Reproducibility of interpretation of the test for antibody-coated bacteria in urinary sediment

The direct immunofluorescence technique for detecting antibody-coated bacteria in urinary sediment is felt to be useful in distinguishing infection of the kidney from infection of the bladder. An independent, blind multiple-reading system was used to measure interobserver variability in the evaluation of slides of urinary sediments for antibody-coated bacteria. Three independent observers agreed unanimously on first reading in 88% of 253 specimens. When compared with the majority opinion, the sensitivity and specificity of an individual reading were 91 and 95%, respectively.