Quantitative signal detection for vaccines: effects of stratification, background and masking on GlaxoSmithKline's spontaneous reports database.

Disproportionality analyses are increasingly gaining acceptance as signal detection tools for drug adverse events. Their application to vaccine adverse events has not been well evaluated. Disproportionality analyses based on the Multi-Item Gamma Poisson Shrinkage principle (MGPS) were applied to spontaneous adverse events reports for five vaccines with different reporting profiles. Sensitivity analyses were used to assess the potential impact of changing key parameters. We used the Company's in-house spontaneous adverse events database. We evaluated the impact of stratification, the comparator dataset and the potential for masking. We conducted a semi-quantitative assessment by comparing the changes in the disproportionality scores and the number of vaccine-event pairs that exceeded an arbitrary threshold as a measure of the impact of any of these choices. The results show that stratification by age and region has a significant impact. The effect of the comparator dataset was dependent on the vaccine being evaluated. The potential for a masking effect was only weakly noticeable. In conclusion, we opt to start with a conservative approach in which we will supplement our primary stratification against the vaccine database with unstratified analyses as well as analyses against the entire database. We recommend that similar studies be performed before introducing disproportionality analyses to a new vaccine adverse events database.