Lack of association between hepatitis B birth immunization and neonatal death: a population-based study from the vaccine safety datalink project

Abstract

Background: There have been no population-based studies of the potential association between neonatal death and newborn immunization with hepatitis B vaccine (HBV).

Methods: As part of the Vaccine Safety Datalink Project, we defined a birth cohort at Southern and Northern California Kaiser Permanente Health Plans of more than 350,000 live births from 1993 to 1998 and ascertained all deaths occurring under 29 days of age. We compared the proportions of deaths among birth HBV-vaccinated and unvaccinated newborns and reviewed the causes and circumstances of their deaths. We performed detailed clinical reviews of all HBV-vaccinated neonates who died and a sample of unvaccinated neonates who died and who were matched to vaccinated deaths for days of life, sex, birth year and site of care. To avoid confounding, we categorized the causes of death as either "expected" or "unexpected" and performed a stratified analysis to compare mortality with immunization status.

Results: There were 1363 neonatal deaths during the study period. Whereas 67% of the entire birth cohort received HBV at birth, only 72 (5%) of the neonates who died were HBV-vaccinated at birth (P < 0.01). We found no significant difference in the proportion of HBV-vaccinated (31%) and unvaccinated (35%) neonates dying of unexpected causes (P = 0.6). Further we could not identify a plausible causal or temporal relationship between HBV administration and death for the 22 vaccinated neonates who died unexpectedly.

Conclusions: A relationship between HBV and neonatal death was not identified.