Table 3 Statistical analyses to be performed

ANCOVA with treatment group and center as qualitative factors and baseline value as covariate

Applicable for analyses at visits 5 and 7 for both treatment comparisons Tr14 injection solution against the comparators dexamethasone and placebo, respectively

• Change from baseline in abduction-rotation pain VAS for active external rotation

• Changes from baseline in ROM for abduction rotation as goniometry in degrees (active external, active internal, passive external, passive internal rotation)

• Changes from baseline in ROM for hand-back range and hand-neck range as distance measurement in cm

• Changes in DASH

Repeated measurements ANCOVA with treatment group, center, visit and treatment-by-visit interaction as qualitative factors and baseline value as covariate

Visits 5 and 7 included in analysis; both treatment comparisons Tr14 injection solution against the comparators dexamethasone and placebo, respectively:

• Change from baseline in abduction-rotation pain VAS for active external rotation

Logistic regression model with treatment group, center and baseline value (positive/negative) as qualitative factors

Applicable for analyses at visits 5 and 7 for both treatment comparisons Tr14 injection solution against the comparators dexamethasone and placebo, respectively:

• Jobe

• Painful arc

Cochran-Mantel-Haenszel (CMH) test for ordered categorical responses stratified by center

Applicable for analyses at visits 5 and 7 for both treatment comparisons Tr14 injection solution against the comparators dexamethasone and placebo, respectively:

• Patient’s global assessment

• Investigator’s global assessment