This article presents a letter to the editor in response to "Analysis of observational studies in the presence of treatment selection bias: effects of invasive cardiac management on AMI survival using propensity score and instrumental variable methods," by T.A. Stukel, E.S. Fisher and colleagues.