Shoulder dislocation and proximal humeral fracture

X-ray · CT

First and second year — the floor first, then every step

Two views in orthogonal planes, always — a posterior dislocation hides on a single AP. Then count the parts of a proximal humeral fracture by Neer and measure the greater tuberosity displacement and glenoid bone loss.

Orient first

  • Anterior dislocation is far commoner; the humeral head sits below the coracoid. POSTERIOR dislocation (seizure, electric shock) can look almost normal on AP — the "light bulb" head and a widened glenohumeral joint are the clues, and the axillary or Y view proves it.
  • A dislocation leaves impaction marks: Hill-Sachs (posterolateral head) and bony Bankart (anteroinferior glenoid rim) with anterior dislocation; reverse Hill-Sachs (anteromedial head, the "trough") with posterior.
  • Neer counts four parts — head, greater tuberosity, lesser tuberosity, shaft; a part counts as displaced when separated by more than 1 cm or angulated more than 45°.

Acquire the study

  • Grashey (true AP) view and an AXILLARY view (or scapular Y if the axillary cannot be obtained); post-reduction films in the same projections.

The manoeuvre

  • Axillary or Y view: the humeral head must sit centred on the glenoid — anterior (under the coracoid) or posterior dislocation.
  • Grashey view: the glenohumeral joint space should be clear; overlap suggests anterior dislocation, widening (> 6 mm) and a light-bulb head suggest posterior.
  • Humeral head for Hill-Sachs (posterolateral flattening) or the trough line (reverse Hill-Sachs).
  • Glenoid rim for a bony Bankart fragment; greater tuberosity fracture (common with anterior dislocation).
  • Proximal humeral fracture: identify each part, and measure displacement and angulation of each.

What confirms it

  • Dislocation: the head is off the glenoid on an orthogonal view.

What licenses you to exclude it

  • A single AP view cannot exclude a posterior dislocation.

The classic misread

  • Missing a posterior dislocation on a single AP view.
  • Calling a greater tuberosity fracture minimally displaced without measuring superior displacement.

Reporting the injury

Classification to use

  • Neer (2-, 3-, 4-part; displacement > 1 cm or angulation > 45°) or AO/OTA 11. Dislocations: direction (anterior/subcoracoid, posterior, inferior/luxatio erecta).

Measurements — and how to take them

  • Greater tuberosity displacement in mm, especially superiorly (displacement of more than about 5 mm is often operative — verify).
  • Glenoid bone loss as a percentage of the inferior glenoid circle on en-face CT.
  • Neer displacement (> 1 cm) and angulation (> 45°) per part; head–shaft angle.

What to report

  • Located or dislocated (direction); Hill-Sachs / reverse Hill-Sachs; bony Bankart and glenoid bone loss.
  • Fracture: the parts, displacement and angulation of each, head-split, medial calcar comminution, greater tuberosity displacement.

How to report it

  • X-ray: "Anterior (subcoracoid) dislocation of the left shoulder with a displaced greater tuberosity fracture."
  • CT: "Glenoid bone loss of approximately 18% of the inferior glenoid circle (best-fit circle, en-face). Hill-Sachs lesion 22 mm wide."

What not to report

  • Do not call a dislocation reduced on a single AP view.

Associated injuries to look for

  • Axillary nerve and artery injury (inferior and fracture-dislocations), rotator cuff tear (older patients with anterior dislocation), greater tuberosity fracture.

What changes management

  • Posterior dislocation — often missed; urgent reduction.
  • Significant glenoid bone loss or an engaging Hill-Sachs — bony stabilisation surgery.
  • 3- and 4-part fractures, head-split, displaced greater tuberosity — operative options including arthroplasty.

Reference values

Each value carries the caveat that keeps it from being misused. Normal limits and diagnostic criteria are kept apart on purpose: a disease cut-off read as a normal range is the more dangerous mistake.

Normal limits

  • Shoulder · Acromiohumeral distanceapproximately 7–14 mm in adults

    True AP (Grashey) shoulder radiograph with the arm neutral. Shortest distance between the undersurface of the acromion and the top of the humeral head.

