RESUMO
A prospective analysis of plain serial radigraphs (PSR), digital subtraction arthrography (DSA), and radionuclide bone scans (RBS) was performed in 56 cemented total hip arthroplasties to evaluate the efficacy and usefulness of each study in the diagnosis of loosening. To avoid selection bias in the evaluation of DSA and RBS, the decision to perform repeat surgery was based exclusively on the clinical history and PSR. Results of each study were compared with intraoperative assessment of the status of components and expressed in terms of sensitivity, specificity, and accuracy. Overall accuracy for the acetabular component by PSR was 66%; by DSA, 93%; by RBS, 46%. Overall accuracy for the femoral component by PSR was 79%; by DSA, 93%; RBS, 50%. Our results indicate that DSA can be recommended as a further analysis in cases with a painful hip prosthesis and no or inconclusive findings on PSR. RBS did not give any useful information and cannot be recommended routinely.
