The Limits of Evidence-Based Medicine in Psychiatry

Abstract

Evidence-based medicine (EBM) has delivered major benefits in somatic medicine, but the situation is less clear in psychiatry. Recent empirical evidence in psychiatry (from within the tradition of EBM) contradicts the fundamental assumption of this approach: the belief that mental health problems, and treatments for them, are best framed in a technical idiom. EBM is essentially an attempt to identify the most successful technical interventions for discretely defined medical conditions. This effectively renders the non-technical, nonspecific aspects of care as being of secondary importance. In this paper, we first consider some of the empirical evidence that challenges this perspective. We look at depression in particular and the results of randomized, controlled trials investigating the effectiveness of modern antidepressant drugs as well as studies looking at the benefits of different forms of psychotherapy. These empirical studies (carried out according to the logic of EBM) indicate that the nonspecific factors are actually of primary importance. To tease out the implications of this, we first consider the work of Thomas Kuhn, who described scientific progress in terms of revolutions and changing paradigms, and then turn to the philosophy of Merleau-Ponty to describe the limitations of scientific accounts of experience. We argue that nonspecific factors have dual significance. They can be seen as Kuhnian anomalies, observations from within the EBM paradigm that are difficult to reconcile with the currently dominant paradigm. At the same time, as Merleau-Ponty argues, they reveal the limitations of scientific knowledge about human beings. They show that culture and meaning play a central role in mental health practice.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 49 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.