Why the Model Matters
There’s a great deal of good intention in the addiction treatment field. Gone are the days when Nurse Rachett got a job so that she could torture people locked up against their will. These days, the staff at addiction treatment facilities will be seen to be caring and well trained. Everyone wants to help. In fact, there’s so much help and so much training that we may be on that part of the curve where the outcomes stop rising with more training and start to go down instead. That’s a counter intuitive idea, so let me explain.
What most of us learn is a basic mental health model of addiction where the illness is caused by the use of drugs and alcohol, and the use may be caused by something else. This combination leaves most people carrying two diagnoses, or being called dually diagnosed. The modern term is co-occurring disorders. This learning is so ingrained that it inculcates our most fundamental view of the patient. This is important because it’s commonly said in the substance abuse field that at least 70% of people with a substance abuse problem have a co-occurring condition.
That’s a big number, but it’s so common as to be generally accepted to the point that it becomes most people’s base rate. The base rate is the guess you'd make at the prevalence of something if you had to make a choice with no other information. For instance, if I told you that a new patient just arrived for evaluation for admission to a substance abuse facility and asked you if he’s likely to have a co-occurring disorder, you’d say yes. You haven’t met him, and you don’t need to. You’re just playing the odds. If you believe the base rate is 70%, then having a co-occurring disorder is what you expect for him. You’ll be surprised if he doesn’t have one.
So what was all that I was saying about too much learning? Well, psychological studies have shown that if you tell a group of people the base rate of an occurrence and then ask them about it, they are guided by the base rate. But if you, in addition, give them extraneous information that has no bearing on the question, they ignore the base rate and move to a 50/50 guess. So getting extraneous information that has no bearing on the issue will confuse people, even experts, and they will forget the base rate.
So you’re probably thinking that such information will decrease co-occurring disorder diagnoses, right? Because 50% is less than 70%. Well, actually, it’s how we got there in the first place.
It was once understood in the addiction treatment field that addiction was an illness, separate from other brain-based illnesses, and that no other diagnosis was necessary. There was a low base rate of co-occurring disorders, just the same base rate you’d find in people without addiction. But so many people had gotten mental health training, and they wanted to do complete evaluations, so that in addition to all the necessary information on how to treat the new patient, there was also a lot of extraneous information. This changed the base rate from low to 50/50 for co-occurring disorders. This is called the Base Rate Fallacy, and when it got started another common bias leaked in.
In their now famous series of papers Kahneman and Tversky outlined a number of heuristics and biases common to people making decisions. Regardless of a person’s experience or expertise, humans showed these biases in their thinking patterns. One of these is confirmation bias. When a person has an idea that something is common, and they receive ambiguous information that could or could not be interpreted as that common thing or something else, they tend to see the thing they know is common. So if you are expecting co-occurring disorders to be as common as occurring in 50% of people, you’ll start to interpret their addiction symptoms as symptoms confirming the presence of a co-occurring disorder. You then start to see them everywhere and they become self-confirming.
So to come back full circle, this is what I mean when I say that we have reached the point where knowing more, and including more in our evaluations and presentations, is actually harming, not helping, outcomes on a global basis. I’m not saying that there aren’t people being helped, but on a global basis averaged over all the patients seeking addiction treatment, we’re not doing as well as we could.
The solution is not to ignore co-occurring disorders, because they do occur. The answer is instead, as with any problem to which we apply TOC, to focus on what’s in front of us. For instance, when a patient comes in for help to stop using heroin, he hasn’t asked us to fix everything that’s wrong, but rather he’s asked for help with his specific goal. Rather than trying to find out everything that’s wrong and focus on all of it (to focus on everything is to focus on nothing), instead we can just focus on what, exactly, is stopping his progress towards his goal today. If it happens to be a co-occurring psychotic disorder and the voices are so loud he can’t hear the counselor, then by all means let’s treat that. But if what’s stopping his progress right now to the next agreed upon goal isn’t a co-occurring disorder then whatever co-occuring disorder he may have can wait. When it becomes the constraint to progress, we’ll tackle it then.
This keeps us focused on the patient and keeps us from being defocused by extraneous information. We keep the base rate in mind, and we do the next right thing. It’s what the patients expect us to do.