Open repair of full-thickness rotator cuff tears via the deltoid-splitting or deltopectoral approach β the technique of choice for complex, retracted and revision tears. Β· intermediate
- Indicated for symptomatic full-thickness tears failing 3β6 months of conservative care β especially complex patterns (L-shaped, massive U-shaped), retracted tears greater than 3 cm, and revision surgery after failed arthroscopic repair.
- Two approaches: the DELTOID-SPLITTING (anterolateral) approach for isolated posterosuperior tears, and the DELTOPECTORAL approach when the subscapularis is involved or combined open work is needed.
- The AXILLARY NERVE enters the deltoid 5β7 cm below the lateral acromion β limit the deltoid split to less than 5 cm from the acromion edge to avoid paralysing the deltoid.
- Tear pattern dictates technique: crescent tears repair directly to the footprint, U-shaped tears need MARGIN CONVERGENCE first, and a double-row suture bridge is the biomechanical gold standard for medium and large tears.
- Deltoid closure is critical in the deltoid-splitting approach β heavy non-absorbable mattress sutures; deltoid failure is a devastating complication with permanent weakness.
When & Why
Indications Open repair is chosen for symptomatic full-thickness rotator cuff tears where direct visualisation and tissue palpation help most. Absolute indications
- Full-thickness rotator cuff tear in a symptomatic patient who has failed conservative management (physiotherapy, NSAIDs, corticosteroid injection over 3β6 months)
- Acute traumatic tear in a young active patient (less than 60 years) with good tissue quality
- Massive or complex tear patterns requiring extensive mobilisation and direct visualisation (large L-shaped, massive U-shaped tears)
- Revision surgery after failed arthroscopic repair, needing superior exposure and mobilisation Relative indications
- Surgeon preference for direct tissue visualisation and palpation
- Concurrent procedures best performed open (extensive subscapularis repair, complex biceps tenodesis, capsular releases)
- Retracted tears (Patte Grade 2β3) requiring aggressive mobilisation including interval slides
- Patient factors: inability to tolerate beach-chair positioning with traction, or concerns about fluid extravasation in arthroscopy Contraindications Absolute
- Active infection (septic arthritis, overlying cellulitis)
- Severe medical comorbidities precluding safe general anaesthesia
- Complete passive stiffness (frozen shoulder) β restore motion first, before repair
- Severe cuff arthropathy (Hamada Grade 4β5) β consider reverse total shoulder arthroplasty instead Relative
- Advanced fatty infiltration (Goutallier Grade 4) β irreversible muscle change, poor repair prognosis
- Extremely poor tissue quality preventing secure fixation
- Significant osteoporosis preventing anchor purchase
- Patient unwilling or unable to comply with a prolonged rehabilitation protocol (6β12 months) Compensation or work-injury status and active smoking are prognostic risk factors for inferior outcomes, not contraindications β counsel accordingly rather than withholding indicated surgery. Which approach, and why The choice between the two open exposures is dictated by which tendons are torn and what else must be done:
For isolated posterosuperior tears (supraspinatus, infraspinatus) and most medium-to-large tears. Smaller 6β8 cm incision, fast to develop, direct access to the subacromial space. The deltoid raphe between the anterior and middle deltoid is split vertically. Cost: axillary nerve risk if the split runs too far distally, and limited subscapularis access.
For subscapularis tears, combined anterior and posterosuperior pathology, and revision or complex reconstruction needing an extensile exposure. Works the internervous plane between deltoid (axillary nerve) and pectoralis major (pectoral nerves); the cephalic vein is retracted laterally with the deltoid. No deltoid detachment, superior anterior access. Cost: longer 10β12 cm incision, cephalic vein at risk, working around the coracoid and conjoint tendon.
Open versus arthroscopic. Both techniques achieve equivalent functional outcomes when performed well by experienced surgeons. Choose OPEN for complex tear patterns needing extensive mobilisation, revision cases, substantial subscapularis involvement, concurrent open procedures, and when direct tissue palpation is wanted. Arthroscopy offers less deltoid morbidity, potentially less stiffness, better cosmesis and a lower infection risk. Surgeon experience and patient selection matter more than the technique itself.
