Where are we going wrong?

The Last Psychiatrist - November 9, 2006

Massacre of The Unicorns

The real problem of a critique of our own cultural models is to ask, when we see a unicorn, if by any chance it is not a rhinoceros.

It’s a convenient fiction that the difficulties with psychiatric diagnosis and treatment are due to incomplete knowledge- if we just knew more about dopamine!—but the real source of the failings is inherent in its structure. Psychiatry fails because it is designed to fail. (continued below...)

Massacre of the Unicorns

The real problem of a critique of our own cultural models is to ask, when we see a unicorn, if by any chance it is not a rhinoceros.[i]

It’s a convenient fiction that the difficulties with psychiatric diagnosis and treatment are due to incomplete knowledge- if we just knew more about dopamine!—but the real source of the failings is inherent in its structure. Psychiatry fails because it is designed to fail.

Semiotics and Psychiatry

The problem is with the signs of psychiatry. Signs are vehicles for conveying information. For example, the term “antidepressant” is a sign conveying the idea of a “drug which treats depression.” While tricyclics and SSRIs are both called “antidepressants” because they treat depression, the term otherwise carries no necessary value except that. Signs can be misused if they are expanded to take on new meaning in an inconsistent fashion. Consider the controversy over “antidepressant induced mania.” Tricyclics have been shown to be associated with higher manic rates, but SSRIs haven’t. So does that mean that most textbooks and articles are wrong—that antidepessants don’t cause mania? Actually, the problem is worse than that. It’s not that antidepressants don’t cause mania—it is that there cannot be any such notion as antidepressant induced mania.

Hydrochlorothiazide (a diuretic) and propanolol (a beta blocker) are both “antihypertensives,” but otherwise are very different drugs with very different side effects; only one makes you pee. There’s no such thing as “antihypertensive induced peeing.”

But this error is itself based on an earlier semiotic error, with the sign of “bipolar disorder,” and its implicit idea of two opposing poles. Depakote will cure and prevent mania; and Zoloft will cure and prevent depression. The treatment of one pole should surely move the patient closer to the next pole; hence, psychiatrists talk about “antidepressant induced mania” but “Depakote breakthrough mania.” See the difference? No? That’s because, in reality, there is no difference in the two manias, and absolutely no reason one should believe one is “induced” and the other “breakthrough,” except that the problem has arbitrarily been defined in this manner, i.e. a) mood stabilizers are a priori preventative, and b) Depakote is a mood stabilizer. But a) is a postulate and b) is an optimistic assumption.

A similar example is the often cited myth that diagnosis of bipolar disorder is frequently missed. A survey[ii] found that 69% of patients were actually misdiagnosed, most often as having regular depression. An average of 4 physicians were consulted “prior to receiving the correct diagnosis.” But who is to say what is the correct diagnosis, when the diagnosis is based on vague and overlapping descriptions (and not on objective pathology?) You can look at this the other way, and say only 1 out of 5 psychiatrists felt it was bipolar disorder, while the other 80% thought it was depression. So it pays to have the last word. Bipolar disorder is “frequently missed” not because it exists and doctors miss it, but because it is defined in a way which allows it, by 80% of doctors, to be legitimately called something else. The only way to say the diagnosis was correct or incorrect, in the absence of objective pathology, is to say that the treatment they received for “bipolar disorder” from the fifth doctor was better (read: safer or more efficacious) than the treatment they received from the first four. This is not evident.

Marco Polo's Dilemma

The problem in psychiatry is the problem in the epigraph at the beginning of the paper. When Marco Polo saw the exotic one horned quadruped, his frame of reference required that it could be none other than a unicorn, even though it did not conform exactly to his prior conception of it. Marco Polo made his observation fit his existing paradigm of zoology. While superficially (and in retrospect) this may seem silly and arbitrary, it is in fact the opposite, Marco Polo believed the only thing he could believe—because the alternative was to believe he had discovered an entirely new, unheard of, creature. This is the semiotic problem in psychiatry, wherein we are forced to interpret clinical signs with our available “encyclopedia,” because we lack both the newer encyclopedia, and its accompanying language, to interpret the signs differently. But here’s the extra credit question: is it really a rhinoceros or a unicorn?

Psychiatry is trying to move away from a symptom based field to a disease (or at least disorder) based science. We are deciding (note word choice) that a psychiatric disease exists a priori, and can present with different symptoms, the way cancer could present as the flu, but really is cancer. The diagnosis (or sign) becomes more important than the individual symptoms, because it demands a specific treatment. But if the paradigm is faulty, what of the treatment?

