Dopesick and information propagation.
My wife and I just finished watching the Hulu mini series Dopesick. First off this was an excellent and captivating series. It approached the opioid crisis from many directions. There were strong story lines regarding Purdue Pharma, the manufacturer of Oxycontin and the US Attorney from the Western District of Virginia. These story lines were were interwoven with the FDA, DEA and the US Department of Justice. These entwined story lines were a mere backdrop to the heartbreaking stories of the individuals and families affected by opioid addiction. It is a series that I recommend you take the time to watch.
I am going to try my best not to label any of the players as evil or saintly. I encourage you to watch it and come to your own conclusions. I will however discuss or motivations and biases. What I am focused on is a piece of information that was quoted several times in the 8 episode series. Although it was quoted often it was very understated until the later episodes. The last half of this clip (Less than 1% ) introduces the "less than 1%" selling point. This clip is the opening of the first episode so no need for a spoiler alert. Warning, spoiler alerts ahead.
The last half of this clip has the US Attorneys questioning doctors on how Oxycontin was "sold" to them by the Purdue Pharma sales people. The claim was that this opioid was safe to use and that less than 1% of patients became addicted to the drug. It wasn't until episode 4 (now I can give a spoiler alert) of Dopesick that someone asked to view the information that this claim was based on. It turns out that this claim was not based on a detailed scientific study but on an observation by a doctor using a newly created patient database. It was not an article but a short letter to the editor. The entire letter was 5 sentences long. Here is an article from the Atlantic with the letter Atlantic Article. Everything was based on this one paragraph letter. I wrote earlier about the smell test Smell Test. An entire market of opioid drugs and pain treatment based on a short letter doesn't really pass the smell test.
In researching this blog I came across this New England Journal of Medicine bibliometric analysis. The authors conclude that the original 5 sentence letter was "was heavily and uncritically cited as evidence that addiction was rare with long-term opioid therapy". Please read the Journal and leave your thoughts about it in the comments. The key phrase to me was "uncritically cited" How could Scientific American, Time magazine and multiple medical schools use this letter without any type of investigation as to the validity of the data.
I am not trying to prove or disprove any specific element of Dr Jick's letter or how it is applied or misapplied. I am more interested on how this 5 sentence letter seemed to take on a life of it's own. How did this observation get so widely cited without any serious peer review or context? Of extreme importance is how we as individuals react when we come across situations such as these.
The first and most obvious reason would be our natural tendency to accept any information that would support our desired outcome. I cannot look into the heart and soul of Purdue employees (executives, medical directors and sales reps) but they do want to sell drugs. To support that goal they will accept and encourage any source that will support that end. This is something that we are all guilty of. We are more likely to believe something that supports or biases. The bibliometric analysis shows that citations of the letter increased as Purdue was bringing oxycontin to market. As a non medical person this correlation of product release and increased citation is concerning. Again does it pass the smell test? As I said we are all guilty of this.
A second factor could be an indolent approach to what appears to be a no brainer. This lethargy coupled with the prestige of the NEJM would allow someone to cite a reference without fully checking the details. This is probably how the information entered into main stream medical schools. Again this is something we all do. We all put different values on sources of information. We do it for newscasts and web sites. Some people believe everything on Fox news where as others have faith in all things CNN. Another issue associated with this is time. This is especially true with doctors. I cannot imagine that a doctor has time to fully fact check everything a pharmaceutical rep tells them. So when they are told that the New England Journal of Medicine says the chances of addiction are less than 1% they will eventually begin to prescribe the medication. To practice what I preach I spent about 2 hours going through the Dopesick episodes to find the one that talks about the letter. If that took 2 hours imagine what it would take for a Doctor to analyze a detailed study.
A third and final factor could be what I would call qualification propagation. The original a letter is referenced in another reputable Journal. Now there are 2 authoritative sources of the information. The next person then cites one or both of the articles and the whole thing snowballs from there. Dr. Jick's original article was referenced in the Journal "Pain"
As for solutions my advice would be to recognize our own biases and try and search out opinions that may differ from our own. We should be diligent about what we use as sources of information and as best as we can not take anything at face value. Do some digging, if there is references cited try and read some of them. Finally just because someone or some entity is important doesn't mean they are right.
Links
Atlantic Article with original Letter
Dopesick timeline very brief, only covers about 30% of the series content but is a nice framework
Have a look at the references and leave a comment on how you fell about them
Comments
Post a Comment