Articles

Study shows antidepressants useless for mild to moderate depression? Not exactly.

As Harriet Hall has written (http://www.sciencebasedmedicine.org/?p=353), psychiatry bashing is a popular media sport. There seems to be a bias against treatment of psychiatric disabilities, and a common claim is that antidepressants are no better than placebo. The New York Times illustrated both the perpetuation of the myth that antidepressants are ineffective, and the increasing and disturbing tendency of major media organizations to confuse the wholesale acceptance of medical press releases with medical journalism.

In Popular Drugs May Help Only Severe Depression The New York Times credulously publicized the findings of a recent study that claimed to show that antidepressants are ineffective in treating mild and moderate depression. Yes, that’s what the study showed, but the study itself is so limited, so fraught with problems, and the conclusions are so misleading that the article is a terrible disservice.

Before we consider what the study showed, let’s think about what kind of evidence we’d need to conclude that antidepressants don’t work.

First, although there are different types of antidepressants, the term used colloquially refers to antidepressants of a specific type, SSRI’s or selective serotonin reuptake inhibitors. There are other, older types of antidepressants that are rarely used today because of their unpleasant side effects. Hence any study that claims to show that “antidepressants” are ineffective, must look at SSRIs.

Second, there are literally thousands of studies of SSRIs, and it would be helpful to aggregate the results. Aggregating results can be done in a type of paper known as metaanalysis. Metaanlysis adds the results of multiple similar studies to find trends that might not be apparent in individual small studies. But a metaanlysis is subject to several important limitations that must always be considered. The most important limitation is that the authors of the metaanalysis choose the papers to be included. Bias can be introduced by examining only papers that have a desired outcome; that can be accomplished by restricting the inclusion criteria in arbitrary ways.

Let’s look at the study, Antidepressant Drug Effects and Depression Severity. According to the abstract:

Randomized placebo-controlled trials of antidepressants approved by the Food and Drug Administration in the treatment of major or minor depressive disorder were selected. Studies were included if their authors provided the requisite original data, they comprised adult outpatients, they included a medication vs placebo comparison for at least 6 weeks, they did not exclude patients on the basis of a placebo washout period, and they used the Hamilton Depression Rating Scale (HDRS). Data from 6 studies (718 patients) were included…

… The magnitude of benefit of antidepressant medication compared with placebo increases with severity of depression symptoms and may be minimal or nonexistent, on average, in patients with mild or moderate symptoms. For patients with very severe depression, the benefit of medications over placebo is substantial.

According to one of the authors was interviewed for the NYTimes article:

“The message for patients with mild to moderate depression,” Dr. DeRubeis said, “is, ‘Look, medications are always an option, but there’s little evidence that they add to other efforts to shake the depression — whether it’s exercise, seeing the doctor, reading about the disorder or going for psychotherapy.’ ”

Let’s go back and compare the paper to the criteria we identified above. The first criterion was to look at the antidepressants currently used in clinical practice. But out of the 6 studies in the metaanalysis, 3 looked at imipramine, a tricyclic antidrepressant that has not been the standard of care for over a decade because of its unpleasant side effects. The other three studies looked at Paxil (paroxitene). Paxil is an SSRI, but it is only one member of the class of SSRIs. Although all SSRIs share the same mechanism of action, they have different profiles of effectiveness and side effects. Therefore, generalizing from Paxil to all SSRIs cannot be justified.

So in terms of clinically relevant information, the paper included only 3 studies of an SSRI. How did the authors whittle down thousands of papers on SSRI effectiveness to only 3? According to the authors:

The criteria for inclusion required studies to be randomized placebo controlled trials of an FDA-approved antidepressant in the treatment of the full range of patients with major or minor depressive disorder … In addition, the studies had to include an ADM/placebo comparison of at least 6 weeks’ duration and HDRS scores at intake and at the end of treatment. Studies were excluded if they excluded patients on the basis of a placebo washout period. The final inclusion criterion was that individual patient-level data had to be available for analysis.

Are these criteria relevant? Certainly, the inclusion of only RCTs is a reasonable criterion. However, it is not clear why the availability of patient level data is a relevant criterion. Most RCTs, from an enormous range of clinical investigations, do not include patient level data, and using that as a criterion is bound to exclude most studies.

Finally, the decision to remove studies that included a placebo washout period also excludes a vast swath of psychiatric studies. That decision is more defensible, however, since there is disagreement among psychiatric researchers about whether a placebo washout period introduces bias into the study. A placebo washout period involves treating everyone in both arms of the study with placebo for an initial period of time, often 3 weeks. People who respond to placebo are then excluded from the study. The theory is that excluding known placebo responders makes it easier to identify real effects.

Others have argued that excluding known placebo responders up front necessarily makes the drug effects look better than they would have. For example, in a traditional placebo controlled RCT, there might be 30% who respond to placebo and 50% who respond to the medication under study, for a difference of 20%. If some placebo responders are identified during a washout period, let’s say 20% of patients, they will be excluded. The final results may be that 10% responded to placebo and 50% responded to the medication under study, for a difference of 40%, making the medication under study look better.

There is one indisputably arbitrary criterion that is acknowledged by the authors. The initial analysis identified 23 studies, but they could only gain access to the data in 6 studies, so they simply ignored the other 17.

In summary, then, by using questionable exclusion criteria, the authors accessed only 3 clinically relevant studies (the Paxil studies), involving only one SSRI. It is not clear that these studies are representative of existing studies on SSRIs, or even if they can be generalized to other SSRIs. Dr. Rubeis’ assertion that for patients with mild to moderate depression there is little evidence that “medications” add to efforts to treat the depression cannot be justified by the findings in his study. I find his claims to be irresponsible. The paper adds to the literature on antidepressants but is so limited that it cannot tell us whether antidepressants are effective for mild to moderate depression.

Posted in: Science and Medicine

Leave a Comment (280) ↓

280 thoughts on “Study shows antidepressants useless for mild to moderate depression? Not exactly.

  1. twisted mentat says:

    I remember one of these sorts of studies coming out about 2 years ago and one interesting thing the University student health people did in response was get the psychiatrist to come in and do a talk about it.

    One thing he said about these studies stuck in my head. As someone committing suicide while testing one of these medications would stop the testing cold and likely put the approval off for years it is likely that the people tested in many of these studies are right on the border of being diagnosed with depression. So it seems there’s a fair chance that some of these medications are more effective then their studies give them credit for.

  2. Tsuken says:

    Thank you very much for looking at this article on SBM. While I haven’t got around to a full deconstruction myself, I mentioned it recently: http://www.tsuken.co.nz/sadness-or-depression-midweek-medicine/ and also linked to another NYT piece where Dr. Richard Friedman discusses this meta-analysis nicely: http://www.nytimes.com/2010/01/12/health/12mind.html?em

    In any case, I really do not understand how anyone can take seriously a study that out of 30 years of research, includes the results of only 6 trials, of 2 antidepressant medications, involving 718 people….

    From. 30. Years. Of research.

    The odd thing is that while I would say the results of this study are not valid, they might well be – at least in part – true. We’re not yet able to differentiate between the different entities that most likely make up our syndromal picture of “Major Depressive Disorder”; it seems likely that the more biological causes would respond better to pharmacological treatments than would a more psychologically-based problem – in the same way that ECT is more effective, the more severe the depression. I suspect that severity might be acting as a proxy for the degree of biological dysfunction, and therefore the more severely ill might well be expected to respond better.

    The above is speculation … but in my opinion none of the three recent meta-analyses purporting to show antidepressants being ineffective for mild to moderate depression can truly provide an advance on said speculation.

    A final note, on tricyclic antidepressants: they are still around, because they can be very useful treatments. The two main reasons advanced for using newer medications instead of TCAs are (1) side effects, and (2) toxicity in overdose. Regarding point (1), all our medications are pretty unpleasant; SSRIs simply have a different side effect load to TCAs – not necessarily better. That depends on the individual patient. Regarding (2), the best way to stop a depressed person killing him or herself is to treat the depression. If a TCA works best for that person, I would argue it’s best to use the TCA and control the supply and access, rather than avoid a treatment we know could be helpful.

  3. Carl Graham says:

    Thanks for this post Amy.

    I would be interested in seeing the evidence that serotonin had been established as causal in depression.

    Once that was established a discussion about SSRI efficacy aside from placebo effects would make sense.

  4. BillyJoe says:

    Two questions:

    Why should a lack of “patient level data” be an exclusion criterion?

    “The initial analysis identified 23 studies, but they could only gain access to the data in 6 studies, so they simply ignored the other 17.”

    Why would you not exclude studies for which data could not be accessed?

  5. “it is likely that the people tested in many of these studies are right on the border of being diagnosed with depression. So it seems there’s a fair chance that some of these medications are more effective then their studies give them credit for.”

    That’s one possibility, and there are others.

    Consider that if a patient failed to respond to several weeks of Paxil treatment (a not uncommon scenario), a provider would not conclude that “antidepressants” don’t work, or even that “antidepressants” don’t work for this particular patient. The provider would either change the dose or try a different SSRI. It often takes several tries to find the right SSRI and the right dose for a particular patient.

    To me, it seems that this study is the equivalent of treating a mild gram negative infection with penicillin. If the patient didn’t get we wouldn’t conclude that “antibiotics” didn’t work. We’d conclude that we chose the wrong antibiotic.

  6. “The above is speculation … but in my opinion none of the three recent meta-analyses purporting to show antidepressants being ineffective for mild to moderate depression can truly provide an advance on said speculation.”

    Agreed!

  7. “Why should a lack of “patient level data” be an exclusion criterion?”

    I’m not sure why that would be a legitimate exclusion criterion in this setting.

  8. It is distressing that the authors would draw the conclusions that they drew from such limited data. However, I find it particularly irresponsible for the authors to publicly claim that the study shows that “antidepressants” are ineffective in mild to moderate depression.

    In other settings, the authors have acknowledged that they “don’t know” if their findings can be generalized, yet they generalized their findings to the greatest possible extent in their own press materials.

  9. Fifi says:

    Dr Tuteur – “It is distressing that the authors would draw the conclusions that they drew from such limited data. However, I find it particularly irresponsible for the authors to publicly claim that the study shows that “antidepressants” are ineffective in mild to moderate depression.

    In other settings, the authors have acknowledged that they “don’t know” if their findings can be generalized, yet they generalized their findings to the greatest possible extent in their own press materials.”

    Dr Novella would have been a much more appropriate SBM blogger to tackle this subject, though an actual psychiatrist would have been more appropriate. What’s distressing is you setting yourself up as an expert on depression and psychiatric studies as a means to drum up yet another controversy when you’re clearly not an expert and there are very distinct dangers with inappropriately prescribing SSRIs to most people with bipolar disorder (and still questions regarding the role of serotonin in depression and the critiques are actually coming from people with much more expertise than yourself).

    Like GPs who don’t have the appropriate training to diagnose and treat mental illness – who actually prescribe a lot of SSRIs for mild to moderate depression, the very conditions that experts have increasing doubts regarding SSRIs’ efficacy over and above a placebo effect – you seem to be overstepping the boundaries of your professional knowledge here. (To be clear, I’m not discounting the usefulness of creating a placebo effect when dealing with depression and anxiety related disorders.)

