What did we do?
We used a natural experiment in Wales where shingles vaccine (Zostavax) eligibility was based on an birthdate cutoff, separating eligible and ineligible cohorts. This gave us a great RDD setting to causally estimate its effect on dementia risk.