Antidepressants don't work? Really?

[whopper]

http://archpsyc.ama-assn.org/cgi/content/abstract/archgenpsychiatry.2011.2044v1

Conclusions To our knowledge, this is the first research synthesis in this area to use complete longitudinal person-level data from a large set of published and unpublished studies. The results do not support previous findings that antidepressants show little benefit except for severe depression. The antidepressants fluoxetine and venlafaxine are efficacious for major depressive disorder in all age groups, although more so in youths and adults compared with geriatric patients. Baseline severity was not significantly related to degree of treatment advantage over placebo.

The above data is more consistent with what we know of depression, and the biological aspects of mental illness.

Frankly, if antidepressants don't work well, so be it. It's not our place to try to defend them as if they are somehow the mark to measure our profession. It's our job to treat people based on science, wherever that brings us. When the article came out in 2010 suggesting antidepressants don't work, I mentioned on the board that if that is the truth, so be it. We must practice based on the truth.

But from my own clinical experience, I haven't noticed antidepressants only working on the severely depressed as suggested before. Much more often than not, I've noticed them working across the spectrum. They aren't the greatest thing in the world, aren't the only way to treat depression, and depression should be treated on a multifactorial level, but that said, I've found patients of varying degrees of depression where an antidepressant helped.

When studies are well-done, I've noticed they usually support what I've noticed going on in real-life. CATIE, for example, has held to that, same with STAR*D. So far I've only seen 2 studies that I felt were reasonably well-done as far as I could tell in terms of design where the results didn't match what I was seeing in real life and that was STEP-BD (the study suggesting that antidepressants don't work in bipolar depression), and the one where antidepressants don't work except in severe depression.

---

Members don't see this ad.

 

[DynamicDidactic]

There are some reservation about the data selection and analysis in this study.

For example study LYAQ is an Eli Lilly's clinical trial registry which compares fluoxetine + Strattera vs. placebo + Strattera yet this analysis examines the efficacy of fluoxetine versus placebo. Furthermore, the patients did not necessarily have depression in that study (and these issues are not noted in the paper).

There is a blog that discuss this paper (and another Gibbons meta) extensively and highlights some major concerns. Its a few posts down.
http://1boringoldman.com/

More importantly

But from my own clinical experience, I haven't noticed antidepressants only working on the severely depressed as suggested before. Much more often than not, I've noticed them working across the spectrum. They aren't the greatest thing in the world, aren't the only way to treat depression, and depression should be treated on a multifactorial level, but that said, I've found patients of varying degrees of depression where an antidepressant helped.

I agree with you that depression can be treated through various interventions and should be target more than one domain. I am also sure that medication has helped many people.

As a student-researcher specializing in depression, I believe that these studies provide a very different point of view than I get as a student-clinician. These studies are trying to identify what effect medication has independent of placebo. Similarly, these studies attempt to quantify the total group effects of medication. Whereas clinicians are biased by the types of patients they see and the settings they work in. Similarly, our memories are likely to be biased with anecdotal evidence that sticks out in our minds.

 

[billypilgrim37]

Same authors had another metanalysis in Archives (or was it AJP?) using a related data set to show that there is no increase in suicidality in kids on prozac and effexor. This guy is either a prophet or a shill. I would place a small bet on the former.

 

[splik]

i haven't read this paper yet, but the lead author is a major BS-merchant if ever there was one. He has previously concluded that antidepressants reduce suicide when the data he was reviewing did not actually support this. You have to be very careful with reading these papers as they are often constructed in a

No one said antidepressants didn't have an effect in non-severely depressed patients. The evidence suggests they no benefit that cannot be explained above and beyond the placebo effect in unipolar depressed patients. That is a different conclusion, and not one that can be examined from your anecdotal experience, from claims data, from correlational data. The gold standard is quite rightly RCTs, and the RCTs are damning - we should not

The real question is, are psychiatric RCTs more flawed and deceitful than those in other areas? I believe they are not. I believe cover-ups, publication bias, pharma burying unfavourable research, massaging the numbers, and over-estimating therapeutic effects and under-estimating adverse effects occur in other areas of medicine too. Take the rofecoxib vs naproxen RCT - this paper found 10x the risk ratio of fatal MI in the rofecoxib group - enough to suggest a causal effect, but alarmingly the authors interpreted it as a cardioprotective effect of naproxen! This drug should never have been on the market! I think because there are so many people against psychiatry (scientologists, the survivor movement) and those who find something morally obtuse about altering our neurochemistry (same groups, plus naturopaths, the neutroceutical industry and other stakeholders) that psychiatric research has been singled out. Well research misconduct, fraud, and misrepresentation are common in all areas of scientific research. Sad but true.

That of course doesn't make antidepressants anymore effective. My experience has been that in patients who don't know they are taking them they usually have no effect. The only remarkable effect I had with an antidepressant in patients who could not have a placebo effect because they did not know they were taking antidepressants, was a man with dystonia who because depressed secondary to tetrabenazine depletion of catecholamines. A little bit of sertraline (which is also a dopamine reuptake inhibitor) and his mood increased significantly. I am also convinced of a rapid improvement of mood in patients with bipolar depression (who then often become mixed, manic etc)

 

[splik]

Same authors had another metanalysis in Archives (or was it AJP?) using a related data set to show that there is no increase in suicidality in kids on prozac and effexor. This guy is either a prophet or a shill. I would place a small bet on the former.

