Prescribing is a form of psychotherapy. There is no such activity as isolated “medication management.” Clinicians who prescribe medication for mental health problems do not “manage” medication. We do not even manage people. Rather, we accompany and collaborate with fellow human beings, discerning together how to face challenges and find ways forward.
One of us (Warren Kinghorn) has worked as a psychiatrist in the U.S. Department of Veterans Affairs (VA) health system for nearly two decades. When I began working at the VA as a psychiatry resident, I came to believe that prescribing medication for combat veterans with PTSD was a relatively straightforward practice. I would interview veterans and ascertain that their symptoms met DSM criteria for PTSD, including reexperiencing of recurrent and intrusive distressing recollections, nightmares, avoidance behaviors, and hypervigilance. I would offer the diagnosis of PTSD to veterans—many of whom, at that time, had not heard of it—and would explain that both medications and psychotherapy were helpful for that disorder. Some patients were interested in psychotherapy, but nearly all were interested in medication. In nearly all cases I would start with prescribing a serotonin reuptake inhibitor (SRI) such as sertraline, fluoxetine, or paroxetine. I would often augment this with an antihistaminergic medication for sleep, such as trazodone. I would speak about community and social support, encourage connection with other veterans, refer for psychotherapy if desired, and then move to the next patient. SRI, sleep medication, psychotherapy, community support, next patient. It became a comfortable rhythm. I felt competent.
It did not take long for me to realize that prescribing medication for PTSD was not that easy. Some of my patients returned after 3–6 weeks and reported that they were feeling better, with fewer PTSD symptoms. Others reported that they were not better. Some disliked the sexual side effects of SRIs because they interfered with actual or desired sexual relationships. Some liked the sexual side effects because it helped them to manage troublesome sexual feelings. Some deeply wanted the rest associated with sleep but resisted taking sleep medication because they did not want to become so sedated that they lost the ability to remain vigilant against threats. Some Black veterans initially greeted me, a white psychiatrist, with suspicion because of past experiences of being overmedicated by white psychiatrists. Some wanted medication to make it possible for them to engage in trauma-focused psychotherapy. Some had no interest in psychotherapy and simply wanted medication to help them feel less anguished. I realized that there was no “right” medication for PTSD. There were only medications that were helpful or not for particular people who lived with trauma in particular ways.
The other of us (Abraham Nussbaum) learned a similar lesson. When I began working with persons with serious mental illness, I was a resident physician involved in schizophrenia research. I helped recruit, screen, and evaluate patients for medication efficacy trials. I learned to administer a battery of instruments—the Abnormal Involuntary Movement Scale, the Brief Adherence Rating Scale, the Positive and Negative Syndrome Scale, the Structured Clinical Interview for DSM-5, and more—and to record a patient’s medication history, usage, response, and efficacy with precision. Listening to the clinicians and the research subjects talk about medications, I began to imagine the conversation of premodern sailors at sea. When crossing the vast sea, they had only the stars by which to navigate. So they talked about the stars—their shape, their movements, their effects on the sea.
Similarly, it struck me that clinicians and patients often spend their time together naming the pharmaceutical stars by which they chart their course. Risperidone helped me get back to work. Olanzapine got me out of the hospital. As they talked about medications, I was struck that they rarely discussed the relationships that formed around their discussions and that none of the rating scales assessed those relationships. We rarely talked about how the medications affected their ability to relate to the other people in their lives or how prescribing medications reinforced larger social relationships of how we treat people with mental illness. Puzzling over this with one day in the resident workroom, a precepting faculty member reminded me that even when we are prescribing medications, we are building and unbuilding relationships.
Her wise comment steered me toward a different career, away from schizophrenia research studies, to a career spent considering how I can be in relationship with the people I meet as patients and the learners I train to care for these patients. For over a decade, I have worked as an adult inpatient psychiatrist and educator at an academic safety-net system. I daily see that my own preceptor was right: the stars that truly set our course are the relationships we establish with patients, not the medications we prescribe in those relationships. So I have gradually realized that we ought to assess our journey by those relationships, rather than by the medications themselves.
