H1N1 vaccination during pregnancy not linked to congenital malformation

In a paper in Annals of Internal Medicine, Jonas Ludvigsson and his colleagues present data on congenital malformations in more than 40 000 offspring to women undergoing vaccination against H1N1 (”the swine flu”). It is already known that the health consequences of influenza are more severe in pregnant women than in non-pregnant women, and that influenza during pregnancy increases the risk of fetal death. Other data have also demonstrated that vaccination against H1N1 influenza protects against influenza disease in pregnant women.
Hence, there are strong arguments to vaccinate pregnant women. However, up until now it has not been clear if H1N1 vaccination especially during early pregnancy can cause congenital malformation.
Today’s paper shows that H1N1 vaccination against influenza does not influence the overall risk of congenital malformation (5.0% of vaccinated offspring and 5.0% of unvaccinated offspring had a congenital malformation). We performed sub analyses of specific malformations (heart malformation, oral cleft, limb deficiency) and while we found no excess risks of congenital malformation in these studies, we had limited power to rule out small risk increases.
The paper on H1N1 vaccination and malformation stands out from earlier research in two ways:

1. It had sufficient statistical power to examine adverse effects from vaccination during pregnancy but vaccination during the first trimester and even in the first 8 weeks. Only in very early pregnancy is malformation likely since most of the organ development takes place early in pregnancy (hence earlier studies looking at vaccination ”at any stage of pregnancy” could have missed a true risk increase).
2. Besides the current paper compared the risk of congenital malformations in siblings to the same mother, child 1 born during a pregnancy when the mother was vaccinated, child 2 born during a pregnancy when the mother was not vaccinated. The researchers found no difference in malformation rates.

This latter comparison is important since it takes several non-measurable factors into account (children to the same mother served as controls), thereby reducing potential confounding.

A Swedish summary of the paper is also presented in ”DN Debatt”