Methodological issues arising from systematic reviews of the evidence of safety of vaccines.

Adaptations to the recognized methods of systematic reviewing are required when addressing questions about safety, particularly about rare and/or long-term serious adverse events. Conducting a systematic review of vaccine safety requires the implementation of novel strategies for locating studies, the use of experimental instruments to assess the quality of non-randomized studies, and the employment of pooling methods for non-randomized data, where appropriate. Standardizing both the indexing of adverse event data on electronic libraries and their reporting would improve the potential of systematic reviews of vaccine to draw accurate conclusions about the safety of a vaccine.