- About
- Blogosphere Buzz
- Archived Posts 2007-2008
- Register for Email Alerts
- The Book: Before You Take That Pill: Why the Drug Industry May Be Bad For Your Health
- Follow me on twitter @dougbremner
- Subscribe to my podcasts
- Rave Reviews for “The Goose That Laid the Golden Egg: Accutane, the truth that had to be told”
- Follow me on goodreads
- Catania! The movie
A recent meta-analysis in bmj of the treatment of people with multiple risk factors for heart disease but without a history of heart attacks looked at 70,388 people. They reported a 12% reduction in overall mortality which was statistically significant.
OK, so far so good.
However there are a number of problems with the conclusions not to mention the multiple pharma conflicts of many of the authors that may have contributed to those distorted conclusions.
In the conclusion section, they incredibly report a 12% reduction in mortality with no mention of differences by sex or diabetes status. They go on to write ”no significant treatment heterogeneity was found between the sexes, in elderly and young people, and between people with and without diabetes.” However inspection of Table 3 shows that there were in fact no statistically significan reductions any endpoint (mortality, heart attacks) in women or diabetics or people over age 65. Furthermore, there was no mention of number needed to treat (NNT), or the fact the number of people saved from a heart attack were not greater than those developed liver problems or severe musle pain (which could lead to rhabdomylosis, possibly fatal). Or that the 12% relative reduction is equal to about a 1% absolute reduction.
This misrepresentation of the results is bordering on fraudulent.
If you enjoyed this article, please consider sharing it!
I’d like to shove this down my statin pushing cardiologist brother-in-law’s throat…
he creeps me out…and you can’t have a conversation with him…so cramming down one’s throat is the only alternative…
or do nothing (which is what I’ll actually do)
shucks.
Hmmm, I’m afraid the details here are over my head, so I’ll take your word for it, Dr. B.
And I definitely see no mention of trials with B vitamins and magnesium, so that’s evidence enough for me!
OK, I hear what you are saying. I think there were no differences among females, or among the elderly, etc., simply because of sample size issues. Similarly, there would have been no differences among left-handed tobacco-chewing train conductors, either. By and large, though, I think the study’s author’s conclusions hold.
But I have a question for you. The authors reported a 12% reduction in mortality. You pointed out that some of the authors had pharma ties. How do you use that pharma connection knowledge to “adjust” the 12% conclusion? Do you just allow it to snap back to 0? Do you cut it in half? Do we pretend the study was not done at all? I’m just mystified as to how, exactly, we are supposed to use the information that several authors had pharma ties.
Thanks.
I would use it to modify my reading of their conclusions which are outlandish
If you told me no author had pharma ties, then I would use their report to construct my personal probability distribution for risk reduction centered at 12%, with some probability on either side of that to account for the imprecision of our knowledge.
If you then said “oops, my bad, I just discovered some of the authors did INDEED have pharma ties” I don’t know how to adjust my probability distribution for that new information. Simply re-adjusting my estimate of the risk reduction to zero doesn’t seem justified to me. If anything, I might move my probability distribution closer to zero, but I’m not sure how much.
[...] More Than Meets The Eye on Most Recent Analysis of Statins for Primary Prevention [...]
I had to get off the Statins I was taking as they were causing my ankles to swell, pain in my knee, even numbness and tingling in my wrist. I’ve been off of them for about two weeks. The swelling went down, the pain in my knee stopped, and my wrist is no longer numb.
My health care provider only believes in medicating me, so I stopped seeing her.