Shoulder — Fast Exam

Tap a diagnosis. Tests first, interpretation underneath.

30-second screen

Run this before any special testRed flags → passive ROM → resisted tests → point to the pain

Red flags? If any → stop, they're listed at the bottom.
Passive ROM, both sides. Is external rotation lost? → capsulitis, and every other test is now unreliable.
Painful arc + Hawkins. Lateral pain, worse overhead → subacromial. Weak screen, but fast.
Resisted tests in three planes — abduction (supra), external rotation (infra), hand-behind-back / internal rotation (subscap). Pain alone is not a tear. Weakness is.
Point to the pain. On top → AC. Anterior, in the groove → biceps. Deep + click overhead → labrum. Medial scapula → scapular / neck.
Apprehension + relocation. Only if the history has dislocation, subluxation or "it feels like it will come out".
Distal to the elbow, or pins and needles? → screen the neck before the shoulder.

The honest caveat: almost every shoulder special test is weak on its own. Use them to shift a pre-test probability you already formed from the history — not to make the diagnosis.

Rotator cuff & subacromial

Subacromial pain syndromeImpingement / bursitis / cuff tendinopathy — lateral pain, worse overhead

Hallmark

Lateral shoulder pain, worse with overhead and behind-the-back reach. Painful arc. Night pain lying on that side.

Tests

Painful arc — abduct slowly. Positive: pain between 60–120° that eases above shoulder height.
Hawkins–Kennedy — shoulder 90° flexion, elbow 90°, then internally rotate. Positive: pain.
Neer sign — stabilise the scapula, passively flex with internal rotation. Positive: pain.
Scapular assistance test — assist upward rotation/posterior tilt as the arm elevates. Positive: pain reduced or ROM improved → the scapula is contributing, and this doubles as treatment.

Next: Hawkins is the best single test here and it's still weak (Sn .58 / Sp .67). If weakness overshadows pain, stop calling it impingement and test the cuff properly.

Supraspinatus tearWeakness on resisted abduction — pain alone is not a tear

Tests

Jobe / Empty can — arm 90° in the scapular plane, full internal rotation (thumb down), resist downward. Positive: pain or weakness.supraspinatus test, pooled: Sn .74 / Sp .77
Full can — same position, thumb up, resist downward. Positive: pain or weakness.Sn 70% / Sp 81% · LR+ 3.75 Better than empty can at ruling a tear in.
Drop arm — passively abduct to 90°, release. Positive: arm drops, or cannot lower slowly and smoothly → full-thickness tear.LR+ 6.45a positive means a lot, a negative means nothing

Next: weakness → tear. Pain with normal strength → tendinopathy. If weakness is new and severe, test range actively before passively — a massive tear is a different conversation.

Infraspinatus / teres minor tearWeakness in external rotation — often surprisingly painless

Tests

External rotation lag sign (ERLS) — elbow 90°, bring the arm to near-full passive external rotation, release. Positive: the arm lags or drops back.Hertel: Sn 70% / Sp 100% · Castoldi: Sn 56% / Sp 98%
Hornblower's / Patte — 90° abduction, elbow 90°, externally rotate against resistance. Positive: weakness.Sp 96% / Sn 17%positive = very likely tear; negative tells you little
Resisted external rotation at the side — elbow at the waist. Positive: weakness → posterior cuff.

Next: isolated external rotation weakness with preserved elevation is the classic posterior cuff pattern. Compare both sides by feel — subtle lags matter.

Subscapularis tearWeak internal rotation — a positive test is meaningful, a negative is not

Tests

Lift-off (Gerber) — hand behind the back at the lumbar spine, lift the hand off. Positive: cannot lift → tear.
Belly-press / Napoleon — hand on the belly, elbow anterior to the mid-axillary line, press and hold. Positive: elbow drifts back, or the wrist flexes to compensate.highest sensitivity of this group
Bear-hug — hand on the opposite shoulder, resist the examiner lifting the elbow forward. Positive: weakness.highest sensitivity of this group
Internal rotation lag sign (IRLS) — take the hand to maximum passive internal rotation behind the back, release. Positive: the hand springs back toward the spine.Hertel: Sn .97 / Sp .96

Next: the whole subscapularis group is high-specificity and low-sensitivity. A positive is a finding; a negative is not reassurance. Belly-press and bear-hug are the most sensitive — start there.

Labrum & biceps

SLAP lesionDeep, diffuse shoulder pain with overhead click — no test diagnoses it

Hallmark

Deep, poorly localised pain; click or catch overhead; throwing athlete or a traction/fall-on-outstretched-arm mechanism.

