Muscle, joint & pain

Shaving the Shoulder Bone Spur, and the Sham Trial That Undid It

Save

Subacromial decompression once seemed obvious: a spur crowds the tendon, so remove the spur. Then surgeons ran the operation against a placebo and found no meaningful difference. Here is what the trials actually showed, what still points toward conservative care, and the specific shoulders where an operation remains the right call.

Last updated: July 2026

Talk to a clinician

Gale can help you find a clinician in your state and request a visit.

Find care →

What subacromial decompression actually removes

Subacromial decompression is keyhole surgery that trims bone and soft tissue from the underside of the acromion — the bony roof of the shoulder — to widen the tunnel the rotator cuff tendons run through. Some surgeons also shave a bone spur or release the ligament that forms the front of that arch. The premise was purely mechanical: a spur or a tight arch pinches the tendon from above, so making more room should relieve the pain. For decades that reasoning went essentially untested against a fair comparison, and it became one of the most common shoulder operations in the world.

The condition it targets, shoulder impingement, is also called rotator cuff tendinitis. It is irritation and swelling of the cuff tendons and the fluid-filled bursa that cushions them, and it produces pain when the arm is raised or reached overhead, frequently with an ache that is worse at night and when lying on that side 1. The operation was built to relieve that pain by uncrowding the space above the tendon. The catch, uncovered only when the surgery was finally tested properly, is that the space and the pain are loosely connected at best — many people have a narrow arch and no symptoms, and many hurt with a perfectly roomy one.

Does shoulder impingement surgery work?

For ordinary, non-traumatic impingement, the strongest evidence says the operation does not work better than a placebo. The CSAW trial randomly assigned patients whose pain had persisted despite non-surgical care to one of three groups: a real decompression, an arthroscopy that entered the joint and looked inside but removed nothing — a placebo procedure — or no surgery and active monitoring. At six months and again at one year, the real operation produced no clinically important advantage over the placebo one, and only a small difference over no treatment that fell below the threshold patients notice 2.

A Cochrane systematic review that pooled the randomized trials reached the same verdict with high certainty: subacromial decompression offers no important benefit over placebo or non-surgical care for pain, shoulder function, or quality of life 3. The operation was no better than a convincing sham. Both the real-surgery and placebo-surgery groups did improve — they simply improved by the same amount. That pattern, a real gain that a fake procedure fully reproduces, is the fingerprint of a treatment riding the natural course of the condition rather than changing it.

Why a fake operation matched the real one

A placebo operation matches a real one when the real one was never treating the true cause. Two forces explain the CSAW result. The first is natural history: most impingement settles over months with time, movement, and the tendon's own recovery, so any group measured later tends to look better whether it had surgery, a sham, or watchful waiting. The second is that the pain was arising from an irritated, deconditioned cuff rather than from a bone that genuinely needed shaving — so trimming the bone changed the anatomy on the scan without touching the source of the symptoms 2.

The shoulder is not alone in this. The knee saw the same reckoning: arthroscopic surgery for a degenerative meniscus tear performed no better than a sham incision in a placebo-controlled trial of middle-aged patients 4. Similar questions now hang over hip impingement surgery vs physical therapy and several other tidy mechanical stories. When a well-designed trial shows an operation only equals a sham, the honest reading is not that the surgery is fraudulent or that surgeons acted in bad faith. It is that the tissue was going to settle either way, and the incision added anesthesia, recovery, and risk without adding benefit.

What the trials do — and do not — say

These trials narrow the claim; they do not abolish shoulder surgery. CSAW and the Cochrane review studied ordinary, degenerative impingement and rotator cuff disease — the aching, gradual, non-traumatic shoulder that makes up most of the caseload. They did not study acute tears from a real injury, dislocating shoulders, or the specific structural problems that carry their own evidence base. A finding that one operation fails for one common problem is not a verdict against all shoulder surgery.

The results also do not mean nothing can be done — the opposite is true, and the sections below cover what actually helps. What the trials retire is a single, specific inference: that a painful shoulder showing impingement on imaging should be decompressed because the picture looks crowded. That chain of reasoning, from scan to scalpel, is the part the evidence broke. The distinction is easy to blur and worth holding onto, because sliding in either direction — surgery never works, or the spur must come out — leads to a worse decision than the evidence supports.

If the operation does not help, what does?

