The Debate Over DSM-5: A Step Backward

You’ve been one of the most outspoken critics of DSM-5, but your ultimate target is really something you call diagnostic inflation. What’s diagnostic inflation, and why should we all be concerned about it?

Allen Frances: Putting aside all the problems with DSM-5, the simple fact is that psychiatric diagnosis has become way too loose. Today, 25 percent of Americans meet the criteria for a diagnosis in any given year. Twenty percent of us take psychotropic medication; that’s one in five people. We now have more deaths in emergency rooms from prescription drugs than from street drugs. Pill popping is rampant, along with all the unnecessary side effects of drugs. The new DSM will just make this worse.

How did psychiatry get itself into that situation?

Frances: Obviously there are any number of causes, but a key turning point goes back over 30 years to the enormous popularity of DSM-III. Before that, psychiatric diagnosis was fairly irrelevant, but once DSM-III became such a bestseller year after year, it attained tremendous influence beyond the therapy room, affecting decisions regarding school placements, disability decisions, courtroom verdicts, who could adopt a child, even who could fly a plane. All sorts of things in society are now determined by diagnosis.

A second major factor was the rise of Prozac in the 1980s and the tremendous profitability of psychiatric drugs. Drug companies, with their expensive marketing efforts, began to have more and more influence in shaping the practice habits of doctors, and later of the public, by convincing people that they were sick.

Third was the insurance company requirement that a doctor give a diagnosis if he wanted to get paid for a patient visit. As a result, doctors began seeing people for only about seven minutes before making a diagnosis. A psychiatric diagnosis is a critical moment in a person’s life, like getting married or buying a house. It shouldn’t be made after seven minutes.

A fourth factor was the drug companies’ market-driven realization that, while there were only about 40,000 psychiatrists, there are more than half a million primary care doctors. So they started pushing psychiatric diagnosis and the prescription of medication onto primary care. They began with the message that psychiatric diagnosis was really simple and the result of chemical imbalances. So increasingly, primary-care doctors began diagnosing in a few minutes and handing patients a prescription or, even better, a free sample. So we’ve gotten into a position where most of the psychiatric diagnosis is being done in just a few minutes, often by a primary-care doctor with little psychiatric training, and with tremendous encouragement by drug companies. This has been the perfect storm that’s led to loose diagnosis and excessive medication.

As the head of the DSM-IV Task Force back in 1994, how did you try to address these problems?

Frances: When I began working on DSM-IV, I was concerned that it would extend even further what was already an excessive diagnostic enterprise.

So we developed strong rules to try to contain the beast. To get a new proposal accepted into DSM-IV, there had to be a rigorous, three-stage process of literature review, data reanalysis, and field trials. And we told the experts that unless the data jumped up and grabbed us by the throat, we weren’t going to make changes.

By and large, we were successful. Of the 94 suggested new diagnoses that were offered, we included only two. But those jumped up, and they did ultimately bite us. The new diagnoses of bipolar II and of Asperger’s became epidemics of their own. Since DSM-IV, the rate of bipolar disorder has gone up by 40 percent, and autism has also gone up by about 40 percent. Even with all our caution, we helped create these epidemics.

The lesson here is that even if you work incredibly hard to limit diagnostic inflation, the power of drug companies to advertise freely can create demand, even if the system has discouraged it. In fact, I’ve been so involved in critiquing DSM-5, not because I’m against psychiatry—far from it!—but because I think accurate diagnosis is so important. That’s why, in addition to diagnostic inflation, I’m equally concerned about the fact that we have one million psychiatric patients in jail because they haven’t received accurate diagnosis and enough treatment or a decent place to live. Only one-third of patients with severe depression today get to see a mental health practitioner. So I’m concerned with both ends of the misallocation—that we’re terrifically overdiagnosing people who don’t need it, and giving treatments to others that will only make them worse.

Do you believe that only people with medical training should prescribe psychiatric drugs?

Frances: The real problem is that most psychiatric drugs are being prescribed by primary-care doctors with very little training in understanding mental illness. I think anyone who’s well trained in diagnosis and understands the proper use of the medication, including psychotherapists, should be able to prescribe. Most of all, I’m concerned about anyone prescribing who does it too quickly and doesn’t understand the value of really knowing the patient, whether that prescriber is a psychiatrist or primary-care doctor or nurse practitioner.