The Operation
The goal is to expose the tear, assess it, mobilise the tendon back to the footprint without excess tension, and re-attach it with a durable double-row construct β then close the deltoid soundly. The exposure is laid out as the first steps below.

Operative sequence β deltoid-splitting approach
- Beach-chair position, 60β70Β° upright; affected shoulder at the table edge for circumferential access; arm free or in a pneumatic holder with slight abduction and neutral rotation.
- Palpate and mark the acromion (anterior, lateral, posterior borders), the AC joint, the clavicle, the coracoid (2β3 cm below the clavicle) and the greater tuberosity.
- Anterolateral incision: from the anterolateral corner of the acromion, 6β8 cm distally in Langer's lines. For the deltopectoral approach, incise from the coracoid 10β12 cm distally over the deltopectoral groove.
- Identify the deltoid fascia and the raphe between the anterior and middle deltoid (a visible slight depression).
- Split the deltoid vertically in the line of the fibres, beginning at the anterior acromion.
- Limit the split to less than 5 cm from the acromion edge β the axillary nerve enters the deltoid 5β7 cm below the lateral acromion.
- Place stay sutures (0 Vicryl) on the split edges for retraction and for later anatomic closure.
- Incise the subacromial bursa; place self-retaining retractors (Gelpi, Weitlaner).
- Perform a thorough bursectomy to expose the undersurface of the acromion and the cuff tear, preserving enough bursa for closure.
- If the acromion is Type 2 (curved) or Type 3 (hooked) or carries an anterior spur (Bigliani classification), perform an acromioplasty: resect 5β8 mm of anteroinferior acromion to a flat Type 1 undersurface, preserving the deltoid origin.
- Size: small (less than 1 cm), medium (1β3 cm), large (3β5 cm), massive (greater than 5 cm or two or more tendons).
- Pattern: crescent (direct repair), U-shaped (margin convergence first), L-shaped (repair the corner first), massive (interval slides Β± partial repair).
- Retraction (Patte): 1 at the tuberosity, 2 at the glenoid rim, 3 past the glenoid.
- Atrophy: Goutallier Grade 0β4 on MRI (Grade 3β4 poor prognosis).
- Integrity of tissue and ROM after mobilisation: the tendon must reach the footprint without excess tension.
- Release adhesions to the acromion and bursa; release the coracohumeral ligament anteriorly.
- For massive retracted tears, perform interval slides: an anterior slide (release the rotator interval between subscapularis and supraspinatus) or a posterior slide (release infraspinatus from the posterior capsule at the glenoid).
- Stay lateral to the glenoid rim β the suprascapular nerve runs at the spinoglenoid notch about 2 cm medial to the glenoid. Limit medial dissection to less than 2 cm beyond the rim.
- For a U-shaped tear, place side-to-side horizontal mattress sutures (#2 high-strength, e.g. FiberWire) along the anterior and posterior limbs, spaced 1 cm apart.
- Tie them to close the U, converting it to a crescent β this reduces the medial-lateral dimension and the strain on the final footprint repair by 20β40%.
- Then repair the converged crescent edge to the footprint.
- Clear the greater tuberosity footprint β superior facet for supraspinatus, middle for infraspinatus, inferior for teres minor β of scar and degenerate tendon.
- Lightly decorticate to a punctate bleeding surface with a curette or burr β enough for biology, without over-removing bone (which weakens anchor purchase).
- Place 2β3 suture anchors at the articular margin, 5β10 mm apart, at a 45Β° deadman angle; use the largest anchor the bone allows (typically 4.5β5.5 mm); threaded anchors grip better in soft bone.
- Pass the medial sutures through the tendon in a mattress configuration, biting about 1 cm from the edge with a minimum 5 mm tissue bridge.
- Tie a sliding-locking knot plus three alternating half-hitches and two backups. Aim for firm tendon-to-footprint contact without over-tensioning (which causes ischaemia and cut-through) β tie medial first, progressing laterally.
- Carry the free medial limbs to two or three knotless anchors placed 10β15 mm lateral to the medial row at the lateral footprint edge, at 45Β°.
- Inserting these captures the sutures and creates a compressive suture bridge β broad tendon-bone contact, even force distribution, and no lateral knots to impinge.