Tests — none rule SLAP in or out

O'Brien / active compression — 90° flexion, 10–15° adduction, full internal rotation; resist downward, then repeat supinated. Positive: pain deep inside on the internally-rotated phase that eases in supination.Sn 63–94% but Sp 28–73%
Compression–rotation — 160° elevation, axial load, rotate. Positive: pain or click.best single pooled test · DOR 6.36 · Sn .43 / Sp .89
Crank — as above but standing/supine at 160°. Positive: click with or without pain.Sn 46% / Sp 72%
Biceps Load II — supine, 120° abduction, full external rotation, forearm supinated; resist elbow flexion. Positive: deep pain. the only test with real utility for isolated SLAP
Dynamic labral shear (O'Driscoll) — 90° abduction, then shear into abduction + external rotation. Positive: deep pain or click.modified version: Sn 72% / Sp 98% · LR+ 31.6

Next: a positive O'Brien with pain on top of the shoulder is AC joint, not labrum — different diagnosis entirely. If SLAP genuinely changes management (overhead athlete, surgical candidate), the answer is MRI arthrogram, not more manoeuvres.

Long head of biceps tendinopathyAnterior pain, tender in the groove, worse lifting and overhead

Tests

Bicipital groove tenderness — palpate the groove directly (best done with the arm internally rotated 10°). Positive: focal anterior pain. Often the single most useful finding here.
Speed's — arm 90° flexion, elbow extended, forearm supinated; resist downward. Positive: anterior/groove pain. Sn 32–65% / Sp 61–75%weak
Yergason's — elbow 90° at the side, forearm pronated; resist supination and external rotation. Positive: groove pain, or a painful click as the tendon moves.Sn 43% / Sp 79%
Upper cut — resist elbow flexion with the shoulder slightly flexed and abducted, forearm supinated. Positive: anterior pain. Sn 73–79% / Sp 78% — most accurate LHB test in several series

Next: Speed's and Yergason's cannot separate biceps from SLAP — a positive in either means "anterior, superior, load-related". Ultrasound answers it quickly and cheaply if the diagnosis changes treatment.

AC joint

AC joint pathologyOA, sprain, osteolysis — pain on top, worse cross-body and bench press

Hallmark

The patient points to the top of the shoulder — the joint line. Worse reaching across the body, pressing, and lying on that side.

Tests

AC joint palpation — direct pressure on the joint line. Positive: focal pain. Its value multiplies when a provoked test reproduces pain at exactly the point of maximum tenderness.Sn 0.48 / Sp 0.60
Scarf / cross-body adduction — shoulder to 90° flexion, then passively adduct across the chest. Positive: pain localised to the AC joint. Pain elsewhere does not count.Sn 64–77% / Sp 26–79% across studies — inconsistent
Paxinos — thumb under the posterolateral acromion, fingers above the clavicle, squeeze. Positive: pain at the AC joint.Sn 0.79 / Sp 0.50
O'Brien, AC pattern — as for SLAP, but the pain is on top in the internally-rotated phase. Positive: superficial, joint-line pain.Sn 0.14 / Sp 0.92 — nearly blind, only a positive helps

Next: the best combinations are O'Brien + AC tenderness in series → specificity 96.7% (both positive) and Paxinos + Hawkins-Kennedy in parallel → sensitivity 93.7% (either positive). Paxinos + O'Brien carries the highest likelihood ratio (2.71). The honest finding from the systematic review: no combination moves post-test probability more than slightly. If your pre-test probability is already clear-cut, tests add nothing — a diagnostic image-guided AC injection is more useful.

Instability

Anterior instability / BankartApprehension with abduction + external rotation — apprehension, not pain

Hallmark

A young patient with a dislocation, subluxation, or the sense that the shoulder "will come out". Traumatic onset. Guarding on overhead reach or throwing.

Tests

Apprehension — supine, shoulder abducted 90°, add external rotation. Positive: apprehension or fear of dislocation — not pain.Sn 92–98% for the apprehension group of tests
Relocation (Jobe) — from the above position, apply a posterior force to the humeral head. Positive: the apprehension is relieved.Sn 65% / Sp 90%
Anterior drawer / load-and-shift — translate the head anteriorly. Positive: excess translation. The most specific of the group.

Next: the commonest interpretation error is accepting pain relief as a positive relocation instead of apprehension relief. Accuracy is above 80% when apprehension defines the positive and below 50% when pain does. Apprehension also needs accounting for the labrum — a positive apprehension test with a negative relocation points more toward SLAP or AC pathology.

Multidirectional instabilityGeneralised laxity, atraumatic, symptoms in several directions

Tests

Sulcus sign — arm relaxed at the side, apply downward traction. Positive: a visible dimple appears below the acromion → inferior laxity.
Apprehension + relocation — as above. Positive: both positive, alongside a positive sulcus, points to MDI rather than isolated anterior instability.

