Reanalyses of case-control studies examining the temporal association between sudden infant death syndrome and vaccination.

ABSTRACT

In this paper we examine different time periods after vaccinations and investigate whether the risk of sudden infant death is different during the post-vaccination period than at other times. Three already published case-control studies are re-examined in this context. Several evaluation approaches are presented. The recently developed self-controled case series (SCCS) method for terminal events, which only takes the cases into account, is used in addition. There is no increased or reduced risk of sudden infant death during the period after the vaccination. The previously reported protective effect seen in case contol studies is based on the inclusion of unvaccinated cases. The results of the case-control analysis of one study is affected by two confounders. The SCCS method for terminal events, in which all time-independent confounders are eliminated, is an alternative to case-control analysis when it comes to the temporal association between exposed time periods and SIDS after vaccination.