Anterior Glenohumeral Instability 

Updated: Feb 26, 2020
Author: Amin H Afsari, DO; Chief Editor: S Ashfaq Hasan, MD 


Practice Essentials

The term anterior glenohumeral instability refers to a shoulder in which soft-tissue or bony insult allows the humeral head to subluxate or dislocate from the glenoid fossa.[1] The function of such a shoulder is compromised. Patients typically experience apprehension, recurrent subluxations, and frank dislocations. This pathology limits many activities, including overhead arm motions, external rotation, and, thus, physical or athletic activities.

The shoulder is the most commonly dislocated joint in the body. When shoulder dislocation occurs in adolescents and children, it has the worst natural history of any injury; the rate of recurrence in later years is at least 70%.

As many as 95% of shoulder dislocations are anterior. Anterior dislocations often lead to recurrent anterior glenohumeral instability. Recurrent anterior glenohumeral instability accounts for the largest portion of the shoulder laxity spectrum. Excessive deviation of the humeral head on the glenoid occurs in all or one of three directions: anterior, posterior, or inferior. Although certainly not life-threatening, recurrent subluxation or dislocation is clearly lifestyle-threatening and can effectively disable an otherwise active individual.[2]

Normal shoulders have a certain degree of laxity resulting from minimal bony restraint of the joint, which in turn allows the widest range of motion of any joint in the body. The result is a tremendous need for competent soft-tissue balance and control. Traumatic damage leads to laxity in the soft-tissue and bony restraints; thus, recurrent subluxation and dislocation ensues. Atraumatic etiologies also exist, but this article focuses on recurrent subluxation and dislocation due to trauma.

Nonoperative and operative therapies both play a role in treatment of anterior shoulder instability.[3]  Both have been studied extensively. Surgical management has gone through an evolutionary process involving multiple methods of fixation. This evolution has led to operative fixation that achieves a stable repair with little restriction of motion.

Reports of anterior glenohumeral instability and its many repair methods date back to Hippocrates' treatise On Joints. Hippocrates described the practice of using cautery to cause the capsule to scar and thus tighten around the joint. He wrote, "[G]rasp the skin at the armpit between the fingers and draw it in the direction towards which the head of the humerus gets dislocated, then pass the cautery right through the skin thus drawn away."

Since Hippocrates' description of capsule repair, a multitude of new techniques have been reported, all aimed at preventing recurrent dislocation or subluxation. These repairs can be divided into anatomic and nonanatomic. Anatomic repairs focus on repairing the structure that has been disrupted. With nonanatomic repairs (eg, Putti-Platt, Magnuson-Stack, and Bristow procedures), an attempt is made to shorten or tighten certain anterior structures, regardless of the specific shoulder pathology.

Early in the evolution of capsular repair, loss of motion was considered acceptable and even desired as a necessary part of stabilizing a shoulder. As new techniques followed, shoulder stability was achieved without significant loss of motion. Postoperative motion restriction is now considered a complication.

In 1923, Bankart described the lesion of traumatic shoulder dislocation in the British Medical Journal, as follows: "The essential lesion is the detachment of the capsule from the fibro-cartilaginous glenoid ligament." He went on to remark that "the only rational treatment is to reattach the fibrous capsule to the glenoid ligament whence it has been torn."[4]

In 1965, du Toit and Roux reported a similar procedure, in which staples were used to attach the capsule to the anterior glenoid rim.[5]  This method had the potential to simplify the procedure, but it also added complications, such as glenoid damage from the staples and loosening.

The Magnuson-Stack procedure was described next, in 1943. This procedure involves transferring the subscapularis from its attachment on the lesser tuberosity to a point lateral to the bicipital groove. The goal was to produce a sling effect on the humeral head. Like the Putti-Platt procedure that would follow, it decreased external rotation by design.

In 1948, the Putti-Platt procedure was described. Here, the subscapularis tendon and capsule are divided longitudinally at the midportion. The lateral free end is attached to the anterior rim of the glenoid, and the medial free end is attached over the lateral end. The Putti-Platt operation led to low dislocation recurrence rates, but it also led to decreased external rotation.

In 1954, Latarjet proposed the procedure that bears his name,[6]  and a very similar operation was later popularized by Helfet in 1958 as the Bristow procedure. This involved transferring the tip of the coracoid process with its muscular attachments. A vertical slit in the subscapularis allowed the osteotomized coracoid to be attached to the anterior glenoid with sutures. This procedure thus provided a bony block to anterior glenohumeral translation. (The term Bristow-Latarjet procedure is commonly used in the English literature.)

For patient education resources, see the Breaks, Fractures, and Dislocations Center, as well as Shoulder Dislocation.


The shoulder joint is a simple structure that provides complex function. It is the most mobile joint of the body, and it is also the joint that is most frequently dislocated. The gross anatomy consists of three main components: musculature, capsule/ligaments, and bone.

