Comparison of culture and PCR for detection of Bordetella pertussis and Bordetella parapertussis under routine laboratory conditions

A PCR assay for the detection of Bordetella pertussis and Bordetella parapertussis was compared with the conventional culture method under routine laboratory conditions. Detection of B. pertussis was based on the amplification of a section of the IS481 insertion sequence and confirmation of positive results was based on a sequence of the pertussis toxin promoter region. Detection of B. parapertussis was based on the amplification of a section of the IS1001 insertion sequence. An internal control was included. Data were available for the period 28 November 2000 to 9 July 2003. In this period, 3096 patients were examined for infection with B. pertussis and B. parapertussis by culture and PCR on the same day. B. pertussis was found in 496 (16 %) patients; 208 (42 %) were diagnosed by PCR alone whereas 17 (3 %) were diagnosed by culture alone. B. parapertussis was found in 64 (2 %) patients. The sensitivity of the PCR was 97 % and of culture 58 %. The specificity of PCR was 93 % when regarding culture as 100 % sensitive. There was a significant relationship between laboratory method and age, as the superiority of PCR was most marked in the age group 0.5-3 years. The PCR assay proved highly sensitive for the diagnosis of pertussis. The specificity estimate of the PCR assay suffers from the influence of a gold-standard method with a low sensitivity. The PCR assay is considered highly specific due to the amplification of two different sequences in two separate assays.