So what are the most important things to consider in making a psychiatric diagnosis?

Frances: First off, the more severe and more characteristic the presentation, the easier the diagnosis. So we can be reliable in diagnosing disorders like schizophrenia, obsessive-compulsive disorder, and major depression. The five percent of the population that has a severe psychiatric disorder can be most accurately diagnosed. In those cases, the rule of thumb is to diagnose quickly and treat them immediately, because the sooner you provide treatment, the better their long-term prognosis.

But when a presentation is milder, often you’re seeing the patient on the worst day of that person’s life. Whatever happens in the next week, they’re usually going to be better. A month later, they may look very different. Making a diagnosis on that first visit may completely miss the point. So the single most important lesson is to take time. You can’t do it in seven minutes, and often you can’t do it in one meeting. Sometimes you can’t do it in a month. When in doubt, wait, watch, provide support, normalize the situation, see what resources there are from within the individual’s own resiliency. All sorts of simple things, including reducing stress, will often make a diagnosis and medication unnecessary.

What are the cases that are most often misdiagnosed?

Frances: We’ll start with kids. In a Canadian study that included over a million kids, the best predictor of attention deficit disorder (ADD) was whether you were born in December or born in January. Especially for boys, being the youngest kid in the class made you almost twice as likely to get the diagnosis of ADD. It’s absolutely crucial with kids to recognize that there are individual differences, developmental differences, family stresses, school stresses, environmental problems, all of which can create behaviors that will be mistaken for psychiatric illness. This is especially true today with the reduction of physical education in schools and the increase in class sizes. Instead of inflicting a medical diagnosis on these kids and treating them with stimulants when they don’t need them, we should be spending money on having smaller classes and more physical education.

We also jump too quickly to a psychiatric explanation with the elderly. There’s a remarkably a high percentage of people in nursing homes who are getting antipsychotic medications, which actually reduce life expectancy. Because the elderly take an average of six medications, it’s difficult to evaluate the degree to which psychiatric symptoms are due to medication side effects. They don’t clear medications and don’t metabolize them nearly as well as younger people do. These medicines often have interactions, and it’s not uncommon for psychiatric symptoms to be caused by the medications that are being given to prevent them. Furthermore, the symptoms may have more to do with the staffing and nursing team than any other factor. If someone gets agitated, for instance, it could be that his bladder’s full, and he has no way of getting to the bathroom.

So you have to understand the environmental factors that are involved before jumping to a diagnosis. The less classical the presentation, the younger or the older the patient, the likelier you are to get overdiagnosis and mistreatment with certain medication.

What changes in the insurance industry would you like to see that would help address the problem you’re describing?

Frances: I think the insurance industry has made a basic error by requiring a diagnosis for reimbursement on the first visit. It would be much better if there were a period of time in which the physician would get paid for evaluating patients without diagnosing them. Of course, the insurance companies don’t want doctors seeing patients for nonmedical problems. While that may seem like a potentially wasteful procedure, the current system is costlier in the long run. Once a person gets a diagnosis, it tends to stick for life. It would be a lot cheaper in the long run to support more detailed evaluations than to have premature diagnoses that often lead to a lifetime of unnecessary treatment.

The placebo response rate for most mild problems in life is 50 percent or more. In other words, if people see a doctor on a particularly bad day of their lives and the doctor says they’re depressed, or have an anxiety disorder, and starts them on a medication, half of those people will have a positive response to that visit—but we don’t know what in particular got them better. Although patients may think the pill made them better, a lifetime of treatment can easily be based on a placebo effect.

You’re a big opponent of drug company advertising. But this is America! Shouldn’t we let the free market do its thing?

Frances: I believe that the single biggest improvement we can make in the mental health system is to place more controls on drug company advertising. Uniformly, when the diagnostic system gets loosened, the major economic effect is a dramatic increase in the sales of medications. And with the dramatic increase in sales of medications comes a tightening on the psychotherapy benefit. So as the diagnostic system gets looser, it paradoxically makes it harder for people to get psychotherapy because so much money gets funneled into the drug companies.

