Radionuclide bone scintigraphy in the detection of significant complications after total knee joint replacement

with complications requiring active treatment. The aim of this study was to determine the usefulness of Tc-99(m)-MDP bone scintigraphy, METHOD: A retrospective study of all patients having a Tc-99(m)-MDP bone sc intigram for a painful knee arthroplasty between 1993 and 1999 was performe d. Bone scintigrams were classified as normal or abnormal by a single obser ver. The results of these investigations were correlated with clinical outc ome. RESULTS: Seventy-five patients with painful knee arthroplasties were referr ed for investigation, A total of 80 bone scintigrams were performed. The av erage patient age was 66.2 years (42 female and 33 male). The mean time per iod between surgery and onset of knee pain was 3 years. A final clinical di agnosis based on arthroscopy, open surgery, and extended clinical follow-up was available for all patients. Forty-three (53.8%) of the scintigrams wer e normal and 37 (46.3%) abnormal. Two patients with a normal bone scintigra m has loose prostheses, Thirteen patients with an abnormal study had normal prostheses on follow-up and these tended to be patients scanned less than a year after surgery. The sensitivity, specificity, positive predictive val ue and negative predictive value of an unequivocally normal or abnormal bon e scintigram was 92.3, 75.9, 64.9 and 95.0%, respectively. The pattern of i sotope uptake in the abnormal studies was not specific enough to reliably d ifferentiate aseptic from septic loosening. CONCLUSION: Radionuclide bone scintigraphy is useful in the assessment of t he painful knee arthroplasty, A negative bone scintigram is reassuring and makes loosening or infection unlikely. (C) 2001 The Royal College of Radiol ogists.