Objectives. - To define the sensitivity and specificity of routine preoperative hip aspiration for diagnosing hip prosthesis infection (HPI) and to separately analyze subgroups with and without a clinical suspicion of HPI before aspiration.

Methods. - From June 1994 to June 1997, all patients scheduled for hip revision surgery underwent aspiration of the hip under image intensifier guidance. Microbiological results were compared between these preoperative specimens and the intraoperative specimens. The reason for surgery was either a clinical suspicion of HPI or pain suggesting loosening.

Results. - The study patients had had multiple surgical procedures. HPI was suspected clinically in 39.4% of cases. Of the 109 patients who underwent aspiration, 54 had true-negative results, nine had false-negative results, and 44 had true-positive results (there were no false-positive results), yielding a sensitivity of 83% and a specificity of 100%. Diagnostic efficiency was 91.6%, positive predictive value was 100%, and negative predictive value was 85.7%. In the subset of 43 patients with a clinical suspicion of HPI, aspiration identified all the causative organisms in 60.5% of cases. Of the 66 patients with no clinical suspicion of HPI, 12 had HPI, and aspiration provided the diagnosis preoperatively in seven of these patients, radically changing their management plans. Restricting routine aspiration to patients whose prosthesis had been implanted within the last 5 years or whose erythrocyte sedimentation rate (ESR) was above 30 mm/h would not have modified our findings.

Conclusion. - Hip aspiration before revision surgery for pain is effective in detecting HPI, which can simulate aseptic loosening. However, this investigation may be noncontributive in patients who have had their prosthesis for more than 5 years and whose ESR is less than 30 mm/h.