Next: this changes management — an isolated anterior repair will not be enough. Also worth a generalised joint laxity / Beighton assessment.

Stiffness

Adhesive capsulitisNo special test. Passive external rotation is lost — that's the diagnosis

Hallmark — this one is not a special test

Both active and passive range are reduced, globally. External rotation is lost first and lost most. Risk factors: diabetes, hypothyroidism. Affects the other shoulder in a significant minority, sometimes years later.

What you actually do

Compare passive external rotation with the other side — with the elbow at the side, 90° flexed. Positive: a clear loss of passive external rotation with an intact rotator cuff is the hallmark sign.
Note the capsular pattern — external rotation > abduction > internal rotation > flexion.

Next: it's a clinical diagnosis — imaging is for excluding alternatives, not confirming this. Focal tenderness is uncommon and argues against it. Every other test on this page is unreliable once passive range is restricted, so establish range first.

Other & referred

Scapular dyskinesisA contributing factor, not a diagnosis — and it doubles as treatment

Tests

Scapular assistance test — apply upward rotation and posterior tilt to the scapula as the arm elevates. Positive: pain reduced or elevation improved → the subacromial space is a real contributor.
Scapular retraction test — hold the scapula retracted, repeat a strength test. Positive: strength improves → scapular position is limiting the cuff.

Next: commonly seen with impingement, labral tears and cuff problems in the general population, not just overhead athletes. Useful in the room because it demonstrates to the patient that proximal control changes their symptoms.

Glenohumeral osteoarthritisCrepitus, global stiffness, older patient, capsular pattern

Hallmark

Older patient, gradual onset, deep pain, crepitus with rotation, global restriction with a capsular pattern. The mirror image of adhesive capsulitis on examination — the radiograph separates them.

Next: plain radiograph is the right first test. Both active and passive range are lost, so special tests won't add much.

Cervical referral / radiculopathyPain distal to the elbow, or pins and needles — screen the neck

Tests

Spurling's — extend and rotate the neck, then apply gentle axial compression. Positive: reproduction of arm pain in a dermatomal pattern.
Neuro screen — C5 deltoid, C6 biceps + thumb, C7 triceps + middle finger, C8/T1 intrinsic hand. Positive: myotomal weakness, dermatomal sensory loss, or a diminished reflex.
Shoulder abduction relief sign — the patient rests the hand on top of the head. Positive: arm pain eases → likely cervical in origin.

Next: shoulder pain that reaches below the elbow, or comes with numbness, is a neck until proven otherwise. Screen before you inject anything.

Cervical root → where it hurts in the shoulderWhich root, where you feel it, and the tell that gives it away

Root	Felt in the shoulder	The tell
C4	Shoulder cap — over the clavicle and upper scapula, base of the neck	Does not go down the arm. C4 is the ceiling of shoulder referral.
C5	Lateral shoulder and deltoid — the classic mimic	Shoulder abduction weak; biceps reflex (C5–6). The best shoulder impersonator of the group.
C6	Anterior shoulder → biceps → lateral forearm → thumb and index finger	Elbow flexion and wrist extension weak; brachioradialis reflex. Crosses the elbow.
C7	Posterior shoulder and scapula → middle finger. The root most often linked to scapular pain	Elbow extension weak; triceps reflex. Crosses the elbow. Also the most commonly compressed cervical root.
C8	Medial forearm, ring and little fingers	Grip weak, no reliable reflex. Rarely the shoulder itself.
T1	Medial arm and elbow	Finger abduction weak. Rarely the shoulder itself.

Neck or shoulder?

Pain below the elbow → neck. Shoulder pathology rarely refers distal to the elbow; C6–T1 radiculopathy often does.
Relieved by putting the arm overhead → neck. Overhead loading makes shoulder pathology worse.
Upper trapezius, between the shoulder blades, or occipital headache → favours neck.
Night pain, cannot lie on that side, worse reaching behind the back → favours shoulder.

Careful: C6 and C7 arm pain is often diffuse and the two overlap — do not expect a clean dermatomal stripe. Scapular pain can precede arm pain by several weeks, so a patient with only interscapular pain today may be declaring a radiculopathy. C4–C6 compression and rotator cuff pathology coexist frequently, and a painful shoulder itself causes periscapular and neck pain through altered mechanics. Both can be true at once.

Suprascapular neuropathyPainless weakness or atrophy — the give-away is what's missing, not what hurts

Hallmark

Atrophy or weakness of the spinati with little pain. Deep, poorly localised posterior pain is possible. Consider in overhead athletes and after trauma.

Next: external rotation and abduction weak, other planes spared, with visible atrophy. EMG/NCS is the next step. It masquerades convincingly as a cuff tear — atrophy is the tell.

