Multidirectional Instability (MDI)

The Stanmore Triangle, developed by Lewis et al., categorises shoulder instability by aetiology and underlying pathology and recognises three distinct phenotypes. It matters in MDI because it separates structural from non-structural causes, which have completely different treatment implications, and it emphasises that "MDI" is not a single entity.

Type I: traumatic, structural. Significant trauma disrupts the capsule of a previously normal shoulder, and the resulting structural laxity allows instability that may be unidirectional or multidirectional depending on the injury pattern. There is a clear injury history, structural damage on imaging (capsular tears, labral injury), apprehension in the direction of the trauma and normal generalised laxity (a normal Beighton score). These patients may require surgery if conservative management fails, and structural repair of the specific capsulolabral injury is appropriate.

Type II: atraumatic, structural. The classic MDI patient, with constitutional capsular laxity and redundancy, often generalised hypermobility (Beighton 4 or more), no trauma history and an insidious onset. Global capsular redundancy, bilateral involvement in 50-80% and a pathognomonic sulcus sign are the pattern. Rehabilitation is first-line: in Burkhead's series about 80% of atraumatic instability, which is what Type II MDI is, responded, against only 16% of traumatic instability. Surgery (inferior capsular shift) comes only after a minimum of 6 months of failed conservative treatment.

Type III: atraumatic, non-structural, muscle patterning. Psychological factors and voluntary muscle inhibition cause the instability. The capsule may be of normal volume, and there is no structural pathology on examination or imaging; aberrant neuromuscular control allows symptomatic translation, and these patients often have voluntary control of their instability, sometimes with secondary gain. Examination under anaesthesia shows normal capsular volume. Management is physiotherapy focused on motor control and neuromuscular retraining, psychiatric evaluation may be required, and surgery is contraindicated because the pathology is not structural.

The sulcus sign. The pathognomonic finding in MDI: inferior translation of the humeral head with inferior traction on the arm, measured as the acromiohumeral interval. Its grade guides treatment and surgical planning.

• Grade I - under 1cm. Mild inferior laxity; may be a normal variant, especially in athletes; usually asymptomatic; no surgical intervention indicated
• Grade II - 1-2cm. Moderate inferior laxity; often symptomatic MDI; a trial of rehabilitation is indicated, and capsular shift may be required if it fails
• Grade III - over 2cm. Severe inferior laxity; usually symptomatic and functionally limiting; indicates significant capsular redundancy and a high likelihood of surgical requirement

Sulcus severity correlates with the degree of capsular redundancy and IGHL incompetence, and a Grade III sulcus predicts a higher failure rate of conservative management and may indicate earlier surgical consideration. Bilateral comparison is mandatory because baseline laxity varies between individuals: a Grade II sulcus may be pathological in one patient and a normal variant in a hypermobile athlete. Document the grade in neutral rotation, the grade in external rotation, the comparison with the other shoulder and whether the sulcus is reducible with muscle contraction, since these influence the treatment algorithm.

Load-and-shift. Quantifies anterior and posterior translation in the horizontal plane and complements the sulcus sign for a complete assessment.

• Grade 0. Minimal translation, no movement of the head relative to the glenoid; normal, intact restraints, not consistent with MDI
• Grade 1. Head translates up to the glenoid rim but not over it; mild laxity that may be a normal variant, particularly common in overhead athletes, so clinical correlation is required
• Grade 2. Head translates over the rim but spontaneously reduces; moderate laxity consistent with MDI, capsular redundancy present, symptomatic instability likely
• Grade 3. Head translates over the rim and remains dislocated; severe laxity indicating significant capsular incompetence, a highly pathological finding

Together the two gradings give a three-dimensional picture of capsular laxity, which is what the diagnosis and the surgical plan rest on.

The Instability Severity Index Score (ISIS) is a validated prognostic tool that predicts the risk of recurrence after arthroscopic Bankart repair. It was developed for anterior instability, but it gives useful prognostic information for the MDI patient being considered for surgery.

Scoring (10 points).

