Ultrasound for diagnosis of infection in revision total hip arthroplasty.

Eighty consecutive patients (85 hips; 43 women; median age, 74 years [range, 33-90 years]) underwent a revision total hip arthroplasty. Preoperatively and guided by ultrasound, biopsy specimens of the joint pseudocapsule were taken, and joint fluid was aspirated for culture. Capsule morphology was investigated with light microscopy. When septic loosening was defined as 2 intraoperative cultures yielding the same microorganism, the prevalence was 12%. Sensitivity of the capsule biopsy cultures was 67%; specificity, 68%; positive predictive value, 22%; and negative predictive value, 94%. Joint fluid was often sparse and always falsely sterile. One specimen from a hip with septic loosening showed histologic changes (> or =3 foci, each with > or =3 plasma cells/high-power field) consistent with chronic infection. The remaining 8 septic loosenings eluded histologic detection, yielding 11% sensitivity. All aseptic loosenings were classified correctly (specificity 100%). Positive and negative predictive values were 100% and 89%. Ultrasound-guided aspiration and capsule biopsy with cultures and histology does not seem to be sufficiently accurate in the preoperative diagnosis of infected total hip arthroplasty.