Thursday, April 02, 2009

In Defense of Psychiatric Diagnoses and Treatments

In Defense of Psychiatric Diagnoses and Treatments

The Boston Globe is out today with a story about conflicts of interest involving American Psychiatric Association guidelines for treating depression, mania, and schizophrenia – disorders for which the market in medications is said to be $25 billion annually. Conflicts abound, it seems, and the consequent implication is that industry influence has led to guidelines that “focus heavily on medications and give relatively little attention to nondrug treatments.”

The lead author of a forthcoming analysis, Lisa Cosgrove from the University of Massachusetts is quoted in the Globe as saying “…the lack of biological tests for mental disorders renders psychiatry especially vulnerable to industry influence." Maybe so, but it’s not just psychiatry. And this new study should not be taken to invalidate psychiatric diagnoses or to undermine the efficacy of psychiatric treatments.

The existence of disease is not predicated on having a biological test. It’s nice when we do have one, but there are many disorders throughout medicine, not just in psychiatry, for which there is no conclusively defining biological test. Think migraine. Think multiple sclerosis. Think chronic pain.

There is a distinguished tradition of clinical research, going back to the late 1950s, wherein disorders like depression, mania, and schizophrenia were studied for their genetic profiles, their clinically meaningful subtypes, their biological correlates, and their responses to treatments. This bottom-up approach yielded useful theories, which in turn led to better theories and to new treatments for depression, mania, and schizophrenia. Like any other branch of science, clinical science has a self-correcting function that distinguishes it from pseudo-science and propaganda. So, we welcome new theories even though most will not survive, and we welcome new diagnostic proposals and we welcome potential new treatments. All are subject to the test of independent confirmation, and nobody is surprised when many fall away.

Since around 1990 that tradition has been corrupted by Pharma and Pharma’s money. There is no shortage of examples regularly in the news. Academic key opinion leaders lost their ethical compass and, with or without a figleaf, devolved into spokespersons for industry. One is reminded of Julien Benda’s term la trahison des clercs. Instead of moving on from undistinguished and problematic new drugs, key opinion leaders work with the marketing departments of Pharma to promote and defend the brand. Remember the PsychNetters? Of course, psychiatry is not the only medical specialty corrupted in this way. Stents, anybody? Orthopedic devices, anybody?

Lost in the marketeering and propaganda since around 1990 is the discipline of clinical therapeutics. As one commentator put it, if there is no way to confirm biologically the existence of disease, then “…we cannot confirm with much confidence that this or that treatment treats mental disease.” It actually is possible to identify effective treatments for mental disease. We did it with lithium. We did it with the early antidepressant drugs. And we did it with the early antipsychotic drugs. There was room for improvement in all of these, but make no mistake – they were dramatic positive developments. Most of the work in the past 20 years, however, put marketing first and clinical therapeutics second. Humbug aside, the goal was not to improve the management of depression or mania or schizophrenia but to jump whatever bar the FDA was setting in order to get a product on the market. Indeed, the really hard clinical research issues were studiously avoided. Along the way, the clinical research enterprise became dominated by academic and for-profit contract research organizations in which, to put it gently, standards of recruitment of patients apparently suffered, so that response rates in placebo-treated patients skyrocketed. As a result, proving that a drug does work became harder, and the overall drug advantage over placebo in treating depression came into serious question. In effect, the corruption of clinical research standards has brought us to an epistemologic quagmire about the efficacy of today’s most popular antidepressant drugs.

As for the charge that psychiatrists and primary care physicians just throw drugs at mental health problems, that is nothing new. It is a reflection of the professional pragmatism that permeates medical practice. Think back to the 1950s when the new wonder drugs called antibiotics were handed out like Pez, mostly for people who did not really need them and who would not benefit from them. The working principle was misguided simplicity and pragmatism, which minimized and discounted the risks. Even today the over-use of antibiotics remains a public health concern. Just as in the 1950s patients with nonspecific upper respiratory symptoms were “given the benefit of the doubt” and were prescribed an antibiotic on the reasoning that it might help them while probably not hurting them, so today patients with nondescript depressive or anxiety symptoms are prescribed an SSRI drug on the same reasoning. The cost of this approach in wasted money, in unnecessary side effects, and in unproductive clinical management is non-trivial.

