Background: Shoulder pain (usually non-traumatic) is the third most common cause of consultations for musculoskeletal pain in primary health care. This article discusses challenges within diagnostics and treatment of non-traumatic shoulder pain.
Material and methods: The article is based on literature identified through a non-systematic search in PubMed and the Cochrane library.
Results: One clinical test alone is generally not sufficient to make a pathoanatomical diagnosis. The incidence of degenerative changes in the normal population makes it difficult to interpret diagnostic images. MR is mainly indicated as part of a preoperative assessment and upon a specific indication. For diagnosis of the rotator cuff, ultrasound is as good as MR. Clinical studies do not show clear effects of specific treatment. This may be caused by methodological weaknesses of the trials evaluated and that patient characteristics and placebo effects contribute to the diagnosis. Injection with glucocorticoids seems to be well founded in patients with adhesive capsulate (frozen shoulder). With subacromial pain, supervised physical exercise and surgical treatment have similar effects, but there is not sufficient evidence to recommend particular treatments for rupture.
Interpretation: Non-traumatic shoulder pain is usually diagnosed adequately from the medical history and a combination of clinical tests; conservative treatment is most often adequate. When diagnostic imaging is required, ultrasound should be the first choice.