Incidence and factors predicting whooping cough due to parapertussis diagnosis among patients referred to general practitioners, Poland, 2009-2011.

Parapertussis leads to similar symptoms as pertussis, both being caused by bacteria from the genus Bordetella. Poland does not routinely diagnose nor conduct surveillance for parapertussis. We estimated parapertussis incidence and determined predictors of parapertussis diagnosis in the Polish population. Between July 2009 and April 2011, we conducted a prospective cohort study among patients attending 78 general practices. We included patients aged ≥ 3 years, with cough lasting >2 weeks, interviewed patients and collected a nasopharyngeal swab. We confirmed cases by real-time PCR. We estimated parapertussis incidence rates by dividing the number of cases by the summed person-time of observation in respective practices. We assessed predictors of PCR-confirmed parapertussis by comparing cases with patients testing negative. Using logistic regression, we calculated odds ratios (ORs) and 95% confidence intervals (95%CI). We identified 78 cases among 1,231 patients meeting inclusion criteria. The incidence rate was 39/100,000 person-years (95%CI 31-49). The highest rates (140/100,000; 95%CI 74-239), were among children 3-5 years of age and the lowest (24/100,000; 95%CI 13-40) among persons aged 20-39 years of age. Boys aged 3-5 years (7.1; 2.1-25.3) and women aged >40 years (4.1; 1.4-11.7) or living in crowded households (4.3; 1.4-12.9) or contacting persons with prolonged cough (2.3; 1.1-4.5) were more likely to be diagnosed. Our results suggest that laboratory diagnosis could be prioritized for children in the preschool age and women aged over 40 who were referred to their GP with prolonged cough. In the absence of vaccine, post-exposure prophylaxis for close contacts of parapertussis cases could an adequate preventative measure.