The effect of changing from whole-cell to acellular pertussis vaccine on the epidemiology of hospitalized children with pertussis in Canada.

BACKGROUND Between July 1997 and April 1998, universal childhood immunization programs in Canada changed from using a whole-cell pertussis to a 5-component acellular pertussis-containing vaccine. To assess effects on pertussis epidemiology of this nationwide change, we analyzed hospitalizations during 1991-2004 using the Canadian Immunization Monitoring Program, Active (IMPACT) pertussis database. METHODS IMPACT is an active surveillance network based in 12 pediatric tertiary-care hospitals across Canada. Characteristics of hospitalized cases of pertussis were compared by type of vaccine received or by birth date (if immunization records were unavailable or the child was unvaccinated). Age-stratified incidence rates were calculated by year and vaccine type. RESULTS Two thousand ninety-six cases of pertussis were admitted to IMPACT centers, 1174 during the whole-cell vaccine program (WCV-P) and 842 during the acellular vaccine program (ACV-P). Pertussis incidence among children <5 years old decreased significantly during the ACV-P, causing an increase in the residual proportion of cases either too young to be immunized (<2 months old: ACV-P 39% versus WCV-P 26.1%; P < 0.0001) or too young for a second dose (2-3 months old: 42.9% versus 34.2%, respectively; P < 0.0001). A significantly smaller proportion of cases (ACV-P 15.1% versus WCV-P 27.3%) occurred in infants who were old enough (4-11 months of age) to have received 2 or 3 doses of vaccine. CONCLUSIONS With ACV-P, pertussis hospitalizations in children 4-59 months old decreased in frequency, consistent with improved vaccine effectiveness, but remained prominent among very young infants. Improved control strategies are needed to reduce infections among infants too young for pertussis vaccination.