The Shoulder Exam: Sensitivity, Specificity, and The Truth

The shoulder examination is a graveyard of eponymous "special tests." From Neer to Hawkins, O'Brien to Speed, Yergason to Jobe—it seems every pioneer in shoulder surgery wanted their name attached to a manoeuvre. Students memorize these lists dutifully, believing that a positive "O'Brien's Test" equates to a SLAP tear diagnosis.

The reality, unfortunately, is far less precise. The literature consistently shows that many of these tests have poor diagnostic accuracy when used in isolation. To be a master clinician, you must move beyond the name of the test and understand its statistical utility: its Sensitivity, Specificity, and Likelihood Ratio.

Visual Element: Physical exam demonstration (vector art) showing the vector forces applied during the Jobe's test.

1. Evidence-Based Medicine 101

Before we dissect the shoulder, let's define our terms.

• Sensitivity (Sn): The ability to rule out disease (SnNout). If a test is 100% sensitive, a negative result means the patient definitely does not have the disease.
• Specificity (Sp): The ability to rule in disease (SpPin). If a test is 100% specific, a positive result means the patient definitely does have the disease.
• Likelihood Ratio (LR): The most powerful metric.
  • LR+ > 10: Excellent for ruling in.
  • LR+ 5-10: Good.
  • LR- < 0.1: Excellent for ruling out.

2. Rotator Cuff Tears

Diagnosing a full-thickness tear is clinically reliable if you use the right tests.

Supraspinatus

• Jobe's (Empty Can) Test:
  • Sensitivity: High (~88%). Good screen.
  • Specificity: Poor (~60%). Pain inhibition from bursitis mimics weakness.
  • Better Alternative: The "Full Can" Test. It is just as valid for strength testing but less painful (less impingement), leading to fewer false positives.
• Drop Arm Test:
  • Sensitivity: Very Low (~8%). Most people with a tear can still hold their arm up.
  • Specificity: Very High (97%). If they literally cannot hold the arm and it drops, it is almost certainly a massive tear.

Infraspinatus / Teres Minor

• External Rotation Lag Sign:
  • Passive ER to near end range, then ask patient to hold.
  • Specificity: High (98%).
  • Sensitivity: Good (97%) for massive tears.
• Hornblower's Sign:
  • For Teres Minor. ER in 90° abduction.
  • High specificity for fatty infiltration/irreparable tears.

Subscapularis

• Lift-Off Test (Gerber):
  • Specificity: High (97%).
  • Problem: Requires internal rotation range. If the patient is stiff, they can't do it.
• Bear Hug Test:
  • Sensitivity: Highest of the group. More sensitive for upper subscapularis tears than the Lift-off.
• Belly Press (Napoleon):
  • Good middle ground. Watch for wrist flexion (compensatory movement).

Visual Element: Table summarizing Sensitivity/Specificity/LR for each cuff test.

3. Labral Pathology (SLAP & Instability)

This is the "Wild West" of shoulder exams. The physical exam is notoriously weak here.

SLAP Tears

• O'Brien's (Active Compression) Test:
  • The Myth: "Positive O'Brien's = SLAP."
  • The Reality: High false-positive rate. It often causes pain in the AC joint.
  • The Nuance: You must ask "Where is the pain?" Deep/Inside = Labrum. Top/AC = AC Joint.
• Biceps Load II:
  • Generally considered robust for SLAP tears in overhead athletes.

Anterior Instability

• Apprehension / Relocation / Surprise:
  • This is a subjective test.
  • Positive: The patient feels APPREHENSION (fear it will pop out).
  • Negative: The patient just feels pain. Pain is not apprehension.
  • Specificity: High (if apprehension is the endpoint).

4. Subacromial Impingement

Impingement is a symptom cluster, not a specific pathology. The tests are provocative—they squash the bursa.

• Hawkins-Kennedy: Highly Sensitive. If this is negative, impingement is very unlikely (SnNout).
• Neer's Sign: Also highly sensitive.

5. The Power of "Clustering"

A single test is rarely diagnostic. Combining tests multiplies their power.

Park et al. Cluster for Impingement

If all 3 are positive, LR+ is 10.5 (Probability >95%):

1. Hawkins-Kennedy Positive.
2. Painful Arc (60-120°).
3. Infraspinatus Weakness.

Cluster for Full Thickness Cuff Tear

If all 3 are positive, Probability is >90%:

1. Painful Arc.
2. Drop Arm Sign.
3. Infraspinatus Weakness.

6. Inspection: The Forgotten Art

Before you start twisting the arm, LOOK.

• Supraspinatus Fossa: Is it hollow? (Muscle atrophy).
• Infraspinatus Fossa: Is it wasted?
• Scapula: Is there winging? (Wall push-up test).
• AC Joint: Is there a step deformity?

Conclusion

The physical exam is a screening tool.

• High Specificity Tests (Drop Arm, Lag Signs): Good for confirming a diagnosis.
• High Sensitivity Tests (Hawkins, Jobe): Good for ruling it out.

Stop relying on a single "positive O'Brien's." Build a clinical picture using clusters, history, and inspection. Treat the patient, not the MRI, but use the exam to decide if the MRI is even necessary.

Clinical Pearl: The Neck Check. Always clear the cervical spine. A C5 radiculopathy looks exactly like a weak deltoid/cuff. Spurling's test is your friend.