An example is the association of bipolar disorder with mood stabilizer. Through a series of laughable twists, psychiatrists came to believe that antiepileptics had special properties in bipolar because they could quell the chaos in the brain in the same way that they calmed seizure activity there. It is now common practice, as defined in numerous “Expert” guidelines and consensuses, that patients with bipolar disorder need to be on a mood stabilizer, specifically lithium or the antiepileptic Depakote. The problem with this is that there has never been a study that found that Depakote is a mood stabilizer. In fact, there is no evidence that any antiepileptics are mood stabilizers. What few studies have been done show no benefit over placebo for this purpose. Remarkably, very few psychiatrists know this. But worse, even when you show them data, they refuse to accept it.

I know, it seems—well, crazy—that psychiatrists would blatantly ignore the absence of data—the opposite of data. Ask your psychiatrist the following question: A patient with a history of bipolar disorder presents depressed. There are only two medications available: Prozac or a new antiepileptic for which no efficacy or safety data yet exist. Which do you use? That the mystery antiepileptic is even considered shows the power of the association, because I haven’t actually told you anything about the antiepileptic. Nothing about efficacy, safety, anything. And a doctor would consider it? It’s considered because it is assumed that it will share the same properties as other antiepileptics, i.e. that it is a mood stabilizer. But no other antiepileptics are mood stabilizers, so why the assumption? And even if one seizure drug was, in fact, a mood stabilizer, why would another one be? If antiepileptics have totally different efficacies with respect to seizure treatment, why should it be any different for bipolar treatment? The power of the paradigm compels us.

Can an antipsychotic be an antidepressant? Sure. Can an antidepressant be an antipsychotic? See? You’re hesitating. It is harder to imagine that an antidepressant can be an antipsychotic—as if there is anything in either term that allows us to predict other actions—because the paradigm has given value to terms that they don’t have.

There’s an analogy in the social sciences: racism.

The Knowledge Trap

But psychiatry is an applied discipline. What’s the harm if assumptions lead to efficacious treatments? This is a trap. Psychiatry has convinced itself that it needs to focus on expanding its knowledge (i.e. data) rather than re-evaluating its postulates and paradigms. New discoveries or information are used to build on a paradigm, not to test it. But now psychiatry is looking to see what it expects to find. And if a discovery flatly contradicts the tradition, then it is ignored or rationalized. How else to explain the comparative absence of articles critically discussing the placebo response? And the even fewer whose proposed solution is not the abandonment of the placebo arm, in favour of “active controls.” Active as defined by whom? Is Depakote an active control?

What Is Modern Psychiatry Seeking?

The fiction is that psychiatry is looking for more efficacious treatments. It is not. It is looking for different treatments; the paradigm does not allow for the creation of better treatments. For example, psychiatry can applaud itself from moving from a “noradrenergic hypothesis” to a “serotonin hypothesis” of depression, but it’s still the same paradigm. While first line medications have changed, they have not changed because of improved efficacy. Nothing has ever found anything to be more efficacious than the previous standard (SSRIs. vs. tricyclics, atypical antipsychotics vs. chlorpromazine, etc.) nor has any “model” been more or less correct than any other. That some medications have less side effects and greater versatility is useful, but a) this is almost never the result of intentional scientific discovery but rather the fortunate by-product of the invention of (yet another) efficacious treatment; b) this greater tolerability in no way reflects the accuracy or inaccuracy of an existing model. That Zoloft is more tolerable than imipramine has nothing to do with the viability of the “serotonin hypothesis.” And yet how many times have I heard that antipsychotics treat depression with no more rigorous explanation than because of their “activity” (note the vague term) on serotonin?

Psychiatry, which seeks to be like physics, becomes instead a caricature of it. It, too, tries to focus on expanding knowledge and not re-evaluating its principles. But unlike physics, psychiatry has no formal principles. They are made up. It is, strictly speaking, not a science but a paradigm, no different than psychoanalysis. It may seem as though Freud concocted the notion of the unconscious out of thin air and developed an entire field around it, but modern psychiatry has done nothing different in concocting the notions of kindling or “upregulation of receptors” as first principles and then constructing an equally arbitrary field around them. That medications help patients has everything to do with the medications and nothing to do with the incense and liturgy that surround them.