Prescribing Happens in Relationship
Prescribing always happens in relationship, and relationships matter for effective prescribing. We believe that strong relationships between clinicians and patients are centrally important for effective prescribing of medications for mental health problems. Psychotherapy researchers have long recognized that the therapeutic alliance is an important common factor for successful outcomes in psychotherapy, associated with large effect sizes (
Flückiger et al. 2018;
Horvath et al. 2011;
Martin et al. 2000;
Wampold and Imel 2015). For many conditions, the strength of the therapist-patient alliance may be a stronger predictor of success in psychotherapy than the specific mode of therapy utilized (
Wampold and Imel 2015).
The clinician-patient alliance also matters for effective medication prescribing. It matters because it is in the context of relationship that patients are willing to disclose the truth of their experience and that clinicians are able to hear their patients’ experiences in nuanced ways. It matters for discerning when medications are needed, what medications are prescribed, and the goals of medication use. It matters for patients’ willingness to take medications (
Sylvia et al. 2013). It matters for clinical outcomes, accounting for 20% or more of outcome variance in some studies (
Krupnick et al. 1996;
Totura et al. 2018;
Weiss et al. 1997). And it matters because it affects the satisfaction of both clinicians and patients with the prescribing relationship, while either enforcing or challenging social conventions about who gets meds and which meds they receive.
From “What to Prescribe” to “How to Prescribe”
This is a different kind of book about psychiatric medication. In a distinction that we learned from psychiatrist David Mintz, we do not focus on
what to prescribe (
Mintz and Flynn 2012). We focus, rather, on
how to prescribe and on how to build strong, effective prescribing relationships.
Most psychopharmacology manuals and textbooks focus on what to prescribe. In our formation as psychiatrists, we have encountered many excellent textbooks related to psychopharmacology. These textbooks are generally organized either by drug class or by type of mental disorder, and cover things that every prescribing clinician needs to know: how particular medications work in the body (pharmacokinetics and pharmacodynamics), how they are dosed, their common side effects, and how they affect the outcomes of particular mental disorders. We continue to learn from these “what to prescribe” resources, and we recommend them to students and trainees.
In contrast, far fewer resources focus on how to prescribe. How do prescribing clinicians talk with patients about medication? How do prescribing clinicians build trust with patients? What psychological and relational meanings do patients associate with medication? What cultural and social factors affect the ways that prescribers and patients relate to medication? As psychiatrists, our training in psychotherapy included nuanced material on how to respond to patients in distress, how to cultivate a strong working alliance, and how to recognize therapy-advancing and therapy-interfering behaviors. While we discussed these issues with our supervisors and colleagues, we noticed that they were rarely addressed in psychopharmacology textbooks.
Several excellent resources fill this gap by focusing on the integration of psychopharmacology with psychotherapy. Even in the early years of modern psychopharmacology, psychiatrists were reflecting on how medication would affect the process of psychotherapy. Psychoanalyst Mortimer Ostow, writing in 1962, emphasized that while he never recommended treating mental illness “of any sort” with medications alone, medications could be helpful in augmenting psychotherapy. “First,” Ostow wrote,
[Medication therapy] is gentle and does not disrupt ego function. Therefore it does not impede psychologic investigation and interpretation, nor does it damage the transference. Second, when properly employed, it actually facilitates the psychotherapeutic work. Third, the chemical substances can be used to achieve quickly though temporarily, the same ultimate therapeutic influence that we achieve more lastingly though slowly and painstakingly by psychologic means. (
Ostow 1962, p. 3)
Other “how to prescribe” resources focus on how to speak with patients about medications in the context of specific important clinical issues.
Shea (2019) provides practical guidance for improving medication adherence (or, in Shea’s term, “enhancing medication interest”).
Gupta et al. (2019) introduce strategies for reducing polypharmacy through the process of deprescribing.
Another stream of “how to prescribe” resources is the “psychodynamic psychopharmacology” of David Mintz and his colleagues at the Austen Riggs Center (profiled in
Chapter 3, “Prescribing Alliances”). Considering prescribing through the lens of psychodynamic theory, Mintz and colleagues emphasize that medications are not only chemical agents but also relational objects. Patients and prescribers can develop relationships with medications, and medications can play important roles in how patients engage, or avoid engaging, in relationships with others (
Mallo and Mintz 2013;
Mintz 2002,
2019;
Mintz and Flynn 2012).
