This post first appeared in the New York Times Magazine.
One day several years ago, I was reaching the end of my first visit with a patient, J.J., who had come to see me for anxiety and insomnia. He was a salesman for a struggling telecommunications company, and he was having trouble managing the strain on his finances and his family. He was sleeping poorly, and as soon as he opened his eyes in the early morning, the worries began. "I wake up with a list of things to worry about," he said. "I just go through the list, and it seems to get longer every day."
A psychiatric interview has a certain rhythm to it. You start by listening to what your patient says for a few minutes, without interrupting, all the while sorting through possible diagnoses. This vast landscape of distress has been mapped into a series of categories in psychiatry's diagnostic manual, DSM-IV. The book breaks down mental suffering into 16 groups of disorders, like mood disorders, anxiety disorders, psychotic disorders, eating disorders and several others. As I listened to J.J. (a nickname that he agreed I could use to protect his privacy), it was clear to me that he had one of the anxiety disorders, but which one? In order to systematically rule in or rule out the disorders, I asked J.J. dozens of questions. "Do you have panic attacks?" "Do you get fearful in crowded situations?" "Have you ever experienced a traumatic event that later caused flashbacks or nightmares?"
Each of J.J.'s answers provided me with a clue, closing off one possibility while opening up others. At its best, when you are working with an intelligent, insightful patient, the process is fun, involving a series of logical calculations, much like working a Sudoku puzzle. Finally, toward the end of the hour, I felt confident that I had arrived at J.J.'s diagnosis. "I think you have what we call ‘generalized anxiety disorder,' " I told him. It may start with a defined series of causes, as was true for J.J., but then it spirals outward, blanketing the world with potential threat. J.J. worried about what the future would bring and experienced a predictable series of physical symptoms: insomnia, muscle tension, irritability and poor concentration.
"I'm going to write you a prescription for a medication called Zoloft," I said, picking up my prescription pad. He asked what was causing his anxiety, and I began one of the stock neurochemical explanations that psychiatrists typically offer patients about low serotonin levels in the brain. The treatment involved "filling up the tank" by prescribing a medication like Zoloft, Celexa or Paxil.
"I specialize in prescribing medications," I said with a smile. I was a psychopharmacologist and specialized in medication rather than psychotherapy. "I can refer you to a good therapist in the area if you'd like."
After J.J. left my office, I realized, uncomfortably, that somehow, over the course of the decade following my residency, my way of thinking about patients had veered away from psychological curiosity. Instead, I had come to focus on symptoms, as if they were objective medical findings, much the way internists view blood-pressure readings or potassium levels. Psychiatry, for me and many of my colleagues, had become a process of corralling patients' symptoms into labels and finding a drug to match.
Leon Eisenberg, an early pioneer in psychopharmacology at Harvard, once made the notable historical observation that "in the first half of the 20th century, American psychiatry was virtually ‘brainless.' . . . In the second half of the 20th century, psychiatry became virtually ‘mindless.' " The brainless period was a reference to psychiatry's early infatuation with psychoanalysis; the mindless period, to our current love affair with pills. J.J., I saw, had inadvertently highlighted a glaring deficiency in much of modern psychiatry. Ultimately, his question would change the way I thought about my field, and how I practiced.
I originally became interested in psychiatry primarily because of my father: he is a psychiatrist practicing in San Francisco. But there was a darker side to my career choice. My mother suffered severe mental illness, with debilitating depressions and paranoid thoughts. One autumn day during my junior year in college, she committed suicide. Psychiatry then became personal, a way for me to come to terms with her illness.
I majored in psychology at U.C. Berkeley, and at U.C. San Francisco I labored through medical school's rites of passage in order to qualify for a psychiatric residency. Eventually, on a steamy July day in 1992, I stood on a Boston street, far from home, gazing at Massachusetts General Hospital (known as M.G.H.), where I was about to start my training.
This was a momentous time at M.G.H. Prozac was introduced four years earlier and became the best-selling psychiatric medication of all time. Zoloft and Paxil, two similar medications, were in the pipeline, and many of the key clinical trials for these antidepressants were conducted by psychiatrists at M.G.H. who were to become my mentors. M.G.H. and other top programs were enthralled with neurobiology, the new medications and the millions of dollars in industry grants that accompanied them. It was hard not to get caught up in the excitement of the drug approach to treatment. Psychopharmacology was infinitely easier to master than therapy, because it involved a teachable, systematic method. First, we memorized the DSM criteria for the major disorders, then we learned how to ask the patient the right questions, then we pieced together a diagnosis and finally we matched a medication with the symptoms.
But learning the formal techniques of therapy was like navigating without a compass. While I learned how to form an alliance with my patients and begin a good dialogue, becoming a skillful therapist requires much more practice than busy psychiatry residencies allow.
