Data & Analysis Code
Synthetic controls for evaluations of vaccine impact
We have been using a method called "synthetic controls" to adjust for unmeasured bias and confounding in time trend analyses of the impact of vaccines. This method has previously been applied to website analytics and various political science and economics analyses. You can find sample datasets and R code here. We also developed an online "point and click" interface that you can find here.
Time series simulator for power analysis of vaccine evaluation studies
Try out this program that we designed, which can be used to evaluate the likelihood of detecting the “true” change associated with vaccination given different numbers of years before and after vaccine introduction, different amounts of noise (between observations or between years), and different levels of expected vaccine impact. The program was built in R and is run using RShiny. This is part of our work from a grant funded by the Bill and Melinda Gates Foundation on PCV impact evaluations.