Major outbreak of pertussis in northern Alberta, Canada: analysis of discrepant direct fluorescent-antibody and culture results by using polymerase chain reaction methodology.

A major outbreak of 5,683 cases of pertussis occurred in northern Alberta, Canada, from December 1989 to January 1991. The outbreak highlighted a number of problems with current methods of pertussis diagnosis. In particular, an exceptionally high proportion of direct fluorescent-antibody (DFA)-positive, culture-negative specimens (88.4%) was identified. We took this opportunity to use polymerase chain reaction (PCR) methodology to examine whether the low culture rates were due to specimens containing dead organisms or whether the DFA results represented high numbers of false-positive results. A set of primer sequences within a Bordetella pertussis-specific repetitive element was used to amplify proteinase K extracts of B. pertussis DNA recovered from 279 submitted slides inoculated at the point of collection with nasopharyngeal material obtained from pernasal swabs. The PCR data corroborated the culture results: 84.6% of DFA-positive, culture-negative specimens were similarly PCR negative. At least three different bacterial species that were significantly cross-reactive with the commercial DFA reagent were identified in clinical specimens and in pure culture, providing one possible explanation for the false-positive DFA results. These results and other limitations of current diagnostic techniques underline the urgent need for a new DFA reagent with improved specificity and a standardized means of measuring the patient antibody response for the diagnosis of pertussis.