DR ALLEN FRANCES - ON 'FALSE EPIDEMICS' GENERATED BY DSM4 -'WE MADE MISTAKES WITH DSM4 WHICH HAVE HAD TERRIBLE CONSEQUENCES FOR CHILDREN.' - " DSM 5 IN DISTRESS" - DR ALLAN FRANCES BLOG - THE LEAD AUTHOR OF DSM4 CHANGES HIS VIEW OF THE DSM.MACHINE

LEAD EDITOR OF DSM4 CHANGES HIS MIND ABOUT IT'S VALIDITY.
"We made mistakes with DSM4 that have had terrible consequences for children."

Bipolar II Revisited-Always Take The Experts With A Grain Of Salt
Their suggestions often do not generalize well to everyday practice.
Published on April 26, 2010 by Allen J. Frances, M.D. in DSM5 in Distress
Nassir Ghaemi
This post is a response to DSM 5 and Bipolar Disorder: Science versus politics by Dr. Nassir Ghaemi

Dr. Nassir Ghaemi has posted a blog on Bipolar II Disorder that is a response to my previous blog. His arguments offer an instructive example of why we should be cautious before blindly following any expert's recommendations on how to change the diagnostic system. The people chosen to work on DSM are (like Dr Ghaemi) experts on one or another psychiatric diagnosis. To become an expert, you must devote yourself to research on the given diagnosis and develop clinical experience with it. Experts usually have considerable confidence in their views and state them assertively in a way that seems to close the argument.  

Of course, experts are absolutely essential to the creation of any DSM, but their recommendations  always need balancing from a broader perspective. It is in the nature of specialization that an expert's research and clinical experiences occur in a rarified, hothouse environment that does not always translate very well to the exigencies of  real world practice.

Experts are also all biased in the same direction - they always worry themselves greatly about false negatives and are relatively indifferent to false positives. In my work on three DSM's over twenty years, never once did an expert make a suggestion that would reduce the boundary of his pet disorder. In contrast , they very often  clamored for expansions that did not make sense in the larger scheme of things.  

Which brings us back to Dr  Ghaemi. He argues that the "science" of Bipolar II diagnosis has now established itself and provides a surprisingly clearcut answer - two days is the best duration requirement for a hypomanic episode (in order to avoid missing patients who fail to meet the higher current threshold of four days). He recognizes, but is relatively untroubled by, the current overtreatment with antipsychotics and mood stabilizers and is not much concerned that his suggestions would increase this problem even further.  He shrugs off my fears about the consequent dire  risks of weight gain and diabetes by characterizing them as somehow "political".

Reading Dr Ghaemi, it sounds like the answer to the duration question for hypomania is definite, neat, and settled. But of course it is not. Decision making about how to define things has to factor in many risk/ benefit complexities that he is blithely ignoring. It is fine for sideline observers to make suggestions based on their own interpretation of the research-that is part of the process. But final decisions for a widely used official nomenclature require a much more thorough and balanced risk/benefit analysis than the one undertaken by Dr Ghaemi.

When it comes to Bipolar II Disorder (or any other disorder in DSM), his dichotomy between "science and politics" is false at both its poles: 1-the "science" will always be much weaker than stated and subject to different interpretations, and; 2-it is not "politics" but rather clinical common sense and DO No Harm prudence to be concerned about the overuse of antipsychotic medications that can cause such life threatening complications.

First the science issue.  The "validators" in psychiatric diagnosis are never unanamous and certainly do not shout out one true answer. Choosing a duration or severity cutoff is always a somewhat arbitrary excercise in interpretation balancing sensitivity (the desire not to miss cases) with specificity (the desire not to overdiagnose and mislabel someone who does not have the disorder). The currently available validators for distingishing unipolar from bipolar disorder work fairly well at the extremes but are very uncertain guides at the extremely fuzzy boundary between them (and will remain so until we have biological tests and a much better understanding of pathogenesis).

Moreover, what information we do have has been mostly derived in specialized research settings with highly trained clinicians and highly selected patients. These  are very different than the settings where most bipolar diagnosis is made. In real life, patients often present with complicated clinical pictures that have to be evaluated in a brief period of time by someone who has not spent his career studying bipolar disorder. Substance abuse (which often disqualifies a patient from participating in a mood disorder study) is a huge and very difficult factor that has to be ruled out before determing that elevated mood or irritability constitute hypomania. And hypomania is easily confused with the return of normal mood in someone who has been in a sustained depression.

The bottom line is -be suspicious whenever an expert confidently asserts there is that there is a best "scientifically proven" cutoff - other thresholds are always equally plausible depending on the respective risks of under and overdiagnosis. Reducing cutoffs based on data in tightly controlled research studies may create a flood of misdiagnosis in the real world.

Next, we get back to whether my concerns about antipsychotic use are well characterized by the term "politics". This touches on an important issue that goes well beyond Dr Ghaemi's careless word usage. In discussions with people working on DSm5, I invariably bring up the fact that their proposals will increase false positives (especially when applied by less expert hands and under the general influence of pharma marketting) and that this will lead to unnecessary and often harmful medication treatment. On numerous occasions, I have heard a dismaying reply that echoes the tone of Dr Ghaemi's piece and his misuse of the word political -some variant of "My job is to follow the science. I can't predict how the proposal might ultimately be misused so that can't factor into my decisions. If problems arise later with the suggestion, let's then deal with them by education on how to properly diagnose and treat".

I have no patience for this naïve and reckless argument. Once the genie is out of the bottle, there is no putting it back. Diagnostic epidemics take on a life of their own that is far outside the control of DSM decision makers -especially when the momentum is fed by the drug companies.

It is crucial that everyone working on DSM5 feel accountable for all the possible misuses that may be encouraged by their suggestions. However well intended the proposal, they should also feel responsible for not having anticipated and prevented its unintended consequences. Falling back on the "science made me do it" simply won't do. The science is never robust or clearcut enough to provide cover or an excuse for any decision that can significantly harm peoples' lives.

By all means, let us use antipsychotics when they are needed for properly diagnosed Bipolar Disorder. But as a profession, we must also do our best to protect our patients from their premature and unnecessary prescription in ambiguous presentations. Pinning what often becomes a lifetime  diagnosis of Bipolar Disorder (and long term antipsychotic treatment) on two days of equivocal mood elevation or irritability makes absolutely no sense.  The stakes are high -obesity and diabetes dramatically reduce life expectancy.