Modelb | Variables | Coefficient Bc | 95% CI | P-valued |
---|---|---|---|---|
Main complaints; mean differences between groups (0–100) | ||||
Model 12.1 | Included: GET | 7.5 | 0.0 to 15.0 | 0.049 |
Excluded: Recruitment strategy | 0.095 | |||
Model 26.1 | Included: GET | 7.1 | −1.0 to 15.2 | 0.084 |
Excluded: Recruitment strategy | 0.047 | |||
Model 52.1 | Included: GET | 9.2 | 1.4 to 17.1 | 0.021 |
Included: Recruitment strategy | −13.3 | −21.1 to −5.5 | 0.001 | |
Excluded: GET * Recruitment strategy | 0.624 | |||
SDQ; mean differences between groups (0–100) | ||||
Model 12.2 | Included: GET | 1.7 | −5.3 to 8.7 | 0.636 |
Excluded: Recruitment strategy | 0.013 | |||
Model 26.2 | Included: GET | 2.3 | −5.9 to 10.6 | 0.578 |
Included: Recruitment strategy | −12.0 | −20.3 to −3.7 | 0.005 | |
Excluded: GET * Recruitment strategy | 0.441 | |||
Model 52.2 | Included: GET | 2.1 | −6.3 to 10.5 | 0.624 |
Included: Recruitment strategy | −12.3 | −20.7 to −3.9 | 0.004 | |
Excluded: GET * Recruitment strategy | 0.418 |
↵a After imputation of the overall mean.
↵b Adjusted analyses of the main complaint instrument (model 12.1, 26.1, 52.1) and SDQ (model 12.2, 26.2, 52.2) after 12 weeks of treatment and during follow up at 26 and 52 weeks respectively.
↵c Regression coefficient for GET (GET=1; UC=0) and for recruitment strategy (Advertisement=1; GP=0).
↵d Statistically significant P<0.01. GET = graded exercise therapy. SDQ = shoulder disability questionnaire.