• Age at surgery: under 20 years, 2 points; 20 or older, 0
• Degree of sport participation: competitive, 2 points; recreational or none, 0
• Type of sport: contact or forced overhead, 1 point; other, 0
• Shoulder hyperlaxity: anterior or inferior hyperlaxity, 1 point; normal laxity, 0
• Hill-Sachs lesion on the AP radiograph: visible, 2 points; not visible, 0
• Glenoid bone loss: defect on the AP view, 2 points; none visible, 0

Risk strata.

• Low (0-2): under 10% recurrence after arthroscopic stabilisation
• Moderate (3-6): 10-20% recurrence; arthroscopic stabilisation reasonable
• High (7-10): over 20% recurrence; consider open stabilisation or augmentation procedures

In MDI. These patients typically score high, being young, hyperlax and often in competitive sport, and that predicts higher failure rates for arthroscopic capsulorrhaphy than in traumatic anterior instability. A patient with MDI and an ISIS above 6 should be counselled about the higher recurrence rate and may be a better candidate for open inferior capsular shift than for arthroscopic capsulorrhaphy. Consider calculating it for every MDI patient being considered for surgery, as an objective input to the choice between arthroscopic and open approaches.

The Burkhead-Rockwood programme. The classic first-line programme for MDI, and its defining result is the one that justifies the whole approach: in the original series it produced good or excellent results in 80% of atraumatic instability, which is what MDI largely is, but in only 16% of traumatic instability. That contrast is the entire reason physiotherapy is first-line in atraumatic MDI and not in traumatic unidirectional instability. The Warby randomised trial later found the Watson MDI programme superior to the Rockwood programme, so Burkhead-Rockwood is best described as the classic, not the proven-best, protocol. It runs in four phases:

• Phase 1 (weeks 0-6), rotator cuff strengthening: internal rotation (subscapularis), external rotation (infraspinatus, teres minor) and anterior, middle and posterior deltoid, at low resistance and high repetition, avoiding provocative positions while strengthening
• Phase 2 (weeks 6-12), scapular stabilisation: serratus anterior (wall slides, protraction), rhomboids and middle trapezius (rows, retraction), lower trapezius (prone Y), scapular control with shoulder motion, and integration of scapular and cuff activation
• Phase 3 (weeks 12-24), proprioception and neuromuscular control: closed chain work (wall push-ups, quadruped), proprioceptive training on unstable surfaces with perturbation, plyometric progression, dynamic stabilisation drills and sport-specific movement patterns
• Phase 4 (weeks 24 onwards), return to activity: gradual return to overhead activity, sport-specific training and conditioning, and a maintenance programme of ongoing strengthening and proprioception that the patient keeps for life

The Watson MDI programme. The highest-level evidence in MDI, the Warby/Watson randomised controlled trial in the Evidence Base, showed that a motor-control and scapular-focused programme was superior to the Rockwood strengthening programme on validated instability scores (WOSI, Melbourne Instability Shoulder Score) and on pain at 24 weeks. Its core principle is that MDI is driven not only by capsular redundancy but by altered scapular and humeral-head neuromuscular control, so it retrains control and positioning first, before loading, in contrast to the strengthening-first Rockwood approach:

• Stage 1, scapular control: restore the correct resting scapular position and dynamic scapular setting (correcting downward rotation, anterior tilt and protraction) as the stable base for the glenohumeral joint
• Stage 2, humeral-head control: retrain deltoid and rotator cuff to centre the humeral head, progressing through increasing range and functional positions while scapular control is maintained
• Stage 3, progressive loading and speed: add resistance, then functional and sport-specific patterns once control is established
• Stage 4, return to sport and maintenance: sport-specific progression and a lifelong maintenance programme

It reframes MDI rehabilitation from "strengthen the cuff" to "restore neuromuscular control of the scapula and humeral head", and it is the programme with Level-II RCT support. Both programmes still require a prolonged, compliant trial of at least 6 months before surgery is considered. Scapular dyskinesis itself is developed in the scapular-dyskinesis topic.