When to image

Choosing the imagingPlain film first for trauma and arthritis · ultrasound for the cuff · MRA for the labrum

Radiograph first — after trauma (fracture, dislocation), and for suspected glenohumeral or AC osteoarthritis, or calcific tendinitis.
Ultrasound — highly accurate for rotator cuff tears and cheap and dynamic. Pooled figures are around full-thickness: Sn 92% / Sp 94% (pooled, 65 studies) for cuff pathology; operator-dependent but excellent in trained hands. Also lets you inject under direct vision.
MRI / MR arthrogram — labral pathology, instability, and surgical planning. MRI Sn 94% / Sp 93% · MR arthrography Sn 95% / Sp 99% — the most accurate
Image-guided AC or subacromial injection — both diagnostic and therapeutic. When the pre-test probability is already clear, this beats more manoeuvres.

Worth remembering: MRI abnormalities are common in symptomatic and asymptomatic shoulders in older adults. Imaging findings do not predict symptoms — treat the patient, not the picture.

Red flags — before you test

Re-route before you testSeven findings that take precedence over any special test

Stop and re-route

Left shoulder pain with chest tightness, breathlessness, sweating or nausea → cardiac until excluded. Emergency.
Fever, or a hot, red, swollen joint → septic arthritis. Aspirate and get bloods. Higher risk: diabetes, immunosuppression, recent injection or surgery, IV drug use.
History of cancer, unexplained weight loss, night sweats, or night pain that wakes them and is not relieved by rest → malignancy. Local urgent referral pathway.
Significant trauma with deformity, or inability to lift the arm → fracture, dislocation, or acute massive cuff tear. Image before you mobilise.
Progressive non-mechanical pain — not provoked by movement and not eased by rest.
New numbness, weakness or pins and needles down the arm, especially bilateral, or any bladder/bowel change → neurological, urgent.
No meaningful improvement after 6–8 weeks of appropriate conservative care → reassess the diagnosis and consider referral.

Built for Ruby — PRM & sports medicine shoulder reference.
Single file, works offline. Edit the <details> blocks directly; each card is self-contained.

Sources. [1] Supraspinatus/Jobe accuracy and hornblower/lag-sign data — ROW cohort study, Am J Sports Med (PMC5218987). [2] Pooled diagnostic performance of shoulder tests — systematic review & meta-analysis (PMC5267375). [3] AC joint examination — Krill et al., systematic review (PMC6396285). Individual test and combination figures taken from its Table 3, not from the abstract. [4] Scarf/cross-body adduction — Chronopoulos et al. 2004, Am J Sports Med (Sn 77% / Sp 79%); the pooled weighted analysis in [3] gives Sn 0.64 / Sp 0.26. The two disagree, which is why the card shows a range. [5] Lag signs — Hertel et al. 1996 (ERLS Sn 70% / Sp 100%; IRLS Sn 97% / Sp 96%); Castoldi et al. 2009 (ERLS Sn 56% / Sp 98%). Subscapularis tests: Burkhart & Tehrany 2002; Barth et al. 2006. [6] SLAP test accuracy — Cook et al., J Shoulder Elbow Surg 2012; Gismervik et al. 2017. [7] Adhesive capsulitis — AAFP 2019; StatPearls; JOSPT clinical practice guideline 2013. [8] Imaging accuracy — AJR meta-analysis of 65 studies (full-thickness tear: ultrasound Sn 92.3% / Sp 94.4%, MRI Sn 92.1% / Sp 92.9%, MR arthrography Sn 95.4% / Sp 98.9%); Cochrane CD009020 (MRI Sn 94% / Sp 93%). [9] Upper cut and dynamic labral shear — Kibler et al. 2009 and the 325-patient cohort study; modified dynamic labral shear Sn 0.72 / Sp 0.98, LR+ 31.57. [10] Long head of biceps tests — Cardoso et al. 2019; JRM systematic review of LHBT examination. [11] Cervical root map — Orthobullets neck & upper extremity spine exam (dermatome, myotome, reflex per level); Physiopedia dermatomes, myotomes and cervical radiculopathy (root compression frequencies: C7 46–69%, C6 18–19%); StatPearls cervical radiculopathy (myotomes and reflexes); Neupsy Key, Differential Diagnosis of the Cervical Spine: Spinal Masqueraders (shoulder vs cervical discriminators). [12] Scapular pain in cervical radiculopathy — scoping review of 81 papers (PMC12226383): scapula involved in 72% of reported distributions; C7 carries the highest proportion of scapular pain descriptions; scapular pain may precede arm pain by weeks.

Diagnostic accuracy figures are pooled from the cited literature and vary widely between studies and populations. Interpret alongside history and pre-test probability. Not a decision-support system, and not a substitute for your own clinical judgement or local protocols.