    All in all, it’s pretty rich that you’re trying to call out the study’s authors all things considered. And that you assume that a press release is written by the actual researchers and not the university PR department yet you don’t even bother to mention the way the SSRIs are advertised and promoted as a panacea by pharmaceutical companies. It makes it appear like you’re simply writing a press release for drug companies about something you apparently have very little knowledge about (especially since you’ve chosen to ignore related studies that are extremely relevant…talk about cherry picking!). It’s as if you have a reflexive doubt of anyone who questions Big Pharma! Including experts in fields you’re not an expert in!

    http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.0020392

    And why are you focusing on the NYT article when the NEJM and many other reputable medical sources have also published on this topic? Why refer to the press release and the NYT article instead of the NEJM one where the authors were actually defending SBM. Really, you’re just distorting the science and using populist sources to further your own agenda that, once again, seems to really have little to do with SBM!

    http://content.nejm.org/cgi/content/abstract/358/3/252

    “Conclusions We cannot determine whether the bias observed resulted from a failure to submit manuscripts on the part of authors and sponsors, from decisions by journal editors and reviewers not to publish, or both. Selective reporting of clinical trial results may have adverse consequences for researchers, study participants, health care professionals, and patients.”

    http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.0050045

    Citation: Kirsch I, Deacon BJ, Huedo-Medina TB, Scoboria A, Moore TJ, et al. (2008) Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the Food and Drug Administration. PLoS Med 5(2): e45. doi:10.1371/journal.pmed.0050045
    “Conclusions
    Drug–placebo differences in antidepressant efficacy increase as a function of baseline severity, but are relatively small even for severely depressed patients. The relationship between initial severity and antidepressant efficacy is attributable to decreased responsiveness to placebo among very severely depressed patients, rather than to increased responsiveness to medication.”

    As noted by another poster already, the older class of tri-cyclic anti-depressants are still used and there’s still controversy regarding the role of serotonin in depression. It really is distressing that you’d exploit this topic to further your own agenda under the pretense that you actually care about people with mental illness. To be clear, I’m not against the use of medication for mental illness at all (or placebos for that matter). What I’m a proponent of is real SBM related to treatments for mental illness not creating the illusion that only pharmaceutical solutions are SBM.

  10. Zoe237 says:

    What do the other 17 studies show for SSRIs for mild to moderate depression? Is there any other basis to the idea (besides this meta) that SSRIs don’t/minimally work for mild depression or that the side effects outweigh the (possibly small) benefits? What does Cochrane show?

    Seems the burden of proof should be on those wishing to show they DO work, given the possible side effects.

  11. Fifi – I disagree with your criticisms of this post. The scope was limited, but appropriate for SBM – focusing on a new study and the press that it is getting. This does not pretend to be an independent systematic review, or even a treatment of the entire question of anti-depressant use.

    Amy’s criticisms of the study, its interpretation, and the mainstream media coverage are appropriate.

    Further, having reviewed the literature myself, my impression is that SSRIs have a small but measurable benefit, either alone or in combination with psychological treatment, for mild to moderate depression. Given that there is a sizable effect for major depression also establishes significant plausibility.

    The literature supports the use of SSRI’s in an individualized strategy. The media, in my opinion, is getting the story significantly wrong. And that is the thrust of this post.

  12. Fifi says:

    Dr Tuteur – “psychiatry bashing is a popular media sport. There seems to be a bias against treatment of psychiatric disabilities, and a common claim is that antidepressants are no better than placebo.”

    This is a strawman since the questions regarding the efficacy of SRRIs are actually coming from within the psychiatric profession who are calling out pharmaceutical companies for practicing pseudoscience. It is you who is actually doing the psychiatry bashing in this instance! Seriously, you’re acting like an apologist for pharmaceutical manufacturers and not a defender of psychiatry or SBM. Another strawman you erect in your first paragraph is that questioning the efficacy of SSRIs or anti-depressants is showing a bias against treatment of mental illness – it’s not, it’s questioning the efficacy of SSRIs or anti-depressants for treating specific conditions. If you were in the least bit familiar with psychiatric and neurobiological research you’d be aware that this happens within the profession all the time and wouldn’t be trying to demonize experts in the field who are actually defending SBM and asking very important questions about best possible treatment for particular conditions and the promotion of pseudoscience by pharmaceutical companies.

  13. Fifi – There certainly are those within the profession raising questions about the efficacy of SSRI’s in mild to moderate depression – as their should be. That, however, does not mean it is a straw man to observe that the media seems to have a bias against the diagnosis and treatment of mental illness. I think there is such a bias. And this relates to the media handling of this research – not the research itself.

    But it is a straw man to suggest that criticism this study or the media coverage is akin to defending pharmaceutical companies. That is dangerously close to the “pharma shill gambit” and is a non sequitur.

    I agree that pharmaceutical companies tend to overstate the usefulness of their own drugs – that’s marketing. It should be viewed with suspicion. They are tightly regulated, but do skirt the lines of this regulation, even resulting in occasional sanctions.

    Consumer marketing is also a separate and complex issue, but one where there is certainly a role for legitimate criticism. But also having nothing to do with this post.

  14. Fifi says:

    Dr Novella – Thanks for weighing in – I still wish you’d actually tackled this topic or, better yet, a psychiatrists since this deals with psychiatry. The media is the media and this is a sexy story – mainly due to how well known Prozac and SSRIs are due to the massive over-prescribing by GPs and the amount of advertising done by pharmaceutical companies.

    I have no objection to discussing this study – though doing so in isolation seems to be a means to cherry pick data to push a certain agenda, particularly considering how many strawmen Dr Tuteur erects at the outset. Since Dr Tuteur is talking about psychiatry bashing, it’s more than a little ironic that she’s actually doing it herself under the guise of promoting SBM. Defending pharmaceuticals isn’t equivalent to defending psychiatry – it’s simply defending pharmaceuticals (a legitimate thing to do but it’s not defending psychiatry) and when it’s also bashing psychiatrists then it’s certainly not defending psychiatry. Discussing one study is also valid, generalizing about it while criticizing experts in a field Dr Tuteur isn’t an expert for generalizing is just hypocritical.

  15. Scott says:

    Fifi,

    Please explain how anything at all you said has any bearing on the fact that a meta-analysis which looked at only two medications is being presented as generalizable to all antidepressants. There is no need for specific psychiatric qualifications to recognize that as a massive flaw.

    Most of your accusations also go far beyond what Dr. Tuteur actually said.

  16. JerryM says:

    Just this week a big survey of anti-depressants was released by a dutch consumer affairs program about side effects of SSRI’s. Many people reported averse effects, which is one thing. What most worried me were the common stories of GP’s not taking the side effects serious when they were reported, saying that they were just part of the depression in the first place.

    A spokesman for the GP’s disagreed with the presenters suggestion that they should stop prescribing anti-depressants, citing lack of psychiatrists and subsequent long waiting lists.

    He did concede that they should make an effort to impress upon GP’s that there are side effects, and that people that report averse changes after beginning their course should be taken more seriously.

    There was one woman who with one SSRI had a worsening of her depression, to the point of suicidal thoughts, but with another SSRI was getting better, so at least that helped to illustrate that the drugs themselves do work, just not every SSRI is the same, and that different people will respond differently to different drugs.

  17. Fifi says:

    Dr Novella, if one is going to critique the media’s handling of this story then one needs to also contextualize it by acknowledging why it’s such a big story (and not blame the study’s authors for what a PR department does). If one’s going to bring up a press release, then one should at least understand that researchers don’t write their own press releases and discuss how and why science gets perverted by PR departments. However, even blaming a PR department would be an incorrect analysis of why this is story with legs.
    It’s such a big story simply because pharmaceutical companies sold SSRIs as miracle drugs and promoted them so heavily to GPs, which resulted in them being over-prescribed. This means that there are a lot of readers with a very personal interest in the subject of SSRIs and that even people who don’t have a person interest are highly aware of SSRIs. While Big sCAM has also helped give this story legs, the reason why the media’s running with it is just as much about how pharmaceutical companies promoted SSRIs to the media, the public and medical professionals (particularly ones who aren’t mental health experts, and as the best treatment for mild to moderate depression). From a media reporting perspective, Dr Tuteur’s blog is much more poorly contextualized and just as unbalanced as the actual article she’s critiquing!

  18. Fifi says:

    Ah, the you can’t call someone on being an apologist for pharmaceutical companies or at least appearing to be biased towards pharmaceutical use because it’s a pharmashill gambit gambit… No, I don’t think Dr Tuteur is taking money from drug companies, I just think Dr Tuteur is biased and reactionary and, in this case, seems to believe that psychiatry is all about drugs – which is just reinforcing the false image of psychiatry promoted by psychiatry bashers! And she likes to stir the pot – which I wouldn’t mind if she was actually cooking up something worthwhile that was nourishing and truly supports SBM (not just gives the illusion that she’s promoting SBM).

    In some unfortunate cases, doctors are pharmashills. (There’s the case of Dr. Lawrence Dubuske who resigned from Harvard’s Brigham and Women’s hospital when they recently changed the rules regarding taking money from pharmaceutical companies, for instance.) I am by no means saying that anyone who prescribes pharmaceuticals or discusses their risks and benefits and reports when science shows benefits is a pharmashill. Medications can be lifesaving when properly prescribed – though they can also be deadly when improperly prescribed. I’m just pointing out that doctors who corrupt SBM for both Big Pharma and Big sCAM do actually exist….they are not unicorns!

    http://carlatpsychiatry.blogspot.com/2010/01/dr-lawrence-dubuske-me-myself-and-irine.html

    Dr. Carlat’s efforts to promote SBM and call out inappropriate influence in psychiatry seem worthy of a post themselves – particularly since he is a psychiatrist and is not alone in trying to defend psychiatry from drug company pseudoscience and meddling. (Sure it’s fun to castigate the woo merchants and media, they’re easy targets, but it’s also good to promote those fighting the good fight and to make it clear that there are many doctors and psychiatrists doing this!)
    http://www.thecarlatreport.com/

  19. David Gorski says:

    The initial analysis identified 23 studies, but they could only gain access to the data in 6 studies, so they simply ignored the other 17.

    Uh, Amy, you can’t do the most powerful form of meta-analysis on studies whose data you can’t get access to. What would you have had the investigators do? It is not an uncommon scenario at all for investigators seeking to do a meta-analysis not to be able to get access to adequate raw data from a significant fraction of relevant studies. The authors pointed out what happened in the methods, and this is exactly what they should have done. Readers then note it as a limitation.

    The rest of the post was a fairly reasonable criticism, but that one sentence stuck out like a sore thumb.

  20. “It’s such a big story simply because pharmaceutical companies sold SSRIs as miracle drugs and promoted them so heavily to GPs, which resulted in them being over-prescribed.”

    I understood the marketing a bit differently. We already had effective antidepressants for very depressed people and psychiatrists were prescribing them. SSRIs are less effective but have fewer dangerous side effects, so they could be offered/marketed to people with mild to moderate depression who would not otherwise have received any treatment at all. They could also be prescribed by GPs. Being able to expand their customer base is great for pharmaceutical companies and they took full advantage.

