Physiotherapy vs. surgery: what the evidence says for common injuries

For much of the past century, a torn knee cartilage or a painful, grinding shoulder led to a fairly predictable conversation: a scan, a referral, and a date for keyhole surgery. It felt intuitive — find the damage on the image, go in and fix it. Then researchers did something unusual. They ran the surgery against a convincing fake, or against a course of guided exercise, and watched what actually happened to patients over the following year.

The results have reshaped orthopaedic guidelines. For several of the most common musculoskeletal complaints, structured physiotherapy — supervised, progressive exercise plus education — performs about as well as an operation, without the risks and downtime of surgery. That does not mean surgery is obsolete; for the right person and the right injury it remains the clear choice. It means the decision is more open, and more personal, than it once seemed.

The knee cartilage trial that changed the conversation

The clearest example is the degenerative meniscal tear — age-related fraying of the C-shaped cartilage in the knee, extremely common in middle age and often visible on an MRI even in people with no pain at all. Arthroscopic partial meniscectomy, trimming the torn portion through keyhole surgery, was for years one of the most frequently performed orthopaedic operations in the world.

In 2013, a Finnish team published a rigorous test in the New England Journal of Medicine. They randomly assigned 146 adults aged 35 to 65 with a degenerative tear and no significant osteoarthritis to either real meniscus surgery or a sham operation — skin incisions, the sounds and sensations of surgery, but no cartilage actually trimmed. Neither patients nor assessors knew who got what. At one year, the two groups had improved by the same amount [1].

A separate Dutch trial, ESCAPE, tackled the same question a different way: it compared keyhole surgery against a course of physiotherapy in 321 patients aged 45 to 70 with non-obstructive meniscal tears. Physiotherapy proved non-inferior to surgery for knee function, and that finding held at five-year follow-up [2]. Reviewing this and related evidence, an international expert panel writing in the BMJ issued a strong recommendation against arthroscopy for degenerative knee disease, concluding the small, short-lived average benefit did not justify the burdens and risks [3].

When the knee is arthritic, not just torn

The pattern extends to osteoarthritis. In a landmark 2002 placebo-controlled trial, patients with knee osteoarthritis who received arthroscopic washout or debridement did no better than those given a placebo procedure [4]. A 2008 Canadian trial reinforced it: arthroscopic surgery added no benefit over optimised physical and medical therapy for moderate-to-severe knee osteoarthritis [5]. Across these studies the message is consistent — for a worn, arthritic knee, keyhole surgery is not the answer, and guided exercise and load management do the heavy lifting.

The shoulder: decompression under scrutiny

Subacromial pain — the catch-all for shoulder pain aggravated by reaching overhead, often labelled "impingement" — was long treated with arthroscopic subacromial decompression, shaving bone and soft tissue to create more room for the tendons. The CSAW trial, published in The Lancet in 2018, put this to the test across 32 UK hospitals with a three-arm design: real decompression, a placebo arthroscopy (a look inside without the decompression), and no surgery [6].

At six and twelve months, the real operation offered no meaningful advantage over the placebo procedure, and the modest edge both surgical groups had over no treatment was small enough to be attributed largely to the effect of undergoing a procedure and the post-operative rehabilitation that followed. For this presentation, exercise-based rehabilitation is now widely positioned as the first-line approach [6].

The ACL: a more nuanced picture

The anterior cruciate ligament (ACL) is where the story becomes genuinely two-sided — and more interesting. The KANON trial randomised young, active adults with an acute ACL tear to either rehabilitation plus early reconstruction, or rehabilitation with the option of delayed reconstruction only if needed [7]. Structured rehab plus optional delayed surgery was not inferior on patient-reported outcomes at five years, and crucially, this strategy meant that around half of the participants never needed an operation at all.

The nuance matters. This does not show that surgery "doesn’t work" for ACL tears — reconstruction is highly effective and often the right choice, particularly for people with recurrent instability (the knee giving way) or those returning to pivoting, cutting sports. What KANON shows is that starting with rehabilitation, and reserving surgery for those who need it, is a legitimate and evidence-supported path for many people, rather than rushing everyone to the operating theatre.

Where surgery is clearly the right call

None of this is an argument against surgery. It is an argument for matching the treatment to the problem. Some situations are genuinely surgical, and delaying can cause harm. Broadly, the evidence and guidelines point to surgery being clearly indicated in cases such as these:

• A "locked" knee that cannot fully straighten, often from a mechanically obstructing (bucket-handle) meniscal tear.
• Acute, high-grade tears in athletes or manual workers whose knees remain unstable and give way despite good rehabilitation.
• Full-thickness rotator cuff tears in appropriate candidates, or shoulder problems with progressive weakness rather than pain alone.
• Fractures, dislocations, complete tendon ruptures, infection, or a joint so worn that joint replacement is being considered.
• Progressive nerve involvement, or "red flag" features that a clinician identifies as needing urgent assessment.

The recurring theme is mechanical blockage, instability, structural failure, or a condition that will worsen without intervention. Where the problem is degenerative wear or non-obstructing tissue changes accompanied by pain, the trials suggest a trial of good conservative care first is reasonable — and frequently sufficient.

How shared decision-making actually works

When two paths produce similar average results, the "best" choice depends on the individual — which is exactly what shared decision-making is designed to handle. Rather than a scan dictating the outcome, a clinician and patient weigh the trade-offs together against that person’s goals, risk tolerance, and circumstances.

1. Clarify the diagnosis and, honestly, how much the imaging findings explain the symptoms — incidental degenerative changes are extremely common in people with no pain.
2. Set out the realistic options, including the expected benefit, recovery time, and risks of each, drawing on the trial evidence.
3. Weigh what matters to the person: their sport or job demands, appetite for surgery, time available for rehabilitation, and other health conditions.
4. For many degenerative conditions, agree a defined trial of structured physiotherapy first, with clear checkpoints and a plan to reconsider surgery if progress stalls.

The strongest evidence rarely says "never operate." It says: for these common conditions, good conservative care deserves a fair trial first, and the operation should be reserved for those it genuinely helps.

A course of physiotherapy also leaves the door open. Choosing rehabilitation first rarely forecloses surgery later, whereas an operation cannot be undone. For a large share of the everyday knee and shoulder complaints that fill orthopaedic clinics, that reversibility — combined with trial evidence showing comparable outcomes — is why guidelines increasingly favour starting with the less invasive path.

None of this replaces an individual assessment. Which injury you have, how it behaves, and what you need from your body all shape the right decision. If you are weighing physiotherapy against surgery, the most useful next step is a conversation with a qualified physiotherapist or doctor who can examine you, interpret your imaging in context, and lay out the options that fit your situation.