Capsuloligamentous structures provide the primary stabilization for the joint. The capsule of the shoulder joint extends from the periphery of the glenoid around the articular surface of the proximal humerus. Within this capsule are three distinct thickenings that constitute the superior glenohumeral ligament (SGHL), the middle glenohumeral ligament (MGHL), and the inferior glenohumeral ligament (IGHL).

The SGHL and MGHL attach proximally at the anterosuperior portion of the glenoid labrum. The proximal attachment of the SGHL has two origins: one at the apex of the labrum that is joined with the long head of the biceps brachii and the other at the base of the coracoid process. Distally, the SGHL attaches just superior to the lesser tuberosity at the edge of the articular surface. The MGHL inserts just medial to the lesser tuberosity.

The IGHL is the key stabilizer of the shoulder, preventing anterior glenohumeral instability. The IGHL attaches proximally to the anterior, inferior, and posterior margins of the glenoid labrum. Distally, it attaches to the inferior margin of the anatomic neck of the humerus.

Together, the glenohumeral ligaments function to limit lateral rotation of the shoulder. Each of the three ligaments is relied upon for stability, depending on the position of the arm. In 1910, Delorme found that the MGHL tightens as the arm is externally rotated or dorsally flexed.[7] If the arm is then abducted, the IGHL becomes the primary stabilizer, with the upper fibers tight at slight abduction and the whole ligament tightening at about 45° of abduction.

The secondary stabilizers of the shoulder joint are the surrounding musculature. This consists of the rotator cuff muscles. The supraspinatus, infraspinatus, teres minor, and subscapularis are intimately associated with the capsule. These muscles provide dynamic secondary stabilization. Conservative therapy focuses on strengthening the rotator cuff muscles to prevent recurrent dislocation.

Of the rotator cuff muscles, the subscapularis is the most important contributor to anterior shoulder stability. With the arm adducted, it tightens with external rotation. Cutting it results in 15-20° of increased external rotation.

At 45° of abduction, the subscapularis becomes taut over the anterior joint surface and ascends so that its inferior margin lies at the inferior margin of the glenoid. External rotation raises it even further and makes it more taut. Finally, at 90° of abduction, the inferior portion of the subscapularis no longer covers the inferior humeral head. It continues to provide anterior stabilization by remaining taut.

The glenoid fossa provides a shallow socket in which the humeral head articulates. It is composed of the bony glenoid and the glenoid labrum.

The labrum is comparable to the menisci of the knee. It is a fibrocartilaginous structure surrounding the periphery of the glenoid. Like the menisci of the knee, it is flexible but constant; when torn, it generally does not heal. The stability of the glenohumeral joint is greatly increased by the labrum, which provides a 50% increase in the depth of the concavity. The bony concavity measures approximately 2.5 mm and has been measured at 5.0 mm with an intact labrum.

In 1923, Bankart claimed that anterior inferior detachment of the labrum from the glenoid was the essential lesion in anterior glenohumeral instability. It has since been proved that this is true in most cases of instability but that the presence of this lesion is not required for instability.

The labrum is closely related to the joint capsule and the glenohumeral ligaments. It is anchored to the bony rim of the glenoid and can be attached to the capsular structures as well.

Finally, it is important to understand the supporting musculature of the scapula that allows such a wide range of motion at the shoulder. A total of 16 muscles move and stabilize the scapula.


In an unstable shoulder, many findings are possible. Each finding can occur alone or in combination with other lesions. These lesions include the Bankart lesion (85% of cases), the Hill-Sachs lesion (77%), anterior glenoid rim damage (73%), capsular redundancy, subscapularis deficiency, and glenoid fossa deficiency. Absence of pathologic findings is also possible.

In 1923, Bankart described the "essential lesion" in posttraumatic anterior glenohumeral instability as the detachment of the capsule from the fibrocartilaginous glenoid ligament. In this lesion, the humeral head translates forward, shearing the IGHL with the anteroinferior labrum from the glenoid rim.

Rowe's review of 28 patients, in which he examined shoulder pathology after traumatic anterior dislocation, disputed Bankart's claim to the essential lesion, stating that "there was no evidence that there is a single essential lesion responsible for the recurrent dislocations of the shoulder." The Bankart lesions occurred in 27-100% of cases.

Other lesions were just as variable. Subscapularis laxity ranged from being present in every case to not being present in any cases. Hill-Sachs lesions were present in 26-100% of cases. Anterior glenoid trauma of all variations occurred in 2-52% of cases.

The next logical question is which of these lesions actually causes the recurrent instability observed after traumatic dislocations. This remains a debated topic. The most accurate conclusion is the one Rowe came to in 1978 when he declared that no single lesion is responsible for the recurrent dislocations of the traumatized shoulder.