    Over-prescribed? I wouldn’t know. Not in my own experience certainly. What does over-prescribed mean? That they are prescribed to people to improve their quality of life — allow them to hold jobs, for instance — even though without them the individual might plausibly have remained alive, albeit on welfare?

    Yes, a lot of people take them. There are various explanations for this. Perhaps misery is the natural human state, a result of the Fall, and given to us by God to bring us closer to Him. Perhaps misery is just the natural human state. Perhaps depression is one of those spectrum conditions that is very common because a little is advantageous (promotes introspection); in the past, perhaps all the moderately and very depressed people would have died from inability to care for themselves, while the introspective and mildly depressed people would have continued the genes to the next generation. Perhaps the isolation of modern life and the associated high demands on the individual create the conditions for depression. Perhaps the inequalities of capitalism do. Perhaps we eat the wrong food and get insufficient exercise. Maybe we were damaged in utero by chemicals in the water supply. Whatever, demand is high and a lot of people take them. And with the exception of the theological account where it would be a sin to use medication to deprive yourself of the full experience of the horror of separation from God, all accounts leave antidepressants at least a theoretical role to play in both saving lives and improving quality of life.

    I mean, what do you do when your beloved hasn’t been able to get out of bed for two weeks? Call a priest? Tell them to get over themselves? Tell them that if they would just be a better person and lose weight, get more exercise, get a job and improve their social life that the depression would go away? Or tell them that they are absolutely right, life sucks and then you die, but in the meantime they aren’t having any fun and they are being a total downer for the household and there are alternatives, so call the doctor already.

    Note that I’m perfectly prepared to believe they are over prescribed. It’s just that observing that a lot of people take them is not sufficient evidence in itself.

  21. Fifi – your point about not putting the issue into a broader context is legitimate. I often think Amy needs to do that more in her posts, and add more caveats to be clear about what she is not saying, and I have expressed this to her.

    However, the tone of your criticism goes way beyond the substance. You make many statements as fact that are little more than opinion. I don’t think you can definitively say why the media is running with this story. My guess is the primary reason is simply that it’s sensational – that’s it. But we can speculate about many other factors.

    Further, you make certain accusations that are not based on the post itself, like that Amy believes treating depression is all about drugs. I don’t see that anywhere in her text. It is outside the bounds of this article – although certainly could have been one of those side issues that would increase the context of the article.

    I think you need to step back an put your own criticism into context, just as you admonish Amy for not doing.

  22. Fifi says:

    Alison – “I mean, what do you do when your beloved hasn’t been able to get out of bed for two weeks? Call a priest? Tell them to get over themselves? Tell them that if they would just be a better person and lose weight, get more exercise, get a job and improve their social life that the depression would go away? Or tell them that they are absolutely right, life sucks and then you die, but in the meantime they aren’t having any fun and they are being a total downer for the household and there are alternatives, so call the doctor already.”

    Of course not and I wasn’t suggesting anything of the kind – do you truly believe it’s a question of medication or not treatment? There are all kinds of medical options and the first step would be getting a diagnosis from a professional that is a specialist in mental health if one is worried about one’s loved ones mental or emotional health. There are also all kinds of relevant questions to ask, such as if there’s a good reason for your loved one’s depression – is it event or context specific. If an emotional response is appropriate to a situation, it’s not actually a mental health issue but a healthy response and the actions needed may be to change the person’s environment or habits. Psychiatry (and psychology) offer treatment options other than just pharmaceuticals – ones that have no or fewer side effects and a lasting effect (CBT, for instance, and for mild depression exercise has been proven to help so being scornful of exercise is misplaced…I’m not in any way suggesting people should “just get over it” or shouldn’t be assisted in implementing needed changes). However, a great deal of SSRIs are prescribed by GPs who aren’t mental health professionals and without any recommendation for other treatments. (The best evidence for SSRIs and anti-depressants is when they’re used in conjunction with therapy.)

    Alison – “Note that I’m perfectly prepared to believe they are over prescribed. It’s just that observing that a lot of people take them is not sufficient evidence in itself.”

    No it’s not but there’s certainly enough evidence and questions from within the medical profession itself to indicate they often are for a variety of reasons that simply have to do with limited resources (and in some cases patients’ preference for a “magic pill”, which is how SSRIs have been sold by pharmaceutical companies until SBM called them on the dangers of prescribing to children, people with bipolar disorders, etc). In Canada and the UK there’s the matter of limited access to public mental healthcare and overloaded mental health systems, in the US there’s the matter of what insurance will cover (usually unlimited pharmaceutical treatments but limited talk or behavioral therapy). Part of the problem is when GPs end up treating conditions they’re not qualified to diagnose or treat (which is not necessarily the GPs fault, as noted above).
    http://www.guardian.co.uk/science/2004/mar/30/drugs.sciencenews

  23. weing says:

    Regarding the pharma shill gambit, the 3 physician authors of the study all have financial ties with various pharmaceutical companies as noted in the financial disclosures. To me it’s not surprising that the more severe the disease, the more evident the response to medication. This study is generalizable only to imipramine and paroxetine.

  24. Harriet Hall says:

    Fifi’s attack on Dr. Tuteur is out of line. Dr. Novella said he disagreed with her criticisms. I not only disagree, I find them offensive.

    Comments like
    “It really is distressing that you’d exploit this topic to further your own agenda under the pretense that you actually care about people with mental illness.”
    are simply despicable.

    “not creating the illusion that only pharmaceutical solutions are SBM.” It is really a stretch of the imagination to think that this post created any such an illusion.

    We get it, Fifi: you don’t like Dr. Tuteur or anything she writes. But please let’s limit the comments to discussing the content of the post, not your personal feelings about the author.

  25. weing says:

    I wonder if there are similar studies comparing CBT and exercise to placebo in patients with mild to moderate depression.

  26. Fifi,

    You can’t both argue that people could get superior treatment that wasn’t an SSRI, and also argue that SSRIs are prescribed because alternative treatments are not accessible.

    That was my point. The choice people have available to them typically is between suffering pointlessly and taking an SSRI. It is certainly possible to buy a book and give yourself CBT and embark on an exercise program, but it’s a lot easier to carry out if you are first able to get out of bed. If people have to do it for themselves, then SSRIs can play a role in enabling them. Also note: combined CBT and SSRIs are more effective than either alone.

    RE prescribing to children: the last I heard, adolescent suicide had been dropping steadily since the introduction of SSRIs. Since they started carrying a black box warning in the US about prescribing them to children, adolescent suicide has been rising. Prescribing them to children is not necessarily across-the-board stupid.

  27. “It is not an uncommon scenario at all for investigators seeking to do a meta-analysis not to be able to get access to adequate raw data from a significant fraction of relevant studies.”

    Why do you need raw data about each patient to do a metaanalysis?

    Most metaanalyses that I’ve read have not involved the inclusion requirement that the authors of the paper hand over their raw data.

  28. “Regarding the pharma shill gambit, the 3 physician authors of the study all have financial ties with various pharmaceutical companies as noted in the financial disclosures.”

    In addition, the disagreement cannot be characterized as those who favor enriching pharmaceutical companies compared to those who do not. The field of psychiatry has been riven by conflicts about the efficacy of drug treatment vs. talk therapy.

    Some insurance companies have favored drug treatment because it is cheaper and faster than talk therapy. Certain providers, on the other hand, have favored spending money on their services as opposed to spending it on medications.

    In other words, both sides have a financial “dog” in the fight. That doesn’t mean that those who favor one treatment over another are insincere or are doing it only for financial reasons. It merely means that the pharma-shill gambit does not come close to characterizing what is at stake.

  29. Fifi:

    Another way of putting it.

    You seem to be deducing that SSRIs are over-prescribed because non drug therapies also exist. This deduction is only valid if superior non-drug therapies are being displaced by inferior SSRIs.

    For instance, if people used to exercise a lot, but now that they can take SSRIs for depression they don’t bother any more. I don’t have any evidence that this is true. Do you?

    Alternatively, now that SSRIs are available, investment in reasearch into effective talk therapy has dried up. My superficial impression is quite the opposite. People know that SSRIs work and are part of a conventional armamentarium, and this makes intuitive sense as well as being intuitively appealing to certain people. This creates the idea of depression as being a treatable disease and not the moral failing that it subjectively feels like. They also hear, from the same popular sources, that CBT and exercise are equally effective for mild to moderate depression. There are a bunch of people who suddenly become much more interested in CBT, creating a market. Research into CBT has blossomed because if it can be classified as science-based medicine then insurance companies can pay for it on the same basis that they pay for SSRIs — because now that insurance companies are paying for SSRIs, cognitive-behabioural therapists want a piece of the pie.

    In the real world, I see no evidence that the availability of SSRIs has caused a drop in exercise. I have a superficial impression that the willingness of insurance companies to pay for SSRIs has resulted in a concomitant increase in the availability of CBT. I could be wrong — I don’t have evidence in either case. But the simple existence of exercise and CBT cannot be used as proof that SSRIs are over-prescribed. You must also establish that people who would otherwise be exercising and seeing a CB therapist are being blocked from doing so and are being prescribed SSRIs instead.

    Also, I repeat, cognitive-behavioural therapy works better when supported with an SSRI. People do get both: it’s not either/or.

  30. Plonit says:

    What do people make of the role of publication bias in proper evaluation of the efficacy of SSRIs?

    http://www.badscience.net/2008/02/619/

  31. moderation says:

    IMHO, there seems to be a lot of merit to the point several posters have made, that if there is over prescribing of SSRI’s by primary care MD’s … it is more the desire to do something for your patient than the influence of marketing. It is painfully difficult to find even a counselor for your patient, much less a psychiatrist or psychologist. If the delay is going to be several months, I can see where the PMD would feel the need to initiate some kind of therapy before then.

  32. Zoe237 says:

    THANK you for the Goldacre article Plonit. Funny how people can have the similar premises and come to different conclusions. The publication bias/file drawer effect was a little scary. I also didn’t realize that there had been previous meta-analysis done. I still don’t know who to believe in this case, haven’t researched it enough, but I think I’d probably try other things first if I were mildly depressed.

    Cue creepy Cymbalta music..

  33. micheleinmichigan says:

    I enjoyed Dr. T’s article. I did not find it to be bias or inflammatory. Not being a research or meta-analysis junkie, I was confused on the point about patient level data, somewhat clarified in the comments.

    The title in the NYT piece is obviously deceptive. One might as well say “Antibiotics are useless in mild moderate infections” Well, what kind of antibiotics, what kinds of infections?”

    Alison
    “I have a superficial impression that the willingness of insurance companies to pay for SSRIs has resulted in a concomitant increase in the availability of CBT.”

    Agreed, the three times I asked for a referral due to depression/anxiety through my insurance, I was referred to a CSW. They each gave me the option to consult with an on site psychiatrist for medication, which I declined twice and was not pushed on. Their treatment plans suggested CBT, cognitive behavior therapy, during which they suggested lifestyle options targeted toward my condition and CBT exercises. I was never offered the option of medication only. If I had asked I’m pretty sure that they would have made it clear that it was not recommended in their treatment plan. This was the approved process by three different insurance companies and three different Mental Health Offices in my area.

