Wieten Presents at CauseHealth Conference

Posted on 10/16/2017 9:40:27 AM

Sarah Wieten, Department of Philosophy, gave a talk titled “Manageable Mechanisms in Medicine: How Confounders Make Amalgamation Difficult” at The Guidelines Challenge: Philosophy, Practice, Policy, a conference organized by CauseHealth and held at Oxford (UK), October 3–4, 2017.

The conference brought together practitioners, guidelines networks, and philosophers of science to address the general problem of how to put the tools of philosophy to use in improving the development and implementation of healthcare guidelines.

An abstract of Dr. Wieten's talk is below:

The current philosophical discussion about causes and mechanisms is diverse and robust. This literature includes debates over causal/non-causal explanations, the possibility of a causality without necessity, and mechanisms as instances of invariance (or at least constant conjunctions) (See Reutlinge, forthcoming; Mumford and Anjum, 2011, Cartwright 2009 for a sample). However, the debate over mechanisms in the philosophy of medicine is unique. Because the history of medicine is riddled with cases of the terrible (and often deadly) consequences of poor mechanistic reasoning, clinical medicine is currently characterized by an aversion to mechanistic approaches. In contrast, paradigms such as Evidence-Based Medicine (EBM) rely on large-scale studies of the efficacy of interventions, without the additional concern with mechanisms which lie behind this efficacy.

In this paper, I argue that in order for mechanisms to be given the important role for intervention selection that the Russo-Williamson thesis (RWT) affords them, additional theoretical work is needed. The RWT is roughly, the claim that evidence from both mechanisms and probabilistic relations is needed to warrant causal claims. This work is needed to narrow the scope of mechanisms in response to concerns raised in EBM manuals and by RWT critics, who contend that mechanisms may be so large when possible confounders (more colloquially, “unlesses”) are included that it is impossible to articulate them in a way that is useful for intervention decision-making. If all “unlesses” are not accounted for, we will draw wrong, often very wrong, conclusions. In short: the worry is that what it takes to make a usable mechanistic claim that allows us to draw roughly reliable conclusions is just too demanding.  Addressing the “size” of the mechanisms in question will help to bring RWT in line with concerns expressed by critics, but disagreements remain. This question is important to address in order to understand the place of mechanisms in guideline creation and other kinds of evidence amalgamation.

I develop this argument in three sections. First, I discuss the points of tension between medicine’s historical distrust of mechanisms and those who insist that mechanisms have an essential role to play in medicine, such as Russo and Williamson of the RWT. I then argue that this tension stems from the assumed impossibility of understanding mechanisms in light of their immense scope. Third, I consider a possible objection to my modification of the RWT.