Twelve adult patients with confirmed tuberculosis of the atlanto-axial spine are presented and a classification proposed. Stage I has minimal ligamentous or bone destruction and no displacement of C1 on C2; the suggested treatment is transoral biopsy and decompression followed by an orthosis. Stage II has ligamentous disruption and minimal bone destruction but anterior displacement of C1 on C2; treatment involves transoral biopsy and decompression, reduction by halo traction, then a posterior C1-2 fusion. Stage III has marked ligamentous and bone destruction with displacement of C1 forward on C2; the suggested treatment is transoral biopsy and decompression, reduction by halo traction, then fusion from the occiput to C2 or C3. One patient died before treatment started; all the others have gone on to solid union with resolution of any neurological deficit. There has been no evidence of reactivation of disease over an average follow-up of 36 months.

Twenty-one patients with spinal brucellosis were reviewed. The disease is difficult to diagnose, and is often confused with spinal tuberculosis. Our study showed that it was best diagnosed by serology and bacterial culture; radiography and scanning were less helpful in the early stages. After only six weeks' antibiotic treatment, there was a 55% clinical and serological reactivation rate: better results were achieved after at least three months of treatment. The adequacy of treatment was best monitored with repeated agglutination titres, and the duration of treatment proved to be more important than the antibiotic agent itself. Surgical intervention was reserved for biopsy, severe neurological impairment, or for spinal stabilisation.

One hundred and sixteen patients with 129 ununited fractures were treated either by rigid internal fixation and bone grafting or, in 14 tibial non-unions, by posterolateral bone grafting. In 37 actively infected cases this was combined with sequestrectomy and appropriate antibiotics. The final success rate was 98.4%. Technical errors or inadequate immobilisation after operation in patients with severe osteoporosis led to some early failures. No patient had a discharge or evidence of osteomyelitis after removal of metal. Rigid internal fixation with or without bone grafting is the treatment of choice for established non-union with bone loss when true synovial pseudarthrosis is present, when malalignment or shortening needs correction, and when prolonged immobilisation is either difficult or would lead to unacceptable stiffness in adjacent joints. Active osteomyelitis is not a contra-indication to internal fixation.