    Just a general point, CBT makes losing 200pounds , 1 pound a week look easy. To explain, CBT encompasses several methods, it may be the process of changing or attempting to change very ingrained and reflexive thought patterns that you have hundreds of times a day. Or it can be the process of exposing yourself to increasing levels of anxiety in order to acclimate yourself, etc. It can be very helpful, even essential, but I can honestly say for each 10 times I attempt to use CBT techniques to lower anxiety, I probably fail 8 times. That is one way that SSRIs can help. They can relieve some symptoms enough to increase success with CBT techniques. Once the CBT techniques become more habitual, those good habits will tend to continue, even after SSRI is discontinued. Of course these good habits often wan after a while or something triggers a flare in symptoms. That is why mental health issues or often chronic. At least that is my experience.

    If I had gone to a GP with the same complaint, I may have been offered Paxil or the like. I’m not sure if my doctor would have, but probably someone in the office might. As I’ve said in another post, I’m not wild about that idea.

    I do think some GP underestimate side effects, I also have known three people who were given SSRI’s by different GPs and not told that they needed to taper off when they stopped. That’s not good. I also think different SSRIs tend to work better with different issues or have different side effects and GP are not as qualified in diagnosing those issues or side effects.

    Once again good post Dr. T. I can not agreed with you on some things, but I feel you are beginning to find your stride.

  34. David Gorski says:

    Why do you need raw data about each patient to do a metaanalysis?

    Most metaanalyses that I’ve read have not involved the inclusion requirement that the authors of the paper hand over their raw data.

    You don’t need the raw data about each patient to do a run-of-the-mill meta-analysis. (I realize now that I may have implied that in my comment. My bad.) However, the most rigorous meta-analysis methods do require the use of the original data if it is obtainable. Investigators calculate the common estimate and its confidence interval using these studies. To do this most accurately it’s best to have the original data from all the studies if they can be obtained. This data can then be combined into one large data file with study as one of the variables. Obviously, it’s not always possible to get this data; often it is not. That’s why, alternatively, investigators may only have the summary statistics obtained from publications themselves. Meta-analysis can be done with these statistics.

    In fact, the reason the investigators decided to look only at studies whose original data they could access was listed right in the methods section of the paper you discussed. It’s not hidden; it’s right there in black and white:

    Because most MDD studies incorporate a minimum baseline depressive severity score as an inclusion criterion, studies of minor depressive disorder (which do not typically have such strict thresholds) were included in this analysis as well. The entry criteria allowed patients to enter these studies with HDRS scores that ranged from the low teens to the upper 30s.11-16 Unlike the data analyzed by Kirsch et al and Khan et al, which contained information only at the level of treatment group and thus could support only standard meta-analytic procedures, the databases from the 6 studies included in the present investigation provided data for a patient-level meta-analysis, also known as a mega-analysis. This approach is more appropriate and more powerful than a standard meta-analysis when original data are available and a fine-grained multivariate analysis is desired.17

    Reference 17 is http://aje.oxfordjournals.org/cgi/content/abstract/145/10/917?ijkey=1ccdf76e59c411bc994e71e5b5f517082832ca5c&keytype2=tf_ipsecsha

    So there you go. They wanted to do a more powerful analysis involving fine-grained multivariate analysis, and that’s why they chose their inclusion criteria. They didn’t just “ignore” the 17 other studies; they tell the reader why they decided not to include them. You can agree or disagree with their reasons, but their reasons are right there in the paper.

    Again, while I agreed with the rest of your post, that one statement stuck out as not being a particularly valid criticism of the article.

  35. micheleinmichigan says:

    FIFI – it’s pretty clear that your opinion of Dr. T falls within a certain framework. While you have a right to your opinion, four or more posts per topic restating the same arguments just gets a little dull and wearing.

    Have you realized yet that if you don’t like Dr. T’s style, opinions, personal philosophy, you could NOT read her articles. It is a very viable option. I do it all the time with the Wall Street Journal. I just don’t read it.

    Or perhaps you could limit your one sided critique to one post? That way any new readers would be appraised of your dislike, but regular readers would not have to scroll past the numerous accusations and defenses.

  36. bluedevilRA says:

    Dr. Tuteur takes way too much flak in my opinion. This was an enlightening post on a controversial topic.

    Thanks for pointing out some interesting flaws in the study, Dr. Tuteur!

  37. Plonit says:

    And thanks to Dr Gorski for pointing out some interesting flaws in Dr Tuteur’s pointing out of some interesting flaws.

  38. micheleinmichigan,

    Interestingly, the one person I know who bought a book and successfully gave herself CBT ran a marathon that year and had also, years previously, lost 100 pounds at a pound a week.

    If insurance companies are using a patient’s refusal to participate in CBT as an excuse to refuse to pay for SSRIs, a cynical person might speculate about their enthusiasm. Interesting.

  39. micheleinmichigan,

    Interestingly, the one person I know who bought a book and successfully gave herself CBT ran a marathon that year — and years previously had lost 100 lbs at a pound a week.

    If insurance companies are using a patient’s refusal to participate in CBT as an excuse to pay for SSRIs, a cynical person might speculate about their enthusiasm. Interesting.

  40. … sorry, that should be “an excuse NOT to pay for SSRIs.”

  41. ” They wanted to do a more powerful analysis involving fine-grained multivariate analysis, and that’s why they chose their inclusion criteria”

    Or they did it simply to exclude studies they wanted to exclude.

    I’d have no trouble accepting your explanation if the authors had emphasized that because of their very strict inclusion criteria, and the fact that those criteria yielded a mere 3 studies of only one SSRI, their results cannot be generalized. Instead the did the opposite.

    Can metaanalysis yield valid results when it includes only a fraction of the existing studies merely because the authors of the other studies refused to participate? I think it is up to the authors of the metaanalysis to demonstrate that the included studies are representative of all the existing studies before presuming to generalize about anything.

    At a minimum the authors of the metaanalysis could have included a data level metaanalysis of all 23 studies to show how their patient level subset analysis compared.

    The authors did not even put the citations for the excluded studies into their paper, but instead relegated them to a web only e-table that I cannot seem to access from the website, making it even more difficult to compare the included studies with the excluded studies. If anyone can gain access to it, I’d be grateful if he or she passed it along.

  42. David Gorski says:

    Or they did it simply to exclude studies they wanted to exclude.

    I’d have no trouble accepting your explanation if the authors had emphasized that because of their very strict inclusion criteria, and the fact that those criteria yielded a mere 3 studies of only one SSRI, their results cannot be generalized. Instead the did the opposite.

    Can metaanalysis yield valid results when it includes only a fraction of the existing studies merely because the authors of the other studies refused to participate? I think it is up to the authors of the metaanalysis to demonstrate that the included studies are representative of all the existing studies before presuming to generalize about anything.

    They did not hide the list of studies or, as far as I can tell, put them in a PDF on the website to make it hard to examine; listing such “peripheral” information in a file on the website is, unfortunately, just the way things are done these days in publishing scientific and medical studies. Readers can judge whether or not they excluded studies that shouldn’t have been excluded; a table lists the criteria and experimental design of each of the 23 studies. Based on my experience publishing studies during the last five years, I’d be willing to bet that the editors of JAMA made the authors list their studies and their table describing them in an extra file on the JAMA website for space reasons; it’s an extra table, and it’s a pretty big table, and the list of citations is a fair amount of text. In fact, I can’t prove it, but based on my experience I’d be willing to bet that the authors probably included the table in the first submission of their manuscript. Editors these days are insisting that more and more stuff be included as “Supplemental Data” or “Supplemental files” instead of being in the paper itself.

    Be that as it may, here are the 23 studies, cut and pasted out of the e-document just for you:

    1. Barrett JE, Williams JW Jr, Oxman TE, et al. Treatment of dysthymia and minor depression in primary care: a randomized trial in patients aged 18 to 59 years. J Fam Pract. 2001;50(5):405-412.
    2. DeRubeis RJ, Hollon SD, Amsterdam JD, et al. Cognitive therapy vs medications in the treatment of moderate to severe depression. Arch Gen Psychiatry. 2005;62(4):409-416.
    3. Dimidjian S, Hollon SD, Dobson KS, et al. Randomized trial of behavioral activation, cognitive therapy, and antidepressant medication in the acute treatment of adults with major depression. J Consult Clin Psychol. 2006;74(4):658-670.
    4. Elkin I, Shea MT, Watkins JT, et al. National Institute of Mental Health Treatment of Depression Collaborative Research Program: general effectiveness of treatments. Arch Gen Psychiatry. 1989;46(11):971-982.
    5. Philipp M, Kohnen R, Hiller KO. Hypericum extract versus imipramine or placebo in patients with moderate depression: randomised multicentre study of treatment for eight weeks. BMJ. 1999;319(7224):1534-1538.
    6. Wichers MC, Barge-Schaapveld DQ, Nicolson NA, et al. Reduced stress-sensitivity or increased reward experience: the psychological mechanism of response to antidepressant medication. Neuropsychopharmacology. 2009;34(4):923-931.
    7. Boyer P, Montgomery S, Lepola U, et al. Efficacy, safety, and tolerability of fixed-dose desvenlafaxine 50 and 100 mg/day for major depressive disorder in a placebo-controlled trial. Int Clin Psychopharmacol. 2008;23(5):243-253.
    8. Cohn CK, Robinson DS, Roberts DL, Schwiderski UE, O’Brien K, Ieni JR. Responders to antidepressant drug treatment: a study comparing nefazodone, imipramine, and placebo in patients with major depression. J Clin Psychiatry. 1996;57(Suppl 2):15-18.
    9. D’Amico MF, Roberts DL, Robinson DS, Schwiderski UE, Copp J. Placebo-controlled dose-ranging trial designs in phase II development of nefazodone. Psychopharmacol Bull. 1990;26(1):147-150.
    10. DeMartinis NA, Yeung PP, Entsuah R, Manley AL. A double-blind, placebo-controlled study of the efficacy and safety of desvenlafaxine succinate in the treatment of major depressive disorder. J Clin Psychiatry. 2007;68(5):677-688.
    11. Feiger AD. A double-blind comparison of gepirone extended release, imipramine, and placebo in the treatment of outpatient major depression. Psychopharmacol Bull. 1996;32(4):659-665.
    12. Feiger AD, Rickels K, Rynn MA, Zimbroff DL, Robinson DS. Selegiline transdermal system for the treatment of major depressive disorder: an 8-week, double-blind, placebo-controlled, flexible-dose titration trial. J Clin Psychiatry. 2006;67(9):1354-1361.
    13. Gastpar M, Singer A, Zeller K. Comparative efficacy and safety of a once-daily dosage of hypericum extract STW3-VI and citalopram in patients with moderate depression: a double-blind, randomised, multicentre, placebo-controlled study. Pharmacopsychiatry. 2006;39(2):66-75.
    14. Hollyman J, Freeling P, Paykel E, Bhat A, Sedgwick P. Double-blind placebo-controlled trial of amitriptyline among depressed patients in general practice. J R Coll Gen Pract. 1988;38(314):393-397.
    15. Lieberman DZ, Montgomery SA, Tourian KA, et al. A pooled analysis of two placebo-controlled trials of desvenlafaxine in major depressive disorder. Int Clin Psychopharmacol. 2008;23(4):188-197.
    16. Liebowitz MR, Manley AL, Padmanabhan SK, Ganguly R, Tummala R, Tourian KA. Efficacy, safety, and tolerability of desvenlafaxine 50 mg/day and 100 mg/day in outpatients with major depressive disorder. Curr Med Res Opin. 2008;24(7):1877-1890.
    17. Liebowitz MR, Yeung PP, Entsuah R. A randomized, double-blind, placebo-controlled trial of desvenlafaxine succinate in adult outpatients with major depressive disorder. J Clin Psychiatry. 2007;68(11):1663-1672.
    18. Mynors-Wallis LM, Gath DH, Lloyd-Thomas AR, Tomlinson D. Randomised controlled trial comparing problem solving treatment with amitriptyline and placebo for major depression in primary care. BMJ. 1995;310(6977):441-445.
    19. Rickels K, Case WG. Trazodone in depressed outpatients. Am J Psychiatry. 1982;139(6):803-806.
    20. Rickels K, Weise CC, Zal HM, Csanalosi I, Werblowsky J. Lofepramine and imipramine in unipolar depressed outpatients. A placebo controlled study. Acta Psychiatr Scand. 1982;66(2):109-120.
    21. Septien-Velez L, Pitrosky B, Padmanabhan SK, Germain JM, Tourian KA. A randomized, double-blind, placebo-controlled trial of desvenlafaxine succinate in the treatment of major depressive disorder. Int Clin Psychopharmacol. 2007;22(6):338-347.
    22. UK Moclobemide Study Group. A multicentre comparative trial of moclobemide, imipramine and placebo in major depressive disorder. Int Clin Psychopharmacol. 1994;9(2):109-113.
    23. Versiani M, Nardi AE, Mundim FD, Alves A, Schmid-Burgk W. Moclobemide, imipramine and placebo in the treatment of major depression. Acta Psychiatr Scand Suppl. 1990;360:57-58.