Several lesions have been suggested. Baker et al devised a system of classification for the Perthes-Bankart lesion, as follows[8] :

  • Type I - Pure capsular lesion
  • Type II - Partial labral detachment
  • Type III - Complete detachment of the inferior glenohumeral-labral complex

Other lesions include attenuation of capsule and capsular ligaments, Hill-Sachs lesions, attenuation of the subscapularis, and humeral avulsion of the IGHL.


The cause of anterior glenohumeral instability can be traumatic or atraumatic. Either mechanism leads to the loss of balance in the surrounding muscular and capsular structures.

Traumatic injury to any one component of the shoulder soft tissue leads to instability. This traumatic insult most commonly comes in the form of an anterior shoulder dislocation.

Atraumatic causes leading to multidirectional instability are not the focus of this article. Such causes include repetitive atraumatic injury, Ehlers-Danlos syndrome, Marfan syndrome, congenital absence of the glenoid, deformities of the proximal humerus, and emotional and psychiatric instability.


In studies of anterior glenohumeral instability, a wide array of incidences have been cited. Most report an average age of initial shoulder dislocation early in the third decade of life. The vast majority (85-95%) of these are anterior dislocations. One fourth of all patients with dislocations present with a family history of the same problem. Rowe's 1956 analysis of 500 shoulder dislocations found only eight occurring in children younger than 10 years; the largest cohort of dislocations occurred in the subsequent decade of life, in patients aged 10-20 years.[9]

Age at the time of dislocation is the most important determinant of future recurrence. Recurrences result from anterior glenohumeral instability and occur most commonly in patients younger than 20 years. Reported recurrence rates in patients younger than 20 years range from 70% to 100%.


Outcomes of anterior glenohumeral repair rely on identification of all of the patient's pathology and choosing the correct repair. With these key factors addressed, patients can expect a low incidence of complications, good range of motion, and stable repairs with a low incidence of recurrence.[10, 11]

Potential morbidities associated with primary anterior glenohumeral dislocations are rotator cuff tear, greater tuberosity fracture, and neurologic deficit. These injuries may occur alone or in combined patterns and are more common in patients aged 60 years or older.[12, 13]

Chalmers et al conducted a systematic review of the literature to critically evaluate eight meta-analyses comparing open and arthroscopic approaches to shoulder stabilization with labral repair and capsulorrhaphy.[14] Two pre-2007 meta-analyses concluded that open stabilization provided lower recurrence rates than arthroscopic stabilization; three 2007 meta-analyses were discordant; and three post-2008 meta-analyses concluded that open and arthroscopic stabilization provided equivalent results.

Two of the meta-analyses were judged to have major flaws.[14] One meta-analysis was independently selected by four authors as providing the highest-quality evidence, and this meta-analysis found no difference in recurrence rates between open and arthroscopic stabilization.

Bliven et al reviewed eight studies (N = 795) to compare the outcomes of Latarjet repair (open only) with those of Bankart repair (open in six studies and arthroscopic in two) in patients with recurrent traumatic anterior shoulder instability.[15]  They found the Latarjet procedure to be associated with a lower recurrence rate, better outcomes as reported by patients, and less restriction of external-rotation motion than the Bankart repair.




Children presenting with a dislocated shoulder may relate a couple of possible mechanisms. These mechanisms occur in a number of ways and are similar to those in adults. Most commonly, the child falls on the outstretched hand, forcing the arm into abduction and external rotation, levering the humeral head out of the glenoid cavity. Activities promoting this injury include contact sports, falls from heights, fights, and motor vehicle accidents. Other mechanisms have been described, including elevation with external rotation and direct blows.

A history of prior dislocations suggests a high likelihood of anterior glenohumeral instability. Studies have shown a 70-100% recurrence rate in various population groups of patients younger than 20 years.

Physical Examination

As with physical examinations of any joint, beginning by observing the shoulder is important. Note any atrophy of the biceps, supraspinatus, or infraspinatus. Gross deformities can also suggest the direction of a dislocation.

Range of motion (ROM) of the shoulder must be tested for restriction or hypermobility. Atraumatic instability generally manifests with hypermobility of the shoulder, whereas traumatic instability typically results in bilaterally symmetric motion. Generalized joint laxity is demonstrated by extending the elbow, wrist, metacarpophalangeal (MCP), and distal interphalangeal (DIP) joints. External rotation can be increased as much as 28° or decreased as much as 14° after dislocation.

Next, the examiner manually assesses translation of the humeral head in the glenoid fossa. The humeral head is grasped in one hand, and the clavicle and scapula are stabilized in the other as the examiner pushes anteriorly and posteriorly. This is known as a shoulder drawer sign. Compared with the unaffected shoulder, the affected shoulder often demonstrates increased laxity. Remember that as much as 50% of posterior translation may be normal.

