Estimating vaccine effectiveness against laboratory-confirmed influenza using a sentinel physician network: results from the 2005-2006 season of dual A and B vaccine mismatch in Canada

Abstract

Introduction: We report a case-control design using a sentinel physician network to estimate vaccine effectiveness (VE) against laboratory-confirmed, medically attended influenza (LC-MAI) and provide results for the 2005-2006 season of dual A and B vaccine mismatch in Canada.

Methods: Participants were patients >or=5 years of age presenting with influenza-like illness (ILI) to a sentinel physician in British Columbia, Canada between November 1, 2005 and April 30, 2006. Cases were participants in whom influenza was identified; controls tested negative for influenza A and B by PCR, R-mix and culture. Isolates were characterized by gene-sequencing and hemagglutination-inhibition (HI) assays. Odds ratios (OR) for LC-MAI in vaccinated versus non-vaccinated persons were derived with adjustment for age and chronic conditions. VE was estimated as [1-OR (vaccinated/unvaccinated)].

Results: The sample included 442 patient visits: median age was 26 years, 10% were >or=65 years, 15% had a chronic condition and 22% received the 2005-2006 trivalent inactivated influenza vaccine >or=2 weeks before ILI onset. Two hundred and six participants were positive for influenza; 107 (52%) had influenza A/H3N2 and 99 (48%) had influenza B/Victoria lineage. Gene sequencing identified mutations away from the vaccine strain at key antigenic binding sites of the hemagglutinin (HA) protein of H3N2 isolates; the neuraminidase (NA) protein was conserved. Based on HI assays, three-quarters of influenza A and all B isolates were mismatched to the 2005-2006 vaccine. Point estimates for VE against LC-MAI were in the range of 50 to 70% for both types of influenza.

Conclusion: 2005-2006 was the third consecutive season of vaccine mismatch based on varying HA for the A/H3N2 component and the third also for the B component since 2001. Vaccine mismatch resulted in diminished VE but substantial cross-protection. More timely detection of drift variants through gene sequencing of isolates facilitates interpretation of VE results. Since it may be more antigenically conserved, the vaccine content and contribution of NA to overall VE should be further evaluated for both A and B components. Infrastructure for real-time epidemiologic assessment of vaccine performance is important annually and in preparation for a pandemic.