Littenberg B, Mushlin A I
Dartmouth-Hitchcock Medical Center, Hanover, New Hampshire.
J Gen Intern Med. 1992 Mar-Apr;7(2):158-64. doi: 10.1007/BF02598005.
To determine the diagnostic performance of technetium bone scanning in the setting of possible osteomyelitis in the foot of a patient who has diabetes or other vasculopathy.
Meta-analysis. Data identification and study selection: To be eligible for inclusion, a report must have used intravenous technetium-99m methylene diphosphonate or a similar agent in humans over the age of 16 years, must have addressed possible osteomyelitis of the lower extremity with ulcer or soft-tissue inflammation in the setting of diabetes, neuropathy, or vasculopathy, and must have allowed the generation of a two-by-two table. A structured search of the MEDLARS database found 296 possibly eligible reports; ten met all the inclusion criteria.
The reported sensitivity and specificity of each report were converted to their logistic transforms and a straight line was fitted by weighted least-squares regression. The line was then back-transformed to yield a summary receiver operating characteristic curve. The false-positive rate of the bone scan is at best in the range of 10 to 20%. This occurs at sensitivities between 70 and 80%. The studies with increased sensitivity also reported sizable increases in the false-positive rate ranging from 20 to over 90%. Even small increases in sensitivity have necessitated large sacrifices in specificity. Seven of the ten studies reported specificities under 70%.
Published data defining the effectiveness of technetium bone scanning for the diagnosis of osteomyelitis in the impaired foot indicate relatively poor performance. In many clinical situations, the specificity of the bone scan will not be high enough to confirm the diagnosis of osteomyelitis.
