Background: Dislocation is a leading and underemphasized cause of failure in revision total hip arthroplasty. Although this fact is generally well recognized, we are aware of no detailed assessments of this problem to date. Our purpose therefore was to evaluate the risk factors leading to instability after revision as well as the expected outcome of various treatment strategies.

Methods: Data were obtained from 1548 revision arthroplasties in 1405 patients who were followed for a minimum of two years (range, 2.0 to 16.4 years; mean, 8.1 years) or until dislocation occurred. Revisions specifically performed because of instability were excluded from the analysis. Risk factors were recorded along with treatment strategies and their success. The statistical relevance of both sets of variables was calculated.

Results: A dislocation occurred after 115 (7.4%) of 1548 revision hip arthroplasties. The use of an elevated rim liner was associated with significant decreases (p < 0.05) in dislocation following revision of femoral and acetabular components. The presence of trochanteric nonunion was a significant risk factor for subsequent dislocation (p < 0.001). Revisions with 32-mm and 28-mm-diameter femoral heads were both more stable than was revision with a 22-mm-diameter head (p < 0.05 for each). Surgery was the initial treatment for twelve of the 115 dislocations. Six of the twelve hips had no further instability. Of the 103 postoperative dislocations initially managed nonoperatively, only thirty-six did not redislocate. Thirty-eight of the sixty-seven hips that had an additional dislocation after closed treatment had repeat surgery for treatment of the instability. Only eleven of the thirty-eight hips were stable at one year after surgery. Overall, at the time of the final assessment, sixty-five (57%) of the 115 hips were stable, forty-one (36%) remained unstable, and the status of nine (8%) was unknown.

Conclusions: The risk factors for instability after a total hip revision are not the same as those after a primary procedure. The extent of the soft-tissue dissection is probably the most important variable since head size and trochanteric nonunion are related to "soft-tissue tension." Modular acetabular components with an elevated rim help to stabilize a hip undergoing a revision procedure.