This study of 10 consecutive patients analysed the benefit of percutaneous scaphoid fixation in minimally or non-displaced scaphoid fractures, looking at time to union, range of movement and strength. In all but one patient, who had an associated distal radius fracture, early mobilisation was the goal. Fixation was by canulated compression screw in all cases.

Clinical and radiological union took place in a mean time of 7 weeks. Mobilisation was possible from 2 weeks postoperatively in nine of the 10 patients. By the time of union, there was minimal loss of mobility. Patient satisfaction was high. There was no need for prolonged immobilisation in a cast.

Between 1997 and 2000, 25 Schatzker type-V and VI tibial plateau fractures were treated at our hospitals with the Ortho-fix ring fixator and followed up for between 10 months and four years. We chose this form of treatment for three reasons. First, because these are usually high-energy injuries, open reduction and internal fixation has to be done on admission or when the swelling has reduced, which can take up to three weeks. Secondly, internal fixation usually requires bone grafting at the same time. Thirdly, there is a high incidence of sepsis following conventional treatment with double plate and bone graft.

There were 20 closed or grade-I compound fractures and five grade-II compound fractures. The mean age of the 22 men and three women was 45.3 years (30 to 71). One patient had a head injury. There were ipsilateral femoral fractures in two patients, one of whom also had a comminuted distal radial fracture on the contralateral side, and one patient had a contralateral tibial plafond fracture. We routinely placed patients on a traction table and reduced the fracture, using Bohler’s method of traction and elastic bandaging. If necessary we made a limited incision and held the fragment with a large fragment screw under C-arm control. Using three or ideally four proximal wires, we made sure that the most proximal went through the fibular head. With local patients the frame was applied within 24 hours of admission. In patients who were referred from other parts of Africa, the frame was applied up to 10 days after the accident, irrespective of the amount of swelling. Continuous passive motion from 0° to 90° was started immediately postoperatively and maintained for a minimum of five days. Patients were then given crutches and mobilised touch weight-bearing in the frame. When the frame was removed, patients were fitted with a DonJoy hinged knee brace with no limitation of knee flexion or extension. All but two patients attained 90° of flexion within 10 days of application of the frame. Four patients failed to achieve full extension. Grade-I pin-tract sepsis developed in 12 patients and grade-II in four. All infections settled with treatment. One proximal wire needed to be resited. The mean time to removal of the frame was 4.5 months. No bone grafting was required.

We concluded that the amount of swelling is not a critical issue and that the ring fixator can be applied within 10 days of the injury. Probably because there was minimal interference with soft tissues, deep-seated sepsis did not occur. Even in osteoporotic bone, wire fixation gives excellent stability. There is still uncertainty about long-term follow-up, development of post-traumatic osteoarthritis and the possibility of sepsis if total knee arthroplasty is undertaken later.