When the hypothesis of a link between vaccination and a possible adverse outcome arises, further investigation is required to confirm or refute the suspicion. Given the rarity of most serious adverse effects, a case-control approach will often be chosen. This paper discusses aspects of the design, analysis and interpretation of case-control studies to evaluate vaccine adverse effects. Potential biases (and how to minimise such biases) in the selection of cases and assessment of vaccine exposure and the potential for confounding are discussed. Finally the increasing use of electronic databases in the evaluation of vaccine adverse effects is considered.
