We studied 57 patients with isolated lunotriquetral injuries treated by arthrodesis, direct ligament repair, or ligament reconstruction. The outcomes were compared by using written questionnaires, the Disabilities of the Arm, Shoulder and Hand (DASH) score, range of movement, strength, morbidity and rates of reoperation. Isolated lunotriquetral injury was confirmed by arthroscopy or arthrotomy. The mean age of the patients was 30.7 years (15.4 to 53.7) and the injuries were subacute or chronic in 98.2%. Eight patients underwent lunotriquetral reconstruction using a distally-based strip of the tendon of extensor carpi ulnaris, 27 had lunotriquetral repair and 22 had lunotriquetral arthrodesis. The mean follow-up was 9.5 years (2 to 22). The probability of remaining free from complications at five years was 68.6% for reconstruction, 13.5% for repair, and less than 1% for arthrodesis. Of the lunotriquetral arthrodeses, 40.9% developed nonunion and 22.7% developed ulnocarpal impaction. The probability of not requiring further surgery at five years was 68.6% for reconstruction, 23.3% for repair and 21.8% for arthrodesis. The DASH scores for each group were not significantly different. Objective improvements in strength and movement, subjective indicators of pain relief and satisfaction were significantly higher in the lunotriquetral repair and reconstruction groups than in those undergoing arthrodesis

Surgical access to the head of the radius is usually performed through a lateral approach. We present an alternative technique through a modified posterior approach which was developed following dissections of 22 human cadavers. An osteotomy of the supinator tuberosity was performed and reflected as a single unit with the attached annular ligament. Excellent exposure of the head of the radius was achieved, replacement of the head was undertaken and the osteotomy site repaired. The elbows were stable and had a full range of movement. The approach was then carried out on 13 patients for elective replacement of the head and was found to be safe and reproducible. In the patient group all osteotomies united, the elbows were stable and had an improved range of supination and pronation. There was no change in flexion and extension of the elbow. Complications included a haematoma and a reflex sympathetic dystrophy.

The modified posterior approach provides excellent access to the head and neck of the radius, gives good stability of the elbow and allows early mobilisation of the joint.

Over a seven-year period we treated a consecutive series of 58 patients, 20 men and 38 women with a mean age of 66 years (21 to 87) who had an acute complex anterior fracture-dislocation of the proximal humerus. Two patterns of injury are proposed for study based upon a prospective assessment of the pattern of soft-tissue and bony injury and the degree of devascularisation of the humeral head. In 23 patients, the head had retained capsular attachments and arterial back-bleeding (type-I injury), whereas in 35 patients the head was devoid of significant soft-tissue attachments with no active arterial bleeding (type-II injury).

Following treatment by open reduction and internal fixation, only two of 23 patients with type-I injuries developed radiological evidence of osteonecrosis of the humeral head, compared with four of seven patients with type-II injuries.

A policy of primary treatment by open reduction and internal fixation of type-I injuries is justified, whereas most elderly patients (aged 60 years or over) with type-II injuries are best treated by hemiarthroplasty. The best treatment for younger patients (aged under 60 years) who sustain type-II injuries is controversial and an individualised approach to their management is advocated.

There are theoretical and practical advantages to modular rather than monoblock designs of prostheses for shoulder arthroplasty, but there are no reported studies which specifically compare the clinical and radiological results of their use. We have compared the results of unconstrained total shoulder arthroplasty for osteoarthritis using both types of implant. The monoblock design was used between 1992 and 1995 and the modular design after 1995. Both had cemented all-polyethylene glenoids, the monoblock with matched and the modular with mismatched radii of curvature. There were 34 consecutive shoulders in each group with a mean follow-up of 6.1 years in the first and 5.2 years in the second.

There were no significant differences in improvement of pain scores, active elevation, external rotation, internal rotation, patient satisfaction, or the Neer ratings between the two groups. Two of 28 glenoid components in the first group and six of 30 in the second met the criteria for being radiologically at risk for loosening (p = 0.25). There were no significant differences in clinical outcome or radiological changes between the first- and second-generation designs of implant for shoulder arthroplasty.