We draw on these resources and will introduce many of them over the course of this book. Our central attention, though, is on what we take to be the animating center of “how to prescribe” resources: therapeutic relationships that move from the “dispenser model” to the “collaborator model.”
From Dispensers to Collaborators
We start with a thought experiment: might effective psychiatric prescribing be done by a vending machine? Imagine, for example, that after undergoing a diagnostic assessment with a nonprescribing mental health clinician, patients open an app on their phones in which they answer questions tied to DSM-5-TR diagnoses (
American Psychiatric Association 2022), complete standardized symptom rating scales related to those diagnoses, and then enter pertinent information about their medical history, family history, and personal characteristics such as gender, age, and weight. The app analyzes the data against the experiences of other patients, runs a search of evidence-based practice guidelines and the latest drug efficacy literature, accepts electronic payment, and then—
clank! thwomp!—medication bottles with 30-day supplies of medicine are awaiting them at a nearby prescription machine. Within a month, patients revisit the app, update their responses to diagnosis-specific rating scales, enter other pertinent information, and then—
thwomp!—the prescription machine then dispenses another month’s supply of medication at different, or perhaps the same, doses.
With advances in computing and machine learning, such a prescription machine may be on the near horizon. A prescription vending machine is already easy to imagine because it bears uncanny resemblance to the way that many prescribing clinicians have been formed to understand their work. We refer to this model as the “dispenser model” (see
Table 1–1).
The Dispenser Model
In the dispenser model, the central role of the prescribing clinician—like the prescription vending machine—is to dispense the right medication for the patient’s condition. Patients present before the dispenser, usually as individuals, with things called “symptoms.” The goal of the prescriber is to provide an accurate label for these symptoms—a diagnosis of a particular mental disorder—and then to identify and to dispense medication that is judged likely, on the basis of published evidence, to reduce or to ameliorate these symptoms. The goal of treatment is symptom reduction, and the prescribing clinician is the expert who can recommend and dispense medication that will reduce symptoms.
The dispenser model is clear that medication is the agent of healing, and so the central task of pharmacotherapy is “finding the right medication” or “getting the medications right.” The prescriber is important as the dispenser and deliverer of medication. A different prescriber might take the place of the first prescriber—covering, perhaps, in an urgent care clinic—but as long as the same medications are prescribed, the treatment will continue unabated. The prescriber-patient relationship is primarily a tool to encourage patients to accept medication, to encourage positive expectation about medication, and to improve adherence to medication.
The dispenser model is often associated with an individual and biomedical view of mental disorder—the assumption that mental disorders are internal to individuals and reflect disease or dysfunction in brain circuits and other body systems (
Zachar and Kendler 2007), though such a view is not strictly required. Because mental disorders reflect dysfunction in neural systems, mental disorders are appropriately treated with medication that modifies the function of neural systems, thereby reducing symptoms. Effective medication works because it corrects dysfunction in brain circuits or other bodily states of affairs, thereby reducing symptoms. When medication is ineffective or does not work, this is because it fails to modify the dysfunctional systems, because it causes unwanted side effects or toxicity, or because it causes physiological and psychological dependence.
The dispenser model has an important place in mental health prescribing. We draw on it frequently in our work. It works especially well in clinical situations where the individual/biomedical model fits well—for instance, in the treatment of delirium or catatonia, and often in contexts such as melancholic depression, bipolar mania, and acute psychosis. But in mental health situations where the problem is heavily influenced by the patient’s past and present relationships, community, and culture—which is to say, in most prescribing contexts—the dispenser model is an insufficient response. In this book we prioritize a more contextual, relational approach to mental health prescribing, which we call the “collaborator model.”