A sulcus sign is demonstrated by pulling inferiorly on the relaxed shoulder. A sulcus observed between the acromion and proximal humeral head is considered a positive finding. This finding indicates that the shoulder has multidirectional instability.

The key finding in anterior glenohumeral instability is a positive apprehension test. The arm is placed in abduction, extension, and external rotation while being stressed in anterior translation. If the patient becomes apprehensive and reports pain, this is considered a positive finding. It is important to note that pain alone does not constitute a positive apprehension test: The patient must report apprehension.

The relocation test involves placing the shoulder in the position of apprehension and applying a posteriorly directed force on the humeral head. The result is considered positive if this relieves the patient's apprehension.

The anterior release or surprise test is sensitive and specific for clinically diagnosing anterior shoulder instability.[16]

Impingement signs must also be evaluated because as many as 10% of patients experience impingement after dislocation. Evaluate for the Hawkins sign, and perform the Neer impingement test.




Radiologic study of the dislocated or subluxated shoulder should include a minimum of three views: true anteroposterior (AP), scapular Y, and axillary. This combination of views provides the best evaluation of the bony structures. Although frequently only soft-tissue injury is present, bony pathology is present in 55% of traumatic dislocations.

True anteroposterior view

Also known as the Grashey view, the true AP view is obtained by placing the posterior surface of the scapula flat onto the radiography film. This results in a view that is 45° oblique to traditional shoulder AP radiography.

A successful exposure demonstrates the glenohumeral joint space, superoinferior head subluxation, joint congruity, joint degeneration, and other articular abnormalities. Disadvantages to this view exist, including anterior and posterior glenoid overlap, which can obscure Bankart lesions. Increased soft-tissue overlap as compared with a traditional AP view lessens the quality of bony detail.

Scapular Y view

The scapular Y view is obtained by aiming the x-ray beam longitudinally down the axis of the scapular spine (see the image below). The humeral head lies directly over the glenoid fossa. The Y shape is formed by the projection of the acromion, scapular body, and coracoid from the longitudinal axis.

Y-view radiograph of the right shoulder shows ante Y-view radiograph of the right shoulder shows anterior dislocation of the humeral head relative to the glenoid fossa

This view may be adequate for evaluating dislocations, but it should never replace the axillary view, which is the most sensitive for detecting subluxations. Like the true AP view, it is a poor choice for evaluating glenoid rim fractures.

Axillary view

The axillary lateral view has had many variations. Originally, as described by Lawrence in 1915, it was obtained with the patient supine, the arm abducted to 90°, and the x-ray beam aimed from inferior to superior with 15-30° of medial angulation, depending on the amount of abduction. The resulting radiograph allows detection of AP subluxation/dislocation and anterior or posterior glenoid rim fractures.

The West Point view is one variation of the axillary lateral view. It places the patient prone with the arm abducted to 90° and hanging over the edge of the table. The x-ray beam is directed 25° medially and anteriorly. This position improves visualization of the anteroinferior glenoid rim. West Point axillary views are the most sensitive for finding osseous glenoid fractures.[17]

Internal and external rotation views

Internal and external rotation views provide oblique visualization of the shoulder joint, with the humeral head overlapping the glenoid rim.

The advantage of these views is the excellent osseous detail of the scapula, clavicle, upper ribs, and soft tissues. The high-quality bony detail is the result of the low density of the surrounding soft tissue. Internal rotation of the arm in the AP view projects the lesser tuberosity medially and the posterolateral aspect laterally, providing a good view of Hill-Sachs lesions. These views are of little value in detecting anterior or posterior dislocation/subluxation.

Stryker-Notch view

The Stryker-Notch view was developed to allow visualization of Hill-Sachs lesions. It is obtained with the patient supine. The hand is placed on top of the head with the elbow flexed. The x-ray beam is directed from anterior to posterior with a 10° cephalic angulation. It provides good detail of the posterolateral margin of the humeral head.

Computed Tomography and Magnetic Resonance Imaging

Computed tomography (CT) is of increasing importance in the assessment of bone defects. In a study of 70 patients with traumatic anterior glenohumeral instability, Delage Royle et al found that regular radiographs had suboptimal sensitivity, specificity, and reliability for evaluation of glenoid and humeral bone loss.[18]  They recommended that CT be considered in the treatment algorithm for accurate quantification of bone loss.

Arthrography has become obsolete since the advent of CT. It is no longer indicated in shoulder dislocation, though a couple of studies have shown 100% sensitivity in detecting capsulolabral pathology with double-contrast computed arthrotomography.

For most authors, magnetic resonance imaging (MRI) is the imaging modality of choice for soft-tissue injury.[19] It has been shown to be 91% sensitive in detecting capsulolabral injury in the early postdislocation period.

