjet procedure is performed with the patient in the beach chair position.
Pre-operatively the patient often receives a regional nerve block.
Once in the operating room the patient is placed under general anesthesia.
We then place a foam bump under the legs of the patient so the knees are flexed.
The bed is maneuvered into the beach chair position.
The patient’s head is secured using a foam pad and tape to ensure it does not lean to either side.
The patient is positioned such that the operative shoulder is free from the bed, but such that their scapula is still stabilized by the bed.
An axillary role is placed behind the scapula of the operative shoulder.
A lateral post is used to prevent the patient’s body from leaning off the bed.
The shoulder is brought through ROM tests, and a load and shift test is performed to assess glenohumeral mobility.
Two U-drapes are placed on the patient, one from bottom up and one from top down.
The arm is hung from an IV pole and the arm is prepped sterilely.
The patient is draped in a standard sterile manner (we like to drape a second mayo stand to allow the arm to rest on this).
The deltopectoral interval is marked out and an incision is made from the tip of the coracoid down to the axillary fold (approximately 7 cm).
Full thickness skin flaps are elevated and care is taken not to damage the cephalic vein.
Deep retractors are placed under the deltoid and pectoralis major muscles, with a third army-navy retractor placed superiorly to aid in visualization (Bovie tip is changed to needle tip).
The coracoid process is exposed by incising the clavipectoral fascia; one must ensure good exposure (you’ll need to dissect the pectoralis minor fibers off the coracoid and reflect them medially) prior to performing the osteotomy.
With deeper exposure, use a Kolbel retractor to hold the deltoid and pectoralis major.
Once the coracoid is exposed mark the distal tip with the Bovie and measure 20 mm proximal to this to ensure a 20 mm bone plug is harvested.
Use a 90 degree oscillating saw for the osteotomy (must place retractor inferiorly and medially to protect the neurovascular structures – cut from medial to lateral).
Free the coracoid by using a Bovie to dissect off any remaining soft tissue and place two drill holes through the center of the coracoid with a 3.2 mm drill bit.
The glenohumeral joint is exposed with a subscapularis splitting (split into 50/50 halves) approach.
It is imperative not to damage the inferior subscapularis as this will facilitate early motion post-operatively.
The anterior capsule is dissected away from the subscapularis bluntly or with a needle tip Bovie.
The capsule is incised vertically, ensuring adequate tissue medially to repair to the coracoacromial (CA) ligament laterally.
The medial glenoid is exposed by dissecting the capsule from it.
A burr is used to clean any soft tissue off of and to decorticate the coracoid and glenoid to enhance bony healing.
The coracoid bone block is then provisionally fixed to the glenoid with a K-wire (to aid with contouring the shape of the coracoid to that of the glenoid, the coracoid is rotated on its side with the tip facing inferiorly in this procedure).
The coracoid should be placed flush with the glenoid rim and below the glenoid equator.
Once in place, the 3.2 mm drill is advanced through the superior coracoid drill hole and the glenoid is drilled bicortically.
A depth gauge measures the length of the screw (usually between 30-40mm) and a partially threaded screw is placed into the hole after placing a washer on the screw.
This step is repeated for the inferior hole and the K-wire is removed.
The lateral capsular flap is repaired to the CA ligament with simple interrupted throws of non-absorbable suture.
The subscapularis is re-approximated with FiberWire using figure-of-eight throws.
The deltopectoral interval is loosely approximated with #1 vicryl, followed by closure of the deep and superficial dermal layer, a running monocryl for the skin.
The patient is placed in a sling post-operatively.
Pearls and Pitfalls of Technique
Ensuring proper mobilization of the capsulolabral complex will facilitate anatomic reduction of the labrum in an arthroscopic Bankart repair.
Pre-drilling the coracoid prior to attempted fixation to the glenoid will allow better control of the bony fragment.
Decorticating the glenoid and coracoid will stimulate bone healing.
Do not hesitate to convert what pre-operatively was thought to be a soft tissue procedure into a bony procedure based on findings at the time of arthroscopy – an attempted Bankart repair in a patient with severe bone loss will lead to recurrent instability.
One cannot assume the musculocutaneous nerve will be out of the operative field; although it often pierces the conjoint tendon 6 cm distal to the coracoid, it can be as close as 2-3 cm, so careful attention must be paid when moving the coracoid and placing retractors.
The contour of the glenoid must be matched with the coracoid transfer; irregular step-offs can cause problems post-operatively including the development of osteoarthritis, so one should adequately visualize the new contour of the glenoid with the coracoid transfer and adjust the coracoid if a step-off is present.
As with any arthroscopic or open shoulder procedure, stabilization procedures carry risks. These include:
Recurrent instability (both subluxation and dislocation)
Loss of external rotation
Other complications such as hematoma formation, non-union, and infection can also occur, but these are much less common.
Patients are initially non-weight bearing through the operative shoulder for 6 weeks.
They remain in the sling for 4 weeks – no pendulums at this time; only light passive ROM (no forward elevation past 90 degrees and no external rotation past 30 degrees).
At 4-6 weeks post-operation patients can perform active assisted ROM with progression to active ROM with forward elevation limited to 140 degrees and external rotation to 45 degrees.
From 6-12 weeks post-operation there is no limit on the patient’s active assisted ROM. They can start light strengthening now but should limit active ROM to below the horizontal.
At 12-16 weeks post-operation patients work on strength and getting back full active ROM with no restrictions.
Full return to sport is allowed at 6 months.
Outcomes/Evidence in the Literature
Arciero, RA, Wheeler, JH, Ryan, JB, McBride, JT. “Arthroscopic Bankart repair versus nonoperative treatment for acute, initial anterior shoulder dislocations”. AJSM. 1994. pp. 589-94. (One of the most highly cited studies from West Point, the authors looked at 36 young, athletic military recruits who had sustained a first time acute, traumatic, anterior glenohumeral dislocation requiring manual reduction. They were divided into two groups based on patient preference: Group 1: non-operative with 1 month of immobilization, Group 2: arthroscopic Bankart repair. They found Group 1 had 80% recurrent instability while Group 2 had 14%. They concluded that in this population, arthroscopic Bankart repair significantly outperformed non-operative management.)
Cameron, KL, Mountcastle, SB, Nelson, BJ, DeBerardino, TM. “History of shoulder instability and subsequent injury during four years of follow-up: a survival analysis”. J Bone Joint Surg Am. vol. 95. 2013. pp. 439-45. (This recent JBJS study looked at young military recruits and attempted to predict if a history of glenohumeral instability was a risk factor for future instability episodes. They followed a class of recruits (710 recruits; 1420 shoulders) for four years. They found that recruits who had a self-reported history of glenohumeral instability at baseline were over five times more likely to sustain an acute instability event.)
Chuang, TY, Adams, CR, Burkhart, SS. “Use of preoperative three-dimensional computed tomography to quantify glenoid bone loss in shoulder instability”. Arthroscopy. vol. 24. 2008. pp. 376-82. (This study looked at whether three dimensional CT scans could accurately quantify the amount of glenoid bone loss in patients with anterior glenohumeral instability and predict pre-operatively which patients would benefit from a bony stabilization procedure. They looked at 25 patients who had undergone a pre-operative CT with three-dimensional reconstruction and attempted to predict, based on estimation of bone loss of greater or less than 25%, if the patient would undergo a soft tissue or bony stabilization procedure. They accurately predicted 24 out of 25 (96%) of the procedures, thereby demonstrating pre-operative CT scans are beneficial in accurately quantifying bone loss.)
Hovelius, L, Olofsson, A, Sandström, B, Augustini, BG. “Nonoperative treatment of primary anterior shoulder dislocation in patients forty years of age and younger. A prospective twenty-five-year follow-up”. J Bone Joint Surg Am. vol. 90. 2008. pp. 945-52. (With one of the longest-term follow ups of any prospective study looking at shoulder instability, the authors published their 25 year data on 227 patients managed with or without immobilization following a primary anterior shoulder dislocation. They showed an overall re-dislocation rate of 57%. They found no difference in the rate of re-dislocation in patients who were immobilized after their dislocation compared to those who were not immobilized. Therefore, they concluded that immobilization is unnecessary after an acute dislocation, as it does not decrease the rate of recurrent instability.)
Rhee, YG, Cho, NS, Cho, SH. “Traumatic anterior dislocation of the shoulder: factors affecting the progress of the traumatic anterior dislocation”. Clin Orthop Surg. vol. 1. 2009. pp. 188-93. (This study examined almost 250 shoulders who sustained an acute anterior shoulder dislocation. They found that the age at the time of initial dislocation (younger than 20) and activity level were reliable predictors in likelihood of re-dislocation.)
Franceschi, F, Papalia, R, Rizzello, G, Franceschetti, E. “Remplissage repair–new frontiers in the prevention of recurrent shoulder instability: a 2-year follow-up comparative study”. Am J Sports Med. vol. 40. 2012. pp. 2462-9. (This study looked at 50 patients with engaging Hill-Sachs and Bankart lesions who underwent either a Remplissage and Bankart repair (25 patients) or an isolated Bankart repair (25 patients) at a mean follow up of two years. There were no statistically significant differences in functional outcome scores or ROM between groups, although both groups improved from baseline. They did note that of the 50 patients, only five had a recurrent dislocation after surgery, and all five of these were in the Bankart only group.)
Randelli, P, Ragone, V, Carminati, S, Cabitza, P. “Risk factors for recurrence after Bankart repair a systematic review”. Knee Surg Sports Traumatol Arthrosc. vol. 20. 2012. pp. 2129-38. (In this systematic review the authors reviewed the literature on outcomes of Bankart repairs for anterior instability to determine pre-operative risk factors for recurrent instability following a Bankart repair. They found that below the age of 22, male gender, participation in competitive sports, and a higher number of pre-operative dislocations all increased a patient’s risk for instability after surgery.)
Shin, SJ, Yun, YH, Kim, DJ, Yoo, JD. “Treatment of traumatic anterior shoulder dislocation in patients older than 60 years”. Am J Sports Med. vol. 40. 2012. pp. 822-7. (This case series examined 67 patients over the age of 60 who sustained a shoulder dislocation and sought to determine the clinical manifestations and functional outcomes of different treatment modalities. They found the most common injury to be a rotator cuff tear, and, interestingly, found that patients treated non-operatively functioned significantly better than those treated with surgery. They also found no re-dislocations in their study.)
Nourissat, G, Kilinc, AS, Werther, JR, Doursounian, L. “A prospective, comparative, radiological, and clinical study of the influence of the “remplissage” procedure on shoulder range of motion after stabilization by arthroscopic Bankart repair”. Am J Sports Med. vol. 39. 2011. pp. 2147-52. (A prospective cohort study comparing Bankart repair to Bankart repair and Remplissage for patients with a Hill-Sachs defect. They found no differences in ROM post-operatively between groups. They did notice that the patients in the Remplissage group did have significantly more posterosuperior pain.)
Yamamoto, N, Muraki, T, Sperling, JW, Steinmann, SP, Cofield, RH, Itoi, E, An, KN. “Stabilizing mechanism in bone-grafting of a large glenoid defect”. J Bone Joint Surg Am. vol. 92. 2010. pp. 2059-66. (This study aimed to determine when a bony procedure needed to be performed for glenoid bone loss. They determined that osseous defects of >19% of the glenoid lead to increased instability after a Bankart repair alone. They also concluded bone grafting these defects increased stability.)
There is no set metric when treating acute glenohumeral instability. As the treating orthopaedic surgeon, you must take all factors into account including patient age and expectations, activity level, bone loss after initial dislocation, etc., and formulate a treatment plan that both the patient (and often times the patient’s parents) and you feel comfortable with. We recommend a more aggressive approach in the young (<20 years old) competitive athlete, especially when out of season and if they participate in at-risk sports such as basketball and football, while we recommend a more conservative approach if the injury pattern is amenable to this in the older (>20 years old), less active patient. Patients with significant glenoid bone loss (>25%) should also be treated more aggressively and will likely require an open stabilization procedure.
Copyright © 2017, 2013 Decision Support in Medicine, LLC. All rights reserved.
No sponsor or advertiser has participated in, approved or paid for the content provided by Decision Support in Medicine LLC. The Licensed Content is the property of and copyrighted by DSM.
- The Problem
- Clinical Presentation
- Diagnostic Workup
- Non–Operative Management
- Indications for Surgery
- Surgical Technique
- Pearls and Pitfalls of Technique
- Potential Complications
- Post–operative Rehabilitation
- Outcomes/Evidence in the Literature