The Collaborator Model
The collaborator model starts with a contextual and relational, not an individual and biomedical, view of human beings and their mental health challenges (and in this sense has similarities with the “contextual” rather than “medical” model of psychotherapy efficacy of
Wampold and Imel [2015]). Humans are fundamentally
biosocial beings. We become who we are not only as our bodies are configured in particular ways but as our bodies engage with other bodies and with our lived world in relationship, community, and culture. While some mental health problems are “more bio” and some problems are “more social,” mental disorder
always shows up at the interface of bodily function and the world of relationships and community. Mental disorders, like human beings, are biosocial to the core. Sometimes mental health challenges can be understood as individual, biological problems that show up in relationship, community, and culture. Sometimes they can be understood as relational and cultural problems that show up in the body and experience of an individual. Usually they can be best understood in light of both perspectives.
Following this biosocial view, the collaborator model understands that humans are not isolated individuals. Rather, we are persons-in-relation, selves who are extended in relationships and community. Understanding people requires not only looking at them but also looking with them at the world that they are experiencing. This requires not only a face-to-face view but also a side-by-side view. The prescribing clinician is not primarily an expert dispenser, but rather a seasoned accompanist and collaborator who walks with the patient through the terrain of his or her life, seeking both to understand the problem (for which diagnoses are useful heuristic guides) and to discern helpful ways forward.
While the collaborator model celebrates symptom reduction, reducing symptoms is not the primary goal of care. The goal, rather, is for the patient to gain or to regain the capacity and power to pursue valued purposes and goals that contribute to their flourishing. The collaborator model prioritizes
empowerment and
agency, and the therapeutic alliance is the clinician’s strongest asset. The alliance serves to activate the patient’s attachment system, secures a bond within which common tasks and goals can be discerned (
Bordin 1979), and promotes the growth of secure and healthy relationships outside of therapy. Healing and recovery come through relationship—in part through therapeutic relationships, and mostly through healthy relationships in life outside of mental health care.
Medication plays an important but different role in the collaborator model. As in the dispenser model, medications may reduce or ameliorate symptoms, but this is not their primary purpose. Rather, in the collaborator model, medications are useful when they reduce or eliminate barriers to the patient’s ability to engage in healthy relationships and to pursue valued purposes and goals. They may also sometimes “carry” relationships as transitional objects (see
Chapter 14, “Strong Emotions: Prescribing in the Context of Borderline Personality Disorder”). Medications are harmful when they increase or create barriers to healthy relationship, when they make it less likely or more difficult for patients to pursue meaningful goals. They are also harmful when they serve primarily to reduce distress in the
clinician (“countertransference prescribing”; see
Chapter 14) (
Mintz and Flynn 2012) or when they are used to ensure the continuity of an otherwise unproductive therapeutic relationship (as a talisman against abandonment) or to avoid the loss of external goods (e.g., disability benefits). Finally, they are harmful when they are used in discriminatory ways (for instance, applied inequitably to different racial groups) or to encourage the patient to conform to harmful social norms (for instance, norms about body image or student achievement).
Commitments of the Collaborator Model: Relationship, Agency, and Story
We present the dispenser and collaborative models as two ends of a spectrum, not as either-or alternatives. Most seasoned prescribing clinicians, including those who work primarily in biological paradigms, draw from each of these models. To be sure, prescribing clinicians need to know the things that are central to the dispenser model and prominently highlighted in the “what to prescribe” literature: mechanisms of drug action, pharmacokinetics, drug-drug interactions, monitoring algorithms, dosing, and so on.
But this is not enough. Prescribing clinicians need to attend to experiences central to the collaborator model: the nature of the patient’s lived world; the way that the patient navigates challenges; the patient’s strengths, goals, and aspirations; and the meaning that patients and clinicians alike attach to medications. There are three central commitments or values of the collaborator model that we will be engaging throughout this book: relationship, agency, and story.
First, the collaborator model is centered foremost on relationship. All medications are prescribed and used in a relational context, and prescribing clinicians cannot understand what medications are for or what they are doing apart from this context. Much of this book will center on the critically important clinician-patient relationship. But prescribers should also attend to the broader relationships that order, or disorder, patients’ lives. Prescribers must attend to the way that mental health challenges may appear differently in different communal and cultural contexts, and the way that particular relational/communal contexts, such as structures of racial and socioeconomic injustice, may precipitate and perpetuate psychological distress. Prescribers must also attend to the way that relationships can soothe, heal, and empower our patients.