Further from the injury, MRI and arthrotomography have been up to 96% sensitive and provide a better depiction of the inferior glenohumeral ligament (IGHL) than does computed arthrotomography. It should be kept in mind that these adjunct studies are necessary only in a minority of patients.

Martins e Souza et al evaluated the accuracy of conventional MRI in determining the severity of glenoid bone loss in 36 patients with anterior shoulder dislocation by comparing results obtained by using conventional MRI with those from arthroscopic measurements.[20] They found that interreader and intrareader correlations of MRI-derived measurements of glenoid bone loss were excellent and that the first and second observers' measurements showed strong and moderate interreader correlation, respectively, with arthroscopic measurements.



Approach Considerations

Indications and contraindications for surgery

The indications for open surgical repair in a patient with anterior glenohumeral instability depend on the individual surgeon. A shoulder in which conservative therapy has failed or any shoulder that has been rendered unstable may undergo open repair.

Frequently, determination of appropriate treatment may be facilitated by classifying patients into one of the following two categories:

  • Young patients with heavy physical demands - These patients may forego conservative therapy and proceed to immediate surgical repair; this recommendation is based on the high recurrence rate and the patient's desire to return to activity
  • Older patients with lesser physical demands - These patients may try conservative therapy before assuming the risks of surgical repair; they have a lower likelihood of recurrence and may not require any treatment beyond conservative therapy

Surgery is also indicated if the patient is symptomatic with activities of daily living or if overhead stability is specifically needed. The patient must understand, however, that overhead stability cannot be guaranteed.

Surgical repair is relatively contraindicated in older patients with low physical demands, who have little chance of recurrence. Conservative therapy, including physical therapy to strengthen the rotator cuff, is indicated in preference to exposing these patients to the risks of operation.

Patients who have been asymptomatic in their activities of daily living also need not be exposed to the risks of surgery. These patients are best monitored for any recurrent dislocations.

A certain contraindication exists in the case of multidirectional instability of the shoulder. In these atraumatic dislocations, patients are able to dislocate and relocate the shoulder voluntarily. Predisposing factors include psychiatric dislocations, laxity due to repetitive injury (such as occurs in competitive swimmers), and congenital collagen abnormalities (eg, Ehlers-Danlos syndrome and Marfan disease). The history and physical examination must be used to identify these patients.

If these patients are treated as patients with unidirectional dislocations, operative therapy will fail. The amount of inferior capsular redundancy in multidirectional instability requires an operative procedure addressing the possibility of future inferior instability.

There is considerable room for improvement in the management of anterior glenohumeral instability, especially with arthroscopic treatment. Studies have shown that any type of shoulder instability can be treated arthroscopically. Advances in techniques and equipment have made treatment somewhat easier. The rate of recurrence in arthroscopically treated patients will continue to approach that of open repairs.

Open repairs will continue to provide good functional results. Postoperative range of motion (ROM) of the shoulder in patients who undergo open repair may not reach that of arthroscopically repaired shoulders.

Nonoperative Therapy

Nonoperative management of pediatric anterior glenohumeral instability must be considered. This treatment consists of a period of shoulder immobilization. Subsequent physical therapy reestablishes ROM and strengthens the rotator cuff.

The first consideration in attempting nonoperative therapy is the age of the patient. Rowe found significantly higher recurrence rates in young patients.[9] The recurrence rate in patients younger than 10 years was 100%, which decreased to 94% in those aged 10-20 years, 79% in those aged 20-30 years, and 50% in those aged 30-40 years.

Marans showed a 100% redislocation rate in 21 patients with an average age of 13 years with open physes.[21] Simonet and Cofield found an 87% recurrence rate in athletes versus a 30% recurrence rate in nonathletes.[22] Mehlman et al suggested that the rate of recurrence is not as high as previously reported[23] ; they found a redislocation rate of 76% in a group of 33 patients aged 2-18 years.

The 76-100% recurrence rate in juvenile patients certainly justifies more aggressive initial therapy. In older, more sedentary patients, immobilization in adduction and internal rotation is a valid option. The period of immobilization is debatable. Hovelius compared early motion with 3-4 weeks of immobilization and found no difference in redislocation rates.[24]

Nonoperative therapy has also been compared with arthroscopic repair. Wheeler and Arciero's 1989 study reported an 80% redislocation rate for nonoperative treatment, using 1 month of immobilization[25] ; in comparison, there was an 86% success rate in those treated by arthroscopic repair.

More important, it seems, is the physical therapy program. This should focus on isotonic strengthening before isokinetic strengthening. The first musculature to rehabilitate consists of the scapula, the serratus anterior, and then the rhomboids. The deltoids follow, and finally, strengthening should focus on the rotator cuff.

No substantial evidence supports physical therapy as a treatment for high-risk patients. Thus, conservative management is reserved for persons with first-time dislocations who are older than 20-25 years and are sedentary.