- The tendon should cover the footprint fully with no gap at rest and stay stable through passive forward flexion, abduction and rotation (if it gaps on motion, the repair is over-tensioned β revise with more release or accept partial repair).
- Biceps: tenotomy (low-demand) or tenodesis (active, prevents a Popeye deformity) if degenerative, subluxed or more than 50% torn.
- AC joint: resect 5β10 mm of distal clavicle if symptomatic.
- Subscapularis: repair to the lesser tuberosity and protect external rotation post-op.
- Re-approximate the deltoid anatomically using the stay sutures as a guide, with heavy non-absorbable sutures (#1 Ethibond) in an interrupted mattress or figure-of-8 configuration and adequate tissue bites.
- Deltoid failure is a devastating complication β take the same care over closure as over the repair itself.
- Layered skin closure.
The axillary nerve exits the quadrangular space with the posterior circumflex humeral artery and enters the deltoid 5β7 cm below the lateral acromion edge. It is the most commonly injured nerve in open cuff repair and its injury paralyses the deltoid. Limit the deltoid split to less than 5 cm from the acromion, split vertically in the raphe, tag the edges with stay sutures, and retract gently. If the nerve is identified, protect it with a vessel loop.
The five danger structures β know each location and how to protect it:
Enters the deltoid 5β7 cm below the lateral acromion; injury causes deltoid paralysis and lateral-arm numbness (regimental badge area). Protect by limiting the split to less than 5 cm, splitting vertically in the raphe, and gentle retraction.
Through the suprascapular notch, then the spinoglenoid notch about 2 cm medial to the glenoid rim; innervates supra- and infraspinatus. Protect by limiting medial dissection to less than 2 cm beyond the rim during interval slides.
Enters coracobrachialis 3β8 cm distal to the coracoid (variable); relevant in the deltopectoral approach. Protect by identifying the coracoid and avoiding dissection more than 3 cm distal during subscapularis mobilisation.
Lies in the deltopectoral groove; injury causes haematoma. In the deltopectoral approach, identify it and retract laterally with the deltoid to preserve venous drainage; ligate only if necessary.
The anchor bed; often osteoporotic in the elderly. Avoid over-decortication and over-torquing anchors; insert at a 45Β° deadman angle and space anchors 5β10 mm apart to prevent fracture and pullout.
In the deltoid-splitting approach, sound deltoid closure with heavy non-absorbable mattress or figure-of-8 sutures, using the stay sutures as a guide, is as important as the cuff repair. Deltoid detachment produces permanent abduction weakness and a poor salvage outcome.
Aim for full tendon-to-footprint contact without blanching the tissue. Over-tensioning causes ischaemia and suture cut-through; under-tensioning leaves gaps that will not heal. After tying the medial row, test passive ROM β the repair must not gap with motion.
Aftercare & Complications
Rehabilitation β balance protection against stiffness. Tendon-to-bone healing takes 12β16 weeks; too-aggressive motion fails the repair, too-conservative motion stiffens it. | Phase | Timing | Immobilisation | Motion | |-------|--------|----------------|--------| | 1 | 0β6 weeks | Sling or abduction brace at all times except exercise and hygiene | Passive only β table slides to 90β120Β° flexion and 30β40Β° external rotation; pendulums. NO active motion | | 2 | 6β12 weeks | Wean sling for activities of daily living | Active-assisted ROM (wand, pulley); progress to full passive ROM; light isometrics under 20% if comfortable | | 3 | 12β16 weeks | None for light activity | Active ROM against gravity; progressive resistance (Theraband, 1β5 lb); cuff and scapular strengthening | | 4 | 16+ weeks | None | Advanced strengthening (5β10+ lb); sport- and work-specific training; plyometrics for athletes | Return to driving at 6β8 weeks (off narcotics), light work at 8β12 weeks, heavy overhead labour at 4β6 months, golf or swimming at 4β6 months, throwing sports at 6β9 months, and contact sports at 9β12 months. For a subscapularis repair, limit external rotation to 0β20Β° for the first 6 weeks. Massive tears may need 8 weeks of passive-only protection with lower ROM and strength goals, and workers'-compensation patients often progress more slowly (maximum medical improvement at 9β12 months). Full biological remodelling takes 12β18 months. Early active motion equals repair failure β patient education and a written protocol are essential.