When decompression fell, the treatments it had overshadowed came back into focus — and they are unglamorous. Standard non-surgical care for impingement is activity modification to unload the aggravating overhead motions, anti-inflammatory measures for flares, a structured course of physical therapy aimed at the rotator cuff and the shoulder-blade stabilizers, and sometimes a corticosteroid injection to quiet a flare enough that the exercise becomes possible 1. None of it is a quick fix, and that is part of why an operation once looked so appealing by comparison.

The rehabilitation is the active ingredient, not a consolation prize. When researchers compared physiotherapy alone against surgery for non-traumatic rotator cuff problems, exercise held its own: at two years there was no significant clinical difference between physiotherapy alone, adding a bone-shaving acromioplasty, or performing a full tendon repair 5. Most irritated shoulders recover with time and coached exercise rather than with a scalpel. The work is slow, and progress is measured in weeks, but the randomized evidence keeps arriving at the same place: the patient, effortful route reaches the outcome the operation was supposed to buy.

What about an actual rotator cuff tear?

A torn tendon sounds like it must be surgical, but many rotator cuff tears are degenerative — a gradual fraying that arrives with age — and they behave much like impingement. In the same randomized trial, patients with a non-traumatic supraspinatus tear did as well with physiotherapy alone as with surgical repair at two years, and the added acromioplasty changed nothing 5. Imaging finds these tears in a large share of older shoulders that have never hurt, which is why a tear reported on a scan does not, on its own, argue for an operation.

That logic has real limits, and this is where the sequence-of-care frame matters. A sudden, high-grade tear from a defined injury — a fall onto an outstretched arm, a heavy load caught wrong — in an otherwise strong, active shoulder is a different situation from slow degeneration, and it is one where earlier repair is genuinely weighed rather than deferred. The word on the radiology report is not the deciding factor. How the shoulder got there, how much strength it has lost, and what it still needs to do carry far more weight than the noun tear.

When shoulder surgery is clearly the right call

Surgery is a later step in a sequence, not a first reflex — but for some shoulders it is plainly correct, and delaying it helps no one. Recurrent instability is the clearest example. After a dislocation, the shoulder can become chronically prone to slipping out of joint, often because of a Bankart lesion — a tear of the labrum, the cartilage rim that deepens the socket — and surgical stabilization is a standard, well-supported option when dislocations keep recurring or the joint feels persistently unreliable 6.

The other clear indications share a theme: a structure that will not recover on its own, or a shoulder that has honestly failed a real course of rehabilitation. Among them are acute full-thickness tears from a discrete injury in an active arm, and shoulders that remain disabling after a supervised, months-long exercise program that was genuinely completed rather than abandoned in the second week. The question is never surgery or nothing — it is where in the sequence this particular shoulder sits. Naming the indication out loud, and matching it to evidence rather than to a scan, is what separates an operation that earns its place from a reflexive one.

How to weigh the decision

The useful frame is sequence of care: what has actually been tried, what the pain really is, and what an operation is expected to change. Three questions do most of the work. Has the shoulder had a genuine, coached course of exercise — not a photocopied sheet, but supervised progressive loading over weeks? Is the problem mechanical impingement, true instability, or a traumatic tear, because those three lead to very different places? And can the surgeon name specifically what the operation will fix, given that decompression for plain impingement has been shown not to change the outcome?

Some shoulders sit in specialized categories that deserve tailored assessment. The overhead athlete's shoulder — the thrower whose pain comes from the extreme, repeated demands of the throwing motion — is evaluated differently from the average painful shoulder, and ordinary shoulder pain when lifting arm overhead is different again. None of this makes surgery wrong or right by default. It makes the timing and the indication the entire question — which is precisely what the placebo-controlled trials taught the field to ask before reaching for the arthroscope.

Common questions

For most people, yes. Ordinary impingement tends to settle over months with activity changes and a structured exercise program aimed at the rotator cuff and shoulder-blade muscles, sometimes helped along by an injection to calm a flare. Randomized trials found that this patient route reaches the same result surgery was expected to deliver, without the operation.

Usually not. Bone spurs and a narrow arch show up on imaging in many pain-free shoulders, so their presence does not prove they are causing the pain. When surgeons tested spur removal against a placebo operation for impingement, shaving the bone produced no meaningful advantage. A spur seen on a scan is not, by itself, a reason to operate.

Impingement is irritation and swelling of the cuff tendons and the bursa in the space beneath the shoulder blade. A tear is an actual defect in a tendon. They overlap: chronic impingement can accompany a degenerative tear, and both often respond to the same exercise-based care. How the injury happened matters more than the label for deciding what to do.