Muscle-patterning (Stanmore Type III) instability. Repeatedly flagged as a surgical contraindication, but its management deserves explicit treatment, because operating on it fails whereas the right non-operative pathway can succeed. Recognise the subtype first. In positional non-structural instability, involuntary subluxation occurs only in certain arm positions and is driven by abnormal muscle recruitment (inappropriate activation of pectoralis major or latissimus, or inhibition of the cuff) rather than capsular deficiency. In habitual or wilful (voluntary) instability the patient can demonstrate it on demand, and there may be secondary gain or psychological factors. The unifying feature is normal capsular volume (Grade 0-1) under anaesthesia despite symptomatic instability when awake, with an abnormal muscle-recruitment (EMG) pattern.

• EMG biofeedback-based neuromuscular retraining (the Stanmore biofeedback approach) is the cornerstone: surface EMG shows the patient their abnormal recruitment, normal firing sequences are retrained, and the aberrant activation that produces the subluxation is suppressed
• Screen for and manage psychiatric comorbidity and secondary gain, with clinical-psychology input where there is a wilful component
• No stabilisation surgery: capsular shift will not correct a control problem and tends to fail; surgery is reserved only for a genuinely coexisting structural lesion
• EUA is the gatekeeper: if laxity is normal under anaesthesia, abandon any planned stabilisation and commit to the retraining pathway

Motivated patients with positional muscle patterning often improve substantially with biofeedback retraining; outcomes are poorer where there is entrenched wilful behaviour or unaddressed secondary gain.

Surgery is considered only after comprehensive conservative management has failed, and strict criteria must be met before operating, because MDI surgery has lower success rates and higher recurrence than traumatic instability repair. Patient selection is the decision.

Absolute requirements. All of these:

• Failed minimum 6 months of structured rehabilitation
• Documented compliance with the therapy programme
• Persistent symptomatic instability affecting function
• Stanmore Type I or II (structural laxity)
• Objective instability on examination (Grade 2-3 laxity)
• Realistic expectations about outcomes and recovery
• Absence of a significant voluntary component
• Absence of psychiatric contraindications

Relative indications favouring surgery. A high-level athlete needing to return to sport, occupational overhead demands, bilateral involvement with functional limitation, a Grade III sulcus sign indicating severe laxity, failure of multiple rehabilitation attempts, and associated pathology that itself needs surgical treatment.

Contraindications. Stanmore Type III (voluntary, muscle patterning), an inadequate rehabilitation trial (under 6 months), poor compliance with therapy, unrealistic expectations, psychiatric comorbidity or secondary gain, medical contraindications to surgery, and a generalised connective tissue disorder with poor healing.

Arthroscopic capsular plication has become increasingly popular for MDI, offering reduced morbidity compared with open techniques, but its outcomes are inferior to open inferior capsular shift, particularly in severe MDI.

Who it suits. The ideal candidate is young with lower functional demands, has Grade 2 laxity (Grade 3 is better suited to open shift), has failed an appropriate rehabilitation trial, is Stanmore Type II and has realistic expectations about the success rate. The relative contraindications are Grade 3 laxity in multiple directions, a high-level athlete needing to return to contact sport, revision after a failed previous stabilisation, generalised hypermobility with a Beighton score over 6, and a significant voluntary component.

Positioning. Beach chair at 30-45° upright, head secured in neutral, the arm free to move through a full range and the table articulated so position can be changed during the procedure. Lateral decubitus with the arm in 10lb of traction is the alternative: better visualisation of the inferior capsule but less physiological for assessment.

Portals.

• Posterior: the standard viewing portal, 2cm inferior and 1cm medial to the posterolateral acromion, established first for orientation
• Anterior superior: just anterior to the biceps tendon at the rotator interval, the working portal for superior and rotator interval plication
• Mid-anterior: through the rotator interval or just superior to subscapularis, the working portal for anterior plication
• Anterior inferior: at the 5:30 position (right shoulder) at the inferior glenoid margin, giving access to the inferior pouch
• Accessory posterior (Wilmington): posterolateral, just lateral to the standard posterior portal, for better access to the posterior capsule

What to look for. The drive-through sign, where excessive capsular laxity lets the scope pass from posterior to anterior without resistance; inferior pouch redundancy viewed from the anterior portal; a patulous rotator interval, widened between supraspinatus and subscapularis; a labrum that is typically attenuated or absent rather than torn; and the articular cartilage, to exclude chondral injury.

