From Popperian science to normal science. Commentary on Sestini (2009) 'Epistemology and ethics of evidence-based medicine'.

In ‘Epistemology and ethics of evidence-based medicine: Putting goal-setting in the right place’, Piersante Sestini argues that evidence-based medicine (EBM) is highly consistent with Karl Popper’s philosophy of science, which famously demarcated science by the falsifiability criterion and highlighted for generations of scientists the inherent fallibility of scientific claims. In this commentary, I will demonstrate, contra Sestini, that Popper and EBM share only superficial similarity. Furthermore, Sestini’s focus on the centrality that formulating the clinical question plays in evidence-based practice instead highlights EBM’s compatibility with the Kuhnian picture of ‘normal science’. Despite Sestini’s claim to find high compatibility between EBM and Popper’s objective theory of conjectural knowledge [1], he illustrates only a thin allegiance between the two. Sestini finds Popperian influence in the EBM criterion that clinicians begin by ‘formulating an answerable question’. While an unanswerable question would certainly fall outside of the purview of Popperian science, it was the untestable hypothesis that interested Popper in his demarcation of science from pseudoscience. A hypothesis is considered scientific if it is testable, that is, falsifiable by way of a possible observational event that is incompatible with theoretical prediction [2]. All other theories, even those making empirical claims, are pseudo-science. For example, broad ideological theories, like Marx’s theory of history and Freud’s theory of psychoanalysis, make empirical claims about the world and have remarkable explanatory power (according to Marxists and Freudians), yet cannot be disconfirmed by singular observational events. The grandness of their claims, and constant supply of confirming evidence – ‘a Marxist could not open a newspaper without finding on every page confirming evidence for his interpretation of history; not only in the news, but also in its presentation – which revealed the class bias of the paper – and especially of course in what the paper did not say’ ([2], p. 46) – seemed to permit such discretion ([2], pp. 43–77). Unlike the logical positivists’ claim that nonscience is non-sense, Popper allowed such metaphysical speculation to be meaningful and informative. They were not, however, scientific claims [3]. Sestini also links the ‘evidence-based’ criterion – the evidence determines the answer to clinical problems – to Popper’s demarcation of science: External clinical evidence both invalidates previously accepted diagnostic tests and treatments and replaces them with new ones that are more powerful, more accurate, more efficacious, and safer, which is as close as it could be to Popper’s criterion of demarcation based on falsifiability [4]. Yet an accurate reading of Popper reveals that theories can only be falsified and not confirmed by any inductive evidence. The ‘replacement’ of old invalidated theories with new ones is merely conjectural. In a dramatic break from the common picture of science as an inductive process, Popper proposed that science progresses through a series of conjectures and refutations [2]. Scientists hold onto their conjectures until damning evidence falsifies them in a methodology of ‘trial and the examination of error’ ([1], p. 18). Once falsified, the conjecture is abandoned, a new one is proposed, and the effort at falsification is repeated. The growth of scientific knowledge, for Popper, results from an enlarging body of falsified knowledge claims. While Popper’s account of science avoids the vexing ‘problem of induction’ for empiricism [5,6], critics have frequently argued that a theory of scientific reasoning devoid of theory confirmation is unworkable for actual scientific practice [7]. What is a Popperian biomedical researcher or doctor supposed to do when faced with competing solutions to a clinical problem – that is, two unfalsified claims? Popper is quite clear that we have no logical grounds for selecting one over the other. It is surprising that Popper is such a hero to scientists given how irrational he made theory choice. Yet Sestini’s comment is also linked to the democratic tenor of the evidence-based programme. EBM was lauded in the early literature as an iconoclastic practice [8]. Much like Popper’s critical science, evidence-based practitioners do not resist disconfirming evidence nor hold onto their theories dogmatically. Popper took this critical attitude even further by encouraging scientists to pursue actively disconfirming observational evidence in order to increase the robustness of their theories. This picture of the proper scientific attitude is hugely attractive. The Popperian scientist’s character traits include creativity, critical thinking and openmindedness [2]. EBM, however, does not demand the critical scientist that Popper envisioned. EBM offers a detailed systematic approach to clinical problem solving that demands rule-following and compliance (albeit those rules may be reasonable!) rather than open-ended inquiry. A re-examination of Sestini’s area of interest, the formulation of the clinical question in evidence-based practice, demonstrates this to be the case.