Thursday, March 29, 2012
Is “Prozac” just “Placebo” misspelled?
A few facts:
In 2010 the top 5 antidepressants were prescribed 115 million times in the USA.
Antidepressants have side effects, the most common being weight gain, sexual dysfunction, and upset stomach. The most severe being suicidal thoughts and fatal birth defects.
Conclusion: These are not benign medications.
About me – I did not read just one article about antidepressants and then write a blog post about it. I have researched this topic. I have read many articles, and read the Newsweek article including the drug companies responses. I have read studies sponsored by drug companies as well as those with no “Pharm” funding whatsoever.
I wrote before about the publication bias in antidepressant research.
If we look at EVERY published and unpublished study submitted to the FDA for a set of antidepressants and compare the drug response with placebo response, guess what we find?
I'll give you a hint...it's depressing.
Everyone in the studies took a "depression rating test" before and after the drug trial.
If patients on Prozac said they got got better by 10 points, those on Placebo got by 8 points. Yes, Placebo had 80% of the effect.
You want the exact numbers?
There is a scale for rating depression called HAM-D which has been used since 1960 for pretty much all antidepressant trials. The higher you score, the more depressed you are.
0-7 = No Depression
8-13 = Mild Depression
14-18 = Moderate Depression
19-22 = Severe depression
>23 = Very Severe Depression
So you take the test before the drug trial. Then you get either an antidepressant or placebo and take it for a few weeks. Then you take the test again. How much lower your score is = your improvement.
Average improvement on antidepressants: 9.6 points (that would move you 1-2 levels up on the scale, like from severe to mild, and from moderate to mild or none.)
So that’s some real improvement
Average improvement on Placebo = 7.8
So the medication works better than placebo – but not much. How do we tell if that difference matters? Does a difference of 1.8 points matter? If you do a statistical analysis – YEP, it’s significant.
If you do a clinical analysis (do the patients notice any real difference?) – NOPE, it's insignificant.
It turns out that to be clinically significant the drug needs to be at least 3 points better than the placebo. (So says the National Institute for Clinical Excellence)
SO WHAT DOES IT ALL MEAN?
Americans are taking drugs for depression and their depression is getting better. If they were getting a sugar pill (but thought it was the drug) they’d do just about as well.
So you tell me - Knowing the side effects and the costs – is it worth it?