PATIENTS AND METHODS

We performed revision surgery for infected arthroplasty in 98 hips (96 patients). In 28 hips infection persisted. Of the remaining 70 hips, 61 (87%) were reviewed after a mean interval of 5.9 years (1 to 17). Infection occurred in 10 % of these after three years and in 26 % after ten years. The infection rate after initially successful revision for septic arthroplasty is higher than after revision for aseptic loosening and the risk of developing infection continues for many years. J Bone Joint Surv (Br] I995:77-B:307-9. Received 4 Ma 1994: A(cept(’d 6 Jo/v 1994 Most papers which report the results of revision surgery for infected total hip arthrop!asty use the term ‘reinfection’. This may imply recurrence of the original infection. It can occur, however, after revision for aseptic failure and it is therefore difficult to distinguish between a recurrence and a new infection. We report our infection rate after revision of infected arthroplasties and compare it with those of several series of revisions for aseptic loosening reported in the literature