He is neither, just full of ****.

 

[splik]

Here is the paper you are referring to - it is not clear whether the data actually supports the claims made here as they have not included it! but since these authors have a habit of interpreting data like performance artists rather than scientists i would not be surprised if the data did not support the findings.

 

[PETRAN]

This is the problem with pharma, it seems to be extremely greedy these days. I have the impression that a lot of pharma-driven "applied" pharmaceutical research is of very low quality (deliberately?) when compared to basic science research.


I like "free markets" but these days, with the extreme US economic neo-liberalism (the "liberalism" has nothing to do with "liberals" as a political group, but with the economic model employed by the states during the last decades. It could be better called "extreme capitalism") i get the feeling that development is regressing simply for the sake of profiting. I mean, when all areas of health services (from the highest pharma CEO to the last health-services provider) are viewed as "business/profit-centric" rather than "patient/anthropo-centric" it is not good at all IMO.


And i get the feeling it is not only health but a lot of things, education, services, energy, industrial production. The only thing that still seems to advance is electronics/information technology (possibly because it is the "newest"/more "virgin" in comparison to others?) and i'm not sure that it advances at the rate it could. Maybe progress in general has started to plateau (like when you go to the gym, you see impressive results in the beginning but as you continue, you need to put more for less), but i think that the economic model has a lot to do with the situation.


Well "free markets" is good but not TOO free IMO (sounds a bit Keynesian but i think it is the truth)


/end rant

 

[splik]

all very well but there aren't free markets in healthcare. If there were the government would be able to negotiate with technology companies and pharma, instead medicare and medicaid pay more for drugs and medical technologies than those on private health insurance plans, and everyone pays more for the latest drugs and technologies than in Canada, Europe, pretty much anywhere else in the world. This is despite US tax dollars contributing to these medical innovations in the first place. If there were free markets in healthcare, then privately provided plans in medicare advantage, would not cost 17c extra for every dollar spent through traditional fee-for-service by medicare.

And if we really had the sort of free markets the doyen of American neoliberalism Milton Friedman wanted you would not need to go to medical school to be a physician. He argued in his doctoral thesis that the regulation of the medical profession restricted the number of healthcare providers and in doing so reduced choice and raised the price of healthcare. In a real market, anyone would be able to provide healthcare (e.g. surgery etc) and the markets would decide as poor practitioners would not receive business. This is not too dissimilar from all the rate-my-doctor and release of mortality statistics etc. now that are supposed to increase patient choice. Many of the arguments about healthcare have been on the central point of - is there something special about healthcare? Is it different from other commodities. There is no doubt in my mind that there is, but it appears that many Republicans and even the Supreme Court disagree.

I also find it amusing that neoliberal physicians resent 'government takeover of healthcare' but as soon as HMOs starting reducing their earning power, or nurse quacktitioners and other health professionals became seen as a threat, they want 'big government' to help them. It is of course not ironic, but highlights the parent-child relationship we have with government. The same bears out in the general public who look at the government with suspicion and disdain most of the time, but during times of uncertainty, and when fear is high (e.g. after 9/11) suddenly people believe in the power of government to look after them, to act as a certainty, to restore moral order.

 

[Ibid]

This is the problem with pharma, it seems to be extremely greedy these days. I have the impression that a lot of pharma-driven "applied" pharmaceutical research is of very low quality (deliberately?) when compared to basic science research.

Why Olanzapine Beats Risperidone, Risperidone Beats Quetiapine, and Quetiapine Beats Olanzapine:

RESULTS: Of the 42 reports identified by the authors, 33 were sponsored by a pharmaceutical company. In 90.0% of the studies, the reported overall outcome was in favor of the sponsor’s drug.

http://ajp.psychiatryonline.org/article.aspx?articleid=178035

 

[whopper]

but the lead author is a major BS-merchant if ever there was one

Very good to know, and something to consider.

 

[billypilgrim37]

Very good to know, and something to consider.

Though a BS-merchant publishing in AJP and Archives. That doesn't mean there aren't issues, but it probably does mean those issues might not be quite so black and white. I read the suicidality meta-analysis pretty closely (I had to write it up for a national newsletter), and the conclusions seemed consistent with the analysis. I do remember the Vioxx paper, and it seemed like one of the lesser sins in a parade of much more fatal ones that occurred prior. I didn't realize that was the same author, and it does perk some red flags, but Gibbons has been part of the FDA side of things for a while. He has several masters, but I don't think it's so clear that he's serving any one of them exclusively.

 

[atsai3]

No one would ever accuse me of being a pharma apologist, but I don't know that I would really place Gibbons in the same category as, say, Keck or Nemeroff. I know Mickey Nardo and others have been beating him up regularly, but Gibbons is a bona fide statistician who cut his chops working with Jim Heckman, and he is still publishing in real journals and not simply the regular slop that gets you tenure in any ol' department of psychiatry.

