Examining the Rotator Cuff: Is the Can Empty or Full?


Source: Itoi E, et al. Which is more useful, the "full can test" or the "empty can test," in detecting the torn supraspinatous tendon? Am J Sports Med 1999;27:65-68.

The authors of this japanese study sought to better define the physical examination tests used to evaluate tears of the supraspinatous muscle-tendon complex. Under consideration were the "empty can" and "full can" tests. The first involves placing the shoulder in 90° abduction in the scapular plane, and then requesting full internal rotation, as would be seen with the emptying of a can. The second is a newer test involving placement of the shoulder in 90° abduction in the scapular plane, and then requesting 45° of external rotation, as with the holding of a full can.1 Two questions were addressed in this study: 1) should the physician use pain, weakness, or both as an indicator of a positive test for supraspinatous tear; and 2) which test is more clinically useful for the detection of supraspinatous tear?

One hundred forty-three shoulders from patients of all ages (range, 13-80 years) consitituted the study population; subjects were questioned regarding pain with the provocative maneuvers and were graded on a 6-point strength scale (5-zero) such as the one used for motor assessment during a neurologic examination. All patients then received high-resolution magnetic resonance imaging of the shoulder, which served as the gold-standard for presence or absence of disease (although Itoi and associates admitted that such imaging is only 95% accurate for detection of full-thickness tears of the rotator cuff). There were 35 full-thickness tears found in the rotator cuff; 19 were in the supraspinatous alone, and the other 16 involved that muscle and at least one other. Not surprisingly, defining a positive test as the presence of pain, weakness, or both yielded the highest sensitivity (empty can 89%, full can 86%). Muscle weakness proved to be the most specific indicator of tear (empty can 68%, full can 74%) and also yielded the highest diagnostic accuracy. Itoi et al concluded that muscle weakness

should be used as the key to interpreting the two tests. Accepting this, the two tests were found not to differ significantly with regard to specificity and accuracy. Itoi et al hypothesized that the full can test may be more beneficial in the clinical setting because it causes less pain; given that the tests are equivalently accurate, the less painful test would be more desirable.

Comment by Richard Harrigan, MD, FAAEM, FACEP

The can tests, full and empty, are a means to detect occult tears of the supraspinatous tendon. Acute tears may be suspected by history, but the more common chronic tears can be difficult to detect in that an antecedent event is often not recalled. Certainly the drop arm test2 is fairly good for the detection of significant tears; if the patient cannot slowly range his arm from 90° abduction to zero, or cannot maintain 90° abduction despite the examiner’s gentle downward tap of the abducted arm, it seems intuitive that a tear exists. The drop arm test is more of an all-or-nothing test, whereas a positive test in this study was defined as any muscle weakness detected during these maneuvers. It makes sense not to use pain as evidence of a full-thickness tear, in that pain during the empty can test may be representative of simple impingement as is seen with rotator cuff tendonitis. As Itoi et al point out in their discussion, pain may negatively influence the assessment of weakness. The empty can test, while no doubt sensitive for impingement, may be flawed as an indicator of tear, in that pain may mimic weakness during provocative testing.


1. Kelly BT, et al. The manual muscle examination for rotator cuff strength: An electromyographic investigation. Am J Sports Med 1996;24:581-588.

2. Hoppenfeld S. Physical examination of the spine and extremities. Norwalk, CT; Appleton and Lange; 1976:33.

Which of the following is correct?
a. Muscle weakness is the most specific indicator of supraspinatous tear in the "empty can test" and the "full can test."
b. The "full can test" is more sensitive for teres minor tears than supraspinatous tears.
c. The "empty can test" is generally less painful than the "full can test" when there is pathology involving the supraspinatous tendon.
d. The "drop arm test" is very sensitive for detection of tears in the teres major tendon of the rotator cuff.