The traditional treatment for a primary anterior shoulder dislocation has been immobilisation in a sling with the arm in adduction and internal rotation. The recurrence rates after the initial traumatic event range from 20% to 94%. However, recent results have suggested that recurrent instability after primary shoulder dislocation may be reduced with immobilisation in external rotation. Since then, controversy exists regarding the position of immobilisation following these injuries. The objective of the present study was to compare immobilisation in internal and external rotation after a primary anterior shoulder dislocation.

Fifty patients presenting to our fracture clinic with a primary traumatic anterior dislocation of the shoulder were randomly assigned to treatment with immobilisation in either internal rotation (IR; 25 patients) or external rotation (ER; 25 patients) for three weeks. In addition of a two-years clinical follow-up, patients underwent a magnetic resonance imaging (MRI) of the shoulder with intra-articular contrast within four days following the traumatic event, and then at three months of follow-up. The primary outcome was a recurrent dislocation within 24 months of follow-up. The secondary outcome was the healing rate of the labral lesion seen on MRI (if present) within each immobilisation group.

The follow-up rate after two years was 92% (23 of 25) in the IR group and 96% (24 of 25) in the ER group. The recurrence rate in the IR group (11 of 23; 47.8%) was higher than that in the ER group (7 of 24; 29.2%) but the difference did not reach statistical significance (p=0.188). However, in the subgroup of patients aged 20–40 years, the recurrence rate was significantly lower in the ER group (3 of 17; 6.4%) than that in the IR group (9 of 18; 50%, p<0,01). In the subgroup of patients with a labral lesion present on the initial MRI, the healing rate of the lesion was 46.2% (6 of 13) in the IR group and 60% (6 of 10) in the ER group (p=0.680). Overall, the recurrence rate among those who showed healing of the labrum (regardless of the immobilisation group) was 8.3% (1 of 12), but patients who did not healed their labrum had a recurrence rate of 45.5% (5 of 11; p=0.069).

This study suggests that immobilisation in ER reduces the risk of recurrence after a primary anterior shoulder dislocation in patients aged between 20 and 40 years. At two years follow-up, the recurrence rate is lower in patients who demonstrated a healed labrum at three months, regardless of the position of immobilisation. Future studies are required in order to identify factors that can improve healing of the damaged labrum following a traumatic dislocation of the shoulder.