Saturday, November 21, 2009

Behavioral medications and the failure of exclusive verifiability

As all of our outcomes are determined by risk, all of our decisions must be informed as strongly as possible by it. The medical community has a long and proud tradition of utilizing the basic tenets of the scientific method to provide confidence of both risk and reward in the application of modern medical practice, including the use of pharmaceuticals. This is not unwarranted; the presence of this tradition is the sole credit upon which we can place millions of lives that have been either improved or outright saved by the more informed application of medicine and medical practice.

I have been on behavioral medications intermittently for nine years and consistently for periods of up to five years. Like any US-sold medication, they were screened to FDA standards to ensure that side effects were identified and quantified to ensure that the users would enjoy benefits to more than offset any of these effects. Unfortunately, the complexity of the human body, our lack of complete understanding regarding its working, and the nature of statistical confidence conspire to create practical limits to the testing process. If you assume that a stated statistical confidence can make us reasonably “certain” of the presence or absence of a specific side effect, then you can also understand that the frequency of various side effects can make us only so “certain” of the presence of any side effect at all, as our testing, by its nature, must be exclusive. Ultimately, the end users of a medication often become the first legitimately comprehensive test group, and even then, such a statement is still technically a generalization. Such is the nature of statistical observance in science.

I feel very confident in regard to the following, which I unmistakably admit to be conjecture. The causes I suggest are guesses, and I am not educated in the field of neuroscience (a single read of my descriptive language will attest to this). The effects that I describe are very real and give me as much (and no more) confidence than any single case can give, which is perhaps not much in the plight of a scientific mind.

I believe simply that the behavioral medication I have taken has had three effects:

1) To directly alleviate the symptoms for which the medication exists; namely, depression and attention-deficit disorder. This is, of course, both expected and great.

2) To promote feedback loops in chemical brain activity that exacerbates the symptoms in the absence of the medication. The testing that the medication went through would statistically deny these effects; however, such a conclusion assumes that the testing period was sufficient to allow brain chemistry to reach a new equilibrium. I could name half a dozen mechanisms that would support a theory of long-duration disequilibrium and rebalancing, but it would do no more good than I can do by evoking the concept: the brain is far too complex in composition and process to play upon with reductionist theories of cause and effect.

3) To promote mental activity that rebalances neural pathways and firing patterns in an imperfect way, so that the end result is not to fix the undesirable behavioral problems but to produce some approximate common effects in a way that interferes with or fails to compliment other brain activities. Here, although the common pop-understanding is that “a behavioral tic is a neural firing pattern”, the truth is that the firing pattern is sufficiently complex that we can more productively think of it as a sequence and interplay of many interrelated firing sequences. Again, in a reductionist fashion, a simple thought might be that a firing pattern has been made to happen “too often”, as in, “repeat too fast”. This is easy to picture, and then we assume that the end result is that the behavioral tic recurs too often. However, imagine instead that our over-active firing pattern is one of the many firing sequences that comprise a much larger firing pattern. What will the effects be? More likely than direct over-activity, the effect will act to diversify the full pattern’s outcome. Poker hands will always be dealt in the same way from a deck organized in a certain sequence, until you begin randomizing the position of a single card throughout the deck on each deal. At this point, the outcome will not marginally diversify - in many cases, it will become completely unrecognizable.

I think that by this point, my lament would be predictable, and so I will resist crossing this discussion into the personal. But the idea of a blind spot in behavioral medicine, due to the enormously more complex workings of the brain than the balance of our physiology, is worrisome and finally unfortunate, and I offer no solutions.

Have the benefits of the medicine outweighed the costs? I don’t expect to ever know.

No comments: