Floating Shoulder Injuries

A floating shoulder is the combination of ipsilateral clavicle and scapular neck (or glenoid) fractures that results in the loss of the bony connection between the axial skeleton and the upper extremity. The term was coined by Ganz and Noesberger in 1975.

Mechanism. High-energy direct trauma to the shoulder: a motor vehicle accident, a fall from height or a motorcycle accident. The force is applied to the lateral aspect of the shoulder or as axial loading through the humeral head, and the components of the superior shoulder suspensory complex (SSSC) fail in sequence.

Associated injuries. With a high-energy mechanism, 80% have additional injuries and the patient often presents in the polytrauma setting, so a complete trauma work-up is mandatory. The injuries to look for:

• Pulmonary: rib fractures (50%), pneumothorax, pulmonary contusion
• Neurological: brachial plexus injury (5-10%), head injury
• Vascular: subclavian or axillary injury, rare but devastating
• Other: spine injuries, abdominal trauma

The superior shoulder suspensory complex. The SSSC is a bone-soft tissue ring that suspends the upper extremity from the axial skeleton, and understanding it is fundamental to managing the floating shoulder. Its components, in order round the ring:

1. Clavicle - medial strut
2. AC joint - connects clavicle to acromion
3. Acromion process - superior link
4. Scapula spine - posterior connection
5. Scapula body and glenoid neck - lateral strut, supporting the humeral head
6. Coracoid process - inferior link
7. Coracoclavicular ligaments - complete the ring back to the clavicle

Two breaks, not one. Like a pelvic ring, the SSSC requires two disruptions to create instability. A single break, an isolated clavicle or an isolated scapular neck fracture, typically remains stable. Two breaks create a floating segment that can displace under the pull of the attached muscles.

The deforming forces. Once the segment floats, these act on the scapula:

• Serratus anterior protracts the scapula
• Pectoralis minor tilts the glenoid inferiorly
• Gravity and the weight of the arm cause medialisation
• Trapezius and levator scapulae elevate the medial scapula, if intact

Why the clavicle matters. The clavicle is the only bony connection between the upper limb and the axial skeleton. Fixing it restores the strut function supporting the shoulder, and through intact CC ligaments it can indirectly reduce the scapula.

View larger

History. The mechanism is high-energy trauma, a motor vehicle accident or a fall from height. Ask about chest pain, dyspnoea and head injury for the associated injuries, and record hand dominance, occupation and functional demands, and comorbidities that affect healing.

Examination. Inspect for swelling, deformity and skin tenting: the shoulder looks shortened and drooping, with a visible or palpable clavicle deformity and ecchymosis over the shoulder girdle. Palpate the clavicle, acromion and scapular spine; the scapular body may be palpable posteriorly. Range of motion is usually severely limited by pain, and the patient cannot actively elevate the arm.

Screening the associated injuries. The chest, the plexus and the circulation are examined as part of the trauma assessment:

• Chest: breath sounds, auscultation for pneumothorax, chest wall tenderness
• Neurology: a complete brachial plexus examination is essential
• Vascular: pulses, capillary refill, expanding haematoma

Differential diagnosis. The table separates the floating shoulder from the patterns that mimic it.

View larger

Radiographs. The initial series covers the chest as well as the shoulder girdle:

• AP chest - pulmonary injury, rib fractures
• AP clavicle - shortening, displacement
• Scapular Y view - glenoid position, GPA measurement
• Axillary lateral - glenoid articular surface
• AP shoulder - glenohumeral relationship

CT is the definitive study. With 3D reconstruction it shows the fracture pattern and displacement, the size and displacement of any glenoid articular involvement, and the medialisation of the scapular neck, and it is the basis of surgical planning: plate contouring and screw trajectories.

Further studies, if indicated.

• CT angiography, if vascular injury is suspected
• MRI for brachial plexus injury, in the subacute phase
• EMG and nerve conduction studies for delayed brachial plexus assessment

View larger

View larger

View larger

The measurements. Four measurements decide how displaced the injury is. The thresholds below are the working ones used on this page; published cut-offs vary, as discussed under Controversies.

View larger

The glenopolar angle (GPA) is the single most-cited number in this topic, yet candidates routinely fumble when asked how it is actually measured. Being able to draw it on demand separates a confident answer from a vague one.

Technique. As originally described by Bestard:

• Obtain a true AP (Grashey) or scapular Y view. A rotated film distorts the angle, so a reproducible projection, or CT or a 3D reconstruction, is preferred
• Line 1 connects the most cranial and most caudal points of the glenoid cavity (the glenoid rim line)
• Line 2 connects the most cranial point of the glenoid to the most caudal point of the scapular body (the inferior angle)
• The GPA is the angle subtended between these two lines

View larger

View larger

Interpretation. Normal is 30-45°, with a mean around 40°. Under 20° signifies significant caudal tilt, or inferior angulation, of the glenoid and is the classic operative threshold. A falling GPA reflects the glenoid rotating and dropping away from the scapular body, the deformity that impairs the concavity-compression and the abduction lever arm of the glenohumeral joint.

Why it drives the decision. The GPA is the parameter that best tracks function. In the Lin randomised study the GPA after bony consolidation correlated strongly with both DASH and Constant scores, and combined clavicle-plus-scapula fixation restored it best. In van Noort's series it was the caudally dislocated glenoid, a low-GPA equivalent, that produced the worst results (Constant 42 versus 85). Restoring glenoid position and GPA, by clavicle fixation and, if the GPA stays low, scapular ORIF, is therefore the true goal of treatment, not simply counting SSSC breaks.

A practical caveat. GPA measurement is projection-dependent and carries genuine interobserver variability on plain films. An isolated borderline value should be confirmed on a reproducible view or on CT before it is used to justify, or to refuse, surgery.

The decision. Count the disruptions, then measure the displacement. A single disruption is usually stable and treated nonoperatively. A double disruption is judged on its displacement: minimally displaced injuries are considered for nonoperative treatment or clavicle fixation alone, and significantly displaced ones for surgical stabilisation.

Clavicle first. Once surgery is chosen, the clavicle is fixed before any decision on the scapula:

1. Plate the clavicle
2. Reassess scapular alignment on intraoperative fluoroscopy
3. If alignment is restored, stop
4. If displacement persists, proceed to scapular ORIF

This stepwise approach avoids unnecessary scapula surgery in most cases.

What the evidence says about clavicle-first. Fixing the clavicle restores the strut function of the SSSC, and through intact CC ligaments this can indirectly reduce the scapula. Be honest about the evidence: no cited study on this page gives a rate for acceptable alignment after clavicle fixation alone. The Dombrowsky systematic review compared operative with nonoperative treatment, and the Lin RCT found that combined fixation restored the GPA best. The clavicle-first strategy is a way of finding out, intraoperatively, which group the patient belongs to; it is not a proven 60-70% answer.

Because roughly 80% of these injuries occur in a polytrauma patient, when to operate is often a harder question than whether to operate. The shoulder girdle must never be prioritised over the resuscitation.

The girdle injury is rarely the emergency. With two exceptions, an open fracture and a vascular injury requiring repair, the floating shoulder itself is not time-critical. The life-threatening companions take priority: tension or simple pneumothorax and haemothorax, flail chest and pulmonary contusion, closed head and cervical-spine injury, and the dangerous mimic, scapulothoracic dissociation with subclavian or axillary disruption. ATLS priorities and the trauma team lead the first hours; the girdle is addressed once the patient is stabilised.

Damage control versus early appropriate care. In the physiologically deranged or borderline patient (persistent acidosis, high lactate or base deficit, coagulopathy, ongoing transfusion requirement, or a severe chest or head injury), definitive shoulder-girdle fixation is deferred. There is no physiological cost to waiting, and a prolonged lateral or prone scapular approach is poorly tolerated by an unstable patient. Once resuscitation endpoints normalise, early appropriate care applies and fixation proceeds.

