Causation and prediction in epidemiology: a guide to the "methodological revolution"

Abstract

There is an ongoing "methodological revolution" in epidemiology, according to some commentators. The revolution is prompted by the development of a conceptual framework for thinking about causation here referred to as the Potential Outcomes Approach (POA), and the mathematical apparatus of directed acyclic graphs that accompanies it. But over and above the mathematics, a number of striking theses about causation are evident, for example: that a cause is something that makes a difference; that a cause is something that humans can intervene on; and that causal knowledge enables one to predict under hypothetical suppositions. This is especially remarkable in a discipline that has variously identified factors such as race and sex as determinants of health, since it has the consequence that factors of this kind cannot be treated as causes either as usefully or as meaningfully as was previously supposed. In this paper I seek to explain the significance of this movement in epidemiology, to understand its commitments, and to evaluate them.

Keywords: Causation; Directed acyclic graphs; Epidemiology; Miguel Hernan; Potential outcomes; Prediction.