A population-based analysis of predictors of influenza vaccination uptake in pregnant women: The effect of gestational and calendar time

Pregnant women are vaccinated against influenza less frequently than other high-risk groups. To design effective vaccination strategies, we must understand how decisions regarding vaccination may vary by trimester and over vaccination campaigns. We used a Cox model indexed by calendar time to estimate the effect of gestational trimester and other factors on vaccination uptake in a large cohort of pregnant women in Catalonia (Spain) during 2008-09 to 2012-13 influenza vaccination campaigns. We analyzed 247,316 pregnancies. Vaccination coverage was 3.7%, 5.2%, 4.8%, 5.6% and 4.6% from 2008-09 to 2012-13 seasonal vaccination campaigns and 8.3% for the 2009 pandemic vaccination campaign. Pregnant women previously vaccinated had higher uptake than women not previously vaccinated and the hazard ratios (HRs) comparing these 2 groups decreased from 10, the first day of seasonal campaigns, to 1.3 the last day. During the pandemic campaign, HRs decreased over the course of the campaign from 8.6 to 1.9. Women in second and third trimester had higher uptake than women in first trimester, with HR = 2.8 and 2.3, respectively, at the start of seasonal campaigns. Influenza vaccination coverage among this cohort of pregnant women was alarmingly low. Our analysis reveals that gestational and calendar time have distinct and interacting effects on vaccination uptake; women in their second trimester and third trimester and previously vaccinated were more prone to be vaccinated, but this effect wanes as the influenza season progresses. (C) 2017 Published by Elsevier Inc.