To assess the effectiveness of a local corticosteroid injection versus physiotherapy-led exercise therapy over a 12 month follow-up period. The secondary aims were the effectiveness at 6 weeks, 3, 6, and 9 months.
A pragmatic randomized controlled trial with two parallel groups was performed. Patients consulting their general practitioner with a new episode of shoulder pain who were eligible for both a corticosteroid injection and exercise therapy could participate. Patients were randomly allocated (1:1) to a corticosteroid injection (triamcinolone acetonide 40mg, possibly in combination with lidocaine 10mg) or a 12-week physiotherapist-led exercise therapy program. Patients were followed for 12 months with 6 questionnaires in total: baseline, 6 weeks, 3, 6, 9 and 12 months. The primary outcome measures were pain and function measured with the Shoulder Pain and Disability Index (SPADI) over 12 months with a minimally clinically important difference of 10 points. A linear mixed model with repeated measures was used to generate estimates of effects (intention-to-treat).
The required sample size of 200 patients was achieved. Their mean age was 61 (sd=12) years and 53 percent was female. The mean SPADI score at baseline was 57 (sd=21). No serious side effects were reported and 7% of patients in the injection group and 20% in the exercise group reported minor side effects. Cross-over in the first three months after randomization was more prevalent in the injection group (n=17) than in the exercise therapy group (n=7).
The exercise therapy group showed a significantly greater improvement in pain and function at 12 months, compared to the injection group (β = 9.5, 95% CI = 2.8 to 16.3).
For the secondary outcomes, the injection group showed a significantly greater improvement in pain and function at 6 weeks (β = -6.7, 95% CI = -13.3 to -0.1), while the exercise therapy group had significantly greater improvements at 6 months (β = 10.6, 95% CI = 3.8 to 17.4) and 9 months (β = 9.0, 95% CI = 2.0 to 15.9).
Despite superiority of a corticosteroid injection at 6 weeks, patients with shoulder pain randomized to exercise therapy had clinically relevant and statistically significant better pain and function 6 to 12 months after treatment initiation, compared to the patients with shoulder pain randomized to the corticosteroid injection.
Future research should focus on the cost-effectiveness of these treatments and the interaction of these trial results with patient preferences for treatment in primary care.
A corticosteroid injection should only be considered for patient with shoulder complaints in primary care if short-term effectiveness is essential. For long-term improvement of pain and function, exercise therapy was more effective than a corticosteroid injection.
Primary Health Care
Conservative Treatment
