A Treatment-Based Statistical Approach to Psychiatry

The DSM is a book that lists mental disorders, published by the American Psychiatric Association. It is also one hell of an ugly beast. In retrospect, it has always been bound to be one because of the conflict of two of its attributes:

1) It was never meant to be a book that explicitly defines different mental conditions. The original intent was for it to simply reduce the number of terms that psychiatrists use to make it easier for them to communicate with one another. As such, the DSM-IV carries the following warning:

“There is no assumption that each category of mental disorder is a completely discrete entity with absolute boundaries dividing it from other mental disorders or from no mental disorder”

2) The DSM is being used as a diagnosis tool. Based on the diagnosis issued the insurance companies can choose to pay/not pay for certain drugs or treatment.

It does not take a behavioral economist to deduce what follows from that. Drug companies are spending large amounts of money to affect definitions that are based on boundaries that are at best highly artificial. That puts a lot of strain on the difficult-as-it-is industry of psychoanalysis.

Moreover, the labels produced by the DSM - “ADHD”, “OCD”, etc. carry a lot of weight in defining the views of a person by himself and others, despite being terribly imprecise. That can have long-lasting negative effects on people’s lives.

The perfect way to approach psychiatry would be to base diagnoses on “absolute” factors, such as deficiencies of certain chemicals in one’s brain. However, that would require a level of understanding of the brain that we do not currently possess and might not possess for a while. Luckily, there is a thing that humans can do when they don’t understand the mechanics of something - use statistics.

There is an approach that would reduce the manipulation of diagnoses for profit and the harmful propagation of vague labels in pop culture, while not requiring an innate understanding of the brain - the Treatment-Based Statistical Approach.

This approach is based on a different organization of knowledge that does not involve diagnoses. A database will need to be created that contains the following data:

1) The behaviors, including self-reported evaluations, that people exhibited prior to seeing a therapist.
2) The treatments they used - whether drug-based or not. Getting a cat may have as large of an impact as some SSRI’s for all we know.
3) The changes in the aforementioned behavior after the treatment, along with any side effects experienced.

Naturally, the job of filling the database will fall on psychiatrists and other mental health professionals based on every interaction with a patient. The data will not contain any information that can be used to identify a particular patient.

With access to such a database people seeking help will not subject themselves to vague labels, but rather see what options they have, what changes they might expect from each option, along with the probabilities of these changes. After that, they (or their parents in the case of minors) will be able to make intelligent choices based on the outcomes and their likeliness. Some people might like the mood improvement by SSRIs and be OK with the possible weight gains. Others might balk at the cognitive downsides of lithium and find it not worth reducing the frequency of their manic episodes. In any case, people will have access to data based on what they can do, instead of getting slapped with a lobbied label that might define their life.

Moreover, such a database would be immensely beneficial to our efforts of understanding the brain, as it would give researchers access to a lot of data that they can use to test their theories.

There are three concerns about this approach:

1) Some people are not comfortable making such difficult decisions as how to manipulate their brains.They still have the option to heed Ice Cube’s advice, go speak to a person of authority and trust their judgment.

2) Self-selection of psychoactive medication can create the potential for its abuse. However, we do not currently have great tools in place to prevent substance abuse. One can order heroin online. While this approach may make matters worse, safeguards can be put in place, limiting the amounts psychoactive medication each person can buy. Moreover, as the world is coming to terms with losing the War on Drugs, we are likely to see governments increasingly letting adults take whatever psychoactive compounds they wish.

3) The issue of Big Pharma using ethically-questionable tools to increase their revenues remains. The integrity of the database will be dependent on the integrity of the psychoanalysts who fill it with data, and we know that it has, at times, been compromised by large amounts of money. The good thing about this approach is that in a centralized database one can filter data points by their creator. Hence, if unethical behavior by a professional is uncovered, the questionable records can be purged in an instant.

And the diagnoses? They can wait until we are able to find the elusive “absolute” boundaries in the brain. It does not seem that they create a whole lot of value for anybody but the APA today.

(There is some perceived value from diagnoses to parents of mentally unstable children. Labels absolve parents of responsibility by reconciling the “we are good parents, but our child behaves poorly” cognitive dissonance. The downside of that is that the child is likely to end up believing that there is something wrong with her.)