In September 2017, James Donahue and coworkers from the Vaccine Safety DataLink reported the results of a study examining the risk of spontaneous abortion following receipt of inactivated influenza vaccine during pregnancy (Donahue JG, Kieke BA, King JP, et al. Association of spontaneous abortion with receipt of inactivated influenza vaccine containing H1N1pdm09 in 2010-2011 and 2011-2012, Vaccine. 2017 Sep 25;35(40):5314-5322.)
The authors found that the incidence of spontaneous abortion in the vaccinated group was indistinguishable from the unvaccinated group in the second and third trimesters. However, a statistically significant increase in spontaneous abortions was observed in the vaccinated group in the first trimester. For many reasons, this result appears to be anomalous:
- The finding was statistically significant for the 2010-2011 season but not the 2011-2012 season.
- The finding was statistically significant only when including women who had been immunized the previous season with influenza vaccine. As the authors state, this was not their hypothesis going into the study, suggesting that only fine-tuning and substratifying of the data retrospectively created this difference.
- Six previous studies failed to show that first-trimester immunization with influenza vaccine increased the risk of spontaneous abortion.
- The finding was confined to only a handful of patients. Fourteen patients in the vaccine group suffered spontaneous abortion as compared with four patients in the unvaccinated group. While statistically significant, these numbers are small, and therefore more likely to be misleading. Epidemiologists refer to this as the curse of small numbers.
- The finding doesn’t make biological sense. While it is true that natural influenza infection during the first trimester of pregnancy increases the risk of spontaneous abortion, events following immunization are quite different than natural infection. The inactivated influenza vaccine, by definition, doesn’t replicate. So it can’t drive the levels of inflammatory responses seen following natural infection. Indeed, the authors state, “the biological basis for our observation has not been established.”
Because of the limitations of this study, the authors concluded, “This study does not and cannot establish a causal relationship between repeated influenza vaccination and SAB [spontaneous abortion].” Unfortunately, the media as well as several articles on the internet referred to this finding as a “signal,” scaring parents unnecessarily.