Second, the collaborator model is oriented toward
agency, the capacity to identify purposes and goals and to act in pursuit of them. Agency (and related concepts such as autonomy and locus of control) is a central good of human life and deeply important to the flourishing of clinicians and patients alike. Many patients come into mental health care because they feel powerless to act or constrained from acting toward meaningful goals (
Nussbaum 2015). But curiously, the dispenser model of prescribing can all too often overlook or minimize patients’ agency—not only by attributing therapeutic power to medications and to the clinicians who dispense them (with the patient the passive receptacle of this power), but also by neglecting agency when we say that medications “work.”
In the dispenser model, when clinicians say that medications “work,” and especially that their use is “evidence-based,” we generally mean that patients receiving medications in randomized clinical trials experienced a greater reduction of symptoms than patients in those trials who were prescribed placebos—with the “symptoms” specified on standardized rating scales. But this means that we have to look at the rating scales themselves to understand the nature of the symptoms for which medications have been shown to “work.” And it turns out that in many commonly used rating scales, measured symptoms tend to focus on states of experience that may be precursors to agency but that do not directly measure patient’s capacity to identify purposes and goals and to pursue them (see
Table 1–2). The dispenser model, that is, inadvertently neglects a domain of human life central to why many patients seek care.
In the collaborator model, by contrast, agency is a primary good. Reducing symptoms is important insofar as this enables patients’ agency. Throughout this book, we will be emphasizing that the most effective prescribing relationships are centered on how prescribers and patients align so that the patient can identify and pursue meaningful purposes and goals.
Third, the dispenser model emphasizes story. Our patients are not primarily bearers of symptoms, but rather inhabitants of stories: individual stories, family stories, communal stories, cultural stories. The dispenser model tells one story so well—the biomedical story—that it can crowd out the other stories that patients bring. One of us regularly asks psychiatry trainees who are deeply formed in the dispenser model to give case presentations without using a single clinical term—and we find that the presentations that emerge are often more nuanced, more phenomenologically descriptive, and more human than anything the biomedical story can offer. While this is not a book on narrative medicine, we believe prescribers are most effective when we are attuned to our patients’ stories, and to our own.
This book reflects our conviction that the collaborator model is a powerful but underutilized resource for mental health training and practice. Furthermore, the collaborator makes clear that prescribing clinicians cannot be replaced, without significant loss, by vending machines.
Learning From Wise Clinicians
This book is a guide for clinicians who prescribe psychiatric medications and want to know “how to prescribe” as well as “what to prescribe” when working with patients with particular forms of mental disorder, offering practical principles for wise and effective prescribing in particular mental health contexts. In writing this book, we draw on published research and on our own experience as clinicians. But we are also very pleased, in each chapter, to introduce readers to expert clinicians whom we respect and who have taught us a lot about building prescribing relationships that help patients to heal and to grow. We introduce the work of these clinicians, quote from original interviews with them that we conducted for this book, and describe their approach to prescribing medication in particular mental health contexts. Just as we prescribe in relationship, we
learn to prescribe through relationship. If nothing else, we offer this book as an introduction to these talented and fascinating colleagues. In each chapter, we introduce the work of one or more expert clinicians as well as specific concepts or skills that we have found useful for building prescribing relationships in these areas (see
Table 1–3).
Finally, a note on terminology: by “prescribing clinicians” or “prescribers,” we refer to any clinician who utilizes medications to treat mental health problems, including nurse practitioners, physicians, physician assistants, and psychologists, who work both in primary care and in specialized mental health contexts. As psychiatrists, we are very aware that in the United States, most psychiatric medications are not prescribed by psychiatrists. It is important to us that the expert clinicians whom we profile in this book include not only psychiatrists, but also psychologists and primary care physicians. We are also aware that, strictly speaking, there is no such thing as a “psychiatric medication” or a “mental health medication.” There are only pharmaceuticals that are used to treat mental health problems and that may also be used for other things. In this guide, we generally speak simply of “medication” unless further specificity is required.