Progressive migration in bipolar arthroplasty for osteoarthritis of the hip secondary to congenital dislocation.

Seventy-one hips with osteoarthritis secondary to congenital dislocation underwent bipolar arthroplasty with autologous bone grafting (grafted group) or without bone grafting (ungrafted group). Clinical and radiographic followup was performed; the average followup was 5.7 years (range, 3-11.3 years). At followup, the mean Merle d'Aubigné and Postel score was 14.3 points. Excellent or good results were achieved in only 50% of all cases. The mean extent of prosthetic head migration was 6.5 mm in the grafted group and 4.7 mm in the ungrafted group. Progressive migration occurred in 68% of the grafted group and in 43% of the ungrafted group. There was a significant negative correlation between the extent of migration of the outer head and the coverage of the outer head by the original acetabulum (R = 0.53). In the grafted group, the revision rate for failure of the outer head of the prosthesis was 60% at 6 years postoperatively. In the grafted group, progressive migration was more extensive and more frequently than in the ungrafted group. Extensive migration was also observed when there was adequate coverage of the outer head by host bone in the ungrafted group. In conclusion, bipolar arthroplasty is not recommended for osteoarthritis secondary to congenital hip dislocation because of the poor clinical outcome, high risk of significant progressive migration, and high revision rate.