- Recognition
- Return of pain and weakness; loss of active elevation; ultrasound or MRI arthrogram shows a recurrent defect
- Prevention
- Optimal tension and fixation, adequate mobilisation, smoking cessation, diabetes control, strict passive-only early rehab
- Management
- Conservative for low-demand or elderly; revision if young, acute, good tissue; SCR or reverse TSA for irreparable
- Recognition
- Progressive loss of passive and active ROM, pain with stretching, develops at 6β12 weeks
- Prevention
- Early passive ROM within 1β2 weeks; appropriate immobilisation (not over 6 weeks)
- Management
- Aggressive PT, intra-articular steroid, manipulation under anaesthesia or arthroscopic capsular release if recalcitrant
- Recognition
- Severe abduction weakness, palpable deltoid defect, lateral shoulder pain
- Prevention
- Heavy non-absorbable mattress or figure-of-8 deltoid closure using stay sutures; protect the deltoid origin; limit split to less than 5 cm; no active deltoid for 6 weeks
- Management
- Acute (under 6 weeks): urgent revision repair with heavy suture, Β± transosseous tunnels. Chronic: poor salvage β prevention is essential
- Recognition
- Early: erythema, drainage, dehiscence. Late: persistent pain, sinus tract
- Prevention
- Prophylactic antibiotics (cefazolin 2 g IV within 60 min), diabetic and smoking optimisation, sterile technique, minimise operative time
- Management
- Superficial: antibiotics and local care. Deep: urgent debridement, culture-directed IV antibiotics for 6 weeks, retain anchors if well-fixed
- Recognition
- Unable to contract the deltoid, absent abduction, numb lateral upper arm (regimental badge)
- Prevention
- Limit split to less than 5 cm; vertical split in the raphe; gentle retraction
- Management
- Most are neurapraxias β observe 3β6 months with EMG; explore and graft if no recovery by 6 months
- Recognition
- Supra- and infraspinatus weakness and atrophy, chronic pain
- Prevention
- Limit medial dissection to less than 2 cm beyond the glenoid rim during interval slides
- Management
- Observation; most leave some permanent deficit β compensatory strengthening
- Recognition
- Pullout: fixation gives way, repair gaps. Penetration: chondral damage, catching pain
- Prevention
- 45Β° deadman angle, largest anchor the bone allows, threaded in osteoporotic bone, medial row at the articular margin not beyond it
- Management
- Pullout: re-place in better bone. Penetration: remove and re-place laterally if symptomatic
Viva & Exam Focus
REPAIRREPAIR β rotator cuff assessment framework
ANCHORSANCHORS β double-row repair principles
Clinical Decision Scenarios
Practise clinical reasoning and management decisions out loud
βA 58-year-old manual labourer has 6 months of shoulder pain and weakness after a fall. MRI shows a 4 cm full-thickness supraspinatus and infraspinatus tear retracted to the glenoid rim with Goutallier Grade 2 fatty infiltration, having failed 3 months of physiotherapy. How would you manage him, and would you choose open or arthroscopic repair?β
βDuring an open deltoid-splitting cuff repair you cannot reduce a retracted supraspinatus tear to the footprint despite subacromial and coracohumeral releases; the edge sits at the glenoid rim. What further mobilisation do you use, and what is at risk?β
βYou have placed the medial-row anchors and passed the mattress sutures. As you tie, what tension are you aiming for, and how do you assess it?β
Indications
- Symptomatic full-thickness tear failing 3β6 months conservative care
- Complex patterns (large L-shaped, massive U-shaped) needing mobilisation
- Revision after failed arthroscopic repair
- Subscapularis tears and concurrent open procedures (biceps, AC joint, capsular release)
Approach choice
- Deltoid-splitting (anterolateral) for isolated posterosuperior tears
- Deltopectoral for subscapularis, combined or revision cases
- Open versus arthroscopic: equivalent outcomes when done well
Danger structures
- Axillary nerve: 5β7 cm below the acromion β limit split to less than 5 cm
- Suprascapular nerve: 2 cm medial to the glenoid β limit medial dissection
- Musculocutaneous nerve: 3β8 cm distal to the coracoid (deltopectoral)
- Cephalic vein: retract laterally with the deltoid
Core operation
- Split in the deltoid raphe, bursectomy, Β± acromioplasty
- Assess with REPAIR; mobilise; margin convergence for U-shaped tears
- Double-row suture bridge: medial-row mattress plus lateral knotless anchors
- Critical deltoid closure with heavy non-absorbable mattress sutures
Tear patterns
- Crescent: direct repair to the footprint
- U-shaped: margin convergence first
- L-shaped: repair the corner first
- Massive: interval slides Β± partial repair
Anchor technique
- 45Β° deadman angle, largest anchor the bone allows (4.5β5.5 mm)