Paradigm shifts do not occur in physics because the principles do not change. Newtonian mechanics will always be useful for prediction because it is correct for the cases in which it is applicable (i.e. for measurable bodies.) It is furthermore not susceptible to political influence. Psychiatry is the opposite. The decision to accept or reject the paradigms in psychiatry are very clearly political, not evidentiary. We as individuals accept the idea that antiepileptics are mood stabilizers because psychiatry has decided to adopt this position, not because the evidence requires us to accept it (in fact, the evidence should require us not to accept it, or at least seriously question it.) No physicist could hope to “practice” physics without having read and understood what came before, without having worked the “block on an incline” problem. But there is no theoretical nor practical requirement to practice psychiatry of reading the papers on, for example, mood stabilization, let alone what came before. All that is required is to know what the current practice is (“Guidelines recommend prescribe antiepileptics.”) This may seem like science, i.e. “scientists have determined that antiepileptics are mood stabilizers, so we will trust their word and prescribe them,” but it is very clearly politics.

The current problem of psychiatry is that it seeks to be something that it is not: science. It may be, at some future date, readily described by scientific principles, but this is assuredly not the case now. It is most certainly a sociological construct, a paradigm, with a shared educational system, shared assumptions, and a mechanism to communicate discoveries (i.e. journals, meetings.) It also has a common language. But it lacks the predictive ability common to other disciplines.

An argument against the notion that psychiatry is an arbitrary paradigm is that it is a reflection of what actually occurs in the brain. This is sophistry. For example, saying the “serotonin system is relevant in mood disorders” is empty because it lacks context. Does it mean that no other system is relevant? Or that it is the most relevant? Or even necessary? Sufficient? What about Wellbutrin? Are you saying psychotherapy alters serotonin? (And not dopamine? Etc)

Why There Can Be No Progress in Biological Psychiatry Using The Current Paradigm

While there is a science of the brain, there is no science of thought. Another way to think of the problem is to question the nominalism of the field. Psychiatry talks about things like mood, emotion, depression. But are those actually real and distinct things? Is there some signal pattern in the brain that is mood that is wholly unrelated to the signal pattern for a thought? Just because something has a name, that does not mean that the thing exists in fact. These may be simply convenient fictions.

To say that because it is known that chemicals can alter what is called mood therefore proves the existence of mood is not satisfactory. The death of a loved one will alter mood as well, often dramatically and to an extreme. How is this possible, if no chemicals are introduced into the system? In fact, nothing new is introduced into the brain except, depending on the paradigm, a) information or b) energy (i.e. converted sound waves.) How does the introduction of a new piece of information trigger an alteration of mood? Why would one thought "My family is dead" trigger the release of some chemicals, but another thought "My family is alive" not trigger that same release? And how does a non-physical entity like a thought trigger a physical reaction? Are thoughts even discrete? Or does the brain operate on a flow of thoughts? Digital or analog?

Linear regression and statistics cannot address all of the problems of chemistry and mood; for one reason because it does not account for thoughts that a patient cannot have while on the drug-- but it is reasonable to think there are some. In other words, if there is some set of thoughts A, a subset B of which are negative and a subset C of which are positive, then is the introduction of a chemical in the brain able to block only subset C, or does it simply block set A? The reverse is potentially applicable: that there are certain thoughts that can be had only when on a drug. (This is obviously evident in the case of perceptions that can be had only on a drug.)

The simple analogy to computer hardware and software illustrates the difficulty in psychiatry. One can understand all of the hardware of a computer system, but this will not explain if or how the computer can run word processing software, video games, instant message or be susceptible to a virus, or make any predictions about the behavior of this software in the real world (for example, no computer technician could predict the writing of this paper, nor, by changing the hardware, alter the content of this paper.) Hardware is finite, but software is infinite, or as infinite as is thought. Without understanding the mechanism of thought, or at least how thoughts or states can affect mood, then a pharmacology of the brain will simply tread water with no progress towards either treatment or diagnosis. One cannot permanently alter mood without at least simultaneously altering thought. In a sense, there is more logic to the psychological approach, or at least in conjunction with medications, because if software (thought) is the problem, more software is the cure. While the applicability of “therapy” in schizophrenia may be debatable, there is no reason as yet to decide that any other approach is applicable to mood disorders such as depression or anxiety, and absolutely no evidentiary reason to assume that pharmacology is the superior approach. It is not necessarily faster nor more consistently reliable.

Myths of psychiatry such as those described above are not isolated examples of poor practice or lack of knowledge, but are the unavoidable manifestations of an artificial paradigm which is arbitrarily derived from unproven assumptions, justified by inappropriate logic. They often lead to ineffective, dangerous, and very expensive treatment. Psychiatry must be more vigilant about its own data. It is necessary to avoid laziness in our education and understand from where comes our knowledge. There are daily diatribes against the influence of pharmaceutical companies; but the effect of pens and detailing is surely much smaller than the effect of misunderstood data, poorly researched axioms, and signs run amok. Psychiatry will not survive as a medical subspecialty if it continues along this path. It will lose its dignity, and worse, it will become irrelevant.