Plication. The inferior capsule is plicated through the anterior inferior portal with multiple horizontal mattress sutures (PDS or FiberWire) placed in an anterior-to-posterior direction; 3-5 sutures are typically required to reduce the inferior pouch significantly, and over-tensioning costs motion. The posterior capsule is plicated from inferior to superior through the Wilmington portal, 2-3 sutures usually sufficing to address the posterior component, and external rotation is checked afterwards. Rotator interval closure, sutured between supraspinatus and subscapularis so that the coracohumeral ligament is drawn to the superior glenohumeral ligament and the interval space obliterated, is mandatory if the sulcus persisted in external rotation on EUA, and it eliminates inferior translation in external rotation.

Technique. Sutures are passed through the capsule by shuttling techniques (suture lasso, penetrator, bird beak), taking adequate tissue purchase with each pass, and are all visualised before tying; tie arthroscopically (SMC, Duncan loop) or use knotless anchors, making sure every knot is well seated. Plicate until the drive-through sign is eliminated but the range is preserved, checking motion after each plication, and stay anterior to the mid-glenoid line.

Before closing. Range is tested against the targets above and restricted motion means suture release. Load-and-shift should now show Grade 0-1 translation, and the drive-through sign should be eliminated or substantially reduced; inability to pass the scope from posterior to anterior indicates adequate volume reduction. Recurrence is higher in patients with severe laxity, generalised hypermobility and high-demand activity, and satisfaction tracks appropriate expectations and an understanding that some residual laxity may persist.

Open inferior capsular shift remains the gold standard for severe MDI, particularly in high-demand athletes and patients with Grade III laxity. It is preferred for Grade 3 laxity in multiple directions on EUA, the high-level athlete returning to contact or overhead sport, revision after failed arthroscopic stabilisation, severe generalised hypermobility (Beighton over 6), and by patient preference after informed consent.

Which approach. Anterior for anterior and inferior laxity; posterior when the posterior component is the primary problem; combined, rarely, for severe bidirectional laxity. The anterior approach is done in the beach chair position at 30-45°, head secured in neutral, arm free and the whole shoulder and upper arm prepped into the field; the posterior or combined approach in lateral decubitus, affected side up on a beanbag or lateral supports, the arm supported but mobile.

Anterior inferior capsular shift. A deltopectoral incision of 8-10cm runs from the coracoid distally along the groove, the cephalic vein identified and preserved laterally. The interval is split, the conjoint tendon (short head of biceps and coracobrachialis) identified medially and the coracoid palpated, and the tendon tagged for retraction. Subscapularis is managed in one of two ways:

• Tenotomy: incise the tendon 1cm medial to the lesser tuberosity, develop the plane between subscapularis and capsule, preserve the inferior subscapularis and tag the tendon with stay sutures for repair
• Split: divide longitudinally between the upper two-thirds and lower third, which preserves tendon continuity and gives direct capsular access but may limit inferior visualisation

The capsule deep to subscapularis looks thin and redundant, and external rotation improves the view of its anterior part. Make a T-shaped or lateral-based capsular incision, the vertical limb along the anterior glenoid margin and the horizontal limb along the inferior capsule at the 6 o'clock position, creating superior and inferior flaps. Then shift:

• Advance the inferior flap superiorly to reduce the inferior pouch and secure it to the superior edge of the capsular incision with interrupted absorbable sutures
• Overlap the superior flap over it (double-breasted) and secure it to the remaining inferior capsule
• A typical shift reduces capsular volume by 30-50%, over a distance of 1-1.5cm; larger shifts risk over-tensioning and stiffness, smaller ones may not reduce the volume enough
• Use interrupted absorbable No. 2 braided suture, knots buried within the capsular layers; non-absorbable suture may cause persistent irritation

Rotator interval closure, if the sulcus was positive in external rotation on EUA, is performed after the shift: the interval between supraspinatus and subscapularis is closed by suturing the coracohumeral ligament to the superior border of subscapularis, which eliminates the sulcus in external rotation. Subscapularis is then re-attached to the lesser tuberosity with suture anchors or transosseous tunnels, tensioned in 30° of external rotation; a secure repair is critical to stability, and anchors in bone are the preferred fixation. The deltopectoral interval is closed loosely, then subcutaneous tissue and skin; a drain is not usually needed.