Surgery becomes the clear choice for a shoulder that keeps dislocating, for an acute full-thickness tendon tear from a real injury in an active arm, and for shoulders that stay disabling after a genuine, supervised course of rehabilitation. The common thread is a structure that will not recover on its own, or conservative care that has truly been tried and failed.

Keyhole surgery leaves smaller scars, but it is still an operation with anesthesia, an infection risk, and months of recovery. The CSAW trial's placebo group had a real arthroscopy that removed nothing and did just as well, a reminder that a smaller incision is not the same as a benefit. Lower visible risk is not the same as no risk.

Related

Say it back

How would you explain this to someone you love?

Two or three sentences, just as you’d say it. Gale reflects back what you focused on — a mirror, not a quiz.

Talk to a clinician

Gale can help you find a clinician in your state and request a visit.

Find care →

When shoulder pain needs prompt attention

  • —A sudden loss of the ability to lift the arm after a fall or a hard pull, often with a pop or a tearing sensation — a possible acute full-thickness rotator cuff tear.
  • —A hot, swollen, red shoulder with fever or feeling generally unwell — a possible joint infection.
  • —Numbness, pins-and-needles, or weakness spreading down the arm into the hand, or deep pain that wakes you regardless of position — signs the pain may be coming from the neck or elsewhere.
  • —A shoulder that repeatedly slips out of joint or feels like it is about to.

A hot, swollen shoulder with fever, or a shoulder that cannot be moved at all after a significant injury, needs urgent assessment — go to an emergency room.

This article explains the evidence on shoulder impingement surgery for general education. It is not medical advice and cannot account for your particular shoulder. Decisions about surgery or rehabilitation belong with a clinician who has examined you.

References

  1. 1.American Academy of Orthopaedic Surgeons (OrthoInfo) (2024). Shoulder Impingement / Rotator Cuff Tendinitis. OrthoInfo — AAOS. link ✓Lay description of shoulder impingement / rotator cuff tendinitis and its standard nonsurgical management (activity modification, NSAIDs, physical therapy, injection).
  2. 2.Beard DJ, Rees JL, Cook JA, et al. (CSAW) (2018). Arthroscopic subacromial decompression for subacromial shoulder pain (CSAW): a multicentre, pragmatic, parallel group, placebo-controlled, three-group, randomised surgical trial. The Lancet. doi:10.1016/S0140-6736(17)32457-1The CSAW placebo-controlled trial: arthroscopic subacromial decompression gave no clinically important benefit over placebo surgery or no treatment for subacromial shoulder pain.
  3. 3.Karjalainen TV, Jain NB, Page CM, et al. (2019). Subacromial decompression surgery for rotator cuff disease. Cochrane Database of Systematic Reviews. doi:10.1002/14651858.CD005619.pub3 ✓High-certainty Cochrane evidence that subacromial decompression provides no important benefit over placebo or non-surgical care for rotator cuff disease.
  4. 4.Sihvonen R, Paavola M, Malmivaara A, et al. (FIDELITY) (2013). Arthroscopic Partial Meniscectomy versus Sham Surgery for a Degenerative Meniscal Tear. New England Journal of Medicine. doi:10.1056/NEJMoa1305189Cross-joint sham-surgery parallel: arthroscopic partial meniscectomy was no better than sham surgery for a degenerative meniscus tear.
  5. 5.Kukkonen J, Joukainen A, Lehtinen J, et al. (2015). Treatment of Nontraumatic Rotator Cuff Tears: A Randomized Controlled Trial with Two Years of Clinical and Imaging Follow-up. Journal of Bone and Joint Surgery (American). doi:10.2106/JBJS.N.01051 ✓RCT evidence that physiotherapy alone matched acromioplasty-plus-physiotherapy and rotator cuff repair at two years for nontraumatic supraspinatus tears.
  6. 6.American Academy of Orthopaedic Surgeons (OrthoInfo) (2024). Chronic Shoulder Instability and Dislocation. OrthoInfo — AAOS. link ✓Lay evidence that recurrent shoulder instability after dislocation (often with a Bankart labral lesion) is treated with surgical stabilization — a clear surgical indication.

6 sources, numbered by first appearance. General health information, not medical advice. AI-assisted editorial content — citations link their sources. Editorial policy

← Back to the health library