We report the preliminary findings of the use of roentgen stereophotogrammetric analysis to evaluate the early migration of five femoral components after revision for mechanical loosening using impacted cancellous allograft and cement. All hips were examined at one week, four to six months and one year after surgery. All the components subsided by 0.4 to 4.9 mm during the first year. In four hips the prosthetic head was displaced 1.1 to 6.9 mm posteriorly. Fixation of the femoral components was less secure than after primary arthroplasty but the incidence and magnitude of early migration were similar to those after revision with cement alone.

We made a study of 49 total hip arthroplasties at five years, comparing the Charnley hip score, the Nottingham Health Profile (NHP) score and the appearance on serial radiographs. Each of the three investigations was performed by an independent observer with no knowledge of the results of the other two studies. The eight patients with radiographic signs of prosthetic loosening had significant reductions in function and quality of life as measured by the NHP questionnaire, but no differences in the mean Charnley hip scores. None of the eight patients had clinical signs of prosthetic loosening sufficient to recommend revision of their THR. The NHP is a relatively low-cost method of providing long-term follow-up of THR.

Roentgen stereophotogrammetric analysis was used to measure the migration of 24 cemented femoral components implanted during revision for mechanical loosening. All hips were examined one week, four months and one year after surgery; 14 hips were also examined after two years. Twenty-one components subsided 0.2 to 5.5 mm during the observation period; in 17 of these, subsidence occurred within four months of surgery. In 16 hips the prosthetic head was displaced 0.7 to 11.2 mm posteriorly. The fixation of the femoral components was less secure than after primary arthroplasty, especially in cases of femoral canal enlargement, when a standard-sized rather than a thick-stemmed prosthesis had been used, and in cases of inadequate cement filling.

Twenty patients were examined by standard radiography and roentgen stereophotogrammetric analysis (RSA) during a two-year period after total hip arthroplasty. Eleven of the acetabular components migrated cranially and three femoral components migrated distally. This migration was most rapid during the first four months after operation. Our findings support the possibility that mechanical loosening is initiated by thermal injury during polymerisation of the cement; the less frequent migration of the metallic femoral component compared with the polyethylene acetabular component may be because the metal acts as a heat sink. Standard radiographs were inadequate for assessment of early mechanical loosening, whereas RSA could reveal migration within four months of the arthroplasty.