Shoulder and elbow pain: which treatments work?


Anna Kelsey-Sugg

The evidence for and against some of the most common treatments for shoulder and elbow pain.

Dr Rachelle Buchbinder has a resume that requires more space than this article has room for. Here's a taste: practising rheumatologist and clinical epidemiologist; active researcher; director of the Monash Department of Clinical Epidemiology at the Cabrini hospital; professor of the Department of Epidemiology and Preventative Medicine at Monash University; and joint coordinating editor of the Cochrane Musculoskeletal Group.

She spoke at the People and Work – Managing Complexities conference in Adelaide about the effectiveness of different treatments for some of the most common elbow and shoulder complaints.

Dr Buchbinder began her talk by discussing rotator cuff disease, the most common cause of shoulder pain. She explained that fifty per cent of people suffering the condition improve within a year, though the symptoms are often recurrent, especially with aging. Treatment aims to relieve pain, and restore movement and function.

The different treatment option, as summarised by Dr Buchbinder, are as follows:

  • Simple analgesics are the first-line treatments
  • Oral and topical anti-inflammatories have benefits supported by only very limited evidence, but they probably provide short-term pain relief  
  • Steroid injections aren't supported by strong evidence but seem to consistently provide two to four weeks' pain relief. There is limited data about the best dosage and frequency  
  • Repeated injections are suggested to be useful. Diabetics need to be aware that blood sugar will be transiently elevated  
  • Guided injections have little evidence to support them  
  • Strengthening exercises might provide short- and long-term benefit but there is little evidence  
  • Non-operative treatments – conflicting studies on heat and ice packs, laser therapy, pulsed electromagnetic field therapy and ultrasound. The benefits lack strong evidence  
  • Acupuncture probably has a transient benefit  
  • Surgery is usually only beneficial for people with persisting symptoms and weakness. It usually takes three to six months to recover from. An exercise program seems just as good in terms of major outcomes.

Adhesive capsulitis (frozen shoulder or painful stiff shoulder) is the second most common cause of shoulder pain. Dr Buchbinder informed her audience that it usually occurs in people of working age and is more common in women. Though many people might believe it to be a result of workplace activity, in the majority of cases it is idiopathic (without a known cause), she said.

Anyone who's had the condition on one arm has a one in three chance of getting it on the other arm as well. The condition is different to rotator cuff disease in that it's self-limiting and it does get better almost completely. Ninety per cent of people improve within two years.

Dr Buchbinder outlined the following treatment options for adhesive capsulitis:

  • One or more steroid injection
  • Short course oral steroids  
  • Hydrodilation, injecting a large volume of fluid into the joint, is well evidenced to improve pain, function and range of movement  
  • Active physiotherapy program at the right time, sequential or combined treatments should be considered, manipulation under anaesthesia, surgery a last resort  
  • Manipulation under aesthesis can provide benefit, but there can be adverse effects, such as fracture and internal tearing

Lateral epicondylitis (lateral elbow pain, tennis elbow) effects 11 per cent of the general population, explained Dr Buchbinder, and is thought to be due to an overload injury at the common extensors. Like adhesive capsulitis it is self-limiting and 80 per cent of people affected by it improve within 12 months.

Treatments options she discussed include:

  • Resting and avoiding effects that aggravate symptoms
  • Ice or heat is not strongly supported by evidence but are used by some  
  • Oral and topical anti-inflammatories might provide short term relief but the trials aren't conclusive  
  • Braces are commonly used and the effect is unknown. They may provide come short-term benefit. They might lead to slightly earlier resumption of activities, but trials are conflicting  
  • Stretching/resistance exercises are usually carried out with other treatment so it's hard to isolate how effective they are  
  • Steroid injections show significant short-term benefits. Evidence isn't conclusive about long-term effects  
  • Ultrasound and acupuncture have conflicting results, with little proven benefit  
  • Shock waves and laser therapy have been found not to be effective  
  • Surgery has only been subjected to limited trials so there is no conclusive evidence.

Dr Buchbinder emphasised the difficulty of drawing conclusions about the effectiveness of our most common treatments, due to a lack of uniformity in how we label and define shoulder disorders. She said there is a wide variation in outcome assessment and timing of outcome assessment, and in the different types of interventions. She said there has also been varying quality of primary trials, with generally inadequate reporting of results. Good news though: according to Dr Buchbinder, quality of trials in this area is on the up. 


Australasian Faculty of Occupational & Environmental Medicine

Australasian Faculty of Rehabilitation Medicine

Published 30 June, 2008 | Updated 09 July, 2013