Surgical Therapy

Indications for repairing an anteriorly unstable shoulder are not obvious. The patient and surgeon must consider the possible outcomes in the light of the patient's activity level and the most probable natural history of the individual's instability. At times, forgoing conservative therapy and proceeding directly to surgical repair may be appropriate.

In the case of a patient who has experienced dislocation fewer than three times, attempting an arthroscopic repair is reasonable. Arthroscopic repair is associated with a lower incidence of shoulder stiffness but has been shown to have a higher risk of recurrence.

Persons with truly recurrent dislocations have dislocated their shoulder more than three times. In these cases, patients may want to risk the possibility of stiffness in return for the stability created by an open Bankart repair.

Many options exist for the surgical repair of anterior glenohumeral instability. Arthroscopy provides the least invasive repair. Open shoulder repairs include the Putti-Platt, Bristow (or Bristow-Latarjet), Magnuson-Stack, inferior capsular shift, Eden-Hybbinette, Trillat,[26] and Bankart procedures.

Arthroscopic repair has potential advantages, including improved cosmesis, less postoperative pain, shorter operating time, decreased blood loss, better preservation of external rotation, and avoidance of subscapularis-related complications.[27, 28]

One study, reviewing data from the American Board of Orthopaedic Surgery (ABOS), noted that the use of open repair has been declining and that there has been a trend toward arthroscopic Bankart repair.[29] This study found the most commonly reported complications to be nerve palsy/injury and dislocation. The rate of nerve injury was 2.2% in the open group, compared with 0.3% in the arthroscopic group; the dislocation rate was 1.2% with open stabilization, compared with 0.4% arthroscopically.

Intraoperatively, arthroscopy provides better visualization of the joint. Although recurrence rates in persons with first-time dislocations have been reported to be as low as 5%, subsequent literature has suggested recurrence rates rivaling those noted with open Bankart repairs. In addition, the operative procedure does require significant training.

Proper patient selection has led to good outcomes in arthroscopically repaired shoulders.[30] Good candidates for arthroscopic repair are patients with the following:

  • Discrete Bankart lesion
  • Well-developed inferior glenohumeral ligament (IGHL)
  • No significant capsular laxity
  • No intraligamentous injury
  • No concomitant intra-articular pathology
  • Unidirectional instability

Patients who require open repair are those with the following:

  • Capsular injury
  • Capsular laxity
  • Bony Bankart lesion
  • Glenohumeral arthritis
  • Rotator cuff tear
  • Poor tissue quality

Preparation for surgery

Several concerns are important preoperatively. The most significant of these is that anterior shoulder instability must be confirmed. The shoulder should be examined again after anesthesia is induced. If the shoulder is found to have multidirectional instability, the operative procedure should be canceled and revised.

Arthroscopic repair may be either intra- or extra-articular. Extra-articular repair is indicated primarily for labral tears; damage that is more extensive requires an intra-articular repair.

The arthroscopic repair can be performed with the patient in either the lateral recumbent position or the beach-chair position. It should be kept in mind that the lateral position has been associated with traction neurapraxia and allows less joint motion. The beach-chair position is convenient if a conversion to an open repair is anticipated.

There are several keys to a successful arthroscopic Bankart repair, including the following:

  • Appropriate patient selection
  • Proper mobilization of capsulolabral tissues
  • Repair of the tissue to the edge of the articular surface (not the glenoid neck)
  • Identification and treatment of capsulolabral injury

Preoperative radiography helps determine the presence of a bony Bankart lesion.

Operative details

Open Bankart procedure

Although the procedure has evolved through the years, the open Bankart repair has remained essentially unchanged from the original description by Rowe in 1978.

First, the patient's shoulder is examined under anesthesia to ensure that the patient truly has purely anterior instability and thus can benefit from a Bankart repair.

An incision is made from the coracoid process inferiorly to the axilla. The deltopectoral groove is dissected, exposing the cephalic vein and retracting it. The coracoid process is osteotomized, allowing the coracobrachialis and short head of the biceps to retract inferiorly.

External rotation of the arm exposes the subscapularis. The circumflex vessels at the inferior border of the muscle can be ligated. The subscapularis is then separated from the capsule. With the arm fully externally rotated, a vertical incision is made in the subscapularis just lateral to the glenoid rim. This provides a large flap for repair of the capsule with adequate external rotation postoperatively.

With the humeral head retracted posterolaterally, the rim of the glenoid and the anterior neck of the scapula are freshened by using a small osteotome. Three holes are made in the anterior glenoid rim, located at 1, 3, and 5 o'clock on the right glenoid and at 11, 9, and 7 o'clock on the left glenoid.

A double-0 suture is passed through each hole and then through the edge of the lateral capsular flap. This is tied while the capsular flap is held to the freshened anterior glenoid rim. With the same suture ends, the medial capsular flap is then tied down on top of the lateral one. This reinforces the capsule at the rim of the glenoid.

