Saturday, November 14, 2015

Reductionism Is Not A Dirty Word...

A recent opinion piece in the New York Times, by George Makari, MD has me shaking my head.  The thesis was that a recent headline grabbing story (what's wrong with that criteria?) on the effects of comprehensive treatment of psychosis as opposed to treatment as usual surprised many and highlighted the problem with reductionism.  He bemoans the fact that the reaction to the story was one of surprise.  He doesn't specify who was surprised.  I certainly was not surprised.  I attended recent meeting and somebody in the audience asked Daniel Weinberger if he was surprised.  His response: "They spent $15 million dollars showing that good treatment is better than bad treatment."    He certainly was not surprised.  I have not heard about Eric Kandel's response, but based on his 1979 paper on plasticity and what happens in psychotherapy - I doubt that he would be surprised.  The exact population of who might be surprised by these findings seems poorly defined at this point in time but I doubt that it included any psychiatrists.

Speaking for myself, I will elaborate on why I was not be surprised.  At one point, I was the Medical Director of a community support program of a group of about 100 outpatients in the State of Wisconsin.  According to the state statutes, access to the program depended on diagnosis and degree of psychiatric disability.  You could only apply if you had a diagnosis of Bipolar Disorder, Major Depression, Schizophrenia,  or Borderline Personality Disorder had significant associated disability or were at high risk for hospitalization.  The clinical goal of the program was to reduce hospitalizations, maintain independent living, and facilitate employment.  The program was staffed by a psychologist, 2 social workers, three nurses and me.  When I arrived, one of the early dynamics was to frame problems in terms of medication needs.  That translated to increasing the dose of a medication (typically an antidepressant or antipsychotic) in crisis situations or other emotional crises.  The patients in the program had chronic problems and symptoms that did not necessarily respond to medication.  One of my first steps was to start to discuss problems and solutions with the patients.  I met with all of the patients and did supportive psychotherapy when possible.  We had team meetings every morning and problem solved around the needs of the patients in the community, how to solve any crises, and how to approach people in ways other than medications.  I tracked the total dose of antipsychotic medication and days of hospitalization as outcome measures.  At the end of three years, the days in hospital had gone down from about 14 days per person to less than 1, and the total dose of antipsychotic medication had gone down a total of 600 mg chlorpromazine equivalents.

My point is obviously that comprehensive care of patients with severe problems results in improved outcomes.  In this case lower doses of medications were used and the patients spent less time in the hospital and more time at home.  My orientation and ability to implement such a program was not an accident.  I was trained by Len Stein, MD at the University of Wisconsin.  Dr. Stein was a pioneer in the area defined as community psychiatry.  He was motivated by realizing that once people were in large state hospitals - it was very easy to warehouse them in overcrowded conditions.  Nobody seems to recognize it but overcrowding and suboptimal conditions were the state hospital equivalent of managed care rationing.  Once your state hospital is on the spreadsheet of a state bean counter with no accountability to patients or their families rationing and fewer and fewer resources are the order of the day.   In a community psychiatry seminar, Dr. Stein projected a slide of a gymnasium-sized room populated by male patients with hundreds of cots aligned edge to edge.  There was no room to walk between the cots.  That was his motivation for moving people out of these state facilities and into their own housing.  When I trained, there were three programs with independent living and quality of life as the primary goals and the staff involved in the programs was very good at it.  My effort just extended that skill set.  Contrary to the "surprising" results of the quoted study - I did the same thing back in 1986!

If it is true that we have known for 30 years that comprehensive care for psychiatric disorders trumps "treatment as usual" what is all of the rhetoric about?  Dr. Makari seems to want to make this into a mind-brain argument.  In other words, the biopsychosocial approach and the uncertain effect it has on the mind as opposed to a brain based approach that looks at specific mechanisms of action and seems to be focused on psychopharmacology.  He points out for example that the highlighted study would possible not qualify for current NIMH funding unless it looked at specific brain mechanisms.  He throws around the word "reductionism".  Anytime reductionistic or reductionism is used rhetorically in the same sentence with psychiatry it is pejorative.  My old psychoanalytic teacher would refer to anyone who talked about brain biology as a "dial twister".  The implication is that the reductionists are somewhat simple minded largely because they cannot accept the uncertainty of dealing with an organ that has poorly defined inputs and outputs.  Kind of a double whammy of rhetoric - you are a unsophisticated reductionist and you really can't see the big picture.  Are things really that simple?  Are these arguments accurate?  Are there problems with equating reductionism with "bad".

