Back and shoulder surgeries are sometimes worse than useless
This page was made after seeing The Economist: Back and shoulder surgery is OFTEN worse than useless
I no longer subscribe to The Economist ($$$)
Claude AI
It's an Economist leader (opinion piece) from September 24 [Economist leader]. A companion science article ran in the September 26 print edition, arguing that randomised trials are showing many surgeries to be unnecessary. It names spinal fusions and rotator-cuff repairs as procedures whose effects match placebo or physiotherapy [Economist, "Jack the knife"].
The "worse than useless" logic is simple. If an operation does no better than a sham or physiotherapy, its risks (infection, anaesthesia, revision surgery, months of recovery) make it a net loss. The evidence for the main targets is strong:
| Procedure | Key evidence | Finding |
|---|---|---|
| Shoulder subacromial decompression (bone-spur removal) | CSAW trial, Lancet 2018; FIMPACT (Finland) | No better than placebo arthroscopy |
| Rotator cuff repair, degenerative small/medium tears | Finnish 3-arm RCT; Norwegian 103-patient RCT | No difference or small differences vs physiotherapy |
| Vertebroplasty (cement into a fractured vertebra) | Cochrane review 2018 | No important benefit vs sham procedure |
| Adding fusion to decompression for degenerative spondylolisthesis | NORDSTEN-DS, NEJM 2021 | Fusion added nothing at 2 years |
| Fusion/laminectomy for non-specific age-related back pain | Multiple RCTs vs exercise | Little or no benefit |
Shoulder. CSAW enrolled patients who had already failed exercise therapy and steroid injections. Removing the bone spur gave no extra benefit over arthroscopy alone, which served as the placebo [CSAW, PubMed]. After this trial, NHS England classed the operation as a "procedure of limited value" [Oxford REF impact]. British operations fell from 28,000 to 5,720 within a few years, while the procedure stayed popular in America [Economist, "Jack the knife"].
For non-traumatic supraspinatus tears, a Finnish trial found no meaningful difference among physiotherapy, acromioplasty, and tendon repair [Finnish RCT summary].
Back. Cochrane found high- to moderate-quality evidence that vertebroplasty gives no important benefit over sham. It also found that open (unblinded) trials had probably overestimated the benefit [Cochrane vertebroplasty]. For people with degenerative spondylolisthesis (one vertebra slipped forward on the next) who were already having surgery, adding fusion to decompression didn't help [NORDSTEN-DS].
A Lown Institute analysis estimated over 200,000 unnecessary back surgeries on Medicare patients in three years, costing about $2 billion. Rates ranged from zero to more than half of procedures, depending on the hospital [Lown Institute].
Broader pattern. A review of 53 sham-controlled surgical trials found that the real operation was no better than the sham in 27 of them [Sham surgery, Wikipedia]. This is the real reason the Economist's argument holds: many operations became standard before anyone tested them against a placebo.
Where the headline goes too far: * "Often" is doing a lot of work. Back and shoulder surgery clearly helps in many situations: cauda equina syndrome, spinal cord compression, fractures, tumours, infection, deformity, shoulder replacement for arthritis, and disc herniation with persistent sciatica (where surgery speeds relief). * For rotator cuff tears, repair results were slightly better at 5 years in one trial. Also, about a third of unrepaired tears grew and had worse outcomes [Moosmayer RCT]. So for some patients, "useless" is too strong. * The CSAW authors stressed that their results don't justify denying other kinds of shoulder surgery once non-operative care has failed [Oxford NDORMS]. * The overuse problem is largest in fee-for-service systems like the US, where surgery brings in revenue. The research mostly comes from Britain, Finland, Norway, and Australia, where usage has already dropped.
Bottom line: For degenerative, non-specific pain, the claim is well supported. Before agreeing to an elective spine or shoulder operation, it's worth asking whether that specific procedure has been tested against a sham or physiotherapy, and what that trial found.
Related in VitaminDWiki
- Back Pain
- 60,000 IU weekly vitamin D beat physiotherapy alone for chronic low back pain - RCT
- Off topic: back muscle pain quickly treated by TENS
- Lumbar Degenerative Disc Disease 3X more likely if poor Vitamin D Receptor
- Rotator cuff and Vitamin D - several studies
- Actual arthroscopic knee surgeries were not statistically better than sham surgeries - several studies