External rotation is then checked; it should extend easily to 25-30° beyond neutral.

Closure is achieved by returning the tissues to their anatomic positions. The subscapularis is sutured to the lesser tuberosity and secured. Thus, the muscle is not shortened, overlapped, or transplanted, as is the case in other techniques.

The coracoid is anchored in place with 0 suture at its base. No separation of the coracoid has been described with the use of suture to anchor it.

Arthroscopic Bankart procedure

The arthroscopic procedure for a soft Bankart lesion also starts with examination under anesthesia. The patient may be in the lateral recumbent or beach-chair position. Access to the joint starts with a posterolateral glenohumeral portal for viewing the joint. A thorough examination is performed before an anterior working portal is made lateral to the coracoid.

The edge of the glenoid articular cartilage is abraded. The capsulolabral complex is dissected and then grasped and advanced over the abraded bone. Anterior reconstruction is then performed with suture anchors and plication of the loose anterior capsule into the labral repair.[31, 32] The most inferior suture is placed first, allowing superior sutures to further tighten the capsuloligamentous complex. The sutures are tightened with the shoulder in internal rotation. ROM can be examined intra-articularly, and the incisions are closed.

Bone graft augmentation may be a beneficial adjunct to arthroscopic Bankart repair in patients who have recurrent anterior shoulder instability with glenoid bone loss.[33]

Arthroscopic Latarjet procedure

The arthroscopic Latarjet procedure appears to be an effective means of treating "off-track" Hill-Sachs lesions, restoring them to an "on-track" state. In a study by Plath et al, a mean persisting enlargement of the glenoid arc of 14% beyond native dimensions remained at a mean of 23 months after this procedure.[34]

Ranne et al reported a modification of the arthroscopic Latarjet procedure for treating anterior glenohumeral instability, in which the detached coracoid was exteriorized through the anteroinferior portal for drilling and shaping.[35] A Coracoid Drill Guide (Arthrex, Naples, FL) was used to help cut the coracoid to the desired size and to make two drill holes in the coracoid for fixation to the glenoid. The Coracoid Transfer Instrument (Acierart, Masku, Finland) was designed to facilitate coracoid transfer and to serve as a pin guide for fixation.

The 10 patients with severe anterior glenohumeral instability who were treated with this technique had only mild-to-moderate postoperative pain, with no postoperative infections or recurrent dislocations.[35] In terms of safety, the operation was found to be comparable to other operations on the coracoid process in the proximity of the brachial plexus.

A retrospective cohort study by Ali et al found the clinical and radiographic outcomes of the arthroscopic Latarjet procedure to be similar to those of the equivalent open procedure.[36]

Open Latarjet procedure

The open Latarjet procedure is being increasingly utilized as a solution to anterior glenohumoral instability due to significant glenoid bone deficiencies of greater than 21%, in which the failure rate of arthroscopic Bankart repair increases significantly.[37, 38, 39]  This technique involves splitting the subscapularis in order to attach a parallel-positioned coracoid process onto the exposed glenoid by using two 3.5- to 4.5-mm screws, as described by Edwards and Walch.[40]

The open Latarjet procedure is reported to prevent recurrent instability in 99% of correctly selected patients.[41]  However, there are several drawbacks to the procedure that may result in complications, the majority of cases presenting with stiffness and increased loss of motion.

Reduction of complications

There are several intraoperative details that a surgeon should be aware of to reduce the risk of postoperative complications. For example, incorrect positioning of the coracoid process may result in a lateral overhang, which can lead to rapid bone degenerative joint disease.[42]  Additionally, overtightening of the screws may cause coracoid fractures; this can be prevented by tightening the screws using the “two-finger” technique. Complications and postoperative rehabilitation can be minimized by employing proper surgical technique. 

Postoperative Care

Postoperatively, the shoulder is kept in an arm sling for 3-4 weeks. The shoulder may be internally rotated and flexed for hygiene maintenance, but external rotation is prohibited.

After 4 weeks, full active flexion is encouraged and assisted with physical therapy. At 6 weeks, internal and external rotation of the shoulder is begun by using 5-lb (~2.25-kg) and 10-lb (~4.5-kg) weights. At 3 months, full and unrestricted weightlifting is allowed. Competitive throwers should be restricted from unrestricted athletics for 6 months.

At 3 months, the patient should have regained 70% of external rotation and elevation of the shoulder. At 6 months, 75-100% of normal motion and strength may be observed.


Successful repair of anterior glenohumeral instability is extremely rewarding and can render a patient fully functional within months. However, as in any surgical procedure, complications are possible. Unsuccessful surgery in this case can be one of the most difficult to salvage.

