Edward Shorter's Comment
Barry Blackwell: The Baby and The Bath Water
Barry Blackwell's thoughtful and learned comment immediately invites the response: (1) What is baby and what is bathwater? And (2) How do we tell them apart?
On (1):
Most of the field probably agrees on the following judgments: psychoanalysis (throw); neurotransmitters (keep); insulin coma therapy (throw); mood stabilizers (keep). These are not really scientificjudgments but cultural opinions based on optics. We cling to neurotransmitters (despite 50 years of failure in drug discovery), because neurotransmitters have made a good marketing trope for the pharmaceutical industry. We throw out psychoanalysis because we find biological treatments more effective, even though the many illnesses caused by stress and misery do not respond to biology; we discard insulin coma therapy because it looks awful (and that nasogastric tube at the end, please!); and we like mood stabilizers because we buy into a theory that a special kind of illness causes mood "instability" and that this illness requires special treatments. All of these judgments are questionable, at least. So, throw vs keep, let's proceed with caution here.
On (2): How do we tell baby from bathwater? How do we decide what to throw and what to keep?
Barry Blackwell rightfully casts doubt upon the usefulness of randomly controlled trials: In theory, they should be a gold standard of evidence; in practice, the influence of the pharmaceutical industry has hopelessly corrupted the integrity of many trial reports -- and therewith the integrity of much of the literature -- and it would be rash indeed to base one's judgment of whether SSRIs represent important drugs for depression on the basis of the trial literature. Here, as Barry Blackwell and others have shown, the literature has been heavily influenced by senior clinicians who have really cashed out in alliance with Pharma (and are now disparagingly referred to as KOLs)
So, how do we sort out baby and bathwater? Here clinical science comes to our aid. Experienced clinicians often have a gut feeling for what works and what doesn't, despite what "the literature" says. Take, for example, the issue: Do the SSRI-style antidepressants evoke suicidality in a subset of depressive patients? The epidemiological literature has been unable to confirm a connection. Yet many clinicians have seen in their practices patients becoming suicidal or homicidal after initiating a course of SSRI treatment. Here there is no question: the temporal relationship is powerful and immediate; It happens to their own patients under their own eyes!
If we aggregate these impressions, what we have is clinical science. It is a science not bolstered by epidemiology, because the relationships are submerged in the great mass of numbers. It is a science not bolstered by genetics, because, when you get right down to it, what diseases are reliably caused by genetics? Family tree, for sure, and in our schizophrenic and melancholic patients there are illnesses all over the family tree. But genetics? So far, little has panned out. "Schizophrenia" does not breed true. Yet we know, in our heart of hearts, that these pathological affinities exist in family trees, and it is not a stretch to call this certainty clinical science. The field needs to give this more thought.
One final comment: Barry Blackwell trashes insulin coma therapy as bathwater, a dangerous antique that, thank God, we are rid of. I think a second opinion is possible on ICT. It definitely helped many patients, and not all these patients would have been relieved by neuroleptics. There was something there, something about the effect of insulin on the brain that you have to see to know that it exists, but after you see it, you know it.
October 26, 2017