Background Recurrent haemarthroses in patients with haemophilia commonly affect the ankles. It can result in haemophilic arthropathy and necessitate arthrodesis. The purpose of this report was to present the results of arthroscopic arthrodesis performed for this condition and to highlight some of the potential difficulties encountered as part of the surgical management. The operative technique is also described.

Method Between January 2001 and May 2004, eight male patients underwent nine arthroscopic arthrodeses. The mean age was 34 years (range 19–44). The patients were identified retrospectively from a surgical database and the radiographs and outpatient notes reviewed. Patients were also contacted to determine their level of satisfaction with the procedure.

Results The mean length of follow-up was 9 months (range 3–18). All patients had united both clinically and radiographically by the time of maximal follow-up and all were satisfied with the result. One patient had minor post-operative bleeding which settled spontaneously. No other significant complications were encountered. Bony cysts were observed in three patients pre-operatively and these all resolved following the attainment of union.

Conclusion Arthroscopic arthrodesis of the ankle is a safe and reliable treatment for haemophilic ankle arthropathy. The union rate is high, the complication rate is low, the risk of disease transmission from patient to staff is lessened and the post-operative rehabilitation regime, including allowing immediate weight bearing is less arduous than with traditional open procedures. Factor requirements are lessened and the length of stay is also reduced compared with open arthrodesis. Collaboration with a haematology unit is essential for a good result to be achieved.

Total elbow replacement can be a valuable option for the treatment of the elbow in haemophilia where there are associated arthropathic changes. We describe the outcome of seven elbow replacements in five consecutive patients with severe haemophilia A (native factor levels < 1%) at a mean of 42 months (25 to 65) after operation. All the patients had excellent relief of pain and improvement in function. One failure was due to infection in an immunocompromised patient with both HIV and Hepatitis C antibodies who was on anti-retroviral chemotherapy. The implant was revised at 30 months in a one-stage procedure and showed no evidence of loosening or infection 35 months later.

Septic arthritis has been regarded as rare in haemophiliacs, but its incidence may have increased since HIV infection has become widespread in these patients. We describe six cases treated at one haemophilia unit over a two-year period and discuss their investigation, diagnosis and treatment. Four of the patients were seropositive to anti-HIV.