Effectiveness of a third BNT162b2 mRNA COVID-19 vaccination during pregnancy: a national observational study in Israel

Our study showed that a third dose of the BNT162b2 mRNA COVID-19 vaccine, given at least 5 months after the second vaccine dose, provides additional protection during pregnancy against hospitalizations with a diagnosis of SARS-CoV-2 infection, and against significant and severe disease. As described above, significant disease was defined by documented hospitalization with moderate COVID-19-related disease, or worse, i.e. COVID-19 related pneumonia justifying hospitalization. Severe disease was defined as a resting respiratory rate >30 breaths per minute, oxygen saturation on room air <94%, or as the need for mechanical ventilation and clinical severe organ failure.

The mRNA vaccines currently available were designed to prevent infection and disease from the wild type SARS-CoV-2 strains. Data from non-pregnant populations demonstrate that the effectiveness of the second vaccine dose declines over time as the humoral immunity wanes and new variants emerge^15,16,17,18. Our data concur with these reports. Previous studies reported 98% effectiveness of the second dose against hospitalization, shortly after vaccination^3,19. We detected reduced effectiveness more than 5 months after the second vaccine dose (61% during the Delta period and none in the Omicron period), findings that might support waning of immunity. In this context, the third dose provided additional protection during the Delta and Omicron periods (97% and 43% protection, respectively) when compared to unvaccinated patients, emphasizing the benefit of vaccine boosting.

We previously showed that a third dose of BNT162b2 mRNA vaccine significantly increased anti-SARS-CoV-2 antibody titers in maternal and cord blood²⁰. In addition, a recent study found that a third booster dose was essential in building neutralizing antibody capacity against the Omicron variant among mothers and neonates²¹. These boosted antibody titers may have provided additional protection from the Delta variant and allowed protection from the Omicron variant.

When focusing on substantial COVID-19 illness, 5 months after the second dose, the second dose effectively protected against hospitalization complicated by significant disease (97%) and severe disease (96%) during the Delta period, but not during the Omicron period. The impact of the third boosting dose was substantial during the Omicron period, effectively protecting against hospitalization complicated by significant disease (97%) and severe disease (94%). To the best of our knowledge, our results present data regarding considerable vaccine effectiveness against severe COVID-19 disease during pregnancy. The fact that vaccines during pregnancy nearly abolish the risk for significant disease has been shown to play a role in patient decision-making regarding vaccination²². Hence, our study might contribute to promoting vaccination uptake among pregnant women.

In the present study, we focused on the impact of COVID-19 vaccine strategy on hospitalization with a diagnosis of SARS-CoV-2 infection, rather than population infection rates. Recorded infection rates may be biased by differential rates of testing in various population subgroups, most notably among unvaccinated patients. Indeed, our data show that unvaccinated pregnant women were considerably less likely to be tested (Table 1). However, while not uniformly executed in all maternity units, routine SARS-CoV-2 testing during maternity admissions was mandatory in most hospitals in Israel. Given the unbiased approach to testing, a finding of positive SARS-CoV-2 during hospitalization represents a better sensor for infection burden, and we therefore assessed and analyzed the data accordingly.

Most previous reports analyzed pregnancy data from a single COVID-19 wave, narrowing observations. We analyzed data from two discrete periods, when two variants having different characteristics were dominant. We focused the time margins on the periods dominated by the Delta and Omicron variants, to present a more comprehensive understanding of vaccine and boosting effectiveness on different viral variants. Indeed, we found substantial differences between the two time periods, which might reflect differences in virulence, ability to evade vaccine-mediated immune protection, and waning of protective titers over time.

The CDC and other health organizations now recommend COVID-19 vaccination for pregnant women to reduce the risks of severe disease and complications^11,23. Similar to the general population, these recommendations include boosting of pregnant women with a third vaccine dose, 5 months following the second vaccine dose. Our findings provide insight into the impact of COVID-19 vaccines during pregnancy and the advantage conferred by the third, boosting dose against serious illness, and serve to reinforce recommendations to vaccinate and boost this population, providing clinicians and policymakers with essential evidence to inform consultation and decision-making.

Research into the long-term consequences of COVID-19 in pregnancy for mothers and newborns is still scarce. Initial, preliminary reports of the long-term effects of maternal COVID-19 infection during pregnancy have suggested worrying adverse neurodevelopmental sequelae^24,25,26. These cases highlight the urgent need for data on measures to limit maternal infection, which may have as yet unknown adverse consequences.

Our study has several limitations. Since the study is based on real-world collected data, no randomization was possible. Individuals opting to decline boosting doses or to refuse vaccination or evade testing may differ in risk-taking behavior, demographic or obstetric characteristics from those opting in. Moreover, vaccinated groups may behave in a more cautious manner that could reduce the chances of infection regardless of vaccination. In addition, natural and hybrid immunity acquired in patients over the course of the pandemic, which may not have been captured by testing in the community, may have had differential protective effects in unvaccinated and vaccinated groups. These are possible sources of bias that are difficult to adjust for in a study like ours, but must be stated and accounted for. We also recognize that other, unaccounted-for individual and group differences in risk factors for severe illness or the likelihood of exposure to the virus, may have impacted our results. Our findings were limited to the BNT162b2 vaccine. We cannot infer whether our observations are relevant to preventing reinfection in convalescent COVID-19 pregnant women, or populations administering other vaccines. The decision to vaccinate during pregnancy is a balance between benefit and effectiveness vs. safety. We did not evaluate COVID-19 vaccine safety, however several other large studies have done so^3,27, and demonstrated a favorable safety profile. Finally, we lack data on variant sequencing of the diagnostic tests used in this study. However, the inclusion of two time periods dominated by distinct variants strengthens our findings.

When compared to eligible non-boosted or unvaccinated pregnant women, those who received a third dose of BNT162b2 had a lower incidence of hospitalization with SARS-CoV-2 infection during the Delta period and considerably lower incidence of COVID-19 related outcomes during the Omicron period. Our data and analyses provide the necessary evidence to support current recommendations to vaccinate pregnant women with the third boosting dose of COVID-19 vaccine.