When my father did his residency at U.C.S.F. in the 1950s, he learned therapy well, because, with few medications available, it was the main treatment psychiatrists could offer their patients. Psychiatric residencies focused on therapy, and many residents extended their training further by enrolling in postgraduate psychoanalytic institutes. When modern medications came on the scene, my father adapted by incorporating them into his therapy practice, as did many of his colleagues. I call this the "golden" generation of psychiatrists, those currently approaching retirement age, who were skilled at offering the full package of effective psychiatric treatments to patients.
The newer generation of psychiatrists, who graduated in the 1980s and afterward, trained in programs that were increasingly skeptical of therapy and that emphasized a focus on medications. M.G.H. was by far the most influential of these modern programs. Graduates of the M.G.H. program and its sister program at nearby McLean Hospital have fanned out throughout the country, becoming chairmen of departments and leaders of the National Institute of Mental Health.
A result is that psychiatry has been transformed from a profession in which we talk to people and help them understand their problems into one in which we diagnose disorders and medicate them. This trend was most recently documented by Ramin Mojtabai and Mark Olfson, two psychiatric epidemiologists who found that the percentage of visits to psychiatrists that included psychotherapy dropped to 29 percent in 2004-5 from 44 percent in 1996-97. And the percentage of psychiatrists who provided psychotherapy at every patient visit decreased to 11 percent from 19 percent.
While it is tempting to blame only the biologically oriented psychiatrists for this shift, that would be simplistic. Other forces are at work as well. Insurance companies typically encourage short medication visits by paying nearly as much for a 20-minute medication visit as for 50 minutes of therapy. And patients themselves vote with their feet by frequently choosing to see psychopharmacologists rather than therapists. Weekly therapy takes time and is arduous work. If a daily pill can cure depression and anxiety just as reliably, why not choose this option?
In fact, during my 15-to-20-minute medication visits with patients, I was often gratified by the effectiveness of the medications I prescribed. For perhaps a quarter of them, medications worked so well as to be nearly miraculous. But over time I realized that the majority of patients need more. One young woman I saw was referred to me by a nurse practitioner for treatment of depression that had not responded to several past antidepressants. She was struggling to raise two young children and was worried that she was doing a poor job of it. Her husband worked full time and was rarely available to help. She cried throughout our initial interview. I started her on Effexor and referred her to a social-worker colleague. She improved initially, but over the years since, her symptoms have waxed and waned. When she reports a worsening of her anxiety or depression, my first instinct is to do one of three things - switch medications, increase her dosage or add another. Over the course of 15 or 20 minutes, this is about all I can offer.
My treatment of this young woman follows the "split treatment" model, but a less charitable description is "fragmented care." Like the majority of psychiatrists in the United States, I prescribe the medications, and I refer to a professional lower in the mental-health hierarchy, like a social worker or a psychologist, to do the therapy. The unspoken implication is that therapy is menial work - tedious and poorly paid.
But over the past few years, research studies have shown that therapy is just as effective as medications for many conditions, and that medications themselves often work through the power of placebo. In one study, for example, researchers did a meta-analysis of studies submitted by drug companies to the F.D.A. on seven new antidepressants, involving more than 19,000 patients. It turned out that antidepressants are, indeed, effective, because on average patients taking the pills showed a 40 percent drop in depression scores. But placebo was also a powerful antidepressant, causing a 30 percent drop in depression scores. This meant that about three-quarters of the apparent response to antidepressants pills is actually due to the placebo effect.
Nobody knows exactly how the mysterious placebo effect works, but it is clear that it has impacts on the brain that can be seen as clearly as medication effects. In one study conducted by pain researchers at the University of Michigan, subjects were given an ache-inducing injection of saline into their jaws and were placed in a PET scanner. They were then told that they would be given an intravenous pain treatment, but the "treatment" was merely more saline solution, acting as a placebo. The PET scan showed that the endogenous endorphin system in the brains of the subjects was activated. The patients believed so strongly that they were receiving effective treatment that their brains followed suit. Presumably, a corresponding brain change occurs when depressed patients are given placebo pills.
Like placebo, psychotherapy has typically been considered a "nonbiological" treatment, but it has become clear that therapy, like placebo, also leads to measurable changes in the brain. In an experiment conducted at U.C.L.A. several years ago, with subjects suffering from obsessive-compulsive disorder, researchers assigned some patients to treatment with Prozac and others to cognitive behavior therapy. They found that patients improved about equally well with the two treatments. Each patient's brain was PET-scanned before and after treatment, and patients showed identical changes in their brain circuits regardless of the treatment.