Neer's T-capsulorrhaphy. The alternative to the lateral-based shift: a T-shaped capsulotomy with vertical and horizontal limbs, the vertical along the glenoid and the horizontal along the inferior capsule. The medial flap is shifted superiorly and laterally and the lateral flap overlapped on it, giving circumferential tightening, a similar volume reduction and possibly a better reduction of the inferior pouch.

Posterior capsular shift. For MDI with a predominant posterior component, or for persistent posterior instability after a failed anterior stabilisation. In lateral decubitus, a 6-8cm incision along the posterior axillary fold, the deltoid split (staying superior to the axillary nerve) and the infraspinatus-teres minor interval identified and split horizontally to expose a posterior capsule that is typically very redundant in MDI. A T-shaped or medially-based capsular incision is made, the lateral flap shifted medially to the glenoid rim and the medial flap overlapped over it, reducing the posterior volume significantly.

Interval closure is an essential component of MDI surgery when inferior laxity persists with the arm in external rotation, because the coracohumeral and superior glenohumeral ligaments within the interval are what restrain inferior translation in that position (the anatomy is in Pathophysiology). Left incompetent, the interval leaves persistent inferior instability despite an adequate capsular shift, and with it patient dissatisfaction and potential surgical failure.

Indications. The absolute indication is a sulcus persisting over 1cm with the shoulder in external rotation on preoperative EUA, the sign of rotator interval incompetence. Relative indications are severe MDI with a Grade III sulcus in neutral rotation, revision MDI surgery, generalised hypermobility with a Beighton score over 6, and a previous capsular shift that failed without interval closure.

Arthroscopic closure. Viewed from the posterior portal, the interval appears as a widened space between supraspinatus above and subscapularis below with the biceps tendon medially. Working through the anterior superior portal, 2-3 sutures are passed from the superior border of subscapularis to the inferior border of supraspinatus, starting medially near the base of the coracoid and progressing laterally toward the bicipital groove until the interval is obliterated, keeping the biceps tendon mobile. Tension matters: over-tensioning restricts external rotation, under-tensioning fails to eliminate the sulcus. Then test the sulcus in external rotation arthroscopically; it should be eliminated or substantially reduced.

Open closure. After the capsular shift, subscapularis is retracted inferiorly to expose the interval, the biceps tendon marking its medial boundary and the superior border of supraspinatus identified. 3-4 interrupted absorbable sutures are placed from superior subscapularis to inferior supraspinatus and tied sequentially from medial to lateral, obliterating the interval space and preserving the biceps centrally.

What can go wrong. Over-aggressive closure restricts external rotation, and closure adds to the stiffness risk of the capsular shift it accompanies: combined, they can cost significant motion if over-tensioned. The long head of biceps must stay free and mobile, never incorporated into the closure sutures, with a clear passage through the interval.

What it achieves. Closure reduces inferior translation in external rotation by 50-70%, eliminates the persistent sulcus in most patients and adds minimal morbidity when combined with capsular shift.

Thermal capsulorrhaphy was popular in the late 1990s and early 2000s as a minimally invasive treatment for MDI, until high failure rates and serious complications led to its abandonment. Its history matters for examinations.

What it was. Thermal energy, from monopolar or bipolar radiofrequency probes or a holmium:YAG laser, denatures capsular collagen and shrinks it. The capsule was "painted" in an overlapping pattern, working systematically from inferior to superior and avoiding neurovascular structures, with multiple passes over redundant areas and the inferior, anterior and posterior capsule covered as needed; tissue was heated to 65-75°C, higher temperatures causing excessive damage and lower ones being ineffective. It was minimally invasive, technically easier than plication and showed immediate intraoperative tightening, and early studies of short-term capsular tightening were promising, so many surgeons adopted it as first-line surgery for MDI.