The practical window. Clavicle-first fixation, and scapular ORIF if needed, is best performed while reduction is still straightforward, typically within the first one to two weeks, before early callus makes an indirect clavicle-led reduction of the scapula harder to achieve. Later fixation remains possible but is technically more demanding. In a marginal patient the two-stage option is legitimate: fix the clavicle acutely, reassess the scapula on imaging over the following one to two weeks, and add scapular fixation only if the glenoid stays displaced, which shortens the initial anaesthetic.

When the plexus is injured. In the case shown below, CT demonstrated clavicle, scapular-neck and proximal-humerus injury, and MRI showed an infraclavicular haematoma obscuring plexus assessment, which emphasises staged orthopaedic and nerve planning in polytrauma.

View larger

Prognostic factors. Good prognostic factors are an anatomic reduction, a single operation (clavicle alone), early mobilisation, a younger patient and an isolated injury without polytrauma. Poor ones are persistent glenoid malposition, intra-articular involvement, an associated brachial plexus injury, delayed treatment, and polytrauma with ICU admission.

The union-rate figures below are conventional teaching ranges rather than single-study measurements. The verified outcome data on this page are the Dombrowsky systematic review, the Lin RCT and the van Noort series, set out under the glenopolar angle and Controversies.

Global epidemiology

• Scapular fractures account for under 1% of all fractures; the floating shoulder (ipsilateral clavicle + scapular neck) is a small subset, reported in roughly 0.1% of fractures.
• No source cited on this page quantifies how often a scapular neck or body fracture has an associated ipsilateral clavicle fracture; the combination is uncommon enough that the largest synthesis available pooled only 371 shoulders from 17 studies.
• Predominantly young men (mean age late 30s) injured in motor vehicle and motorcycle crashes and falls from height; the high-energy mechanism explains the high rate of associated thoracic and neurovascular injury worldwide.

Registry note: There is no dedicated implant registry for the floating shoulder (it is a fracture-fixation, not an arthroplasty, problem), so evidence rests on small series and a single small RCT rather than registry data - a key reason equipoise persists.

High- vs limited-resource practice variation

• High-resource: routine CT with 3D reconstruction for planning, precontoured clavicle and scapular plates, intraoperative fluoroscopy to confirm GPA, single-stage combined fixation when needed.
• Limited-resource: reliance on plain films (scapular Y plus axillary), conservative management for all but grossly displaced injuries, and clavicle-only fixation favoured to minimise theatre time and the morbidity of the posterior Judet approach.
• Across settings the converging principle is the same: reduce the glenoid/restore GPA and treat the patient's overall trauma burden first.

The floating shoulder is a classic viva topic precisely because the evidence is thin and genuine equipoise exists. Be ready to argue both sides.

Is the injury inherently unstable? Goss's biomechanical concept predicts instability from any double disruption, yet van Noort's multicentre series found the injury is not inherently unstable and that good function follows conservative care unless the glenoid is caudally dislocated. The modern view is that instability depends on displacement, not on the count of disruptions.

Operate or not? Dombrowsky's systematic review found comparable mean Constant scores (≈80% of ideal) for the operative and nonoperative groups, but the small RCT (Lin et al.) favoured fixation for GPA and DASH/Constant. Selection, not a blanket rule, drives outcome.

Clavicle alone or combined? Clavicle-first with intraoperative reassessment is widely taught, but indirect reduction of the scapula through intact CC ligaments is unreliable. Some surgeons argue that a displaced scapular neck should be addressed directly rather than hoping the clavicle plate corrects the GPA.

Which radiographic threshold? Cited GPA cut-offs vary (under 20° vs under 22°), as do medialisation (over 10mm vs over 20mm) and angulation (over 40° vs over 45°) thresholds. These numbers are expert-derived, not validated prospectively; treat them as guides, not absolutes.