The most commonly reported complication of open Bankart repair is persistent instability.[37] Reported rates of recurrent instability range from 3% to 50%. The cause of failure should be determined.

The most common adverse effect of anterior instability is loss of motion. Frequently, decreased external rotation of the humerus is desired to prevent recurrent dislocation; thus, the loss of motion is not reported as a complication. Complications of decreasing the normal shoulder ROM can be serious. In athletes, the late cocking position is disabled, thus decreasing the velocity of their throw. More extensive reductions in external rotation can cause posterior translation of the humeral head.

Iordens and Lieshout found that Putti-Platt repairs in 51 patients resulted in tight anterior structures, which led to glenohumeral arthritis.[43] This procedure is not the only repair that may result in tight anterior structures: The classic Bankart repair has also been shown to cause limitation of motion and, thus, clinical and radiographic arthritic changes. How much limitation in motion is required to cause late degenerative arthrosis remains to be determined. Studies to date have suggested that small limitations do not cause significant degeneration.

Excessively tight anterior capsules must be addressed so as to prevent premature arthritic changes. The subscapularis can be released and sutured back into place with the arm externally rotated to the desired position. For greater restrictions, requiring more than 20° of correction, the subscapularis is dissected in the coronal plane, creating a superficial and deep layer. The medial edge of the deep layer of the subscapularis can be sutured to the lateral edge of the superficial layer.

An incorrect diagnosis prior to surgery undoubtedly leads to failure. This can occur when atraumatic multidirectional shoulder instability is confused with traumatic unidirectional anterior instability. In the case of the former, an inferior capsular shift is in order, as opposed to the latter, for which an open or arthroscopic Bankart repair is indicated.

Voluntary instability is a contraindication to any surgical repair mentioned above. These patients can dislocate their shoulders using muscular contractions or arm positioning. A study by Rowe found that most people who dislocate their shoulders voluntarily have "significant psychiatric histories." Nonoperative treatment is necessary in these patients.

Nonanatomic repairs have been shown to result in higher postoperative instability rates. In the original Bankart repair, the anterior glenoid is essentially reconstructed to its original anatomic shape. Procedures that fail to reconstruct the anterior glenoid have been shown to fail more frequently.

Bony Bankart lesions, if large enough, leave a glenoid fossa that lacks concavity. Defects involving greater than 20% of the articular surface must be addressed. Bony lesions involving less than 20% of the glenoid can be resected with the capsulolabral mechanism reattached to the remaining anterior glenoid.

One study assessed the factors associated with the presence, size, and type of glenoid bone defect in patients with anterior shoulder instability.[40] Computed tomography (CT) revealed a glenoid bone defect in 72% of the 161 patients. The defect was significantly associated with recurrent dislocation, increasing number of dislocations, timing from first dislocation, and manual work; a critical defect was associated with number of dislocations and age at first location; and bony Bankart lesion was associated with male sex and age at first location.

Osteochondral defects of the humeral head may predispose a shoulder to continued instability. A Hill-Sachs lesion alone infrequently causes instability, but when it is coupled with a Bankart lesion, it creates an easily subluxed or dislocated shoulder. Preoperative CT of the Hill-Sachs lesion helps determine the need for repair. The humeral lesion can be prevented from perching on the anterior glenoid by imbricating the anterior capsule and thereby decreasing external rotation.

Other methods of filling in a Hill-Sachs defect have been described, including transfer of the infraspinatus with a portion of greater tuberosity, humeral osteotomy, and humeral hemiarthroplasty or osteochondral allograft for defects involving more than 40% of the head's articular surface.[44, 45]

Further treatment using so-called remplissage (filling in) of the Hill-Sachs lesion defect with rotator cuff tendon has been studied. A study by Nourissat et al found no significant statistical difference in ROM between patients treated with arthroscopic Bankart repair alone and those treated with Bankart repair and remplissage.[46] The recurrence rates were identical in the two groups, and one third of patients experienced posterosuperior pain.

Another cause of surgical failure is rupture of the subscapularis. Patients may present after a traumatic event or with persistent postoperative pain, weakness, or continued instability. Physical examination reveals increased external rotation and positive liftoff test results. MRI may be helpful in confirming this complication. Operative repair includes mobilization of the tendon, which frequently retracts underneath the conjoined, and reattachment to its insertion.

Hardware placed in or around the glenohumeral joint can always cause complications. Loosening and failure have been shown to occur at any time point after surgery. Zuckerman found that the most common culprits were screws or staples that were placed for a coracoid transfer or anterior capsular plication.[47] Most of these failures necessitate reoperation. Chondral damage was found in 41% in Zuckerman's report.

Long-Term Monitoring

Follow-up care can be scheduled on an annual basis after patients have returned to full activity and strength. At these visits, any symptoms of subluxation or dislocation should be explored. Shoulder ROM, strength, and stability are examined, and a radiographic examination is also performed.