Shoulder Special Tests: A Practical Guide for Physiotherapists

Before any special test: form your hypothesis

Special tests have modest diagnostic accuracy in isolation. Their value comes from being applied to a pre-test probability you already estimated. After a thorough subjective examination - mechanism, irritability, pattern of pain, functional limitations, red flags - you should be able to predict the most likely structural and movement-based diagnoses before you touch the patient.

From there, choose 2–4 tests per hypothesis and interpret them as a cluster. A single positive test rarely changes management; a converging cluster does.

Rotator cuff tendinopathy / tear

Use the Jobe (empty can) test, external rotation lag sign, and the drop-arm test. Pain reproduction with weakness on resisted abduction in the scapular plane is a strong indicator of supraspinatus involvement; an external rotation lag suggests infraspinatus tear.

The Hawkins-Kennedy and Neer tests are sensitive but not specific - they will be positive in most painful shoulders, including frozen shoulder and labral pathology. Use them to rule out subacromial pain when negative, not to confirm it when positive.

Anterior shoulder instability

The apprehension and relocation tests, when both positive, have high specificity for anterior instability - particularly in patients under 40 with a traumatic dislocation history. Add the surprise/release test to strengthen the cluster.

Apprehension is the patient anticipating dislocation, not pain. A pain-only response with apprehension testing is more suggestive of internal impingement than true instability.

Labral pathology (SLAP, Bankart)

No single test is reliable for labral tears. The O'Brien active compression, Speed's test, and biceps load II perform best when clustered. Even then, MRI arthrogram or arthroscopy remains the reference standard. Use the cluster to decide whether imaging is warranted in a patient who has not responded to a structured trial of conservative care.

AC joint

Cross-body adduction, the active compression (O'Brien) test, and direct palpation of the AC joint together raise specificity substantially. A positive cluster in the absence of trauma should still prompt consideration of degenerative AC joint pain rather than ligamentous injury.

Documenting and re-testing

Record the cluster you used, each test's result, and the patient's reported pain or apprehension. Re-test the most clinically meaningful one or two at each follow-up - they become your asterisk signs and let you and the patient see objective change.

Key takeaways

• Choose tests to confirm or refute a pre-test hypothesis, not as a checklist.
• Single positives rarely change management; clusters do.
• Hawkins-Kennedy and Neer are sensitive but not specific - best for ruling out.
• Apprehension + relocation + surprise has high specificity for anterior instability.
• Pick 1–2 cluster tests as asterisk signs to track outcomes over time.

Frequently asked questions

Which shoulder special tests should I always include?

There is no fixed always-include list. Choose tests based on the hypothesis your subjective examination suggested. For most painful shoulders without trauma, a rotator cuff cluster (Jobe, external rotation strength, drop-arm) plus Hawkins-Kennedy is a reasonable starting screen.

Are shoulder special tests reliable?

In isolation, most have only modest sensitivity and specificity. Reliability and diagnostic value improve substantially when tests are clustered for a single hypothesis and interpreted alongside subjective findings and functional testing.

What is the difference between apprehension and pain on the apprehension test?

True apprehension is the patient anticipating dislocation and guarding against it. Pain alone - without apprehension - is more suggestive of internal impingement, labral pathology, or rotator cuff irritation than true anterior instability.