In contrast, there’s no psychotherapy advertising industry. Psychotherapists are mom-and-pop operations. There’s no budget for advertising the value of psychotherapy. The drug companies are spending $70 billion advertising their products. In the battle for the airwaves, the drug companies are in absolute control, in spite of the fact that psychotherapy for mild to moderate conditions is just as effective as drug treatment, with longer effects. So it’s probably more cost effective.

But from the insurance companies’ perspective, isn’t it cheaper for patients to take pills than to go into psychotherapy?

Frances: No it’s not. It just seems more costly because there’s more upfront expense. A patient may need reimbursement for 12 therapy visits over three months. But if you step back and think about a lifetime of being on unnecessary medication, which could have been avoided if psychotherapy had been the initial treatment, the lifetime costs are much less to keep people away from medication.

The trouble is insurance companies usually have year-to-year contracts, so they’re not thinking about the trajectory over the lifetime of the individual. They’re thinking about how they can reduce costs in this quarter, and that leads to decisions that are neither economically smart for people nor good for them clinically.

Another prominent critic of DSM-5 has been Thomas Insel, the director of NIMH, who announced that they weren’t going to use it as the template for funding psychiatric research. What do you think of that decision?

Frances: I think it makes absolute sense for NIMH to try to develop simpler questions and, instead of studying heterogeneous DSM diagnoses, to find simpler dimensions for study. So I’m completely on board with the decision. But I was tremendously against the way it was announced, and the overall stress at NIMH, which is overpromising biological advances in psychiatric treatment. We’ve gone through 30 years of an amazing revolution in neuroscience and genetics. But so far, that’s contributed not one iota to caring for a patient. We don’t have a single biological test for mental illness, and our treatments are no better than they were 60 years ago.

The fact is that the brain is the most complicated thing in the universe. We shouldn’t expect quick hits in understanding something like schizophrenia. Schizophrenia won’t be one disorder: it’ll be hundreds of disorders, many different causes that have a common pathway. It’s going to be the task of lifetimes of researchers to tease all this out.

NIMH is promising quick results when we know from the past 30 years it can’t deliver them. Meanwhile, we have a million psychiatric patients in prison because they didn’t get adequate care in the community and don’t have a place to live. A million psychiatric patients were discharged from psychiatric beds in the last 50 years, and a million prison beds—the exact same number—were added for psychiatric patients. This is a shame. NIMH should be in the forefront of trying to help patients now, not overpromising for patients in the future. We need to do biological research, but we also need to be doing research into health services, social context, psychological factors—and NIMH has made itself into a brain institute.

Do you see any indication that the processes leading to diagnostic inflation are changing?

Frances: The drug companies and the health insurance industry have a lock on Washington, DC. But 25 years ago, Big Tobacco ruled the roost: they owned Washington. Who’d have thought they’d eventually be restrained in the way they have? The single thing that would most transform the public health of this country and save us billions of health dollars would be to control drug company marketing. Only in this country can they advertise to consumers. The freedom to distort the medical enterprise through advertising to consumers is unbelievable. The marketing to doctors is distorting the medical enterprise, and the American Journal of Medicine agrees with me.

The DSM system needs to be tightened. DSM-5 is an absolute step in the wrong direction. No one needs to use DSM-5 unless your institution requires it. I think we should take the DSM system out of the American Psychiatric Association (APA). It shouldn’t be that a small psychiatric association controls something that has such an enormous impact on so many societal decisions. You need people who have expertise far beyond psychiatry to determine the long-term consequences of these decisions. You can’t trust the experts in the field. Of course, you need their input, but you can’t have them calling the shots. I think the new drugs in psychiatry are much less dangerous than the new diagnoses. We spend a lot of time making sure the drugs are safe and effective, but no time making sure that new diagnoses are safe and effective.

We need something like the Food and Drug Administration for our field, a broader regulatory mechanism that’s not within the purview of one psychiatric association, particularly because the APA has tremendous financial conflicts of interest. They spent more than $25 million on DSM-5. (We spent $5 million on DSM-IV.) They expect to make hundreds of millions of dollars. The whole thing shouldn’t be a commercial venture. It should be a public trust. At the same time, we need to focus attention away from coming up with more diagnoses and on the whole mental health enterprise, and those people who are suffering greatly, the people who really need help but aren’t getting it.

Allen Frances is author of Saving Normal and leading DSM-5 critic