So, the issues are more nuanced than today’s Boston Globe story would suggest. For good and for ill, psychiatry is not so very different from the rest of medicine. Corruption and compromise respect no subspecialty boundaries. We can have reliable and valid knowledge of disease without a biological test. We can show efficacy of treatments for disorders that lack a biological test. But when simplistic pragmatism and cutting corners dominate in therapeutics and in clinical trials, then we are in trouble.


margaret soltan said...

Barney: Thanks for that corrective to my post, which does, now that I look at it, makes more than it should of the absence of biological markers.

Marilyn Mann said...

Can you elaborate on this part:

"standards of recruitment of patients apparently suffered, so that response rates in placebo-treated patients skyrocketed."



Anonymous said...

Great post.

Functional MRIs have shown various structural brain abnormalities with various mental illnesses, so such diagnostic testing has the ability to confirm at least some mental illnesses.

Clearly, the U.S. is an over-medicated society, with psychotropic drugs in particular. And with children taking these drugs, that's were my greatest fear presents itself regarding these types of drugs.

Anonymous said...

In a March 30 AP story: Study urges depression test for teens, Lindsey Tanner highlights the standards set by the US Preventive Services Task Force of a once a year doctor visit and a simple questionnaire to determine if a teen is clinically depressed.

While of course, (sarcasm) the panel does not want drug treatment alone, the reality is with a shortage of child specialist, kids will be given happy pills because teenagers are not suppose to be moody.

Once again I see children being pushed into what could easily turn into a lifetime of pharma support. Get them started young, eliminate the stigma of taking a medication, and they are well on their way to taking whatever pharma produces. Helicopter parents want only what is the best for their children and if a drug treatment can give them an edge so be it. The long term effects are discounted in the hope of a short term gain.

Children have just become pawns in the latest pharma game. Various psychological problems are being touted as a means to generate revenue for the drug companies.

But hey, I am just a dumb business guy looking in from the outside.

Steve Lucas

hankm said...

Steve Lucas is absolutely right.

I should add that while much of what Dr. Carroll is saying is also correct, I don't follow the point that because there is no scientific testing for migraines or chronic pain either, we can therefore also legitimize calling psychiatric disorders "medical."

Like psychiatric disorders, these are symptoms of an underlying condition not yet found. Symptoms are not disease.

What's worse about psychiatric disease is that very powerful psychotropics are used to treat these vague, overlapping and very subjective psychiatric diagnoses that are created by committee consensus.

And because everybody has problems, almost anyone can be sold a psychiatric drug.

To expand on Steve's point, it is important to remember that psychotropics are designed to subdue behavior considered undesirable.

In this politically correct world of today, the ramifications of mandatory teen depression screening should scare anyone.

Bernard Carroll said...

Margaret: it is I who thanks you for getting me to think about the Boston Globe story yesterday.

Marilyn: in the early days of antidepressant clinical trials, the drug-placebo difference in response rates was around 30% to 40%. We saw response rates of 60% to 70% with active drug and 20% to 35% with placebo. So, the Number Needed to Treat to response (NNT) was around 3. Nowadays it is not uncommon to see placebo response rates north of 50% and drug response rates around 65%. So in contemporary trials the NNT is more like 8-10. The UK regulator known as NICE (National Institute for Clinical Excellence) regards a NNT of 10 as only marginally clinically meaningful in the acute treatment context. It means you would have to treat 10 patients with a drug in order to obtain 1 response that would not have occurred anyway with placebo.

The nadir for response to placebo was recorded in some of the Corcept Therapeutics trials of RU486 for psychotic depression. Rates as high as 80%-85% were observed.

hankm: I was making the opposite point, viz., just because there is no biological test for a disease, that does not make it any less of a disease. I think you are quite right to remind us that symptoms are not disease. That issue underlies many of the scientific reservations about DSM-III and DSM-IV.