Why it was abandoned.

• Failure: initial success rates of 80-90% at 1 year fell to 30-50% at 3-5 years, the capsule stretching out as the thermal injury healed with scar that elongated over time
• Chondrolysis: thermal energy killed articular cartilage, producing rapid glenohumeral arthritis in 1-5% of patients, many needing arthroplasty at a young age
• Capsular necrosis: excessive energy caused full-thickness necrosis rather than controlled shrinkage, leaving massive capsular deficiency and permanent instability
• Axillary nerve injury: energy conducted to the adjacent nerve caused neuropraxia or permanent damage in a small percentage
• Severe stiffness: over-aggressive application caused dense scarring and adhesive capsulitis needing manipulation or arthroscopic lysis

The biology. The immediate collagen denaturation and shrinkage healed by scar formation rather than capsular regeneration, and scar gradually elongates under physiological stress. The depth of penetration was uncontrolled, so superficial application was ineffective and deeper penetration risked the cartilage; there was no safe therapeutic window. Thermal injury also impaired the normal healing cascade, compromising capsular blood supply and killing cells, and the tissue that resulted was biomechanically inferior to normal capsule.

Now. No current role in MDI or any instability treatment; it is mentioned in historical context only, and modern MDI surgery uses capsular plication, which preserves tissue, with superior outcomes and no risk of chondrolysis. Patients treated in the past may present with late complications, and documentation of informed consent is critical for surgeons who performed it. The lesson is that long-term outcomes must be established before a new technique is widely adopted, that the biology of the healing response has to be understood, and that immediate intraoperative results do not predict long-term clinical outcome.

Stiffness is the most common complication after MDI surgery, occurring in 10-20% of patients. The risk is inherent in tightening a lax capsule: the balance between stability and motion is the difficulty of the operation.

Risk factors.

• Patient: a tendency to stiffness (previous adhesive capsulitis), diabetes mellitus, female sex, age over 40 and an underlying connective tissue disorder
• Surgical: over-tensioning the shift, aggressive rotator interval closure, combined anterior and posterior procedures, thermal capsulorrhaphy (historical) and inadequate intraoperative assessment of range
• Postoperative: prolonged immobilisation (over 6 weeks), non-compliance with rehabilitation, delayed initiation of range-of-motion exercises, and pain limiting participation

How it presents. Restricted motion affecting daily life: forward flexion under 120°, external rotation under 30°, internal rotation unable to reach the back. Pain at end-range is common, patients describe feeling "tight" and "restricted", night pain may occur from capsular inflammation, and quality of life is significantly affected.

Prevention. Intraoperative assessment of range before final closure, with suture release if the targets are not met, and tensioning to Grade 0-1 rather than complete elimination of laxity; some laxity must be preserved for motion. Modern protocols immobilise for 4-6 weeks rather than the historical 8 (the protocol in Rehabilitation uses 6 weeks), and gentle passive motion starts at 6 weeks, since delay beyond that point raises the stiffness risk significantly. Counsel patients about the risk, the importance of rehabilitation compliance and the potentially slow recovery of motion.

Treatment. Conservative management first: intensive physiotherapy with passive stretching, heat before stretching, joint mobilisation, NSAIDs for pain and inflammation and a home programme, for a minimum of 3-6 months. If that fails, manipulation under anaesthesia at 4-6 months postoperatively, once healing allows, gentle and followed by intensive physiotherapy, accepting the risk of capsular re-injury and recurrent instability. Arthroscopic capsular release is reserved for refractory cases that fail manipulation: selective release of contracted capsule and adhesions, needing an experienced surgeon because it risks destabilising the previous stabilisation, with a 70-80% success rate for improving range. Stiffness remains a challenging complication needing prolonged rehabilitation and possibly further intervention.

