Safety and Effectiveness of Vaccines in Pregnancy
MOTT-DISSERTATION-2020.pdf (3.252Mb)(embargoed until: 2021-05-01)
MetadataShow full item record
CitationMott, Katrina. 2020. Safety and Effectiveness of Vaccines in Pregnancy. Doctoral dissertation, Harvard T.H. Chan School of Public Health.
AbstractOver the last 15 years, changes in the recommendations for vaccination in the U.S. have increased the prevalence of exposure during pregnancy, despite limited clinical trial data examining their safety in pregnant women.
In this dissertation, we first investigated the effectiveness of providing Tdap during pregnancy on the prevention of infant pertussis infection using data from the Medicaid Analytic eXtract (MAX) and IBM MarketScan claims databases. Inverse probability of treatment-weighted Cox proportional hazards models were used to calculate hazards ratios. We also stratified the models to investigate the protection provided by this vaccine for preterm infants specifically. We found that Tdap vaccination in pregnancy provides a 36% (95% CI 0%, 59%) reduction in hazard, and a much stronger protection (pooled HR = 0.10) for preterm infants.
Amid new concerns of the safety of influenza vaccination in pregnant women, in the second chapter, we investigated the association of influenza vaccination in early pregnancy with spontaneous abortion using a case-time-control study design. This study was conducted in the MAX database, with livebirth pregnancies frequency-matched to spontaneous abortions on calendar month and year of last menstrual period (LMP) to control for seasonality of vaccination. We found no evidence of an increased association between influenza vaccine and spontaneous abortion (OR=0.93, 95% CI (0.87, 0.99)) across 10 influenza seasons.
In the third chapter, we conducted a simulation to estimate the potential bias when conducting studies of spontaneous abortion and vaccines or other point exposures due to missing information on gestational age at spontaneous abortion from claims databases. We found that there is minimal bias when the trend of exposure remains relatively constant during the gestational period of interest, but bias increases with stronger trends in exposure. In all scenarios we simulated, using a distribution of gestational ages to match to livebirths at the same gestational time leads to decreases bias compared to using a single average gestational age.
Our first two studies support the current CDC recommendations for Tdap and influenza vaccination during pregnancy. Our findings show the importance of careful study design when investigating benefits and risks of exposure in pregnant women.
Citable link to this pagehttp://nrs.harvard.edu/urn-3:HUL.InstRepos:42676014