In this video, a narrator describes the study, and animations illustrate the narrative.
[Text on screen] Logo of animation company: SciAni Science Animated
Video title: Frozen Shoulder: Which treatment should you choose? UK FROST Frozen Shoulder Trial
[Narrator]: Frozen shoulder is a common cause of shoulder pain and stiffness. About 8% of men and 10% of women of working age are affected by this disabling condition. They may struggle with basic daily activities. It can also cause sleep disturbance.
The exact cause remains unknown. Current treatments aim to relieve the symptoms. For those with persistent symptoms, specialist care in hospital commonly involves a steroid injection with physiotherapy, manipulation of the shoulder under anaesthesia, or keyhole surgery. Manipulation and keyhole surgery are also followed by physiotherapy to regain shoulder movements.
The UK Frozen Shoulder Trial (UK FROST) compared the effectiveness of the three care pathways. It is the largest trial of its kind, involving over 500 patients randomly assigned to one of the three groups.
All three treatments led to significant improvement, but none were clearly clinically superior using patient-reported questionnaires over a 12 month follow-up. Those who had steroid injection with physiotherapy (the only non-surgical care pathway) were more likely to need further treatment. Keyhole surgery carried higher risks and costs. Manipulation under anaesthesia was the most cost-effective treatment to the NHS, providing the best value for money for the benefits gained by patients in terms of their improved quality of life. Thanks to this trial, patients can now discuss the most appropriate treatment option for them with their doctor.
[Text on screen] Logos of supporting organisations: NHS Greater Glasgow and Clyde; NIHR; University of Oxford; BESS; The University of York; NHS South Tees Hospitals NHS Foundation Trust
Study details: Chief investigator: Professor Amar Rangan; Trial Manager: Dr Stephen Brealey; Acknowledgement: The trial management team would like to thank all the trial participants, collaborators, NHS hospitals, BESS, Funder and independent members of the trial oversight committees that contributed to the trial; Trial Sponsor: South Tees Hospitals NHS Foundation Trust http://www.southtees.nhs.uk; Trials Unit: University of York Trials Unit https://www.york.ac.uk/healthsciences/research/trials/; Funding/Support: This project was funded by the NIHR Health Technology Assessment programme (project number 13/26/01); Disclaimer: The views expressed are those of the author and do not necessarily reflect those of the HTA programme, the NIHR, the NHS, or the Department of Health and Social Care. The full study results are published in The Lancet: http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)31965-6/fulltext

View commentaries on this research

This is a plain English summary of an original research article

A so-called frozen shoulder is painful and stiff for months and sometimes years. People with the condition may be unable to move their arm or shoulder and the pain may disturb their sleep.

The three treatments most often offered by the NHS are physiotherapy, manipulation of the shoulder under general anaesthesia, and a form of keyhole (minimally-invasive) surgery called arthroscopic capsular release.

The largest comparison of these treatments to date found that, after a year, all three approaches were effective, but they had different benefits and drawbacks. Physiotherapy could be delivered more rapidly but was more likely to require follow-up treatment. Surgery carried higher risks and costs. Manipulation under anaesthesia was the most cost-effective treatment, but waiting times were longer than for physiotherapy.

The findings should help patients and clinicians decide on the most appropriate treatment. The study could reduce the use of surgery for people with a frozen shoulder, and make considerable cost savings for the NHS.

What’s the issue?

Frozen shoulder occurs when the soft tissue around the shoulder joint becomes inflamed and scarred. The cause is often unknown, but when the tissue shrinks to become tight around the joint, it can make movement so painful and stiff that people may struggle with basic daily activities and have disturbed sleep.  Around one in 10 people in their 50s have frozen shoulder, and most are treated in the community. Those who do not improve are referred to hospital.

There is little good quality evidence on which of the three treatments commonly used in NHS hospitals (physiotherapy, keyhole surgery and manipulation under anaesthesia) leads to better outcomes. However, hospitals commonly offer keyhole surgery and manipulation because they are widely believed to lead to faster recovery. Both of these procedures involve anaesthesia, and therefore carry increased risks and costs. Many hospitals supplement them with steroid injections, but its use is variable.

The new study aimed to compare the effectiveness and cost-effectiveness of the three techniques and provide evidence to guide future practice.

What’s new?

The UK-wide trial looked at the treatment of 503 adult patients with frozen shoulder at 35 NHS hospitals. Each was randomised to receive either:

    • structured physiotherapy (up to 12 weekly sessions including a home exercise programme and written advice on pain management) plus a steroid injection
    • manipulation under anaesthetic with steroid injection followed by physiotherapy, or
    • keyhole surgery followed by physiotherapy.

People in all three groups had improved shoulder function after one year and many had nearly full function. No treatment was clearly superior to the others in improving function or reducing pain, but each had different benefits and drawbacks.


    • was the cheapest option
    • had the shortest waiting times (an average of 14 days), which was valued by patients
    • had no serious adverse events such as chest infection
    • had the highest rates of people needing further treatment (15%).

Manipulation under anaesthetic:

    • had a longer waiting time than for physiotherapy (57 days on average)
    • improved quality of life
    • had two serious adverse events reported
    • had lower rates of people needing further treatment (7%).