Recurrent instability after MDI surgery occurs in 10-20% of patients, higher than the 5-10% after repair of traumatic anterior instability. Several kinds of failure contribute.

Why it recurs.

• Patient selection: a Stanmore Type III patient was operated on, a voluntary component was missed, expectations were unrealistic so a fair result is perceived as failure, generalised hypermobility (Beighton over 7), or non-compliance with restrictions
• Technical: inadequate capsular volume reduction, missed rotator interval pathology, under-tensioning of the shift, failure to address every direction of instability, inadequate fixation of the shift, or subscapularis repair failure after an open procedure
• Biological: capsular stretch-out during healing, poor tissue quality in connective tissue disorders, an inadequate healing response, progressive laxity in the hypermobile, or new trauma
• Postoperative: early return to provocative activity or overhead sport, non-compliance and inadequate rehabilitation

How it presents. Return of instability symptoms, often in the same directions as before surgery, either gradually (progressive stretch) or acutely (traumatic re-injury), and either milder than the original presentation or equally severe depending on the mechanism. Examination shows a recurrent positive sulcus sign and increased load-and-shift translation. Timing is informative: early recurrence (under 6 months) suggests technical failure or non-compliance, late recurrence (over 1 year) progressive stretch or new trauma.

Work-up. Document the timing and mechanism (gradual versus traumatic), compliance with restrictions, rehabilitation participation and activity at the time of recurrence. Examine both shoulders, quantify the laxity with sulcus grading and load-and-shift, and look for stiffness, which suggests over-tensioning in another direction. MRI assesses capsular integrity, distinguishes disruption from stretch, looks for new labral pathology and, after an open procedure, checks the subscapularis. EUA may be needed to quantify true laxity and separate structural recurrence from muscle patterning.

Management. Start again with rehabilitation focused on rotator cuff and scapular strengthening; the success rate is lower than in primary MDI (40-60%) but it is a reasonable first attempt. Modify activity, avoiding provocative positions and overhead work, which may mean a change of sport or occupation, and accept some laxity with functional adaptation. Revision is reserved for significant functional limitation after a failed adequate repeat rehabilitation trial, with structural failure identified on imaging, documented compliance with restrictions and realistic expectations of what revision can achieve.

Revision surgery. Planning identifies the failure mechanism (under-correction, technical error, trauma), evaluates the previous repair on MRI, quantifies current laxity under anaesthesia and counsels about the lower success rate and higher complication rate than primary MDI surgery (70-80% versus 85-90% for primary surgery). The options are arthroscopic revision when the index procedure was arthroscopic with an identifiable technical error, conversion to open when it was arthroscopic with inadequate correction, open revision with a repeat shift and augmentation, and allograft capsular augmentation for severe deficiency. Scar tissue complicates the dissection, the capsule is of poorer quality than at primary surgery, a more extensive shift or augmentation may be needed, the rotator interval is closed if it was not before, and immobilisation is longer (6-8 weeks).

Nerve injury is uncommon but significant, and the axillary nerve is the one most at risk because of its proximity to the inferior capsule during capsular shift.

The anatomy. The axillary nerve arises from the posterior cord of the brachial plexus (C5-C6), courses along the inferior capsule of the glenohumeral joint and passes through the quadrangular space approximately 1-2cm inferior to the glenoid rim, lying along the anterior-inferior capsule at about the 6 o'clock position anteriorly and 7 o'clock posteriorly on the glenoid face. It is exposed during inferior capsular plication and suture passage through the inferior capsule arthroscopically, during the inferior shift and retractor placement in the open procedure, and historically by thermal energy.

How it presents. An acute injury declares itself when the nerve block wears off: the patient cannot abduct against gravity, cannot initiate abduction from the adducted position, has posterior deltoid weakness affecting extension, and deltoid atrophy develops within 2-3 weeks if the muscle is denervated. Numbness over the lateral shoulder, in the distribution of the superior lateral cutaneous nerve (the sensory branch of the axillary nerve), may be the only finding in a partial injury. Some injuries are only recognised during rehabilitation when weakness persists despite therapy, and EMG confirms the denervation.

