Clinical Education

Special Tests

There are a multitude of ‘Special tests’ for the shoulder, two of the most commonly used and best known tests are Hawkins Kennedy and Jobe’s Empty Can test.

Special tests are still taught commonly in physiotherapy and medical schools and used widely in clinical practice with too much importance placed on findings.

Tests all tend to have poor sensitivity or specificity.

Certain special tests may rule in or rule out certain pathology.

Special tests are of limited value as diagnostic tools for specific pathology.

Essentially the advice is to be cautious using special tests for the shoulder and instead use the history and other physical examination findings to form a working diagnosis.

The two mostly commonly used tests are detailed below:


Originally described in 1980

Procedure: forward flexing the humerus to 90° and forcibly internally rotating the shoulder.

This manoeuvre drives the greater tuberosity farther under the coracoacromial ligament similarly reproducing the impingement pain.

Typical values (many reported):

  • Sensitivity 85%
  • Specificity 40%
  • Positive likelihood ratio = 1.3 approx.
  • Negative likelihood ratio = 0.4  approx.


It may rule out a shoulder pathology if negative, however the negative likelihood ratio is not quite at useful range (<0.33).

A positive test does not rule in impingement as 1.3 is too low (>3 to be useful), the low specificity many patients who do not have impingement will also test positive.

Jobe’s Test (Empty can)

Jobe and Jobe described the “supraspinatus test in 1983

Procedure: “The supraspinatus test is performed by first assessing the deltoid with the arm at 90° of abduction and neutral rotation. The shoulder is then internally rotated and angled forward 30°; the thumbs should be pointing toward the floor. Muscle testing against resistance will clearly demonstrate a weakness or insufficiency of the supraspinatus secondary to a tear or pain associated with rotator cuff impingement.”

Typical values – many reported:

  • Sensitivity 86%
  • Specificity 50%
  • Positive likelihood ratio 1.72
  • Negative likelihood ratio 0.28

Commentary: Similar to Hawkins Kennedy – it may rule out shoulder pathology but a positive test does not rule impingement in.