- Medial row 2β3 anchors at the articular margin, lateral row 2β3 knotless 10β15 mm lateral
- Mattress bites 1 cm from the edge, minimum 5 mm tissue bridge
Post-op
- Passive ROM only 0β6 weeks (protect the repair)
- Active-assisted 6β12 weeks, active at 12 weeks, strengthening after 3 months
- Full recovery 6β12 months; subscapularis repair β limit ER to 0β20Β° for 6 weeks
Complications
- Re-tear 10β30% (massive tears, age over 65, smoking)
- Stiffness 5β15%; deltoid detachment 1β3% (devastating)
- Infection 0.5β2%; axillary nerve injury 0.5β1%
Background & Evidence
Epidemiology. Full-thickness rotator cuff tears are common and increase with age, found in a substantial proportion of asymptomatic older adults and in the majority of those over 70. Most degenerative tears are managed non-operatively; surgical repair is reserved for the symptomatic tear that fails structured conservative care, and for acute traumatic tears in active patients, where earlier repair limits retraction and fatty infiltration. Classification (reference). Tears are described by size, pattern, retraction (Patte) and fatty infiltration (Goutallier) β the same axes used intra-operatively in the REPAIR framework:
- Dimension
- Less than 1 cm
- Dimension
- 1β3 cm
- Dimension
- 3β5 cm
- Dimension
- Greater than 5 cm or two or more tendons
- Tendon edge position
- At the level of the greater tuberosity
- Tendon edge position
- Retracted to the level of the glenoid rim
- Tendon edge position
- Retracted past the glenoid (medial)
- Muscle-to-fat ratio
- Normal muscle
- Prognostic note
- Good
- Muscle-to-fat ratio
- Some fatty streaks
- Prognostic note
- Good
- Muscle-to-fat ratio
- More muscle than fat
- Prognostic note
- Repairable
- Muscle-to-fat ratio
- Equal muscle and fat
- Prognostic note
- Guarded β discuss expectations
- Muscle-to-fat ratio
- More fat than muscle
- Prognostic note
- Largely irreversible β poor prognosis
Outcomes. Structural healing on post-operative imaging falls with tear size: small tears 90β95%, medium 80β90%, large 70β85%, massive 50β70%, revision 40β60%. Around 80β90% of patients achieve good-to-excellent pain relief and satisfaction, 70β85% a functional range of motion, and 60β80% functional (rarely normal) strength. Poor-outcome predictors are large or massive tear size, age over 65, chronicity over 6 months, Goutallier Grade 3β4, poor tissue quality, smoking, diabetes, workers' compensation, and rehab non-compliance. Open versus arthroscopic β evidence. Meta-analysis (Morse 2008) shows no significant functional difference between all-arthroscopic and mini-open deltoid-splitting repair, with lower infection and stiffness favouring arthroscopy. Surgeon experience and patient selection outweigh the technique. Single- versus double-row. Double-row suture-bridge constructs are biomechanically superior β higher load to failure, larger footprint contact, better gap resistance (Park 2007: 77.6% versus 39.6% footprint contact). Clinically the functional advantage is modest (Sheibani-Rad 2013 meta-analysis), so double-row is preferred for medium and large tears while a well-executed single-row repair remains acceptable for small tears or poor tissue and bone. Technique evidence. Margin convergence reduces strain at the tear edge by 20β40% (Burkhart 1996) and is mandatory for U-shaped tears. Routine acromioplasty is not required β recent RCTs show no outcome difference, so it is selective (Type 3 acromion or anterior spur). Modern suture anchors match the biomechanics of historical transosseous tunnels and are faster to place. Guidelines, registries and global practice | Body / source | Position relevant to cuff repair | |---|---| | AAOS (US) Clinical Practice Guideline | Surgical repair is an option for symptomatic full-thickness tears; routine acromioplasty not required; non-operative care reasonable for many degenerative or irreparable tears | | BOA / BESS (UK) | Structured physiotherapy first for atraumatic degenerative tears; earlier surgery for acute traumatic tears in active patients to limit retraction and fatty infiltration | | AO Foundation / international consensus | Tear pattern, retraction (Patte) and fatty infiltration (Goutallier) drive reparability and technique selection, not chronological age alone | Global practice varies: arthroscopic repair predominates in well-resourced settings; open repair remains essential where arthroscopy access or training is limited, and for complex, revision and subscapularis cases everywhere. Double-row constructs cost more for a modest, inconsistent functional gain, so single-row remains defensible in resource-limited settings. For irreparable tears, the realistic salvage pathway (superior capsular reconstruction, tendon transfer, reverse TSA) depends on regional availability.
