In 2017, a British trial published results that changed NHS policy within months.
It examined one of the most commonly performed operations in orthopaedics, and it did something almost never done in surgery. It compared the operation against a placebo operation.
The results are worth knowing if anybody has mentioned shoulder surgery to you.
The Trial
CSAW, which stands for Can Shoulder Arthroscopy Work, was coordinated by the University of Oxford and run at 32 hospitals across the UK involving 51 surgeons.
It recruited 313 patients, all of whom had experienced shoulder pain for at least three months despite having already tried physiotherapy and steroid injections.
They were randomly assigned to one of three groups:
Arthroscopic subacromial decompression, the real operation, which involves removing bone and soft tissue through keyhole surgery to open up the space under the shoulder blade.
Arthroscopy only, meaning the surgeon performed a keyhole procedure and looked inside without removing anything. This was the placebo surgery.
Or no treatment, with one reassessment appointment three months later.
Patients did not know which of the two surgical groups they were in.
The Results
There was no difference in outcome between the two surgical groups at any time point.
The real operation and the placebo operation produced the same results.
Both surgical groups did somewhat better than the no treatment group, but the trial reported that this difference was not clinically important, meaning it fell below the threshold at which a difference actually matters to a patient.
The authors’ interpretation, published in The Lancet, was that the findings call into question the value of shoulder decompression surgery for this group of patients.
NHS England responded rapidly to the publication.
And It Was Not a One Off
Six months later, a Finnish trial called FIMPACT published similar findings.
That trial randomised 210 patients to decompression or diagnostic arthroscopy, which again served as placebo surgery.
It reported no statistically significant difference in clinical outcomes between the two. Both groups improved significantly, and the decompression added nothing.
A five year follow up of that trial has also been published.
Two independent placebo controlled surgical trials, in two countries, reaching the same conclusion. That is an unusually strong position for evidence in surgery, where placebo controlled trials are rare for obvious practical and ethical reasons.
And three systematic reviews have separately indicated that decompression is not superior to exercise therapy in patients with these symptoms.
The Honest Caveats
We would rather give you the full picture than a selective one.
A notable proportion of participants did not receive the intervention they were assigned, reported at 23 percent in the decompression group, 42 percent in the arthroscopy group and 12 percent in the no treatment group. That complicates interpretation, though the authors concluded it did not change the findings.
Both surgical groups did improve. The question the trial answered is whether the specific bone and tissue removal added anything, and the answer was no.
And this applies to this specific operation for this specific presentation. It does not speak to other shoulder surgery, including rotator cuff repair or surgery after trauma, which are different questions entirely.
What This Means If You Have Shoulder Pain
Shoulder pain affects around 14 in every 100 adults in the UK, so this is not a niche concern.
The practical implications:
Surgery for this presentation is not the straightforward solution it was once considered. If it is being discussed, these trials are reasonable to raise.
Conservative management deserves a proper trial, which means a structured, progressed exercise programme rather than a handful of sessions and a sheet of exercises.
And the fact that people improve substantially with placebo surgery tells you something important about recovery in this condition. Shoulder pain of this type frequently improves with time and appropriate management.
The Terminology Problem
Worth knowing, because the words have changed and the change reflects the evidence.
The condition has traditionally been called impingement, on the basis that the tendon was being pinched in a narrow space, which was the rationale for making the space bigger.
Researchers and clinicians have argued that impingement does not adequately explain this pain and is not a valid term.
Which follows logically from the trials. If widening the space produces no benefit over not widening it, the explanation that the space was the problem becomes difficult to sustain.
Alternative terms now used include rotator cuff related shoulder pain, or simply subacromial shoulder pain, pending better understanding of the underlying mechanisms.
If your diagnosis was given as impingement, that is not wrong exactly. It is a term the field is actively moving away from, and the reason is relevant to your treatment.
What the Evidence Does Support
Exercise therapy, which performs comparably to surgery across several systematic reviews, costs considerably less and carries no surgical risk.
A properly delivered programme means progressive loading, targeted at the rotator cuff and the muscles controlling the shoulder blade, built up over months rather than weeks.
Addressing the mid back, since thoracic mobility influences how much range the shoulder has to find on its own.
Load management, reducing the provocative activities temporarily while capacity is rebuilt.
Patience, since this is strength work and it behaves like strength work. Months rather than weeks.
And steroid injection, which has a role for short term symptom relief and is a conversation with your GP. The evidence generally supports it as a window rather than a solution, and the participants in CSAW had already had injections before entering the trial.
Questions Worth Asking
If surgery has been raised:
What specifically do you expect the surgery to achieve for me?
How do my circumstances differ from the patients in the CSAW and FIMPACT trials?
Have I had a properly progressed course of exercise therapy, and for how long?
And what would happen if I waited?
Those are reasonable questions, and a surgeon who engages with them is giving you what you need to decide.
Symptoms That Need Prompt Assessment
See your GP promptly for:
Inability to lift the arm at all following an injury.
True weakness rather than pain limited movement.
Significant loss of movement in all directions.
Numbness, tingling or weakness spreading down the arm.
Severe unrelenting night pain.
Shoulder pain with fever, unexplained weight loss, or any history of cancer.
And seek emergency care for shoulder or arm pain accompanied by chest discomfort or pressure, breathlessness, sweating, nausea, or pain spreading into the jaw.
Give Conservative Treatment a Proper Go
The trials above compared surgery against placebo and against no treatment. What they did not test is somebody receiving a well designed, properly progressed rehabilitation programme, which is the option with the best balance of evidence, cost and risk.
Body Zest offers a free discovery visit at no cost and no obligation. You will get a thorough assessment of your shoulder, shoulder blade control and mid back, a clear explanation of what is driving your symptoms, and a programme progressed properly rather than handed over on a sheet.