What the Doctor Ordered

Every now and then — I confess that the interval may be more like six months than three — I consult to a kindly older gentleman. He talks for an hour, updating me on his family life, which by any standards has had its share of sad events. He is a born helper, this patient of mine, and, as is often the case with helpers, the horizon he sees is darkened with visions of setbacks to those he loves. I value these sessions, value the privilege of seeing humanity through compassionate eyes, At the end of each meeting, I write out a prescription for 25 milligrams of trazodone.

I understand that 300 to 400 milligrams, or even more, is the active dose you read about in the literature. Half a 50-milligram Desyrel barely suffices as a hypnotic. But my patient does not use the medication for sleep alone. He finds it makes him at once less anxious and less sentimental, less vigilant over the constant threats to his world. On numerous occasions we have tried to eliminate the trazodone; but despite my doing the usual things that ought to make the enterprise work — increasing the frequency of meetings, reframing the patient's cognitions about impending disaster-he finds himself, off medicine, depressed and panicky in ways that plagued much of his earlier life.

Prescribing an antidepressant in homeopathic doses is sure to arouse skepticism: It is just what general practitioners do, the very thing we specialists scorn. I confess I have never been terribly self-critical about this transaction — whatever the specific role of trazodone, the proceedings are doctorly enough to satisfy my standards. But the little doubt and shame I might feel were assuaged this past summer by a conversation I had with a senior diagnostician, one of those "psychiatrist's psychiatrists" whom we consult when conventional courses of treatment fail.

We were talking about intractable cases, the ones where the usual pharmacologic regimens, from SSRI and lithium to tricyclic and neuroleptic, have been tried without success, and where psychotherapy barely makes a dent. The diagnostician told me that his practice in these instances is to pull gravely at his chin and to recommend resuming what the local doctor had the patient on in the first place (and often for many years), usually 50 milligrams of Elavil. This expedient, he said, generally works, even if the disorder in question involves psychotic delusions. And it has the added advantage of making the consultant look like a homespun miracle-worker.

Here is the question, then: how do the 25 milligrams of trazodone or 50 milligrams of amitriptyline perform their magic? To call them placebos is all very well, but why were the higher doses of combined medicines not placebos? Why, with the patient I was trying to wean off medication, did my hopeful expectancy about more frequent visits not outperform my skeptical prescribing of a serotonin agonist in minute amount? Might low-dose antidepressants sometimes have direct pharmacologic effects?

Human beings are remarkable for their biological diversity, a truth that plagues drug developers. When I interviewed pharmaceutical researchers for my recent book, they told war stories about particular subjects who gave false leads — people in whom drugs that later proved panicogenic appeared anxiolytic. The letters columns in our journals are full of anomalies, most recently reports of patients sedated by stimulating antidepressants.

One colleague who repeatedly reminds me of the importance of individual variation is John K. Pearce, a Harvard-based evolutionary psychiatrist, author of Exiles from Eden and Ethnicity and Family Therapy, and a prolific correspondent. Regarding psychotherapy, Pearce takes seriously the possibility that symptoms and personality differences may not arise from inner conflict and repression; they may not even result from injury or constitute illness. Peculiarities are often just the phenotypic evidence of genotypic or cultural variability.

Regarding pharmacotherapy, Pearce believes that the gravest conceptual error is a failure to apply the lessons of population biology. Groups vary. The people family practitioners treat are different from the people who show up in mood disorder clinics of tertiary-care hospitals. Pearce finds it unsurprising that general practitioners should have success with low-dose SSRIs while university consultants have trouble bringing patients round on monoamine oxidase inhibitors supplemented by amphetamine. The primary care psychiatrist whose patients seem always to be on insufficient doses of medication may be acting on experience.

Of course, extrapolating from one's own small store of successes is dangerous. One wishes, in the ideal case, to make choices based on research. But from the population biology point of view there are precious few studies whose results apply directly to the office practice of psychiatry.

This point was driven home to me in a recent conversation I had with the director of one of those tertiary-care mood clinics. The program in question evaluates 150 patients a month. The director, who is also a pharmacologic researcher, requires 300 to 400 new subjects each year for his clinical studies. Of the 1800 patients seen annually, fewer than 20 ever make it into the research. The real patients tend to be too young or too old, or to have comorbid illness, or to have partial syndromes that do not meet diagnostic criteria, or just to prefer not to submit to experimentation when their insurance makes well-tested remedies available.

Ninety-nine percent of patients in the research trials are recruited by newspaper advertisement. No one knows just what characterizes people who respond to ads. They differ from the clinic population in terms of their symptom complexes-neither too much nor too little illness-but they also differ in a variety of other ways, most likely including their level of integration into mainstream social institutions.