Keyhole surgery:

    • was the most expensive option, costing £1733 more per person than physiotherapy
    • with the longest waiting time (72 days on average)
    • had slightly better outcomes in shoulder function and pain than the other options, but this improvement was unlikely to be meaningful
    • had the greatest risk of serious adverse events (8 reported)
    • had the lowest rates of people needing further treatment (4%) of the three treatments.

Overall, the study found that manipulation under anaesthetic provided the best value for money of the three options.

Why is this important?

The study shows that a programme of structured physiotherapy – which can be provided at low cost to the NHS – is as effective as the other options, and many people value the shorter waiting times. However, the structured programme of up to 12 sessions, combined with steroid injection, is not currently offered by the NHS. There is considerable variation in the physiotherapy offered, and steroid injection is not usually given. The physiotherapy pathway used in the study would therefore need extra funds to deliver.

This information will help clinicians and patients decide together on the most appropriate treatment. For example, they show that for patients who wish (or need) to avoid a general anaesthetic, physiotherapy and a steroid injection is an effective alternative. The authors argue the results should lead to more selective use of keyhole surgery, thus reducing the cost and risks for patients.

What’s next?

The results should change practice, the study authors say, particularly to favour manipulation under anaesthetic over keyhole surgery.

The study did not examine a fourth treatment option called hydrodilatation, in which the steroid is injected with a large volume of saltwater (saline) to distend the shoulder, followed by physiotherapy. The technique was rarely used when this study was planned but has grown in popularity since then. The authors say its effectiveness should be formally tested in a trial.

You may be interested to read

The full paper: Brealey S, and others. Surgical treatments compared with early structured physiotherapy in secondary care for adults with primary frozen shoulder: the UK FROST three-arm RCT. Health Technology Assessment 2020;24 

Further details are included in: Rangan A, and others. Management of adults with primary frozen shoulder in secondary care (UK FROST): a multicentre, pragmatic, three-arm, superiority randomised clinical trial. The Lancet 2020; 396: 977-89. 

Summary of treatment options: Rangan A, and others. What is the most effective treatment for Frozen Shoulder? BMJ 2016; 354: i4162 

Further research studying the experiences of trial participants and healthcare professionals: Srikesavan C, and others. Experiences and perceptions of trial participants and healthcare professionals in the UK Frozen Shoulder Trial (UK FROST): a nested qualitative study. BMJ Open 2021;11:e040829


Funding: This project was funded by the NIHR Health Technology Assessment programme.

Conflicts of Interest: One researcher has received funds from a medical device company. 

Disclaimer: NIHR Alerts are not a substitute for professional medical advice. They provide information about research which is funded or supported by the NIHR. Please note that views expressed in NIHR Alerts are those of the author(s) and reviewer(s) and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care.

  • Share via:
  • Print article


Study author

I was surprised to see that manipulation under anaesthesia was the most cost-effective intervention and that the surgical treatments were not clinically superior to physiotherapy. That is likely to be because we standardised the physiotherapy treatment pathway, with up to 12 sessions and a steroid injection. That specific pathway is not currently available on the NHS. If we had tested current usual care in the NHS with non-standardised physiotherapy (not including a steroid injection), I believe that the surgical treatments would have been clearly superior to physiotherapy. The study findings therefore only apply if the physiotherapy pathway follows that used in the study.Amar Rangan, Shoulder Surgeon, James Cook University Hospital, Middlesbrough 

Specialist physiotherapists 

The intervention was a standardised pathway of up to 12 sessions of physiotherapy in 12 weeks AND a steroid injection in a secondary care setting. The pathway was informed by systematic review of effective treatment for frozen shoulder, evidence-based guidelines and a delphi survey of shoulder physiotherapy specialists across the NHS. The physiotherapy programme included mobilisation, advice, education, home based exercises and some supervised exercises plus the disallowing of immobilisation, some electrotherapy and 'alternative therapies'. The primary care pathway that resulted in referral to secondary care was not described. Current practice in some primary and secondary care physiotherapy services is to review frozen shoulder patients once every 3-6 weeks, in part related to resource constraints. Provision of weekly, face to face sessions for up to 12 weeks may prove challenging, particularly in the current COVID-19 climate. This study supports the pathway for frozen shoulder in physiotherapy being appropriately accessible, coordinated with steroid injection and available for 12 weeks to allow patients who prefer not to have surgery the best chance for a good outcome. However, one possible, clinical question arising from this study is, how does this study impact current physiotherapy management of primary frozen shoulder in primary care?” Sharon Morgans and Cathy Barrett, Shoulder Specialists, Central Health Physiotherapy, London, and Members of the Musculoskeletal Association of Chartered Physiotherapists (MACP).  

Member of the public 

From the perspective of patient, partner and family, this study has uncovered more detail about the pros and cons of each of these treatments. This should allow patients and health professionals to make a more considered and informed decision. The risks, potential complications and potential benefits all need to be explained to patients in plain language. It needs to be spelt out that if a particular treatment is not successful, they may well have to have another one of the options anyway. This research could be followed up with a study into how future shared decision making on these three treatment pathways pans out. That's where the proving ground of the implementation will be.Matthew Moore, Public Contributor, Nottingham

Member of the public 

It would be interesting to revisit the three groups after three, five and ten years to see if the outcomes change. It seems from the detail that although physiotherapy alone is quite effective, the longer term effectiveness of this and the other treatments still needs to be demonstrated.Pamela Law, Public Contributor, Northampton 
Back to top