 

[splik]

well i don't think gibbons has made any money from lily for this study as fluoxetine and venlafaxine are off-patent now. however he reportedly makes money in the courts. you probably cant make as much as a clinician, but many a statistician has taken money from drug companies or the tobacco industry. Ken MacCrae was a british medical statistician whose expert testimony won the case for pharma in claiming 3rd generation OCPs do not increase the risk of venous thromboembolism, even though the evidence is clear they do. Even the father of modern epidemiology Richard Doll was found to have taken large sums of money, not from drug companies, but companies wanting Doll to provide testimony that their products did not cause cancer, respiratory disease etc.

 

[splik]

David Healy has a pop at Gibbons in his blog, but on the paper I attached above that bp37 mentioned on ADs and suicide. By repeatedly making claims that are unfounded and unsubstatiated by the data, and selectively quoting the literature, he undermines the veracity of other claims he may make which may be grounded.

We are it seems, unable to have a reasonable discussion about antidepressants (whether it be around suicidality or efficacy) without rhetoric, hyperbole, statistical gymnastics, selective quoting of the literature, and gold standard statistical methods (e.g. RCTs) when they don't tell us what we'd like. This is true for both opponents and proponents.

 

[whopper]

Unfortunately true. Add to the histrionicism that thankfully hasn't hit this thread yet of the typical psychology vs psychiatry hate volleys occur when medication efficacy is brought up.

Industry-funded studies could be real. While pharm companies hands in the pot certainly begs the question of bias, one cannot discount EVERYTHING brought out by them. As for doctors taking plenty of money as court-witnesses, my personal opinion is so be it as long as the testimony they gave was intellectually and factually accurate. Often-times, to do an appropriate job as an expert witness requires tremendous work, training, and preparation. If they made money, even a lot, I got no problem so long as the information was accurate and not misleading.

But when it's not honest that's the problem, and when more money's involved--that just makes it look more evil.

I do think it's entirely appropriate to point out speciifc authors as having some type of funding because these things need to be out in the open and transparent, and beg the question of bias within the article.

 

[kugel]

Unfortunately true. Add to the histrionicism that thankfully hasn't hit this thread yet of the typical psychology vs psychiatry hate volleys occur when medication efficacy is brought up.

Industry-funded studies could be real. While pharm companies hands in the pot certainly begs the question of bias, one cannot discount EVERYTHING brought out by them. As for doctors taking plenty of money as court-witnesses, my personal opinion is so be it as long as the testimony they gave was intellectually and factually accurate. Often-times, to do an appropriate job as an expert witness requires tremendous work, training, and preparation. If they made money, even a lot, I got no problem so long as the information was accurate and not misleading.

But when it's not honest that's the problem, and when more money's involved--that just makes it look more evil.

I do think it's entirely appropriate to point out speciifc authors as having some type of funding because these things need to be out in the open and transparent, and beg the question of bias within the article.

But any time someone is consistently getting money from a source that has an axe to grind, you should wonder whether their judgement is being influenced. Doesn't matter whether the source is industry, plaintiffs, prosecutors, political parties, or one individual. If you are being paid to argue on one side and you do it long enough, you are likely to become biased. If you know (or even suspect) that your cash will be cut off if you argue against the source of cash, then you are almost bound to become biased. Nobody's brain is completely above conditioning, anymore than it's possible to choose to abstain from oxygen.

The general pattern doesn't prove it's true in any one instance, but it's prudent to wonder.
Even about ourselves.
I've been working for public mental health agencies a long time. I KNOW I've taken on attitudes and beliefs that are beneficial to the existence of those agencies. I sometimes try to imagine the opposing viewpoints - and I usually can't even imagine what they are any more.

 

[whopper]

I mentioned this before, a guy I know who is in a namebrand university gave some pretty BS testimony in front of my eyes in court concerning a child-rape case.

Very reprehensible. No surprise the guy was one of the drug dinner hookers, er cough cough speakers. If one were to truly do their homework, read my posts about the times I posted about it, you could probably narrow it down to 2-3 people given that the namebrand place is close to where I did residency.

 

[aphenomenon]

I know there are a billion recent pieces on the subject, but here is the last one that I read and liked:

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3313530/
Antidepressants and Advertising: Psychopharmaceuticals in Crisis

As the efficacy and science of psychopharmaceuticals has become increasingly uncertain, marketing of these drugs to both physicians and consumers continues to a central part of a multi-billion dollar per year industry in the United States. We explore how such drug marketing portrays idealized scientific relationships between psychopharmaceuticals and depression; how multiple stakeholders, including scientists, regulatory agencies, and patient advocacy groups, negotiate neurobiological explanations of mental illness; and how the placebo effect has become a critical issue in these debates, including the possible role of drug advertising to influence the placebo effect directly. We argue that if and how antidepressants "work" is not a straightforward objective question, but rather a larger social contest involving scientific debate, the political history of the pharmaceutical industry, cultural discourses surrounding the role of drugs in society, and the interpretive flexibility of personal experience.