    As for the point of doing the summary statistics-level meta-analysis first, it is quite possible that they could not do it and make the results comparable for various outcomes. Personally, I consider concluding that the investigators are cherry picking studies, which is in essence what you are accusing them of, to be leaping to conclusions on the basis of no evidence, particularly given that it’s spelled out in black and white exactly how the studies were chosen. You might disagree with how the studies were chosen, but there’s nothing there to indicate cherry picking that I can find.

  43. Thanks for the list! I’ll take a look at them.

    What criteria do you think I should use in determining if the cherrypicked the studies? I want to give the authors the benefit of the doubt.

  44. micheleinmichigan says:

    “If insurance companies are using a patient’s refusal to participate in CBT as an excuse to pay for SSRIs, a cynical person might speculate about their enthusiasm. Interesting.”

    Did I say that? I didn’t mean to say that. I meant to say that the CSW didn’t push SSRI over therapy and it seemed they might discourage SSRI without therapy.

    As to your one friend with the self-help book. Wow, I’m impressed. I did get some good effect practicing techniques from one self help book recommended by a CSW, really was feeling free from anxiety for about a month. Unfortunately the positives effects waned. Still good techniques but anxiety seems to adjudge even to CBT therapy.

  45. micheleinmichigan,

    No, you didn’t say that! I said that. I don’t know if anyone’s ever been denied drug therapy for minor/moderate/major depression or anxiety on the grounds that they weren’t interested in CBT. It’s just that your story suggested that possiblility to me — that it might happen — which had never occurred to me before.

  46. micheleinmichigan says:

    sorry typo, anxiety adjusts to CBT therapy

  47. micheleinmichigan says:

    It’s just that your story suggested that possiblility to me — that it might happen — which had never occurred to me before.

    Oh good, sometime I don’t know what I’m say. ;)

    I don’t have any reason to believe that they would have ultimately denied SSRI without therapy. To much second guessing, that.

  48. Okay, I just finished reviewing the abstracts for all 23 studies. The differences between the included studies and the excluded studies are dramatic.

    Of the 6 included studies 5 showed no benefit for antidepressant treatment of mild to moderate depression. The sixth study did not mention HDRS scores in the abstract and I haven’t yet obtained the paper.

    Of the 17 excluded studies 15 DID show effectiveness of the antidepressant treatment of mild to moderate depression. Two studies showed no difference compared to placebo.

    By insisting on access to patient level data, the authors effectively excluded EVERY study that showed improvement of mild to moderate depression by treatment with antidepressants. And considering that fully 65% of the 23 studies showed antidepressants to be effective, there is no possible justification for Dr. DeRubeis’ public claim:

    “The message for patients with mild to moderate depression is, ‘Look, medications are always an option, but there’s little evidence that they add to other efforts to shake the depression …”

  49. micheleinmichigan says:

    Regarding the idea that study authors who don’t read the PR department Press Release.

    I guess I’m the only one who thinks that would be a great opportunity to get a job in their PR department and get creative with the press releases.

    “meta-analysis shows the uselessness of meta-analysis” and “95% of women think the author should shave off the ponytail” or of course the classic “Study shows that Bob is a Butthead”

    I actually don’t believe any of those statements, but I’m pretty sure it would teach researchers to read their press releases.

  50. pmoran says:

    “By insisting on access to patient level data, the authors effectively excluded EVERY study that showed improvement of mild to moderate depression by treatment with antidepressants. And considering that fully 65% of the 23 studies showed antidepressants to be effective, there is no possible justification for Dr. DeRubeis’ public claim:”

    Well, I don’t know. Even before this study, I was concerned as to whether it is a good thing that doctors are using drugs like this like lollies, for common, mild, depressive states.

    Any negative studies are a worry, because known biases tend to favour positive results. One important bias that is not mentioned even in relation to these selected studies is whether patients were able to tell that they were taking an antidepressant through their milder side effects.

    One aspect — an effect size of 0.2 is LESS than that of placebos overall (0.3 — see Hrobjartsson et Al) , with presumably fewer ill effects depending on the one chosen. We cannot know for sure what the effect sizes were for mild depression in those studies that have not offered primary data.

  51. JoshS says:

    I credit the fact that I’m even alive (and, thankfully, relatively happy , sane, and productive) to SSRIs. Right upfront: I know that my anecdotal experience is not data, and that I’m not immune to the placebo effect. But as someone who’s been on them for almost 20 years — and whose life went from a hellish cycle of suicide attempts, debilitating panic attacks and sleep entirely disrupted by obsessive compulsive problems, to a normal, productive life — I get weary reading about how SSRIs are “overprescribed” or perhaps ineffective for milder mental problems. That may well be, and yes, I’m interested in factual efficacy data, whether or not I happen to like that data.

    But there are people out there who have suffered from severe, horrifying mental debilities, people like me, who have been saved from indescribable suffering and likely suicide through SSRI therapy. There’s no way to say that without sounding melodramatic, but it’s true. I would most certainly be dead by my own hand had my family not forced me to try medication. Given how mentally tormented I was with depression, panic, and OCD, I think I would have chosen death rather than continue suffering that way.

    People with those severe problems don’t seem to rate much attention anymore. If SSRIs are discussed publicly, it’s usually to talk about how they’re “misused,” “over-prescribed,” “over-hyped,” etc. It’s easy to look down our societal nose at the worried well who get a pill they don’t need, but why is the conversation about SSRIs so focused on that? Why isn’t more attention paid to the tremendous difference they’ve made for people with severe mental illness?

    Yes, it’s possible that my response to SSRIs is all placebo. We are, every one of us, capable of fooling ourselves. But given the severity of my illness, given the fact that no drug therapy of any sort did a bit of good, given how I gradually (almost unknowingly) became a *normal* person and “woke up” a few weeks into SSRI therapy, I find the placebo hypothesis unlikely. There are many people out there like me who thank fate, the universe or their God for these medications.

    That’s enough self-indulgence from me. I just wanted to remind everyone there’s a crucial place for psychoactive medications, even if there’s legitimate doubt about their scope of efficacy.

  52. micheleinmichigan says:

    JoshS – Thanks for your stories. I think you make some incredibly important points. The stories of all the people helped by anti-depressants often are not told because the social stigma that still remains attached to mental health issues.

  53. JoshS says:

    given the fact that no drug therapy of any sort did a bit of good</blockquote.

    CORRECTION: I meant to write, "given the fact that no non-drug therapy of any sort did a bit of good. . .”

    That is to say, no talk-therapy of any stripe worked. The only thing that worked for me was medication.

  54. JoshS says:

    given the fact that no drug therapy of any sort did a bit of good

    CORRECTION: I meant to write, “given the fact that no non-drug therapy of any sort did a bit of good. . .”

    That is to say, no talk-therapy of any stripe worked. The only thing that worked for me was medication.

  55. Zoe237 says:

    I have to agree with pmoran, and am confused by why Dr. Tuteur is (possibly) assigning nefarious intentions to the authors. This is not the only analysis that has been done showing little to no effect- Ben Goldacre mentioned another in the linked to article in 2008.

    Furthermore, it is certainly plausible that the risks of SSRIs for mildly depressed patients outweigh the benefits. It is also possible, that due to drug study funding, and publication bias, that people think SSRIs are more beneficial than they really are (again, for mild depression). Somebody mentioned antibiotics- these are drugs that shouldn’t be used for every last infection, and using them as such has lead to loads of unintended consequences. I am also curious, after reading Dr. Hall’s critique mentioned in the blog, why the media apparently lied about SSRIs being correlated with suicide in adolescents. Is there a reliable media source of medical/scientific studies for the layperson (besides this site!)? I guess NYTimes ain’t it.

    So at this point, I don’t know who to believe. I guess I’d start by looking at RR and confidence intervals in the 17 other studies. Is is possible that severe depression is caused by chemical imbalances and mild depression is caused by something else, therefore explaining the (possible) reduction in efficacy?

  56. “Furthermore, it is certainly plausible that the risks of SSRIs for mildly depressed patients outweigh the benefits. It is also possible, that due to drug study funding, and publication bias, that people think SSRIs are more beneficial than they really are (again, for mild depression).”

    It is important to separate the analysis of this specific study from the larger issue of whether SSRIs are effective for mild to moderate depression.

    My point about this study is pretty basic: you can’t use three studies on Paxil to draw ANY conclusions about the effectiveness of SSRIs as a class on mild to moderate depression.

    In terms of the larger issue, the effectiveness of all SSRIs, the clinical question is not pharmaceuticals vs. nothing; rather it is medication vs. talk therapy. Both sides have vested personal and financial interests in the outcome. Both sides have the unfortunate tendency to look at it as a zero sum game, assuming that money given to pharmaceutical companies is money taken away from psychologists and other mental health professionals who don’t have prescribing power.

    It seems to me that the data in total indicate that treatment of depression (whether mild, moderate or severe) needs to be individualized. Moreover, the combination of both modalities can be more effective than either alone.

    Three studies on one SSRI cannot offer much guidance on the real clinical issues.

  57. Zoe237 says:

    “Both sides have the unfortunate tendency to look at it as a zero sum game, assuming that money given to pharmaceutical companies is money taken away from psychologists and other mental health professionals who don’t have prescribing power.”

    Yes, I hadn’t considered the financial interests on the other side. Goldacre mentioned the media having a “good stamp” and a “bad stamp.”

