Evaluation and validation of a monoclonal immunofluorescent reagent for direct detection of Bordetella pertussis

Abstract

An outbreak of pertussis in Manitoba, Canada, provided an opportunity to evaluate the recently developed monoclonal antibody (MAb) BL-5 for the direct detection of Bordetella pertussis. The MAb recognizes a lipooligosaccharide epitope. A total of 1,507 consecutive nasopharyngeal swabs for culture and companion smears for direct fluorescent-antibody (DFA) detection were evaluated at Cadham Provincial Laboratory between September and November 1994. The cutoff for DFA positivity was four fluorescing organisms with morphology characteristic of B. pertussis. PCR analysis for B. pertussis DNA was performed on a subset of 100 smears by eluting material from the slides after DFA examination. In comparison with culture, the sensitivity, specificity, and positive and negative predictive values of BL-5 were 65.1% (41 of 63 samples), 99.6% (1,438 of 1,444 samples), 87.2% (41 of 47 samples), and 98.5% (1,438 of 1,460 samples), respectively. The sensitivity of culture compared with PCR was 45.5% (10 of 22 samples) for the subset of 100 specimens tested by both procedures. An expanded "gold standard" of positivity by culture or PCR for these 100 specimens resulted in DFA sensitivity, specificity, and positive and negative predictive values of 32.3, 97.1, 83.3, and 76.1%, respectively. The utility of MAb BL-5 for direct detection of B. pertussis in a clinical laboratory setting has been demonstrated by this investigation.