We undertook a prospective, controlled trial which compared two rehabilitation programmes for 86 patients who sustained two-part fractures of the proximal humerus. Patients were randomised either to receive immediate physiotherapy within one week (group A) or delayed physiotherapy after three weeks of immobilisation in a collar and cuff sling (group B).

At 16 weeks after the fracture, patients in group A had less pain (p < 0.01) and had greater shoulder function (p < 0.001) than those in group B. At 52 weeks, the differences between the groups had reduced. Although group A still had greater shoulder function and less pain, there was no statistical difference when compared with group B. By analysis of the area under the curve, an overall measure up to the 52-week period, group A experienced less pain as measured by the SF36 general health questionnaire and had improved shoulder function.

Our results show that patients with two-part fractures of the proximal humerus who begin immediate physiotherapy, experience less pain. The gains in shoulder function persist at 52 weeks which suggests that patients do not benefit from immobilisation before beginning physiotherapy.

Between 1995 and 2000, 19 consecutive patients with fractures of the distal humerus were treated by primary total elbow replacement using the Coonrad-Morrey prosthesis. No patient had inflammatory or degenerative arthritis of the elbow. The mean age at the time of injury was 73 years (61 to 95). According to the AO classification, 11 patients had suffered a C3 injury, two a B3 and two an A3. One fracture was unclassified. Two patients died from unrelated causes and one was unable to be assessed because of concurrent illness.

The mean time to follow-up was three years (1 to 5.5). At follow-up 11 patients (68%) reported no pain, four (25%) had mild pain with activity and one had mild pain at rest. The mean flexion arc was 24° to 125°. The mean supination was 90° (70 to 100) and pronation 70° (50 to 110). No elbow was unstable. The mean Disabilities of the Arm, Shoulder and Hand score was 23 (0.92 to 63.3) and the mean Mayo elbow performance score was 93 (80 to 100). Of the 16 patients, 15 were satisfied with the outcome.

Radiological evaluation revealed only one patient with a radiolucent line at the cement-bone interface. It was between 1 and 2 mm in length, was present on the initial postoperative radiograph and was non-progressive at the time of follow-up.

Primary total elbow arthroplasty is an acceptable option for the management of comminuted fractures of the distal humerus in elderly patients when the configuration of the fracture and the quality of the bone make reconstruction difficult.