“It would be good to conclude by recommending a short book, What Is Science?, that does things the right way. It takes a robustly objective view of the relation of evidence to conclusion, explains what laws of nature are, briefly shows how measurement, data, statistics, and mathematical models work in science, states which parts of science are well-established and which not, illustrates with engaging episodes in the history of science, and ends with some colorful rudenesses on postmodernist solecisms concerning science. Unfortunately, it does not exist.”[iii]

---

[i] Eco, Umberto (1998). Serendipities: Language and lunacy (William Weaver, Trans.). New York, NY: Harvest.

[ii] Hirschfeld RM, Lewis L, Vornik LA. Perceptions and impact of bipolar disorder: how far have we really come? Results of the national depressive and manic-depressive association 2000 survey of individuals with bipolar disorder. J Clin Psychiatry. 2003 Feb;64(2):161-74.

[iii] Franklin, James. Thomas Kuhn’s irrationalism. The New Criterion Vol. 18, No. 10, June 2000

April 11, 2007

The Trouble With Psychiatry-- "Not Even Wrong"

I recently read Martin Gardner's review in the New Criterion of Lee Smolin's The Trouble With Physics, and Peter Woit's Not Even Wrong: The Failure of String Theory.

I am almost finished Smolin's book, but I wanted to make a comment about Gardner's piece. Writing in the New Criterion, he should have appreciated a wider view of the books, that they speak to more than physics. They're just as much about psychiatry.

Because I suck at writing, I'm a much better speaker, I often have difficulty getting my ideas-- which are already baffling-- across to people indoctrinated in psychiatric mythology. But Gardner does a great job, so I'm simply going to quote him verbatim. I don't think he'll mind. Just substitute the word "psychiatry" anywhere you see "physics."

He sees string theory as not a theory—only a set of curious conjectures in search of a theory. True, it has great explanatory power, but a viable theory must have more than that. It must make predictions which can be falsified or confirmed.

Consider that the terms "borderline" and "narcissist" which have supposedly no predictive power, yet give you more information than the epidemiologically valid and "reliable" diagnosis "bipolar."

In a chapter on sociology, Smolin introduces the concept of “groupthink”—the tendency of groups to share an ideology. This creates a cultlike atmosphere in which those who disagree with the ideology are considered ignoramuses or fools. Most physicists tied up in the string mania, Smolin believes, have become groupthinkers, blind to the possibility that they have squandered time and energy on bizarre speculations that are leading nowhere.

The last part is key-- who among psychiatrists are willing to say, "holy crap, we just made this all up as we went along!"

The other book Gardner reviews is Not Even Wrong, the title coming from Wolfgang Pauli's quote that a theory was so ridiculously unscientific that it was "not even wrong." Quoth Gardner: "By this he meant it was so flimsy it couldn’t be confirmed or falsified."

Gardner quotes another writer (Glashow), in an indictment of why the masturbatory nature of psychiatry is detrimental to its own practitioners. Replace "string theory" with "bipolar model:"

Until string people can interpret perceived properties of the real world they simply are not doing physics. Should they be paid by universities and be permitted to pervert impressionable students? Will young Ph.D’s, whose expertise is limited to superstring theory, be employable if, and when, the string snaps?

And closing with a quote from the physicist Gerard ’t Hooft:

Actually, I would not even be prepared to call string theory a “theory” rather a “model” or not even that: just a hunch. After all, a theory should come together with instructions on how to deal with it to identify the things one wishes to describe, in our case the elementary particles, and one should, at least in principle, be able to formulate the rules for calculating the properties of these particles, and how to make new predictions for them. Imagine that I give you a chair, while explaining that the legs are still missing, and that the seat, back and armrest will perhaps be delivered soon; whatever I did give you, can I still call it a chair?

I wish I could write better, more clearly. I can't. But at least read Gardner and Smolin-- if the problem of loose science and groupthink occurs in the hardest of the sciences, physics, it most certainly occurs in psychiatry. The first step in recovery, of course, is admitting you have a problem. We may never get past this step. The second step is to stop pretending you're something you're not, stop trying to make judgments and pronouncements about which you know nothing: Wovon man nicht sprechen kann, daruber muss man schweigen.