About that study claiming anti-depressants are no better than placebos...

[hedgehog]

Supposedly, the study is a meta-analysis of all the data submitted by various drug manufacturers for various SSRIs. I always thought that a drug had to be proven safe and effective before it would be approved for sale by the FDA. How is it then, that government examiners, on every occasion for the twenty or so SSRIs out there, failed to notice that the drugs worked no better than placebos?

Here we have a class of drugs with a range of known side effects. We also know that the effect on a particular patient can not be predicted with certainty, that often a patient will try several different drugs or combinations to achieve relief. If the drugs are no more effective than placebos, why haven't the thousands of doctors and patients using them noticed? Two of the more common side effects are loss of libido and weight gain. Doesn't the fact that people continue to take these drugs despite these side effects suggest that something other than placebo effect is involved?

[Liberal Veteran]

1. Not really thousands...more like MILLIONS of people using SSRIs

Myself included.

[godai]

2. Before approval, efficacy must be shown in 2 clinical trials.

Clinical trials often fail to show efficacy and they can be repeated and looked at independently. Sometimes, for example, the trial wasn't large enough to show a statistical effect over placebo. When a drug repeatedly fails to indicate efficacy in trials, the drug company will drop it as not effective.

Anti-depressants are effective, with numerous side effects but the new study leads England to conclude that non-drug treatment may be the long term preferred treatment, dropping the drug treatment when the patient is stabilized.

[pinto]

3. From what I read, they did a sub set study about effectiveness on "mild"

and "severe" depression, and found the meds to be more clearly effective on severe depression, much less so on mild depression in relation to a placebo. Bottom line, iirc, was that standard one-on-one talk therapy may well be a better approach for people with mild, or transient, depression, with medication reserved for more intractable situations.

In that light it made sense to me. :shrug:

[Tandalayo_Scheisskopf]

4. Has anyone thought to look for...

The fingerprints of the $cientologists on this report? Just sayin', is all...

[Hannah Bell]

5. That meta analysis included...

some unpublished studies obtained through FOA.

One of the problem with research today is that when the pharmas foot the bill & the results don't go their way, they can just choose not to publish the results. So there's right off a bias toward postive research.

Another bias is the fact that trials of specific drugs are increasingly conducted by reseachers with ties to drug cos in multiple ways. They're more likely to find positive results - I'm not saying it's conscious, just saying that's the case, according to comparisons of research done by people w/o ties.

Finally, some people do show improvement with certain drugs. Some people will also show improvement with placebo + something to give it a "side effect". They'll swear up & down it helps them.

People have been noting poor results from SSRI's for some time. I read a study showing the same thing nearly 4 years ago. I can show you a drawer full of studies saying cholesterol meds don't work as claimed, either.

Why or why not these claims get public attention is another matter.

[hedgehog]

6. Your first statement is exactly what provoked my post. These studies

may have been unpublished, but they must have been submitted to the FDA if they were obtained via the FOA. So, did the FDA not read these studies? Did the FDA ignore these studies? Is it possible that the studies were so poorly designed that they could be interpreted several ways? Something about this story doesn't smell right to me.

[godai]

7. FDA reviewed the studies.

FDA can review and then put aside failed studies (no statistical difference from placebo), as long as 2 studies showing effect (better than placebo) are submitted. Those are the rules.

[hedgehog]

8. Depression is very difficult to diagnose or to quantify. I can see

where the design of a study can change the results of the study. In this case, I have to wonder if the researcher at Hull deliberately or unconsciously devised his study to prove that SSRIs are no better than placebos.

[godai]

9. They combined all the studies they could get their hands on.

This is valid and is called a meta-analysis. Combined, apparently there was not enough statistical effect to show a positive effect compared to placebo. But, there is no doubt that SSRIs, for example, can be effective in many patients needing them. Many individual studies have shown this. There are side effects however and, if non-drug treatment is sufficient (may not be in a lot of patients), this would certainly be drug-side-effect free.

[hedgehog]

10. Meta-analysis of garbage produces garbage, maybe?

If the people involved weren't really depressed, then they wouldn't repsond to drugs meant to treat depression.

[Hannah Bell]

13. Yes but...

one of the basic criteria of a published study would be selection of subjects based on some accepted criteria of depression. Why are you thinking the subjects weren't "depressed"?

[RedCappedBandit]

11. If there is no doubt that SSRIs

can be effective, then it's pretty obvious that this specific study is *not* reliable OR valid.

[Hannah Bell]

12. It's a meta-analysis.

Basically a review & statistical test of previous studies.

Not much "design" there.