Monday, October 16, 2017

Addicted to Bad Statistics - a Growing Epidemic

OK, references keep popping up to this CDC study on the dangers of patients receiving prescriptions for opiate pain-relievers becoming chronic opiate users. Specifically, they make reference to a finding that patients receiving a 1-day prescription for opiates only have a 2.6% chance of still being on opiates 1 year later while those given an initial 30-day prescription have a 45% chance of being on opiates 1 year later... So, I look up the study to see the data and conclusions for myself. Apparently not enough people sit through classes on research traps and simple mistakes.

This is the most telling point which should jump out at anyone interested in citing the research to make recommendations on opioid prescriptions or policy:

The findings in this report are subject to at least five limitations... Second, the extent to which chronic opioid use was intentional ... is not known. ... Third, information on pain intensity or duration were not available, and the etiology of pain, which might influence the duration of opioid use, was not considered in the analysis. [ https://www.cdc.gov/mmwr/volumes/66/wr/mm6610a1.htm (emphasis mine)]

They also note that longer-term prescriptions were much less common than short-term prescriptions. So... the obvious conclusion would be that doctors tend to write prescriptions for longer-terms for patients expected to need the pain relief for a longer period of time and... not surprisingly, they do. The study also notes that those receiving longer-term prescriptions tend to (among other things) be older and have a previous pain diagnosis. (duh!)

In other words, doctors are already likely exercising judgement by providing short term prescriptions to most people and only giving 30-day prescriptions to a much smaller number of people with longer-term conditions. But the researchers did not have data available to exclude this possibility ---and to their credit, say so. The warning in the study, however, is completely ignored by everyone citing it even though the data is useless for the purposes they are trying to make of it.

Garbage in; Garbage out

The capital-P Problem is that the purpose they are generally trying to make of it is this: recommending that doctors stop providing 30-day prescriptions to people with chronic conditions because of the higher danger of long-term use! This is usually linked to a claim that 80% of heroin users get their start with prescription opioids (another bad statistic I'll save for another day) and that reducing opioid prescriptions will therefore reduce future heroin use.

Question: what will long-term pain sufferers likely turn to when they cannot get legal prescriptions from their doctor to treat their valid medical conditions? In all likelihood, they will turn to black market sources, with no quality control, which has a higher chance of leading to addiction and overdose as well as feeding our larger issues with drug smuggling.

Is there a very large recent increase in drug overdoses? Yes. Is this a problem? Yes. Does stupid misuse of incomplete data help point to a solution? No.

And the solutions we persistently fail to consider...

You cannot willy nilly stop prescribing opiates to people with long-term pain unless and until you can offer other solutions. To do so is inhuman. Unfortunately, for many people who suffer long-term pain (including myself until recently), other solutions are few and non-existent. Opiates are more effective for short-term pain and lousy for long-term pain--- this is well-known. Using opiates to control long-term pain simply sucks less than being in pain for months on end with no other alternative.

Ironically, better alternatives for long-term pain are either difficult to get FDA-approval in the US or are flat illegal. Marijuana, for instance, shows long-term efficacy for chronic pain for some classes of patients and is much less dangerous than opium-based medications, but it is currently defined in law as having no valid medical use. Nor am I making a pro-marijuana case here, but rather the case that it is disingenuous to say people should stop using opiates while we are actively foreclosing alternatives.

No comments:

Post a Comment