Utility of three-phase skeletal scintigraphy in suspected osteomyelitis: concise communication.

Three-phase skeletal scintigraphy, consisting of a radionuclide angiogram, an immediate postinjection "blood-pool" image, and 2--3 hr delayed images, was performed on 98 patients with suspected osteomyelitis. This procedure was evaluated by first interpreting only the delayed images, next the combination of "blood-pool" and delayed images, and finally the three-phase study. There was no change in the sensitivity (12/13 = 0.92) for detecting osteomyelitis, but the false-positive rate for osteomyelitis decreased from 0.25 (21/85) to 0.06 (5/85). In 21 of 64 patients (33%) with abnormal studies, the "blood-pool" image and/or the radionuclide angiogram led to a more accurate scintigraphic diagnosis. In 12 patients (19%) the "blood pool" alone was enough to achieve the correct final diagnosis and was used most often to identify noninfectious skeletal disease. In 9 patients (14%) the radionuclide angiogram was required for an accurate interpretation and was considered essential most often in cases of soft-tissue infection. Both radionuclide angiography and "blood-pool" imaging appear to augment the specificity of skeletal scintigraphy in patients with suspected osteomyelitis.