Prosthetic graft infection: limitations of indium white blood cell scanning.

The lack of a rapid, noninvasive, and accurate method to confirm or rule out prosthetic graft infection continues to constitute a compelling and vexing clinical problem. A host of adjunctive diagnostic techniques has been used in the past, but early promising results subsequently have usually not yielded acceptable sensitivity (reflecting false negatives) and specificity (reflecting false positive) data. White blood cell (WBC) indium 111 scanning has recently been added to this list. The utility and accuracy of 111In WBC scans were assessed by retrospective review of WBC scan results in 70 patients undergoing evaluation for possible prosthetic graft infection over a 7-year period. Operative and autopsy data (mean follow-up, 18 months for survivors with negative scans) were used to confirm the 22 positive, 45 negative, and three equivocal WBC scans. The false positive rate (+/- 70% confidence limits) was 36% +/- 6% (n = 8) among the 22 patients with positive scans (44% +/- 6% [11 of 25] if the three equivocal scans are included as false positive), yielding a specificity of 85% +/- 5% and an overall accuracy rate of 88% +/- 4% (80% +/- 5% and 84% +/- 5%, respectively, if the three equivocal cases are considered as false positive). All three patients with equivocal scans ultimately were judged not to have prosthetic graft infection. As implied by the high accuracy rate, the sensitivity of the test was absolute (100% [14 of 14]); there were no false negative results. (ABSTRACT TRUNCATED AT 250 WORDS)