  58. Thanks, JoshS. And… yes.

    “Even before this study, I was concerned as to whether it is a good thing that doctors are using drugs like this like lollies, for common, mild, depressive states.”

    pmoran, it’s the patients who use SSRIs. If a doctor is using them it’s because they are a patient. Nobody is forced to take them. In our puritan anglo societies most people are reluctant to have recourse to drugs for what is subjectively felt as a moral failing. (See JoshS, and he was not suffering from “mild” depression.)

    No doctor I met with was handing out SSRIs like lollies. I actually found them quite hard to get. I was seeing a psychotherapist who actively blocked my access. My GP told me that there was nothing wrong with me that a nice long psychoanalysis wouldn’t fix; she also observed that it was too bad that I obviously didn’t have the self-discipline to see it through. And that ended that. When I eventually got a prescription for SSRIs because I was no longer able to keep it together and I was obviously disabled even to the casual observer, my pharmacist openly wondered why I would need them.

    It would be nice to have some evidence that doctors handing out SSRIs like lollies is a problem.
    1) That they do indeed hand out SSRIs like lollies.
    2) That the harm caused by making SSRIs available to people who might not need them outweighs the harm that might be caused by making them less available to people who do.

    It’s true, a lot of people take them. But you can’t deduce from that fact alone that they shouldn’t.

  59. “Both sides have the unfortunate tendency to look at it as a zero sum game, assuming that money given to pharmaceutical companies is money taken away from psychologists and other mental health professionals who don’t have prescribing power.”

    Certainly true in my experience. My psychologist used various methods to prevent or discourage me from using medication. When I finally did, I started functioning better and taking more control of my life. I also found talk therapy completely unengaging. When I told my psychotherapist this she said it was normal when taking SSRIs and that I should consider not taking them. When my doctor asked if I was seeing a psychotherapist I said yes, but that now that I was on SSRIs it was completely unengaging. My doctor suggested that I stop seeing the psychologist.

    In this case, they were both getting paid and they still put me in the middle of their battlefield.

    This was the first time I had been given permission not to engage in psychotherapy because “everyone knows” that it’s what you’re supposed to do. Psychotherapy didn’t help me, which I assumed was my fault. Whether because I wasn’t working hard enough, or because I wasn’t advocating for myself appropriately, or because I hadn’t shopped properly for just the right match for me. I’m really not sure how someone with major depression trying to survive day to day is supposed to do all these things, but all the patient-oriented literature is clear: you are supposed to see a psychotherapist, and if they don’t help you it’s because you’re doing it wrong.

    Anyway, though I resented being the monkey in the middle between my psychotherapist and my GP, my GP did give me permission to quit psychotherapy. Which I did, to my great relief.

    The huge advantage of SSRIs over talk therapy, from my perspective, is that they give control to the patient. Don’t like the meds? Nasty side effects? Not working? Then don’t take them. It’s the meds’ fault, not yours, and your doctor will prescribe something else or refer you to a specialist. Or you can just stop. Easy-peasy. Meds don’t make you feel guilty. They don’t tell you that if you are friends with your mother that you are in denial because all thirtyish women want to complain about their mothers. They don’t take your money and then spend your fifty minutes telling you about their new baby, or what it was like for them to be poor when they were a young newlywed. They don’t angrily demand to know who told you that you were incompetent when you were a child, when your experience was one of glowing expectations. Most of all, they don’t tell you that if you are more depressed and less functional than ever before in your life that it’s because you’re doing such wonderful work in therapy and it’s a sign of progress. When you take meds, getting worse is always a sign that something’s wrong.

    Talk therapy has side effects too, and someone who is depressed may not be able to deal with them effectively. SSRIs have side effects that are much less emotionally charged, more concrete, and might be more easily dealt with by someone who is not coping with anything very well.

    … anyway, I got sidetracked. Yes, MDs and psychologists fight over patients. Even when they are both getting paid.

  60. DanaUllman says:

    I think that this blog proves that anyone can be deemed to be using the straw man argument no matter what side of the argument you take.

    I love it when “skeptics” use their own jargon on themselves, and then, they defend themselves by calling the other side something similar to what they have been called.

    For those of you who believe that Prozac is effective for moderate or severe depression will benefit from knowing about a randomized double-blind trial that compared Prozac with homeopathic medicines…it showed non-inferiority and greater safety to the homeopathic medicine.

    So, either Prozac works and homeopathy works for moderate or severe depression OR neither works. What is it going to be?

    ecam.oxfordjournals.org/cgi/content/abstract/nep114

    I love this challenge for you because you are all officially in-between a rock and a healing place…

  61. lillym says:

    I have some anecdotal evidence about this situation and also about GP’s diagnosing mental illness and prescribing anti depressants.

    As I mentioned in another post I have bipolar disorder, so I don’t fall into the mild-moderate category of depression. However, many of my family members do. This is kind of long winded, please bear with me:

    I started having emotional and behavior problems when I was about 10 years old, by the time I was 14 (in 1987) I was severely depressed, I was barely attending school and my depressions were so bad I was barely functional, it was all I could do to get out of bed some mornings, I went days without taking a shower because I didn’t have the energy, I pretty much sat and cried, but there were times when I was overwhelmed with black rages that just came over me. At this point I’d been seeing a psyhcologist for about 4 years, not the same one I was on my third.

    The best thing they could come up with is if I wanted to I could be like other girls, and if Iloved my Mom I wouldn’t keep doing things to make her cry. I’d also been tested for about every illness you can think of, gone to multiple specialists, and even had been evaluated at a children’s hospital, my mom was desparate to find out what was wrong with me. Finally she took me to a new GP, he evaluated me, gave me one of those mood checklists, and told me he could help me. He said he thought I had clinical depression and he told me there was hope. No cure, no magic pill, but medicine that would make me feel better but I’d need therapy. He recommended a therapist and I went to see. And sat in her office and cried for 8 consecutive visits.

    She diagnosed me with atypical clinical depression started me on anti depressants and they seemed to help.

    Obviously the GP and the therapist misdiagnosed me, I had pediatric bipolar disorder. But I can’t blame them, pediatric bipolar disorder didn’t come on the DSMV until 2 years after I graduated high school and even after that I ran into therapists who didn’t believe children could have mental illnesses.

    The antidepressants helped more than they hurt me, it got me stable, made me able to work through some issues and able to live a mostly normal life.

    But it also got my family help. My therapist talked to my parents and my brother, she took extensive family history, she talked to my paternal grandmother. She ended referring my father to a psychologist who diagnosed him with depression. He wasn’t unable to get out of bed, he has a PhD and a successful career but he had to work much harder at it because of his depression. He started taking antidepressants and his life got easier, it got better, he got better.

    I now know that 2 of my father’s siblings have also been treated for mild-moderate depression, I have 2 first cousins who have been treated for depression. And we’re fairly sure that my paternal grandfather (and his mother) had some form of mental illness.

    Both my brother and mother have used anti depressants for short term depression. Both of them have been in situations where changes needed to happen but because of the depression they couldn’t make those changes. With a combination of medicine and therapy they were able to work through those issues and end both medicine and talk therapy.

    My maternal grandmother also had a rough several years where she needed a combination of medicine and therapy. She never wanted me to be on anti depressants and told my mother that it was a mistake. However, after my grandmother’s experience she apologized, she said she understood.

    On my mother’s side I know of 2 cousins who have had mental health issues treated and one who is in serious denial.

    Over my 20+ years dealing with various psychiatrists, GPs, psychologists, and therapists I’ve seen good ones and I’ve seen ones who should have had their licenses revoked. The good ones have always wanted me to be on as little medicine as possible.

    When I was first diagnosed with bipolar disorder 10 years ago I was started on anti anxiety meds, mood stabilizer, and anti psychotic medication. I went from having multiple anxiety attacks per day to not needing any anti anxiety medication for nearly 7 years. I no longer get depressive episodes several times a year so I don’t take the mood stabilizers.

    But when I’m off the anti psychotics my mind still races, I can’t concentrate, and I begin to start the cycle of self doubt and recrimination, but I’m tapering to a lower and lower dose of that.

    It took about 3 years just to get to the right mood stabilizer and anti psychotic. I had terrible side effects with some I tried and others didn’t work.

    But I’m here now healthier than I have ever been.

    I’ve seen my family members have better, happier, more productive lives because of anti depressants. I’m fairly sure that if my great grandmother had access to mental health services she wouldn’t have been the bitter, needy, mean spirited woman she turned out to be.

    I want to say that I’ve had some really awesome therapists and some really good talk therapy and I’ve had bad therapists. I’ve had good psychiatrists and bad pyschiatrists.

    I haven’t tried CBT, I want to, but the last time I checked around either the therapists offering it didn’t accept my HMO or they weren’t taking new patients or I was considered to healthy for the program.

  62. lillym says:

    Alison,
    I want to say how sorry I am you had that experience. I’ve never been put in the middle by my GP and my therapist and psychologist.

    There was probably a better choice of therapist for you and the therapist and your GP should be working as a team for you.

    I’ve had to fire therapists because they weren’t working out. And then I had to find a new one and luckily I only had to do it a few times.

    It’s like trying to find the medicine that works, the first or second aren’t right so you try again.

  63. David Gorski says:

    I have to agree with pmoran, and am confused by why Dr. Tuteur is (possibly) assigning nefarious intentions to the authors.

    Certainly that’s how I interpreted it, namely that Dr. T was assigning nefarious motives to the researchers with regard to how they designed the study, although I agree with her that in their public statements they went beyond what the study could justify. Let’s put it this way. I’m a researcher; I get a bit touchy when I see other researchers seemingly accused of shadiness or deception on the basis of no evidence and based on not knowing enough about meta-analyses to understand that wanting to use patient-level data is a perfectly legitimate thing in doing a detailed meta-analysis (hence my reaction). If Dr. T had simply stuck to pointing out that only 6 of 23 studies ended up being included because of the inclusion criteria, that would have been fine, but she had to go beyond that and insinuate that it was done in order to cherry pick negative studies.

  64. srdau says:

    Amy,

    Just one nitpick. In paragraph 4 you define SSRIs as “selective serotonin receptor inhibitors”. They’re actually selective serotonin reuptake inhibitors.

  65. Plonit says:

    Dr. T was assigning nefarious motives to the researchers with regard to how they designed the study,

    ++++++++

    But that’s Dr Tuteur’s MO – surely you knew that before you made her part of the SBM team. See how she does the same thing over exclusions here….then refuses to concede the point in the comments….

    http://homebirthdebate.blogspot.com/2008/10/new-cochrane-study-promptly.html

    What she doesn’t seem to get is that researchers for the most part decide their method, including exclusion/inclusion criteria – and in the case of Cochrane publish their protocol – before they even do the literature search and start crunching the numbers.

    In the case discussed above, they needed the data in order to do the analysis they wanted to do. So, they approached authors for the data and were declined. They had no way of knowing in advance which authors would decline to share the data on which their findings were based. As it turns out there seems to be some sort of correlation between having positive findings and being unwilling to share your data. That could be interpreted more than one way, if we are willing to assign nefarious intentions.