In depression treatment, too, pills and therapy each lead to brain changes, but in this case they appear to be intriguingly distinct. In studies by Helen Mayberg, a professor of psychiatric neurology at Emory University, depressed patients given cognitive behavior therapy showed decreased activity in the frontal lobe, the brain center that might be responsible for the overmagnification of life's problems that leads to depression in some patients. And they showed increased brain activity in parts of the limbic system, a brain region associated with strong emotion. But Mayberg found that when patients were given medication, their brain activities changed in the opposite direction, stimulating the frontal lobe and damping down the limbic system. "Our imaging results suggest that you can correct the depression network along a variety of pathways," she said.
Clearly, mental illness is a brain disease, though we are still far from working out the details. But just as clearly, these problems in neurobiology can respond to what have traditionally been considered "nonbiological" treatments, like psychotherapy. The split between mind and body may be a fallacy, but the split between those who practice psychopharmacology and those specializing in therapy remains all too real.
After I saw J.J., I decided that I wanted to try to change my approach to treatment. I gradually began to carve out room in my schedule for longer visits with my patients. I endeavored to do what's called "supportive" therapy, a technique favored by many therapists and involving basic problem solving and emotional support. It is a bit like what a friend would do for another friend offering advice in times of trouble, but more elaborate and with an accompanying raft of studies endorsing its effectiveness in psychiatry.
When I started to probe my patients, I realized I barely knew most of them. I had exuberantly documented their moods, sleep habits, energy levels and whether they had suicidal thoughts, but I didn't know what made them tick as people. For example, I had treated Jane (her middle name), a health care administrator in her 40s, for depression and bulimia for many years, focusing on a complicated combination of medications like Effexor, Provigil and Xanax. Then she had a depressive relapse. This time, rather than simply adjusting her medication, I asked her what was going on in her life. I found out that her boss had recently given her an impossible assignment to complete and had berated her when the results were not to his liking. It seemed clear to me that her depression was partly triggered by the fact that she blamed herself for her boss's poor communication skills and managerial lapses. I encouraged her to question that assumption each time it popped into her mind. Over the next few visits, she improved - relating better to her boss, able to take his criticism with a grain of salt and feeling more confident.
But while my simple therapeutic suggestions were helpful, as Jane and I continued to explore her work issues in detail, she said that she often lacked focus on the job. Knowing that poor concentration can be a symptom of depression, I asked if she associated her distractibility with periods of sadness. But she said there was no such correlation with her moods.
I wondered if she might have an adult version of attention deficit disorder. Indeed, carefully reviewing her years in school, I saw she was always easily distracted but overcame this problem by working extra hard, and managed to achieve good grades. Even now, her job performance was consistently rated as excellent, but the extra work this entailed sapped her energy.
I decided to prescribe her a version of Ritalin, the standard treatment for A.D.D. The next month, she said she felt that this medication had turned her life around. Not only did she feel more focused and productive at work, but she was more apt to get things done at home, which in turn enhanced her mood, indirectly improving her depression.
Is Jane's story an argument for psychiatrists providing both medication and some sort of psychotherapy? I think it is. This does not mean that dedicated psychotherapists are not crucial - they are, because they can provide the in-depth therapy that psychopharmacologists will never have time to deliver. When our patients need more from us than just medication, however, we should be prepared to provide it. Oddly, managed-care companies discourage us from doing psychotherapy, arguing that it is cheaper to have psychiatrists do 20-minute medication visits every three months and to hire a lower paid non-M.D. for more frequent therapy visits. But the few studies that have analyzed the economics of these arrangements have found that integrated treatment actually saves money. Mantosh Dewan, the chairman of psychiatry at SUNY Upstate Medical University in Syracuse, found that when psychiatrists do both medication and psychotherapy, the overall amount of money paid out by insurance companies is actually less than when the treatment is split between psychiatrists and psychotherapists. When patients see only one provider, they require fewer visits overall.
It may be time to consider whether the term "psychopharmacologist" is actually doing damage to the field of psychiatry. The American Society of Clinical Psychopharmacology defines "psychopharmacology" on its Web site as "the study of the use of medications in treating mental disorders." But in a recent article, three Harvard psychiatrists (interestingly, two of them from Mass General) have pointed out that no other medical specialty has carved out a separate "pharmacology"subspecialty. Good doctoring, they remind us, involves perfecting all the skills relevant to healing and deploying them when needed.
During my mother's last months, she isolated herself from her family, so I don't know what kind of treatment she was receiving before her death. But I do know what kind of treatment I would have hoped for her. She needed medication to combat her paranoia and the emotional pain of her depression. She needed someone who could expertly probe her thought process, in order to understand the fateful logic that led her to conclude that the only solution was to end her own life. She needed treatment that was intensive and exquisitely coordinated.
Such care is not always capable of saving damaged lives. But it is the best that we can do. It's what we owe our patients - and ourselves.
Copyright Daniel Carlat