Patterns of Propensity: A Review of Heinrichs’ How Psychiatrists Make Decisions

A few years ago, I had a rather testy email exchange with a senior official at a government health agency. We were debating the significance of a study of antidepressant efficacy, which concluded that about 15% of participants have a substantial antidepressant effect beyond a placebo effect in clinical trials. This physician (who was not a psychiatrist) argued that psychiatrists are often misled into believing that antidepressant treatment is pharmacologically effective for a given patient, when supposedly the psychiatrist is merely observing a placebo effect. While not denying that placebo effects are real and significant, I argued that experienced psychiatrists are quite capable of distinguishing a placebo response from one that is pharmacologically based. I gave the example of a patient of mine who had virtually no response to 2 robust antidepressant trials, then showed a remarkable response to a third agent from a different antidepressant class—a monoamine oxidase inhibitor—after about 4 weeks of treatment. I argued that nothing in my work with this patient over more than a year would have led me to conclude that this dramatic remission represented a placebo response. My interlocutor was not impressed, arguing that my conclusion was not scientific.

He was basing his skepticism on the results of this large study, examining 232 randomized, double-blind, placebo-controlled trials of drug monotherapy for major depressive disorder.¹ I was basing my conclusion on dozens of clinical observations, including sequential scores on the Beck Depressive Inventory over more than a year of treatment, as well as on my patient’s subjective reports. A philosophical chasm separated our positions—one that is explored in detail in Douglas W. Heinrichs’ well-reasoned book, How Psychiatrists Make Decisions: The Science of Clinical Reasoning. In my view, among the most important claims of Heinrichs’ book is the assertion that “a science of the individual is possible,” and that “the generation of particular models to represent concrete pieces of reality is in fact the central scientific activity” of good psychiatric care. A corollary of this claim is that evidence-based medicine (EBM) by itself “fails to provide an adequate framework for translating research findings into good clinical care of the individual.”² Indeed, Heinrichs argues, some strong proponents of EBM “degrade the epistemological status of clinical judgment by seeing it as fundamentally unscientific.” My interaction with the government health official certainly confirmed this impression.

In the preface to his book, Heinrichs notes that the book “grew out of reflections on [his] experience of over 40 years as a practicing psychiatrist” in various capacities, and his expertise is certainly evident in this book. Heinrichs indicates that his primary audiences for this book are “psychiatric residents and those responsible for the education of future psychiatrists.” However, he hopes the book will also be of value to psychiatrists in any stage of their careers as well as to “those philosophers with an interest in the conceptual underpinnings of psychiatry.” That is a very broad audience, and I suspect that among this wide array of readers, many will find some chapters of the book more relevant to their needs than others. Chapter 10, for example, deals with such thorny philosophical matters as, “are the laws of nature literally true?” and “logical empiricism and the demarcation problem.” I suspect this chapter—important though it is—will resonate mainly with philosophers of science. On the other hand, the detailed clinical cases Heinrichs presents should be of great practical interest to psychiatrists at any level of training, and nicely demonstrate the utility of what Heinrichs calls the “patterns of propensity (POP) model.” This term requires a bit of unpacking.

In simplest terms, the POP model “offers a hypothesis to explain how the patient’s current situation developed and can generate a set of predictions as to the impact of potential interventions.” A propensity is defined as “an element or state of affairs that, when present in a system of interacting elements, has a disposition to bring about certain results.” Heinrichs uses “propensity” in preference to “cause” in order “to remain neutral as to how strictly deterministic the interactions between elements are and what room should be left for human choice and existential freedom.” We might think of propensities as tendencies to bring about a certain result, rather than so-called hard causes.

Importantly, POP models “are always about a specific concrete patient, not a group or category of patients.” Hence, my treatment-resistant depressed patient—not depressed patients in general—would be the recipient of the POP model. Heinrichs describes the POP model as “an imaginative act regarding a specific individual, but one disciplined and informed by the general knowledge base of psychiatry,” and notes that “all models must be viewed as provisional.” Heinrichs recognizes the value of considering biological, psychological and social factors with respect to any given patient, but regards the ubiquitous biopsychosocial model (BPS) as “not a model at all,” and finds it “vacuous and short on particulars.” In this respect, Heinrichs follows Nassir Ghaemi’s critique of the BPS.³ For Heinrichs, the POP model represents “pluralism at work,” rather than mere eclecticism.

The case studies Heinrichs presents are accompanied by detailed diagrams that illustrate the complex interaction of numerous elements, such as the patient’s coping skills, parental and familial factors, current demands on the patient, intrinsic anxiety modulating mechanisms, etc. Critically, for Heinrichs, the POP model is organized around the patient’s goals and develops “through a process of exchange between psychiatrist and patient leading to some shared perspective that becomes the basis for the clinical work.” In essence, the POP model mirrors how expert clinicians already think, and represents the evolution of the BPS concept into something operational, testable, and clinically powerful. In some respects, the POP model has affinities with the DSM’s concept of the “case formulation”—a widely neglected component of the DSM-5 diagnostic process—though the POP model is more specific and testable.⁴

There is much more valuable material in Heinrichs’ book than can be summarized in this brief space, and additional aspects of the book may be gleaned from the author’s guest posting on Psychiatry at the Margins.³ Suffice it to say that I recommend How Psychiatrists Make Decisions to any clinician who wants to gain an appreciation of how experienced psychiatrists actually reason about their patients. Equally important: Heinrichs’ book reaffirms the scientific legitimacy of careful clinical observation—a point that has gotten lost in the fixation on evidence-based medicine.

Note: All quotes are from Dr. Heinrichs’ book unless otherwise indicated.

Dr Pies is professor emeritus of psychiatry and lecturer on bioethics and humanities, SUNY Upstate Medical University; clinical professor of psychiatry, Tufts University School of Medicine; and editor in chief emeritus of Psychiatric Times (2007-2010). He is also part of the creation of the PRISM diagnostic instrument, which can be found here.

Psychiatric Times is hosting a writing contest for the PRISM diagnostic instrument! See the author guidelines here.Submissions are open until March 31, 2026.

References

1. Stone MB, Yaseen ZS, Miller BJ, et al. Response to acute monotherapy for major depressive disorder in randomized, placebo-controlled trials submitted to the US Food and Drug Administration: individual participant data analysis. BMJ. 2022;378:e067606.

2. Heinrichs D. The science of clinical psychiatric reasoning. Psychiatry at the Margins. May 25, 2025. Accessed February 3, 2026. https://www.psychiatrymargins.com/p/the-science-of-clinical-psychiatric

3. Ghaemi SN. The biopsychosocial model in psychiatry: a critique.Existenz. 2011;6(1).

4. Pies RW. Poor DSM-5—so misunderstood! Psychiatric Times. March 23, 2021. https://www.psychiatrictimes.com/view/poor-dsm5-so-misunderstood