Map of life expectancy at birth from Global Education Project.

Thursday, June 01, 2017

Bad Science


This is not really news -- there have been less rigorous critiques before -- but this research letter in NEJM has gotten people's attention. In 1980, Jane Porter and Hershel Jick of Boston University published a one paragraph letter in the Journal. They said they had reviewed the medical records of 11,882 hospitalized patients who received opioids, and found evidence of addiction in the files of only four of them.

Note that these are hospitalized people who presumably receive only a short-term course of opioids; and that there is no reason to think that subsequent addiction or opioid use disorder would wind up in the records of that particular hospital, where most of the people will very likely not be seen again once they make their way into the world.

Nevertheless, this single letter to the editor -- not peer reviewed, not even really research -- was cited 608 times. Most of the citations were used to support the proposition that prescribing opioids for chronic, long-term pain, was not dangerous. More than 80% of the citations did not even say that the patients were hospitalized. The authors of the new letter give some examples, e.g. "The medical evidence overwhelmingly indicates that properly administered opioid therapy rarely if ever results in 'accidental addiction' or 'opioid abuse.'" This assertion is based on the single citation of the 1980 letter.

What they don't tell us is that this all happened in the context of heavy promotion of the long-acting opioid OxyContin by Purdue Pharmaceuticals, which was spending lots of money to promulgate these claims. They turned out to be false, and the result is the epidemic we face today.

So yes, science can be corrupted, by money as in this case, and in other ways. I spend a lot of time on this blog complaining about it. The good news is that the truth comes out in the end, but sometimes it takes too long. I should also make it clear that some scientific findings are simply not in doubt. Climate science is not corrupted by money -- on the contrary, the big money has been trying to corrupt it, and failing, for decades. So don't draw the wrong conclusions. But we need to keep our critical thinking faculties sharp.

Update:  I should have noted that the state of Ohio is suing Purdue and four other drug companies for causing the opioid epidemic.

No comments: