How to Avoid Mortal Time Bias in an Observational Study

An effective way to avoid both immortal time bias and selection bias is to kind of imagine that you were designing a randomized controlled trial. So, patient shows up in the emergency department, we screen them for eligibility, and then we randomize them to drug A or drug B,. And that all happens on the same day. It's called the target trial approach where your first step in designing your study is to design kind of your ideal target trials.