The fundamental goal of policy analysis is to understand the effect of an intervention on patients, health care professionals, or hospitals. Although a randomized controlled design is considered the criterion standard for isolating a causal effect, such a design is rarely plausible in health policy evaluation. As such, researchers often rely on observational data to evaluate a policy’s consequences.¹ For example, recent analyses of the association of the Patient Protection and Affordable Care Act (ACA) with outcomes for surgical patients have shown important gains in insurance coverage, timely access to care, and increased rehabilitation facility access after discharge.^2-4 While analyses of observational data alone cannot determine causality, valuable evidence regarding their impact can be drawn when applying best analytic practices from across academic disciplines.¹