Question: In patients with frozen shoulder, does manipulation under anesthesia confer benefit in addition to that obtained from home exercises?
Design: Randomized (allocation concealed), blinded (outcome assessor) controlled trial with 1-year follow-up.
Setting: 3 regional hospitals in Finland.
Patients: 125 patients (mean age, 53 y; 71% women) who had a frozen shoulder with =140° in elevation and =30° in external rotation. Exclusion criteria included arthritis, osteoarthritis, traumatic bone or tendon changes in the affected shoulder, and rotator cuff rupture. Follow-up was 81% at 3 months and 63% at 1 year.
Intervention: Patients were allocated to receive manipulation under general anesthesia plus home exercises (n = 65) or home exercises alone (n = 60). Manipulation was done with the patient under brief general anesthesia. The physician lifted the arm and positioned it in flexion and abduction while the scapula was supported against the thoracic cage. After stretching the shoulder into flexion, the elbow was flexed to a right angle and the arm was rotated into internal and external rotation. Home exercises involved instruction from physiotherapists in two sessions and written instructions for a daily training program. The program included pendulum exercises for the arm and stretching techniques for the shoulder joint.
Main outcome measures: Range of motion, shoulder pain, disability (shoulder disability questionnaire), and working ability.
Main results: Results are presented for the 6-week and 3-month assessments because patient follow-up was <80% after 3 months. The study had 80% power to detect a clinically significant difference of 1.5, on a 0-to-10 pain intensity scale, between groups. At 6 weeks, the manipulation and home-exercise groups did not differ significantly for any outcomes (Table). At 3 months, the manipulation group had a greater degree of shoulder flexion, but no significant differences were observed for any other outcomes (Table).
Conclusions: In patients with frozen shoulder, manipulation under anesthesia did not confer additional benefit to that obtained with a home exercise program.
This study by Kivimäki et al. confirms that a home-based stretching exercise program is the primary treatment for management of primary frozen shoulder (adhesive capsulitis). Regardless of treatment after enrollment in the study, almost all patients were free of pain and had near-normal motion of the shoulder after 6 months of treatment.
Despite the evidence provided in this well-designed and performed randomized clinical trial, manipulation under anesthesia or arthroscopic capsular release may have a role in the management of refractory primary or secondary frozen shoulder. By most standards of clinical practice, surgical management is considered only after at least 6 months of a well-prescribed and performed home-based stretching exercise program. In this study, patients had not received previous treatment and had an average duration of symptoms of 7 months at the time of diagnosis. Less than 10% of enrolled patients had diabetes, and the study excluded patients with other forms of secondary frozen shoulder. A randomized clinical trial is needed to define the clinical value of manipulation under anesthesia or arthroscopic capsular release in patients who have primary frozen shoulder and in whom nonoperative treatment has failed or in patients with systemic diseases (diabetes or thyroid abnormalities) or frozen shoulder after surgery. Such patients may have a different prognosis and natural history.
In conclusion, Kivimäki et al. showed that manipulation under anesthesia does not add any clinical advantage for treatment of pain or loss of motion when used as part of the initial management of primary frozen shoulder and therefore should not be used for this indication.