Anonymous said...

As Dr. Carroll noted in his follow-up, the problem is with the overly broad definitions of psychiatric disorders in the DSM. Having these imprecise definitions is advantageous for the pharmaceutical industry. It is more profitable for them to sell ten million pills with 10% efficacy for a widespread disorder than 1 million pills with 100% efficacy for a narrowly defined one. As the psychiatric research funding comes mostly from the industry, this situation is unlikely to change. The government should take a larger role in financing psychiatry research as it did in the 1960s.

Pain vicodin said...

Excellent item, really good people who truly believe in this solution, mainly because it supports that medicine evolved satisfactorily and helps many people with this information, since many can not visit a doctor for lack of money that the economy is much less due to across the world.

Doug Bremner said...

Here here, good post. Psychiatric diagnoses are based on symptoms reported by patients. There are not and will not be diagnostic tests for these disorders because even though 66% of patients will be captured by a certain biological marker there will always be someone outside of this group who will have the symptoms but not meet the criteria of the biological marker.

Anonymous said...

Okay fine. So your solution is?
Seriously, I have a waiting room of depressed patients. What do I change TODAY to fix this (perceived) problem?
And I say perceived because I think psychiatry is doing the best it can with what it has. I'm lucky enough to work within a health care environment where I can easily refer to psychologists for individual therapy, so there's no pressure to prescribe because it's the only option.
But, in the many areas where psychotherapy resources are scarce, what are we supposed to DO now? Forget the academic mental masturbation for a minute, and try to find a real world pragmatic solution that can be employed today? Most psychiatrists simply can't start doing hour-long therapy with everyone. Besides going out of business, that would only make them (as a resource) MORE scarce (there are only so many shrink-hours, right?) and though care for a minority might improve, many more would then simply have no care (or, at least, an abysmally long wait).

Cheers, all. I enjoy this debate. Our profession needs it.

Bernard Carroll said...

Teufelhunden (Devil Dogs?): what you say underlines my expression ‘epistemologic quagmire.’ Over the past 30 years we have thrown away hard won insights into which kinds of depressed patients have meaningfully better responses to drugs than to other treatments. Working with the nominalist convention called generic major depression doesn’t help you clinicians to maximize therapeutic skill. The parallel problem is that the antidepressant drugs introduced since 1985 examined only generic major depression for establishing efficacy. That is why the efficacy data are underwhelming, as the recent debates illustrate.

A further complication in the past 10 years is the cynical promotion of ‘treatment resistant depression’ as a vehicle for marketing antipsychotic drugs to nonpsychotic depressed patients. Here again, the efficacy data are underwhelming, while the toxicity data are alarming.

So, in clinical practice, we see on the one hand many depressed patients receiving psychotropic drugs when they have a low probability of meaningful specific benefit. On the other hand we see plenty of patients who need an effective antidepressant drug being prescribed agents that don’t work very well. I learned this business in the 1960s using tricyclic antidepressants, monoamine oxidase inhibitors, and lithium. No drugs with comparable clinical impact have appeared since the 1960s, although heavy marketing has steered clinicians away from these early agents. Hint: they still work.

So, what do you do? In your space you have no choice but to be pragmatic. It is up to the academic research community to get its head out of the wrong place and to reframe the issues so that all those research dollars can actually make a difference.

Discover and Recover said...

Antipsychotics (older ones) were successful?

Dream on -

The majority of people diagnosed with "schizophrenia" recovered during the days of Dorothy Dix... The Moral Era of psychiatry (well over 100 years ago)... long before the pharmacological approach took root.

The past 50 years of "evidence-based" psycho-pharmacology has been an abysmal failure... in EVERY sense... With major depressive disorder, bipolar, schizophrenia... all of it -

There are many non-drug options that work and work well. It's time we started using them, and closed the chapter on pharmacological psychiatry... once-and-for all!

Duane Sherry, M.S.