We present 2 year results of a prospective natural history study of Type III Acromioclaviclular joint dislocation (ACJD) treated non-operatively. Previous natural history studies are compromised by inconsistent definitions of the grade of injury and non-validated scoring tools; they do not identify which patients will have ongoing symptoms. This trial documents the strength and subjective recovery over time, and identifies risk factors for poor outcome and need for surgery.

Patients with Rockwood Type III ACJD received a standardised rehabilitation protocol (6 sessions of physiotherapy). Clinical assessment was performed at presentation, 3, 6, 12 and 24 months after injury, including isometric strength testing, pain (VAS/10) and subjective scores – Subjective percentage of normal (SPON), American Shoulder and Elbow (ASES), Oxford Shoulder Score (OSS) and Quick Disabilities of Arm Shoulder and Hand (qDASH).

28 male patients were recruited, 26 reached 12, and 9 to 24 months follow up. Two required surgery and one emigrated. The mean age was 39 (15 to 67). Initial mean pain was 3/10, SPON 51% (6–95) and strength was 76% of the other side. By 3 months mean subjective recovery was 70% and strength 90%. Strength recovered to 99% of normal by 12 months but subjective scores remained at mean 90%; by 24 months subjective scores were mean 94%. 2 patients had subjective scores <80%.

There was a wide range of initial subjective scores and weakness after Type III ACJD. 95% of strength had recovered by 6 months and subjective scores recovered to 94% of normal by 24 months. 4 patients (14%) did poorly with 2 requiring surgery. Low initial subjective score and inability to abduct the arm at presentation were risk factors for a poor outcome.