Of course there are major problems.  The first is the statement that inherent to the proposition that mental illness is a brain disease is "the implication that psychological and social events somehow are not also brain events."  This is a serious misreading of the definition of plasticity or experience dependent changes in the brain.  When I give my neurobiology of the brain lectures. I use Kandel's original New England Journal of Medicine article that discusses brain changes in a patient and a therapist conducting psychotherapy and how those changes are associated with brain plasticity.  I give further examples - weightlifting,  playing the violin, and how the typical stream of consciousness is profoundly altered by drug addiction.  There is no neuroscientist or biological psychiatrist I know who would suggest that psychological and social events are not brain events and there are numerous experimental paradigms that look specifically at how these events occur in the brains of animals.

The second aspect of Dr. Makari's argument has to do with reductionism.  His specific comment is:

"With luck, studies like Dr. Kane’s, which undermine these suppositions, will help move us away from such narrow thinking and embolden the substantial community within psychiatry that has never accepted such reductionism."

The suppositions in this case are that mental illness is a brain disease and that social or psychological events have no brain representation.  The argument is based on that false premise.  But further the use of the term "reductionism" is instructive here as previously noted.  By definition reductionism applies to many proposed etiologies of psychiatric disorders.  Those etiologies can be studied at a molecular level or at a higher level.  Schaffer (2) says that a model is reductive if it "employs standard biochemical and molecular entities to account for psychiatric symptoms and disorders".  Non-reductive models discuss "causal connections at higher levels of aggregation."  He illustrates these definitions by looking at Kendler's non-reductive account of major depression.  Kendler has used path analysis to look at clinical variables relevant to psychiatric disorders and although I do not have access to the one used in the book, here is a typical example.  The model looks at life stages, familial factors and psychological factors and all are higher levels of aggregation than molecular mechanisms.  At the reductive side of things he examines Harrison and Weinberger's proposed genetic susceptibility genes for schizophrenia.  At the time the book was written the author limited the discussion to 5 genes.  He also looked at the continuum of psychiatric genetic models ranging from basic and advanced genetic epidemiology being non-reductive, gene finding partially reductive, and molecular genetics fully reductive.  It seems perfectly logical to me that the study of brain biology proceeds in the same way that the biology of all living organisms proceeds.  The difference is that we are studying an infinitely more plastic organ with significant computational power.  There is clearly a lot of phenotypic heterogeneity that is unexplained in psychiatric diagnostic categories.  It is highly unlikely that refining diagnostic descriptors or applying clinical methods will lead to any significant change in the diagnostic or treatment process.  I don't understand the reluctance to go after more specific mechanisms or treatments.

The idea that a molecular or clinical focus in psychiatry is the problem with psychiatric services is also misleading.  As I hoped to point out by my mental health center example, psychiatrists know all about comprehensive care but they are rarely able to provide it.  They have known about how to provide it for decades.  State asylums became overcrowded and not therapeutic due to the financial management of the system by state governments.  The bean counters have moved out of the asylum and they are now integrated at every level in the health care system.  They all have a very strong bias against the comprehensive treatment of mental illness.  They insist that patients with severe psychiatric problems do not get comprehensive evaluations, that they are discharged before they have been adequately treated, and that any associated addictions are poorly treated.  They do not have the same biases against people hospitalized for medical or surgical illnesses.  They have in effect, moved the poorly run, overcrowded asylum model into the general health care system.  Any comprehensive care for severe mental disorders in such a system is an advertising phenomenon rather than reality.

The reductionism argument is good for New York Times opinion pieces.  It may sell a few more papers or get a few more clicks online.  Unfortunately it perpetuates an old pattern of blaming people and psychiatrists in particular for the shortcomings of a non-system of mental health care in this country that is set up to favor large health care businesses.  You can blame psychiatrists all you want for that - but until people realize that the real problems are the product of business and politics - and not the scientific interests of psychiatrists - nothing will change.

George Dawson, MD, DFAPA


1.   George Makari.  Psychiatry’s Mind-Brain Problem.  New York Times.  November 11, 2015.

2.  Scaffner KF.  Etiological Models in Psychiatry - Reductive and Nonreductive Approaches in  Philosophical Issues in Psychiatry.  Kenneth Kendler, Josef Parnas (Eds), The Johns Hopkins University Press, Baltimore, 2008:  pp 48-98.


Image is Microscope 1 by Bill3t Hughes on Flickr.  Reposted as noncommercial via Creative Commons License on 11/14/2015.  The original work is not modified.

No comments:

Post a Comment