Discussion

Main findings

This study provided a unique prospective assessment of accurate pregnancy and infant data with confirmed maternal vaccination status. The analytical framework treated pertussis vaccination as a time-varying exposure and computed time at risk windows for each of the time sensitive outcomes of interest and adjusted for a comprehensive set of confounding factors to reaffirm that maternal pertussis vaccination is safe for both the mothers and their newborn infants.

Strengths and Limitations

The major strength of this study is the prospective cohort design that recruited a large number of nulliparous women with singleton pregnancies at low risk for obstetric complications at two major maternity hospitals. Vaccinated women in our study were more likely to engage in healthy lifestyles i.e. pregnancy micronutrient supplementation, regular exercise, non-smoking than unvaccinated women. The analysis framework used herein adjusted for putative risk factors, including psychosocial factors, to mitigate the impact of any ‘healthy vaccinee bias’ on our findings. However, we cannot rule out the possibility of residual confounding. Our use of Cox proportional-hazards models accounting for time-varying vaccine exposure during pregnancy minimized the introduction of immortal time bias in our data.25 The potential for this bias arises because the opportunity for vaccination increases the longer a woman remains pregnant.28Furthermore, many studies use the earliest recommended maternal pertussis vaccination time (i.e. 28 weeks’ gestation in this case) as a cut-off point to restrict their data but pregnancies must survive within the recommended timeframe (i.e. 28-32 weeks’ gestation) to be eligible to receive the vaccines. Hence, immortal time bias may also be present in studies of maternal pertussis vaccination evaluating adverse pregnancy outcomes that develop in mid-to-late pregnancy, including PE and GH, where the bias may attenuate the true relative risk. Our analytic approach used 32 instead of 28 weeks’ gestation as the cut-off point in order to allow all women to have had the chance to receive the recommended pertussis vaccination in that optimal 4-week window. This may have reduced the introduction of immortal time bias in our analyses. Additionally, including follow-up time during which pregnancies are no longer at risk of some adverse outcomes (e.g. gestation after 37 weeks considered for PTB outcomes) can lead to incorrect estimation of the effect of maternal vaccination on the outcome of interest but our time-to-event analytic approach minimized the risk of these biases occurring. Another major strength of our study is confirmed maternal vaccination status. As this study was a secondary analysis, we did not conduct an a priori power analysis to show sample size adequacy. A post-hoc power analysis was not conducted as there is a rich literature, in both medical and statistical journals, warning against post-hoc power calculations.29, 30 However, the cohort was originally powered on the basis of 25% of women affected by pregnancy complications (preeclampsia, small for gestational age birth, spontaneous preterm birth and gestational diabetes mellitus).19

Interpretation

Our Cox proportional hazards models accounting for the time-dependent nature of exposure to vaccination during pregnancy, thereby avoiding the introduction of immortal time bias to our analyses, found no association between maternal pertussis vaccination and PTB, spontaneous PTB, nor PPROM, reaffirming the conclusion from previous systematic reviews.11-14
In keeping with previous studies11, 12, our findings demonstrated no association between maternal pertussis vaccination and gestational hypertensive disorders. The new Australian guidelines recommend pregnant women to receive a pertussis vaccine from 20 weeks of gestation rather than 28 weeks’ to maximize the opportunity for vaccination to protect all infants, including preterm infants.31 Administration of maternal pertussis vaccination from 20 weeks’ can align with other key routine antenatal visits such as morphology scanning and gestational diabetes testing, potentially improving the uptake of pertussis vaccination among pregnant women. However, there is a need for continued surveillance and monitoring to confirm that a broader window for pertussis vaccination during pregnancy is safe for the pregnant mother and the newborn.
Our study demonstrates that maternal pertussis vaccination was not associated with chorioamnionitis. In contrast, three large retrospective studies15-17conducted in the USA showed receipt of pertussis vaccination during pregnancy was associated with a small but significant increase in risk of developing chorioamnionitis. These studies15-17 used commercial health data and ICD codes for identifying chorioamnionitis from electronic medical records with no clinical case definition. Coding for commercial reasons, such as insurance claims, is potentially subject to favoring more severe diagnoses and might be prone to poor external validity, selection bias, confounding and misclassification bias.32 Furthermore, these studies did not find an association with an increased risk of PTB, which is an expected major clinical sequel of chorioamnionitis and most women with chorioamnionitis had at least one pre-existing risk factor for this complication. This suggests the observed relation between receipt of pertussis vaccination during pregnancy and chorioamnionitis was unlikely to be casual, and is probably more reflective of residual confounding affecting the results in these studies.
In our study, we found an association between maternal pertussis vaccination during pregnancy and longer gestation. There are no known biologically plausible direct effects of pertussis vaccination on pregnancy duration but women who remained pregnant longer have more opportunity to have received pertussis vaccine in the late third trimester of pregnancy. This may have created a spurious relationship between pregnancy duration and time-varying pertussis vaccine exposure during pregnancy because linear regression analysis is not suited to include both the event and time aspects in the model.
Our study provides further assurance that pertussis vaccination during pregnancy is not associated with any adverse birth outcomes including LBW or SGA births, consistent with previous findings.11-14 Receipt of pertussis vaccination during pregnancy was not associated with increased risk of perinatal outcomes including admission to the neonatal care unit, respiratory distress, Apgar scores <7 nor need for mechanical ventilation at birth compared with infants born to unvaccinated women.11-14