Skip navigation
Acute Pain | Chronic Pain | General

Ultrasound therapy for musculoskeletal disorders

Clinical bottom line:

Ultrasound is not effective for reducing pain or functional disability associated with lateral epicondylitis (tennis elbow), soft-tissue shoulder disorders or ankle distorsions. Based on three trials, ultrasound showed some benefit for general improvement of symptoms. The NNT for ultrasound for 'at least satisfactory general improvement of symptoms' compared with placebo at 4 to 8 weeks was 6.0 (95% CI 3.2 to 40). There was no convincing evidence for ultrasound for degenerative rheumatic disorders, heel pain, myofascial trigger points, supraspinatus tendinitis or lateral epicondylitis. However, more trials of higher quality are required to confirm this. Ultrasound in combination with exercise therapy did not offer any further benefit.

Therapeutic ultrasound is often used in the treatment of musculoskeletal disorders. In Canada 94% of physical therapists use ultrasound. Ultrasound is assumed to have thermal and mechanical effects on the target tissue resulting in an increased local metabolism, circulation, extensibility of connective tissue and tissue regeneration.

Systematic review

Van der Windt DAWM, van der Heijden GJMG, van den Berg SGM, ter Riet G, de Winter AF, Bouter LM. Ultrasound therapy for musculoskeletal disorders: a systematic review. 1999; Pain 81; 257-271.

Date review completed: July 1997

Number of trials included: 38

Number of patients:

Control group: active and placebo (sham ultrasound)

Main outcomes: improvement in pain or functional disability

Inclusion criteria were randomised controlled trials of ultrasound for pain and/or restriction of range of motion associated with musculoskeletal disorders; ultrasound alone or as adjuvant to exercise therapy; placebo (sham ultrasound) or active comparisons; any language.

Reviewers provided a descriptive summary of trials taking quality into account. Evidence for ultrasound was assessed according to quality and quantity of data from trials. Data were pooled where trials represented clinical homogeneity with respect to diagnosis, intervention and outcome measures. Only high quality trials were considered for data pooling. For general improvement outcomes (dichotomous information on numbers with symptom improvement), differences in success rates between study groups were calculated with 95% confidence intervals, together with the NNT point estimate. For non-dichotomous information, differences between groups in change from baseline or differences in post-treatment scores were calculated. Effect sizes were calculated as the difference between the mean (change) scores in the compared groups divided by their pooled standard deviation. We have calculated the relative benefit with 95% confidence interval for the pooled data as a measure of statistical significance, together with the NNT with 95% confidence intervals. We have not reported on individual effect sizes from trials as individual trials are either too small or of insufficient quality to provide useful clinical information.


Included trials were of mixed quality and blinding status, and were generally of small size. Eighteen trials included a placebo control. Number of treatment sessions was not stated.

Lateral epicondylitis (tennis elbow).

Data were pooled from three high quality, placebo-controlled trials in 185 patients looking at outcomes over 4 to 8 weeks. Two of three reported no difference, but the pooled relative benefit showed some benefit of ultrasound compared with placebo for general improvement in symptoms, 1.4 (95% CI 1.04 to 2.0). The NNT for ultrasound for 'at least satisfactory general improvement of symptoms' compared with placebo at 4 to 8 weeks was 6.0 (95% CI 3.2 to 40). The remaining three trials were of poor quality, and reported a significant benefit compared with no treatment (one trial), and inconsistent results compared with other therapies (four comparisons). Evidence suggests that ultrasound may be of benefit for treating tennis elbow.

Soft-tissue shoulder disorders.

None of the seven trials showed any benefit of ultrasound over control. This included five placebo comparisons, three no-treatment comparison, and three active comparisons. Evidence suggests that there is no benefit of ultrasound for soft-tissue shoulder disorders.

Degenerative rheumatic disorders.

Ten trials were identified. These trials were mainly active comparisons, and findings were mixed. However, trials were of insufficient size to measure this adequately, and quality of trials was poor. There is therefore a lack of evidence for ultrasound in this area.

Ankle distorsions.

Three of three high quality controlled trials showed no benefit of ultrasound over placebo, suggesting that ultrasound is ineffective.

Temporomandibular joint pain or myofascial pain.

Four low quality trials reported no benefit in most cases. There is therefore a lack of evidence for ultrasound in this area.

Other diagnoses.

Seven trials considered patients with a variety of musculoskeletal disorders (heel pain, myofascial trigger points, supraspinatus tendinitis or lateral epicondylitis). Description of results was unclear, but there was no convincing evidence for ultrasound in these disorders.

Ultrasound in combination with exercise therapy.

Of the included trials, 13 considered ultrasound in combination with exercise therapy. The four high quality trials all reported no benefit of ultrasound plus exercise compared with exercise alone, suggesting no additional benefit of ultrasound.

Adverse effects

Reviewers did not report on adverse effects.

Further reading

Related topics