A 52-year-old recreational tennis player comes in with shoulder pain. He can’t reach behind his back anymore, and overhead serves are gone. You run through the usual special tests: Neer’s is positive. Hawkins-Kennedy is positive. Empty can test is positive and weak.
So — rotator cuff tear, right? Maybe. Based on the numbers alone, you genuinely can’t be sure. Here’s why, and here’s how a five-minute ultrasound scan in the same visit answers the question these tests only guess at.
The Problem with “Positive” Special Tests
Orthopedic special tests are everywhere in clinical training. We learn them early, perform them daily, and often make real decisions — order imaging, start therapy, consider injection, consult surgery — based on whether they’re “positive” or “negative.”
The uncomfortable truth is that many of these tests aren’t nearly as accurate as they’re taught to be. Some perform only slightly better than a coin flip. Understanding which tests are reliable, and which aren’t, changes how confidently we can use them — and shows exactly where ultrasound closes the gap.
Quick Refresher: Sensitivity and Specificity
Sensitivity tells you how good a test is at catching true positives. A very sensitive test, when negative, helps rule a condition out.
Specificity tells you how good a test is at correctly identifying true negatives. A very specific test, when positive, helps rule a condition in.
A good test needs both. Most orthopedic special tests have neither — which is the real issue.
The Data: Common Tests, Real Numbers
| Joint | Test | Target Pathology | Sensitivity | Specificity |
|---|---|---|---|---|
| Shoulder | Neer’s Test | Subacromial impingement | ~75% | ~48% |
| Shoulder | Hawkins-Kennedy | Subacromial impingement | ~80% | ~45% |
| Shoulder | Empty Can (Jobe’s) | Supraspinatus tear | ~45–75% | ~50–55% |
| Shoulder | Speed’s Test | Biceps/labral pathology | ~32% | ~75% |
| Knee | McMurray’s Test | Meniscal tear | ~53% | ~59% |
| Knee | Lachman Test | ACL tear | ~85% | ~94% |
| Hip | FADIR | Labral tear / FAI | ~96% | ~11% |
| Wrist | Phalen’s Test | Carpal tunnel syndrome | ~68% | ~73% |
Two results are worth pausing on.
FADIR is 96% sensitive but only 11% specific. It almost never misses true hip pathology — but it’s positive in nearly everyone with an irritated hip, torn labrum or not. A positive FADIR tells you to look further. It doesn’t tell you what’s wrong.
McMurray’s test sits close to 50/50 for detecting meniscal tears. That’s barely better than guessing. And here’s the part that matters clinically: ultrasound doesn’t fix this one. The meniscus sits deep between the femur and tibia, largely shielded from the ultrasound beam by bone and the joint line itself. Ultrasound can sometimes show secondary signs — a meniscal cyst, extrusion, or surface irregularity — but it cannot reliably see the central, weight-bearing portions of the meniscus the way MRI can. For suspected meniscal tears, MRI remains the right next step, not ultrasound.
Back to the Shoulder: Where Ultrasound Actually Shines
Unlike the meniscus, the rotator cuff sits just beneath the skin and deltoid — shallow, accessible, and ideally positioned for ultrasound. This is where the technology moves from “helpful” to “nearly as good as MRI.”
Published studies comparing ultrasound to MRI and arthroscopic findings consistently show ultrasound performing in the same range as MRI for full-thickness rotator cuff tears — often above 90% for both sensitivity and specificity in experienced hands. For partial-thickness tears, accuracy drops somewhat for both modalities, but ultrasound still holds its own.
Back to the tennis player. His three special tests were positive, but as the table above shows, that combination still leaves real diagnostic uncertainty — impingement, tendinopathy, partial tear, and full tear can all produce a similar pattern of positive tests.
The ultrasound probe goes on his shoulder in the same visit. Within minutes, the supraspinatus tendon shows a focal, full-thickness gap at the footprint, with fluid tracking into the subacromial-subdeltoid bursa. The contralateral shoulder, scanned for comparison, looks normal.
No ambiguity. No two-week wait for an MRI. A clear structural diagnosis, made in the room, during the first visit — and a clear path toward either surgical referral or a regenerative injection approach, depending on tear pattern and goals.
The Honest Caveat
Ultrasound isn’t magic, and where it works depends entirely on what’s being looked at. It excels at shallow, accessible structures — rotator cuff tendons, biceps tendon, superficial nerves, bursae, ligaments around the elbow, ankle, and wrist. It struggles with deep or bone-shielded structures — the meniscus, the inside of the hip joint in larger patients, deep posterior shoulder labral tissue.
It’s also operator-dependent. Image quality and interpretation accuracy scale directly with the training and experience of the person holding the probe — a true limitation, and one reason this should never fully replace a thoughtful clinical exam and a knowledgeable radiologist or MRI when the structure in question calls for it.
Bottom Line
Special tests still matter — they guide where to look and what to suspect. But many of them, taken alone, leave real diagnostic uncertainty: a positive FADIR, a borderline McMurray’s, a cluster of positive shoulder tests. The right next step depends on which structure is in question. For the meniscus, that’s still MRI if mechanical signs or complex tear. For the rotator cuff, point-of-care ultrasound can answer the question as accurately as MRI, in the room, during the same visit — turning “probably a tear” into a confirmed diagnosis before the patient even leaves the office.
This page is educational and general in nature and does not replace an individualized evaluation — see our medical disclaimer.