We report the outcome of 32 patients (37 knees) who underwent hemicallostasis with a dynamic external fixator for osteoarthritis of the medial compartment of the knee. There were 16 men (19 knees) and 16 women (18 knees) with a mean age at operation of 54.6 years (27 to 72). The aim was to achieve a valgus overcorrection of 2° to 8° or mechanical axis at 62.5% (± 12.5%). At a mean follow-up of 62.8 months (51 to 81) there was no change in the mean range of movement, and no statistically significant difference in the Insall-Salvati index or tibial slope (p = 0.11 and p = 0.15, respectively). The mean hip-knee-ankle angle changed from 190.6 (183° to 197°) to 176.0° (171° to 181°), with a mean final position of the mechanical axis of 58.5% (35.1% to 71.2%).

The desired alignment was attained in 31 of 37 (84%) knees. There were 21 excellent, 13 good, two fair and one poor result according to the Oxford knee score with no correlation between age and final score. This score was at its best at one year with a statistically significant deterioration at two years (p = 0.001) followed by a small but not statistically significant deterioration until the final follow-up (p = 0.17). All the knees with Ahlback grade 1 osteoarthritis had excellent or good results. Complications included pin tract infections involving 16.4% of all pins used, delayed union in two, knee stiffness in four, fracture of the lateral cortex in one and ring sequestrum in one.

In conclusion, hemicallostasis provides precision in attaining the desired alignment without interfering with tibial slope or patellar height, and is relatively free of serious complications.

Isolated dislocations of the navicular are rare injuries; we present our experience of six cases in which the navicular was dislocated without fracture. All patients had complex injuries, with considerable disruption of the midfoot. Five patients had open reduction and stabilisation with Kirschner wires. One developed subluxation and deformity of the midfoot because of inadequate stabilisation of the lateral column, and there was one patient with ischaemic necrosis. We believe that the navicular cannot dislocate in isolation because of the rigid bony supports around it; there has to be significant disruption of both longitudinal columns of the foot. Most commonly, an abduction/pronation injury causes a midtarsal dislocation, and on spontaneous reduction the navicular may dislocate medially. This mechanism is similar to a perilunate dislocation. Stabilisation of both medial and lateral columns of the foot may sometimes be essential for isolated dislocations. In spite of our low incidence of ischaemic necrosis, there is always a likelihood of this complication.

Neglected fractures of the femoral neck, common in young adults in underdeveloped countries, may be complicated by nonunion or avascular necrosis (AVN). We treated 52 cases by open reduction, fixation by compression screw and a free fibular graft. The mean delay between injury and operation was 5.1 months. Of 40 fractures assessed at a mean of 58.8 months (24 to 153), 38 were found to be united and two, owing to surgical errors, were not. Seven of eight heads which were avascular before operation revascularised without collapse, while seven others developed AVN after the procedure. At the last follow-up considerable collapse was apparent in five femoral heads, and 11 hips had developed coxa vara. The fibular graft had fractured in four cases. The hip had been penetrated by the screw in six cases and by the graft in three. Hip function was excellent in seven patients, good in 21 and fair in seven. Five patients had poor results. Incorporation of the fibular graft was seen after four years: in many cases the graft had been almost completely resorbed.

We recommend this procedure for the treatment of neglected fractures of the neck of the femur in young adults to reduce resorption of the neck, AVN and nonunion.