Which Shoulder Special Tests Are Accurate?
Many health care practitioners rely on orthopaedic special tests to evaluate specific pathologies within the shoulder. Unfortunately, many of these tests have proven to be unreliable and misleading. For example, the empty can test to evaluate a rotator cuff tear (mainly the supraspinatus muscle) is a common special test that is used widely in clinical practice and is still being taught in school. However, recent evidence has shown limited diagnostic value for this test, with a poor ability to differentiate structural lesions of the rotator cuff. In layman’s terms, many shoulder tests will provoke symptoms (high sensitivity) but cannot differentiate the cause of the symptoms (poor specificity).
If these common tests can’t be trusted, then which ones can?
Current evidence has shown the diagnostic utility of only a handful of orthopaedic special tests for shoulder pathology. This post will highlight the most useful tests and provide statistical values to back up those claims.
To understand what makes a special test useful, we must look at four statistical variables:
- Positive Likelihood Ratio
- Negative Likelihood Ratio
Sensitivity is the ability of a test to correctly identify those with the pathology. Specificity is the ability of a test to correctly identify those without the pathology.
In isolation, these values may not be as useful to the practicing clinician. This is where likelihood ratios are helpful. Likelihood ratios take both the sensitivity and specificity values and produce a number that gives a better diagnostic probability of a certain pathology.
A positive likelihood ratio tells us how much we should increase the probability of a particular pathology based on a positive test result. A negative likelihood ratio tells us how much we should decrease our suspicion of a particular pathology based on a negative test result.
In medicine, a positive likelihood ratio (LR+) of >5 is used to rule in a certain condition. Similarly, a negative likelihood ratio (LR-) of <0.2 is used to rule out a certain condition. Therefore, we can use LR+ to diagnose a condition and LR- to screen for a condition.
A study simplifying likelihood ratios outlined a method to estimate the percentage change in probability of a condition based on the likelihood ratio.
The following tests have proven to be valid in identifying specific shoulder pathologies. Click on the name or picture to read about the research behind each test.
Sensitivity: 72 Specificity: 96 LR+: 18 LR-: 0.29 Use: Diagnose
Sensitivity: 81 Specificity: 92 LR+: 10.1 LR-: 0.1 Use: Diagnose / Screen
Sensitivity: 92 Specificity: 89 LR+: 8.4 LR-: 0.08 Use: Diagnose / Screen
Sensitivity: 53 Specificity: 94 LR+: 8.3 LR-: 0.5 Use: Diagnose
Sensitivity: 82 Specificity: 86 LR+: 5.7 LR-: 0.2 Use: Diagnose / Screen
Sensitivity: 72 Specificity: 98 LR+: 36 LR-: 0.29 Use: Diagnose
Sensitivity: 86 Specificity: 91 LR+: 9.7 LR-: 0.14 Use: Diagnose / Screen
Sensitivity: 40 Specificity: 98 LR+: 20 LR-: 0.61 Use: Diagnose
Sensitivity: 80 Specificity: 88 LR+: 6.7 LR-: 0.23 Use: Diagnose / Screen
Sensitivity: 60 Specificity: 92 LR+: 7.23 LR-: 0.43 Use: Diagnose
Rotator Cuff Tear (RTC)
Sensitivity: 81 Specificity: 89 LR+: 7.4 LR-: 0.21 Use: Diagnose / Screen
Sensitivity: 46 Specificity: 49 LR+: 7.2 LR-: 1.1 Use: Diagnose
AC Joint Pathology
Sensitivity: 72 Specificity: 85 LR+: 4.8 LR-: 0.3 Use: Diagnose
Sensitivity: 94 Specificity: 84 LR+: 5.9 LR-: 0.07 Use: Diagnose / Screen
Sensitivity: 84 Specificity: 99 LR+: 84 LR-: 0.16 Use: Diagnose
Future of Special Tests
Not every shoulder special test was included in this article. If you are curious how some of the more common shoulder tests hold up to scientific scrutiny, check out this systematic review.
We don’t know the accuracy of clustering these special tests, perhaps because no study has been done to date. However, a recent study attempted to use the best available clustered tests and clinical prediction rules to inform clinical practice. They concluded that no one clinical tests is definitive for diagnosing shoulder pathology.
One other point that should be mentioned is the relationship between shoulder pathology on imaging and presence of symptoms. Many people are walking around with asymptomatic rotator cuff tears and studies have shown that athletes can compete painfree with evidence of labral tears. These factors are likely what prevent any orthopaedic special test from being 100% accurate. As always, no test should be interpreted in isolation to a thorough subjective and objective examination.