References
Arthroscopic vs mini-open repair: equivalent outcomes
- Meta-analysis of five Level IβIII comparative studies (minimum 1-year, mean over 2-year follow-up) using validated shoulder outcome scores
- No significant functional-score difference between all-arthroscopic and mini-open (deltoid-splitting) repair
- No difference in complication rates between the two approaches
Single-row vs double-row: similar clinical outcomes
- Meta-analysis of five Level I RCTs comparing single- and double-row repair (homogeneous, no significant heterogeneity)
- No significant difference in ASES, Constant or UCLA scores between constructs
- Despite superior double-row biomechanics, a consistent clinical functional benefit is not demonstrated
Transosseous-equivalent suture bridge maximises footprint contact
- Cadaveric pressure-film study: a 4-suture-bridge transosseous-equivalent construct restored 77.6% of footprint contact versus 39.6% for a conventional double-row
- Mean interface pressure was higher for the suture bridge (0.27 MPa) than the double-row (0.19 MPa, P=0.002)
- Bridging medial-row sutures laterally optimises tendon-to-bone contact and compression
Margin convergence reduces strain in massive tears
- Seminal description of margin convergence: side-to-side repair of the free margins of a U- or L-shaped tear before footprint fixation
- Converging the margins markedly reduces strain at the tear edge, so weaker bone-tendon fixation becomes adequate
- Lower strain also reduces cuff mechanoreceptor stimulation, contributing to pain relief
High re-tear rate in large and massive tears; function can persist
- 18 patients with tears over 2 cm repaired; recurrent defects on ultrasound in 17 of 18 at 12 months
- Despite non-healing, ASES scores improved (48.3 to 84.6) at 1 year with good pain relief and restored elevation
- Results deteriorated by 2 years, with fewer patients maintaining high scores
Goutallier fatty infiltration predicts repair outcome
- Established the 5-stage (Grade 0β4) classification of muscle fatty degeneration on imaging in 63 cuff-repair patients
- Fatty infiltration worsens with time; infraspinatus degeneration did not regress after repair and often progressed
- Greater degeneration correlated with worse pre- and post-operative external-rotation function and worse repair outcomes
Double-row has higher load to failure than single-row
- Biomechanical study: double-row repairs had a 30β50% higher ultimate load to failure than single-row
- Superior gap resistance under cyclic loading
Repair integrity correlates with functional outcome
- Correlated repair integrity on ultrasound with functional outcomes over time
- Healed repairs had better strength and function; pain relief could occur even with a re-tear
- Emphasises patient selection and managing expectations
Subscapularis transposition for chronic cuff tears
- Classic description of open rotator cuff repair techniques, including subscapularis transposition for massive irreparable tears
- Historical gold-standard reference for open cuff reconstruction
Early arthroscopic repair β outcomes at 2 and 3 years
- Early arthroscopic repair outcomes demonstrating feasibility and challenges
- Drove the evolution of both arthroscopic and refined open techniques