  66. “So, either Prozac works and homeopathy works for moderate or severe depression OR neither works. What is it going to be? ”

    Either or? Why do you suggest those are the only two choices? Why do you imply that ONE study is definitive?

  67. David Gorski says:

    In the case discussed above, they needed the data in order to do the analysis they wanted to do. So, they approached authors for the data and were declined. They had no way of knowing in advance which authors would decline to share the data on which their findings were based. As it turns out there seems to be some sort of correlation between having positive findings and being unwilling to share your data. That could be interpreted more than one way, if we are willing to assign nefarious intentions.

    Indeed.

    That’s actually a rather interesting observation and one that I hadn’t thought of. Why would the researchers who had more positive results be less willing to share their data than the ones whose studies were mostly negative? One possible reason is that perhaps the ones who had positive results were funded by drug companies, and the company wouldn’t allow release of patient-level data. Other possibilities, such as a random quirk, also present themselves.

    Be that as it may, unless Amy can demonstrate that they did a little post hoc rearranging of their methodology in order to make sure to exclude the positive studies, I consider her criticism about the study’s inclusion criteria to be off base, even as I point out that I do agree with her when she criticizes the investigators for going far beyond what this study can support when talking to the press.

  68. psychability says:

    I appreciate Dr Tuteur’s deconstruction of this research methodology as well as the honest criticism of her methods by her colleagues and others.

    This discussion highlights the impossibility of simplifying the factors in antidepressant response to the degree that we can glean useful clinical information. As psychiatrists wanting science-based credibility, we have shot ourselves in the foot by seeking sensationalism, getting into bed with drug companies and promoting quick fixes for complex psychosocial problems. The diagnostic criteria used in psychiatric research are based on artificial constructs and have been broadened to a degree that I hypothesize many of the subjects with mild “major” depression have a condition that is far removed from that of the patient who is unable to get out of bed or feed himself.

    As the 2013 release of new diagnostic criteria approaches, I would like to see a rational science-based evaluation of the factors that are important in diagnosis and treatment decisions. We are seeing lots of political and other agendas, but what does science actually know about psychiatric illness?

  69. “that’s how I interpreted it, namely that Dr. T was assigning nefarious motives to the researchers with regard to how they designed the study”

    I never used the word nefarious. That was someone else’s claim. Nor would I use the word nefarious to describe the way that data can be sliced and diced to show what the researchers hoped it would show.

    Publication bias is not nefarious; financial ties to industry (if appropriately disclosed) are not nefarious; even cherry picking studies for inclusion is not nefarious. In many cases, it is simply human nature. The people who write papers that are affected with these issues are not evil people planning to deliberately hookwink the population. They believe what they write; they don’t think of themselves as either biased or having an agenda. But one of the reasons we insist on financial disclosures in scientific papers is because we understand that human beings are often unaware of how incentives can influence actions.

    I looked at this study the same way I look at any study. The first question is whether the data in the study justifies the conclusions. As I said in the post, although the authors technically “showed” that their metaanalysis finds that Paxil has no effect on mild to moderate depression, their ultra strict inclusion criteria (which left them with only 3 studies) means that their findings are not generalizable in any way.

    Although I understand why someone might choose patient level analysis over group level analysis in constructing a metaanalysis, I personally cannot see how that makes this paper more reliable. Indeed, I think it makes the paper far less reliable since the majority of studies were excluded by this requirement. My feeling is that if the authors were unable to obtain patient level data for the majority of the studies, they had no business writing a paper that claimed to be a metaanalysis of the existing research. The raw data was unavailable TO THEM, but it exists. They are not entitled to simply leave it out.

    Everyone can judge the quality of the paper for himself or herself. I simply think that it is critical for people to understand what was in the paper and what was left out. As for me, I acknowledge that I don’t think it was a coincidence that ALL 15 papers that showed a beneficial effect of antidepressants in mild to moderate depression were excluded. And we haven’t even addressed the issue that 118 OTHER papers were excluded because they used a placebo washout period. As the sheer volume of those papers indicates, use of a placebo washout period is standard in research on antidepressants.

    “in their public statements they went beyond what the study could justify.”

    We agree on the claim at the heart of my post. The study does not show that “antidepressants” are ineffective for mild to moderate depression.

    It is the authors’ public behavior that made me examine the paper very carefully. I have no trouble saying that the authors were highly irresponsible and clearly pushing a personal agenda in generalizing from such a tiny study, in The New York Times, no less. Newsweek has devoted a cover story to the “findings” of this paper, even though we agree that the paper is not generalizable. The authors have done some serious damage and journalists helped them do it.

    Are the authors motives nefarious? I doubt it. I suspect that they genuinely believe that antidepressants should not be used to treat mild to moderate depression. However, that’s not what their paper showed; it didn’t really show anything at all.

  70. “In paragraph 4 you define SSRIs as “selective serotonin receptor inhibitors”

    Thanks! I am terrible at proof reading my own work.

  71. “That’s actually a rather interesting observation and one that I hadn’t thought of. Why would the researchers who had more positive results be less willing to share their data than the ones whose studies were mostly negative?”

    The explanation might be very simple. They’re professional rivals.

    Prior to my clinical training, I did research in two rather arcane areas. In both areas, tiny as they were, researchers were divided into rival groups who could be counted upon to disagree violently with each other in print and at professional meetings. In the world of “publish or perish” and tight research funding, scientists can be competitive and cutthroat in pushing a personal agenda.

  72. lillym,

    “I’ve had to fire therapists because they weren’t working out. And then I had to find a new one and luckily I only had to do it a few times.

    It’s like trying to find the medicine that works, the first or second aren’t right so you try again.”

    I saw more than one as well: I think four over the course of three years. I was limited in who I could see because my income was so low.

    I went to a family counselling clinic where I could see someone for $10. I’m not sure what her qualifications were, but I suspect the counsellor I saw didn’t have a university degree. I went to a couples therapist with my ex, who I was able to pay for because my ex chipped in. She was terrific, but then she moved out of town. I went to a CB clinic at a local hospital where I was interviewed by the head of the clinic. I wept during the interview, he listened carefully, and said that he thought I’d been depressed since I was fourteen and once the objective tests came back proving it he could refer me to psychiatrists who could prescribe medication for me. I said yes, I’d be interested in that. Feeling hopeful I went to fill out the stack of objective tests. “How has your libido changed in the last six weeks?” Well, I used to be very horny a long time ago but even though I’m in a relationship I haven’t had sex in four years and I don’t miss it. So, no change. “Do you cry all the time?” No, I lie on the couch for many hours at a time trying to will myself out of existence, or at least to stop breathing. So, no. “How do you feel right now?” Well, I’m doing a test and I love doing tests, and I’ve just met with someone sympathetic who offered me hope, so actually I’m feeling pretty good. Etc.

    I was referred to an MSW who couldn’t understand what I was doing there. Most of who they saw were people with phobias, and I wasn’t complaining about phobias. I claimed to be depressed but the objective tests proved I wasn’t. So he had me listen to a relaxation tape and rate my stress level before and after. I didn’t feel stressed before and I didn’t notice any change in stress after, but it did make me cry. He thought that was odd. He gave me the tape to take home, and it still made me cry. Which he still thought was strange. So after that he filled time talking about himself.

    I stopped going, but when I became worse a few months later I called the CB clinic again and asked to see someone else. I was on the waiting list for months until my sister realised that I was trying to figure out how to kill myself and called them and asked, are you aware she’s trying to figure out how to kill herself?

    And I ended up with a PhD in psychology, a woman who, as far as I could tell, gave me very limited CBT and eventually took me out of the program so she could see me at home and charge me less and free us up to talk about my mother. At this point my income was about $100/week, and I was paying her a sliding scale of $35/week as well as paying my share of rent and utilities. My ex wouldn’t help me out with groceries, but she would make extra when she was cooking for herself and let me eat with her. I was/am vegetarian, and she ate meat-and-two-veg. So as not to starve, I ate meat. Bitterly. And when I got money I paid my therapist for weeks in advance so that I wouldn’t be tempted to divert it into groceries. That was fodder for therapy discussions about not taking responsibility for myself and making her into my banker.

    If I wasn’t getting appropriate therapy, how was I supposed to know? I was seeing someone because my own judgement on how to live and think was obviously not serving me well. How could I use my own impaired judgement to decide that my therapist was not helping? That’s at the core of my issue with talk therapy: people who don’t trust their own judgement are to blame for not firing the people they have hired to help them. My psychologist had been assigned to me by a prestigious hospital clinic and was obviously successful. If she thought I was doing good work, who was I to question that? (Fortunately I was on the ball enough that when she suggested I have a child to give me focus in life I was able to tell her that was a very bad idea.)

    Meds, besides being paid for by provincial universal medication insurance, were so much simpler to deal with. And they changed my life immediately – at least, I was able to change my life. Within a year I got well-paying work and moved to a better apartmen. I took lovers. A year after that I asked my ex to move out. All those things. I didn’t need to spend years of misery arguing with a therapist about whether my problems were the result of unresolved conflicts with my mother, or the result of my refusal to be accountable. I took the magic little pills and I was able to do the things I already knew I needed to do but hadn’t been able to.

    In terms of talk therapy, the best thing I did was attend a six-week workshop for people at risk for depression where we learned about different therapies for depression. It was fun and informative. I got to meet other depressed people and we didn’t hate eachother, so maybe I wasn’t so awful after all. And I kept thinking “Oh, so that’s what she was trying to do! Why didn’t she just say so?” It wasn’t therapy but it was enormously helpful. When I got married I asked people to donate to this workshop series in lieu of wedding presents. (Yes, I am now happily married and a homeowner.)

    No, meds are not for everyone. But neither is psychotherapy. Even the much lauded CBT can be hard to get, even when you go to a clinic that’s supposed to offer only that. And advocating for yourself when you question your own judgement – advocating for yourself with someone whose job is to question your judgement – can be an impossible task. It certainly was for me. I really don’t understand how other people can do it so easily, though clearly they do. (And other people can’t understand why I’m not trying to go off my meds, so that makes sense.)

    But the moral hierarchy of Therapy Good, Meds… Inferior and a Marketing Ploy is one I simply can’t accept. The therapies I was able to get for myself did not have an acceptable risk/benefit profile and the meds I was able to get for myself did. That’s not me settling for second-best. That’s me self-advocating.

  73. micheleinmichigan says:

    Alison – “Talk therapy has side effects too, and someone who is depressed may not be able to deal with them effectively. SSRIs have side effects that are much less emotionally charged, more concrete, and might be more easily dealt with by someone who is not coping with anything very well.”

    That is an excellent point.

    Can I say, your therapist was worth *#@%? It make me angry just hearing about it. I can’t even imagine how you felt. And I don’t think that’s uncommon.

    I agree with you, it can be really hard and or impossible to find the right match, depending upon finances, insurance situation or local availability and when you are dealing with depression or anxiety…

    The other reality there are a lot of good therapist (not the one Alison saw, who deserve a special purgatory) out there that are just not experienced with some conditions. They don’t recognize the borderline versions and they don’t know the best forms of therapy.

  74. micheleinmichigan says:

    Alison I also can see how we came across my statement up-thread about the treatment plan with or without medication from a different angle.