    Narrowing under about 7 mm suggests a large or massive cuff tear with superior migration. ⚠️ ARM POSITION AND ROTATION CHANGE IT — measure on a true AP with the arm at the side, and never on an abducted or externally rotated film.

    X-ray · MRI

  • Glenoid · Glenoid versionroughly 2–7° of retroversion

    Axial CT or MRI at the mid-glenoid level, corrected to the scapular plane. The angle between the glenoid face and a line perpendicular to the scapular body axis (Friedman line).

    ⚠️ SCAPULAR-PLANE CORRECTION IS THE MEASUREMENT. An uncorrected axial slice gives a systematically different figure, which is why arthroplasty planning uses reformatted or 3D data. Increasing retroversion with posterior humeral head subluxation defines the Walch B glenoid.

    CT · MRI

  • Glenohumeral joint · Glenohumeral joint-space width

    maintained and concentric on a true AP (Grashey); a single millimetre normal is less honest than concentricity and comparison with the other side

    A non-Grashey AP overlaps the joint and under-calls loss. Superior migration is the cuff-tear arthropathy pattern (see acromiohumeral-distance).

    X-ray

Diagnostic criteria

  • Shoulder · Critical shoulder angleapproximately 30–35°

    True AP (Grashey) radiograph. The angle between a line joining the superior and inferior bony margins of the glenoid and a line from the inferior glenoid margin to the most lateral point of the acromion.

    Above roughly 35° is associated with degenerative cuff tears, below about 30° with glenohumeral osteoarthritis. ⚠️ INVALID ON A NON-TRUE AP — glenoid rim overlap makes the landmarks unidentifiable, and the measurement is highly sensitive to it.

    X-ray

  • Glenoid · Anterior glenoid bone loss

    none; surgically significant above roughly 20–25%

    En-face (sagittal oblique) 3D CT of the glenoid with the humeral head subtracted. A best-fit circle is drawn on the inferior glenoid; loss is expressed as a percentage of that circle's diameter or area.

    ⚠️ THE METHOD CHANGES THE NUMBER. Diameter-based and area-based loss are not interchangeable, and thresholds published for one do not transfer to the other. Read with the Hill–Sachs lesion — the glenoid track concept, not either lesion alone, determines engagement.

    CT · MRI

  • Glenoid labrum (Bankart) · Bankart and bony Bankart — how to report them

    a named injury pattern: anteroinferior labral avulsion (soft Bankart) with or without a glenoid-rim fracture (bony Bankart) — qualitative, not a millimetre diagnosis

    ALPSA, GLAD and HAGL are different labels, not synonyms of Bankart. Name the lesion the surgeon will look for.

    MRI · CT

  • Humeral head · Hill–Sachs lesion — how to report it

    absent; when present, report location, depth and whether it is on-track or off-track relative to the glenoid track — a single depth cut-off is less useful than the track concept

    A posterolateral impaction is the Hill–Sachs; do not invent a millimetre that "needs surgery". Engagement is a glenoid-track calculation, not a depth alone.

    X-ray · CT · MRI

See it on real cases

Direct links to Radiopaedia — the reference article and worked cases with their images. Each opens on Radiopaedia.

Key papers

Reviews and guidelines from RSNA, ESR and related journals. Each opens at its DOI.

  1. Acute shoulder trauma: what the surgeon wants to know ↗Sandstrom CK, Kennedy SA, Gross JA · RadioGraphics 2015RSNA · PubMed
  2. Imaging in anterior glenohumeral instability ↗Bencardino JT, Gyftopoulos S, Palmer WE · Radiology 2013RSNA · PubMed
  3. Imaging signs of posterior glenohumeral instability ↗Shah N, Tung GA · AJR 2009ARRS · PubMed
  4. Traumatic shoulder injuries: a force mechanism analysis of complex injuries to the shoulder girdle and proximal humerus ↗Sheehan SE, Gaviola G, Sacks A, et al. · AJR 2013ARRS · PubMed

More searches

More in Trauma