Prevention. Arthroscopically the safeguard is the 6 o'clock rule under Surgical Technique, with every suture visualised before tying and no extra-capsular structure caught. In the open procedure, avoid aggressive inferior retraction, use handheld rather than self-retaining retractors and reposition them frequently, and stay within the capsular layers rather than dissecting deep to the capsule where the nerve runs.

Management. If injury is suspected intraoperatively, remove the sutures in the inferior capsule and, in an open procedure, explore to make sure the nerve is not incorporated. Recognised postoperatively, the initial management is observation, since many neuropraxias recover spontaneously over 3-6 months, and early exploration is avoided. EMG and nerve conduction studies at 3-4 weeks establish a baseline and are repeated at 3 months (fibrillation potentials mean denervation, motor unit recruitment means reinnervation); physiotherapy keeps the passive range while waiting, though deltoid strengthening is impossible until reinnervation. Explore if there is no EMG evidence of recovery by 4-6 months, repairing a transected nerve (rare) or performing neurolysis if it is entrapped; with no recovery by 12 months and complete denervation confirmed, consider a trapezius-to-deltoid transfer to restore abduction.

Prognosis. Neuropraxia (stretch or traction without disruption) recovers fully in 80-90% over 3-6 months; axonotmesis recovers in 50-70% over 6-12 months; neurotmesis needs surgical repair and the prognosis is guarded. Counsel about the prolonged recovery, the chance that it is incomplete and the functional impact in the meantime.

Infection. Rates after shoulder surgery are low (under 1%), and diabetic patients are at higher risk. Prevention is preoperative cefazolin, sterile technique and a short operating time. A superficial infection is treated with oral antibiotics; a deep infection needs arthroscopic or open irrigation and debridement with culture-specific intravenous antibiotics, may compromise the integrity of the capsular repair, and may require revision of the repair.

Chondrolysis. Cartilage death with rapid-onset arthritis in a young patient, the devastating complication of thermal capsulorrhaphy and rare with modern plication. The causes are thermal energy (historical), prolonged intra-articular anaesthetic infusion through pain pumps, which led to intra-articular catheters being abandoned, and unknown idiopathic cases. It presents with progressive pain, stiffness and crepitus starting 3-6 months after surgery, rapid joint space narrowing on radiographs and cartilage loss on MRI. There is no effective treatment beyond symptomatic management with activity modification and NSAIDs; arthroplasty may be needed at a young age, with poor outcomes. Prevention is not using thermal capsulorrhaphy, not using intra-articular pain pump catheters, and using only FDA-approved intra-articular substances.

Subscapularis failure. After open shift with subscapularis takedown, failure of the tendon repair causes anterior instability and loss of internal rotation strength. Preoperative subscapularis atrophy, poor tissue quality, inadequate fixation, early aggressive motion and non-compliance are the risk factors; it presents with anterior pain, internal rotation weakness and positive lift-off and belly-press tests, with discontinuity on MRI. Revision repair is possible if identified early; a late presentation may need a pectoralis major transfer to restore internal rotation, and the functional impact is significant. Prevention is secure fixation with anchors or transosseous tunnels, tensioning in 30° of external rotation, and protection with gradual range progression in early rehabilitation.

Heterotopic ossification. Abnormal bone in the soft tissues, rare after shoulder surgery, presenting as progressive stiffness and pain with radiographic calcification; the risk factors are trauma, prolonged immobilisation, genetic predisposition and neurological injury. Observe initially, NSAIDs may reduce progression, and excision is reserved for mature ossification (over 12 months) causing significant functional limitation.

Adhesive capsulitis. True frozen shoulder, distinct from post-surgical stiffness, is inflammation and fibrosis of the capsule that resists treatment: progressive stiffness in all planes, night pain, restricted active and passive range, and a natural history of 12-24 months. Treatment is intensive physiotherapy, NSAIDs and intra-articular corticosteroid injection, manipulation under anaesthesia after 6 months if that fails, and arthroscopic capsular release for refractory cases.

The key findings below reflect what each paper actually reports, not the conclusion it is usually quoted for.