Decompression surgery does not reduce shoulder pain any more than placebo surgery for people with shoulder impingement - when the tendon rubs and catches in the joint, according to the first placebo-controlled trial in shoulder surgery published in The Lancet.
Furthermore, although both types of surgery were slightly more effective at reducing subacromial shoulder pain compared to no treatment, the difference was small and not likely to result in a noticeable effect.
In the USA, shoulder pain accounts for 4.5 million visits to the doctor each year, and 2.4% of all GP visits in the UK. For subacromial shoulder pain, treatment options include subacromial decompression, which is one of the most commonly used surgeries in orthopaedics. In England, the number of surgeries has risen from 2523 in 2000, to 21355 in 2010.
The authors of the study say the findings question the value of this operation, and that patients considering undergoing the operation should be informed.
"Over the past three decades, patients with this form of shoulder pain and clinicians have accepted this surgery in the belief that it provides reliable relief of symptoms, and has low risk of adverse events and complications. However, the findings from our study suggest that surgery might not provide a clinically significant benefit over no treatment, and that there is no benefit of decompression over placebo surgery." says co-chief investigator Professor Andrew Carr, National Institute of Health Research (NIHR) Biomedical Research Centre, University of Oxford, UK. 
The study involved 32 hospitals and 51 surgeons across the UK. In the study, 90, 94 and 90 people who had suffered shoulder pain for at least three months despite having physiotherapy and steroid injections underwent decompression surgery, placebo surgery or no treatment, respectively. They completed questionnaires about their level of pain and function at six months.
Decompression surgery is a keyhole surgery that involves removing a small area of bone and soft tissue in the shoulder joint to open up the joint and prevent rubbing or catching when the arm is lifted. In the placebo surgery, surgeons conducted a procedure to look inside the joint where the joint was inspected but no tissue was removed.
Both surgeries were completed as keyhole procedures to ensure that patients were not aware of which surgery they had had. Surgery participants also had one to four physiotherapy sessions afterwards, while those having no treatment only had a check-up appointment three months after the start of the trial.
Six and 12 months after they entered the trial, the participants completed questionnaires rating their symptoms, including pain (from 0-48, with a higher number meaning less pain). Overall, symptoms diminished in all three groups from the start of the trial.
At six months, people who had had decompression surgery and those who had had placebo surgery rated their pain and function at a similar level, with no statistical difference (32.7 points and 34.2 points, respectively).
Comparatively, both forms of surgery showed a small benefit over no treatment (rated at 29.4 points in the no treatment group), however, the difference is unlikely to result in a noticeable difference in symptoms.
The authors suggest that the difference could be attributable to a number of factors, including a placebo effect related to surgery, a nocebo effect related to having no treatment, other unintended effects of the placebo surgery, or because people undergoing surgery were given physiotherapy and told to rest.
"Our findings call into question the value of shoulder decompression surgery for this group of patients, and should be communicated to patients and doctors considering this type of surgery. In light of our results, other ways to treat shoulder impingement could be considered, such as painkillers, physiotherapy and steroid injections." says co-chief investigator Professor David Beard, University of Oxford, UK. 
During the trial, six participants had a frozen shoulder (two people in each treatment group) related to the study, and one person in the placebo surgery group underwent decompression surgery for pain.
The authors note some limitations to their study, including that some patients did not continue with their assigned treatment (23%, 42% and 12% of decompression, placebo surgery and no treatment participants, respectively). For example, some patients assigned to surgery did not undergo treatment as their symptoms improved, and other patients assigned to no treatment chose to undergo decompression surgery. However, the authors found that the effects of this do not impact on the results.
In addition, there were long waiting times within the study meaning some people who had only recently had surgery and may have still been recovering were compared with people who had had no treatment for six months, which may affect the results. On average, people assigned to decompression surgery waited 90 days for their surgery, and people assigned to placebo surgery waited 82 days.
The study did not look at the recurrence of pain after a year, but the authors state that it is unlikely that one group would show larger improvements long-term when they have not done so in a year after surgery.
Writing in a linked Comment, Dr Berend Schreurs, Radboud University Medical Center, the Netherlands, says "The findings send a strong message that the burden of proof now rests on those who wish to defend the standpoint that shoulder arthroscopy is more effective than non-surgical interventions. Hopefully, these findings from a well respected shoulder research group will change daily practice. The costs of surgery are high, and although the low occurrence of complications might suggest that the surgery is benign, there is no indication for surgery without possible gain."
NOTES TO EDITORS
This study was funded by Arthritis Research UK, the National Institute for Health Research Biomedical Research Centre, and the Royal College of Surgeons. It was conducted by researchers from National Institute of Health Research (NIHR) Biomedical Research Unit at the University of Oxford, University of Oxford, and University of Bristol.
 Quote direct from author and cannot be found in the text of the Article.
IF YOU WISH TO PROVIDE A LINK FOR YOUR READERS, PLEASE USE THE FOLLOWING, WHICH WILL GO LIVE AT THE TIME THE EMBARGO LIFTS: http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(17)32457-1/fulltext?elsca1=tlpr