Supervised rehabilitation comparable to single advice session after nonoperative treatment of displaced proximal humerus fracture: a randomized controlled trial

Background: More older adults with displaced proximal humerus fracture (PHF) are managed nonsurgically and referred to supervised rehabilitation, but its effectiveness is unknown. The primary objective of this trial was to investigate if supervised rehabilitation is superior to a single session of advice on patient-reported shoulder pain and function at 6 months postinjury, in older adults with displaced PHF treated nonoperatively. Methods: This is a pragmatic, single center randomized controlled trial conducted in an orthopedic outpatient clinic at a Danish university hospital. Eighty adults aged ≥60 with displaced PHF treated nonoperatively were screened between May 5, 2022 and November 11 2023. Patients were excluded if they presented with competing injuries. Sixty patients (mean age 75 years, 44 (73%) female) were enrolled. All patients received a session of advice on gradual return to daily activities. After randomization, 30 patients were allocated to referral to the experimental supervised rehabilitation by physiotherapists in the municipalities (usual care). The remaining 30 patients did not receive further training (comparator group). All patients received routine clinical follow-up with radiographs at six weeks and six months. The primary outcome was patient-reported shoulder function measured with Oxford Shoulder Score 6 months after injury. A 95% confidence interval excluding differences greater than 10 units between groups was prespecified as indicating the absence of a clinically meaningful difference. Secondary outcomes included Oxford Shoulder Score at 12 months, quality of life assessed with European Quality of life-5 Dimensions-3-Level at 6 and 12 months, conversion to surgery, and adverse events at any time point. Analyses were on an intention to-treat basis adjusted for age, sex, and fracture type. Results: Fifty-six (93%) completed the primary outcome score (n = 29 allocated to advice and supervised rehabilitation, n = 27 allocated to advice only). The least squares mean Oxford Shoulder Score at 6 months was 33.9 in the supervised rehabilitation group and 33.8 in the advice-only group. Consequently, in the main analysis, there was no difference between groups (mean difference, 0.05; 95% confidence interval, −4.5 to 4.6 P = .98), potentially corresponding to a negligible effect size <0.01. All secondary outcomes aligned with the primary endpoint. At 12 months, 4 patients had converted to surgery (3 supervised rehabilitation vs. 1 single advice session). Adverse events were similar across the 2 groups. Conclusions: The self-reported outcomes of supervised rehabilitation were comparable to those of a single advice session. Routine referral to supervised rehabilitation may not be necessary for all patients.