What’s Wrong With What We Think
It’s usually our assumptions, not our logic, that fails us
I received an email yesterday announcing the results of a “new study” that will be presented as a poster at the meeting of the American Academy of Pediatrics. It seems that teens given opioids after surgery have more of a likelihood of becoming persistent users compared to those who aren’t.
I don’t know where to start. I can’t find a face-palm emoji.
First let’s look at the question of the study. Does persistent use matter? There are at least three reasons for persistance of use of an opioid: pain that persists, physical dependence on the opioid, and addiction involving the opioid. What matters isn’t, does opioid use persist, but why it persists. The three different groups need three very different treatments. They have 3 very different trajectories if untreated and three very different sets of concomitant social and healthcare problems. So it’s not a very worthwhile question to be asking.
Now let’s look at the way they asked. They selected from a large population opioid naive patients between 13–21 years old who had undergone of of 13 common surgeries. As a control they used a random sample of 3% of the same population, also opioid naive, who did not undergoe a surgery. Then they looked for persistence which they defined as opioid prescription fills between 90–180 days out from index time. They were surprised to find that 4.8% of those that had surgery were still getting opioids while only 0.1% of the controls were. They were surprised to find that only 0.1% of people not using an opioid and had no reason to be put on one were getting opioids 6 months later? Really? Do they think the stuff is in the water supply? This is not a control group for any useful question on the population.
So what was this study? Well, unfortunately it was the most common of studies we see today. It roughly falls into the “We have all this data we don’t know what to do with and can you make a study out of it so we can show a return on the investment” kind of study. You can tell these studies because they have these factors in common: there’s usually no problem identified (except it’s loosely associated with a recent headline), it’s retrospective, there’s no hypothesis, and the conclusion is always, “We need more studies.” If any for profit business ran this way, it would be shut down by the FTC.
There will be those of you who think I’m too harsh. After all this is probably a study by students or residents (hence the poster) who need the practice. Yes, that’s true. So let’s give them good practice, not teach them bad practices. My beef isn’t with the student researchers, but with the academic system that has failed them.
Is there anything we can learn from this data? Yes there is. From other studies we know that about 10% of the general population gets energy when given opioids. The other 90% want to go to sleep, hence the name narcotic. These 10% of people with an inborn abnormal response are the ones most at risk for addiction because there’s a reward for them from opioids that others don’t get. In spite of 10% of young people having surgery getting energy from opioids, only half of that number were still taking opoioids 6 months later. That tells us that our system is good enough to get half the people at risk for addiction off opioids in 6 months. Now we need to ask ourselves, “is that good enough?”
I doubt it, but what is needed are more discriminating studies asking the right questions: who is it who stays on opioids, what do the opioids fix that was wrong with them, how else can we fix what was wrong, and how could we do that before the surgery? By seeing addiction as based in the drug, we lose sight of it being a biological largely genetic brain illness, and we lose our best chance at prevention and treatment. By allowing studies like this to get into the literature, even a medical society’s throwaway journal, we are misdirecting our efforts and delaying the day when we solve this problem once and for all.
********************
Dr Wetsman is retired from the practice of addiction psychiatry and currently studying data science. He has released his serial Ending Addiction. He also has something else in the fire — more to come.
