What pitfalls should you watch for when running experiments in a hospital setting, such as interference, novelty effects, or sample ratio mismatch?
