Evaluate the positive predictive value of labeled-leukocyte scintigraphy performance in the diagnosis of bone and joint infections

Objective. - Evaluate the positive predictive value of labeled-leukocyte scintigraphy in the diagnosis of bone and joint infections with or without orthopaedic implants. Materials and methods. - Retrospective study of 80 patients suspected of bone and joint infections, with or without orthopaedic implants, who underwent bone scintigraphy, labeled-leukocyte scintigraphy and colloid bone marrow scintigraphy. This preliminary study focuses on nonnegative scintigraphy results performed from 2005 to 2010. The final diagnosis conclusion was based on clinical, biological, bacteriological data, imaging techniques, and follow-up. The patients were classified as true or false-positives. The influence of the different parameters on the exam's diagnostic performances was assessed by a statistical analysis. Results. - Among the 80 scintigraphies performed, 55 were positives and 25 were inconclusive. Among the 55 positive scintigraphies, there were 33 true-positives, 10 false-positives and 2 inconclusive (10 patients were lost sight). Positive predictive value was 76.7%. There was no significant difference between sub-groups: with or without orthopaedic implants (P = 0.385), suspected bone infection location (axial or peripheral skeleton) (P = 0.425), with or without antibiotherapy (P = 0.311), technical way of labelling leukocyte (isolated of granulation or mixed leukocytes) (P = 0.892), with or without colloid marrow imaging (P = 0.529). Conclusion. - Our results indicate that positive predictive value was good and different parameters did not significantly modify diagnostic performance of labeled-leukocyte scintigraphy. (authors)