[Influence of distinct criteria for selecting patients for swabbing on estimation of the effectiveness of the influenza vaccine].

OBJECTIVE

To estimate the effectiveness of the influenza vaccine under different criteria for selecting patients for swabbing.

METHOD: A case-control study was performed of laboratory-confirmed cases (n=909) and negative controls for influenza (n=732) in the 2010-2011 to 2012-2013 seasons in Navarre (Spain). The adjusted vaccine effectiveness was estimated by including all swabs from patients with influenza-like-illness and selecting only the first two cases per physician and week.

RESULTS

The first two patients per physician and week were less frequently vaccinated against influenza (7.9% vs. 12.5%, p=0.021) and less often received confirmation of influenza (53.6% vs. 66.4%, p <0.001) than subsequent patients. These differences decreased after adjustment for covariates. The effectiveness of the influenza vaccine was 49% (95% CI: 23-66%) when all swabs were included and was 55% (95% CI: 27-72%) when we selected the first two swabs per week and physician.

CONCLUSION

The selection of the first two patients per physician and week may bias assessment of the effectiveness of the influenza vaccine, although this bias was small in the seasons analyzed.