Skip to content
View commentaries on this research

This is a plain English summary of an original research article. The views expressed are those of the author(s) and reviewer(s) at the time of publication.

A corticosteroid steroid injection into the shoulder provides some short-term pain relief for adults with rotator cuff disorders.

This review compared injection of corticosteroids (‘steroids’) with injection of local anaesthetic or placebo. The average improvement in pain relief at two months was calculated as moderate using standardised techniques. The effect wore off by three months.

Given the temporary benefits, it may be worth considering other treatments including physiotherapy alongside a steroid injection. Information given by an injecting physiotherapist, for example, regarding the expected duration of pain relief could also help manage the patient’s expectations for recovery in these painful conditions.

Why was this study needed?

Shoulder disorders are common and affect about three in ten adults at any one time. Rotator cuff disorders are the most common cause of shoulder pain. The tendons in the shoulder can be vulnerable to injury or tear, getting trapped (impingement), or just gradual degeneration of the tendon as a normal part of aging. Giving a corticosteroid injection into the joint is one option to treat pain.

Despite the widespread use of corticosteroid injections, past trials have had mixed findings and their use continues to be debated. The last review on the topic was conducted in 2010, and since then four new trials have been published.

What did this study do?

This systematic review and meta-analysis identified 11 randomised controlled trials including 726 adults with rotator cuff disorders. One of the trials compared corticosteroid injection with placebo injection of salt water, and the remaining trials compared with injection of local anaesthetic. Three trials used repeat injections.

All trials assessed pain at one, two and three months after the injections using a 0 to 10 visual analogue scale (VAS), with 10 indicating severe pain and 0 indicating no pain.

The number needed to treat (NNT) was calculated as the number of adults who would need to have an injection in order for one to experience a decrease in their pain score to mild or less (3.4 or lower on the VAS).

The researchers excluded three trials considered low quality. There was also substantial variation in the interventions and the way the trails were undertaken and this might lessen the applicability of the results to usual UK practice.

What did it find?

  • Corticosteroid injections did not significantly reduce pain compared with control at the final three month assessment (Hedges’ g effect size 0.23, 95% confidence interval [CI] -0.09 to 0.56).
  • At one month, corticosteroid injections had a small to moderate effect on pain compared with the control (Hedges’ g 0.44, 95% CI 0.15 to 0.73). However, less than 1 in 1000 who received a corticosteroid injection would meet the definition of mild pain at this point, so it was not possible to calculate a NNT.
  • The largest effect on pain was seen at two months (Hedges’ g 0.52, 95% CI 0.27 to 0.78). At least five adults with rotator cuff tendinosis would need to be treated with corticosteroid injections for one person to experience only mild pain at two months.
  • Repeated corticosteroid injections were no more effective than giving a single corticosteroid injection at all assessment points up to three months following injection.

What does current guidance say on this issue?

There is no NICE guidance on use of corticosteroid injections for treating rotator cuff disorders.

2010 guidance from the American Academy of Orthopaedic Surgeons on the use of corticosteroids for rotator cuff tears is inconclusive due to a lack of compelling evidence. It suggests practitioners use individual judgement and consider future publications and patient preference in their decision making.

What are the implications?

Corticosteroid injections provide moderate pain relief for adults with rotator cuff disorders up to two months after injection. There is no evidence of any effect after this time.

Steroids are unlikely to affect the long-term progress of these disorders but they may provide some short term relief. Other treatments such as exercise and physiotherapy also show some benefit. The cause of rotator cuff disorder, severity and duration of pain varied among the participants in these trials. Therefore, it is unclear if corticosteroid injections could be of universal help. A tailored offering including other therapies remains a pragmatic approach for now, while alerting patients to the short lived benefit.

Citation and Funding

Mohamadi A, Chan JJ, Claessen FM, et al. Corticosteroid injections give small and transient pain relief in rotator cuff tendinosis: a meta-analysis. Clin Orthop Relat Res. 2016. [Epub ahead of print, 28 July 2016].

No funding information was provided for this study.


American Academy of Orthopaedic Surgeons. Clinical practice guideline on the diagnosis and treatment of osteochondritis dissecans. Rosemont (IL): American Academy of Orthopaedic Surgeons (AAOS); 2010.

NHS Choices. Shoulder pain. London: Department of Health; 2014.

Produced by the University of Southampton and Bazian on behalf of NIHR through the NIHR Dissemination Centre

  • Share via:
  • Print article
Back to top