    I was making some assumptions based on my experiences. I think in my area (near university with teaching hospital, regular hospital and VA hospital) CBT is widely available and insurance programs seem to send patients to more CBT type clinical social worker and psychiatry offices within the hospital network. My wait has never been more than 2-4 weeks (and I always considered that too long). The CBT that has been recommended is some talk, but mostly a set of exercises on reframing your thinking, stopping all or nothing thinking, etc it is really quite dry stuff. Like I said it’s hard (for various reasons), but it can be very useful.

    There are many other therapists, counselors, CSW and psychologist in our area that are in private practice (have never been an option with my insurance). I did see one psychologist years ago with an ex for couples counseling. She gave me the single most misguided piece of advice that I have ever received in my life. Luckily, I dumped both the ex and the counselor.

    And even with CBT there are different approaches for different problems. I found if you have issues with ruminating, little charts analyzing and reframing specific thoughts (one feature of CBT) are useless. It’s like trying to count cars on the freeway (well, a Michigan freeway where everyone is going 85 mph).

    And TESTING – testing is ridiculous. I never know what to say on those tests. My answer is never an option. I wish they had an essay section.

    This is why I get uptight when people start going on about how other people should be treated for any psychiatric disorders mild or severe. I do have strong opinions about supervision when perscribing medication. I think patients should be aware of side effects, dangers and how to taper off. I think that doctors should evaluate patients for possible ill effects. The patient is not always the best judge of that.

    Beside that,this really should be between the medical professional and the patient. That is hard enough. It is not for public opinion to decide.

  75. micheleinmichigan says:

    “I haven’t tried CBT, I want to, but the last time I checked around either the therapists offering it didn’t accept my HMO or they weren’t taking new patients or I was considered to healthy for the program.”

    Actually there are some excellent self-help books in the CBT arena. They are pretty dependent on your particular issue though, there are book tailored toward depression, ocd, anxiety, etc.

    The nice thing about books is that you can work at your own pace and pick and chose methods that speak to you without the social pressure of working with a therapist. Of course they do tailor a plan to met your specific goals like a good therapist might, but I have found them to be very useful (could just be a personal preference)

  76. micheleinmichigan says:

    DON’T tailor a plan, sorry.

  77. David Gorski says:

    I never used the word nefarious. That was someone else’s claim. Nor would I use the word nefarious to describe the way that data can be sliced and diced to show what the researchers hoped it would show.

    Oh, come now. You sure implied that their motives were less than pure in picking the studies they did. Certainly it came across that way to me and some others.

    Although I understand why someone might choose patient level analysis over group level analysis in constructing a metaanalysis, I personally cannot see how that makes this paper more reliable.

    Argument from personal incredulity. Just because you personally cannot understand it is not an argument against it.

  78. David Gorski says:

    The explanation might be very simple. They’re professional rivals.

    Granted, that is also a possibility that I should have pointed out.

  79. DanaUllman says:

    Because people at this site disparage homeopathy because of the uncertainty about its mechanism of action, does anyone want to comment on the mechanism of action on SSRIs and depression? There is a LOT of uncertainty about whether there is ANY real relationship between serotonin and depression.

    It seems that people may be betting on the wrong horse.

  80. Scott says:

    The situations aren’t in the least comparable. Homeopathy has no possible mechanism of action that doesn’t require all of physics, chemistry, and biology to be completely discarded. The idea that a chemical which definitely has physiological effects may have beneficial effects by some insufficiently ascertained path is many, many, orders of magnitude more plausible.

  81. “Because people at this site disparage homeopathy because of the uncertainty about its mechanism of action, does anyone want to comment on the mechanism of action on SSRIs and depression?”

    First of all, we do know about the mechanism of action of SSRIs. Second, we know that homeopathy is quackery because it fails to meet virtually ALL criteria for effectiveness.

    Showing that a substance “works” involves far more than a study that suggests effectiveness or a possible mechanism of action. It also involves a thorough elucidation of the pharmacokinetics and pharmacodynamics of the substance. Moreover, the purported mechanism of action must be consistent with what we know about the laws of physics and chemistry.

    Claiming that dilution makes a substance more effective is like claiming that gravity makes objects lift skyward. It is simply impossible in light of the basic laws of physics and chemistry.

    If you wanted to show that homeopathy “works,” you’d need to present detailed data on the pharmacodynamics and pharmacokinetics of homeopathic “remedies” AND you would need to explain how it could possibly work without violating known laws of chemistry and physics.

  82. micheleinmichigan,

    Yes, absolutely, there are plenty of good ones. And I’m sure the ones I saw were able to offer perfectly good psychotherapy to other people in other situations. It’s just that it can be difficult or impossible for a mentally ill person to identify bad psychotherapists or inappropriate psychotherapy. If your fourth-grader complains that math is useless, you don’t conclude that math is useless or that her teacher is necessarily complete crap. If your mentally ill patient complains that psychotherapy is useless, does that mean that psychotherapy is useless? that the psychotherapist is complete crap? or that your mentally ill patient is behaving like a fourth-grader?

    The local hospital with the reputable CBT clinic is part of McGill Medical School. Lacking other input, I had to assume I was behaving like a fourth-grader.

    I think the biggest problem for me was not having clear goals for psychotherapy. I knew I wanted my competent self back but I didn’t have concrete goals like sleeping or managing anxiety or coping with trauma that I could ask someone to help me with. I just presented myself: I’m depressed. Help me. And they would say, you’re not depressed.* Help yourself.

    I had the same fuzzy goal for medical treatment: I wanted my competent self back. Meds addressed the fuzzy goal. That was it.

    I actually have a talk therapist now who I see from time to time, usually in the winter when my beloved is at his hypercritical worst and I am at my hypersensitive worst. Now I have a clear goal for talk therapy: I want her to be nice to me. I told her that when we first met. She was surprised but went along with it. When I need someone to be nice to me I call her up, pay her $75, and we talk about things I feel inadequate about and how I should meditate and why I’m not getting more exercise and what I could do about it. And then I feel cheered up and competent and go home and tell my beloved where he can stuff it. It’s perfect. Because I’m clear about what I want it’s easy for me to identify whether I’m getting it or not. Not coincidentally, I am not incapacitated by mental illness at this time.

    _________
    * And I would think to myself, Well, I don’t look depressed to you because I have atypical depression, commonly assiciated with bipolar disorder, and you only see me when I’m sitting chatting with you in your nice sunny office. You don’t see me home alone, locked in the dark trying to die. But I couldn’t say that because teaching a psychologist about atypical depression would be presumptive on my part and saying I had it would be self-diagnosis, which is against etiquette when dealing with professionals. (When I finally did get a psychiatrist of my own, years later, she diagnosed me with Bipolar II. Unprompted. So I hadn’t been wrong.)

  83. daedalus2u says:

    I looked at the authors of the included studies and noticed that of the 3 SSRI studies, two of them have authors that were also authors of this meta-analysis. They only got patient level data from one outside group for the SSRI.

    To me, that is a pretty thin meta analysis to draw sweeping conclusions from.

    As someone who has been depressed all of my life, figuring out if you have “depression” can be quite difficult if that is all you know because it is your “normal” state. You learn to hide it and deny it. It wasn’t until I was in my 40′s and on meds that worked quite well that I began to appreciate how non-depressed people feel every day.

    Figuring out if what you have is “mild” or “severe” is simply not something that a naive depressed person can figure out on their own, particularly if that depression has existed since childhood.

    Allison makes a number of excellent points.

    What basis is there for the idea that too many antidepressants are prescribed? Suicide is the 11th leading cause of death. Is that an indication of too much treatment or too little?

  84. daedalus2u:

    “I looked at the authors of the included studies and noticed that of the 3 SSRI studies, two of them have authors that were also authors of this meta-analysis. They only got patient level data from one outside group for the SSRI.”

    Ooh, good catch!

  85. Fifi says:

    weing – “I wonder if there are similar studies comparing CBT and exercise to placebo in patients with mild to moderate depression.”

    Not that I’ve seen, it’s difficult or impossible to create a placebo equivalent for CBT or exercise (and one would think unethical to do so in psychiatric terms since patient/doctor trust and relationship are of significant importance in terms of mental health – mental health issues very much involve the realm of emotions and our relationship to the world or, to put it another way, our social and physical environment and how we interact…after all, not functioning well in our environment and coping with our emotions are significant symptoms of many mental illnesses and mood disorders). This highlights the difficult of doing studies on conditions where emotion plays a significant role, as does the mind/body relationship, and the placebo effect.

    So far the overall evidence is inconclusive regarding exercise for depression but there are also a lot of positive outcomes for specific studies in specific populations (and the general expert opinion seems to be that more studies need to be done). Either of us could easily find evidence to support either a positive or negative correlation between exercise and depression (though my bias and personal experience leads me to lean towards a positive effect, taking into consideration that the mere act of doing something regularly and a sense of achievement may also contribute to the beneficial effects of exercise on depression, as well as there being a neurobiological component).

    The most common clinical trials compare the outcomes of treatments that include CBT and medication vs medication alone. The ones below – with conflicting results – focus on treatment resistant depression in adolescents.
    http://bmjcom.highwire.org/cgi/content/abstract/335/7611/142
    http://www.ncbi.nlm.nih.gov/pubmed/9294380?dopt=Abstract
    http://www.ncbi.nlm.nih.gov/pubmed/18314433?dopt=Abstract

    Not surprisingly, in light of recent discoveries regarding the neuroanatomy of anxiety and depression – which seem to indicate that SSRIs would actually be effective for anxiety but not depression – the outcome in a trial for panic disorder and CBT/SSRI combined, SSRI and CBT alone seem to indicate that SSRIs are very effective for panic disorders. I’m a big fan of Dr Nutt and this is an excellent paper – in my non-professional opinion – that also addresses some of the issues and challenges of conducting a trial of this nature.
    http://www.cnsforum.com/commenteditem/6A752532-0081-45BE-BA2F-2FEF42E90BB9/default.aspx

    It’s worth keeping in mind that we’re still really in the early days of understanding neurobiology and the wide variety of mental and mood disorders and their causes, and the most effective treatments. Also, since so much of psychiatry is about improving a patient’s subjective experience – much like chronic pain treatments – there are many more confounding variables than with studying other forms of illness (and the efficacy of a placebo is not a negligent or unimportant factor, and not as neutral as it may be in other types of clinical trials).

    And, yes, this is a subject I’m passionate and intensely curious about having grown up around neurobiological research (and being well aware of how radically our understanding of the brain and mind has changed over the past 30 years) and psychology, with an awareness of how our own mind and cognitive processes create biases and our personal narratives influence how we understand the world and others. (And also working in the arts, which is all about communicating subjective experiences, cognition and culture/environment for some of us geekier types.) I find it frustrating when people make simplistic statements regarding mental illness or attack people asking relevant and highly appropriate questions. The history of psychiatry is certainly not without some ugliness – mainly due to ignorance and biases and not intentional cruelty, but cultural prejudices have led to some horrific practices historically – so it’s one area where being cautious and humble regarding making generalizations and grand claims is called for. Being able to say “we don’t know”, as the study’s authors did, is doing so and certainly not something for which they should be castigated. While it’s tempting to “tidy up” psychiatry or be reductive by trying to study only neurobiology, it’s not a field where you’re just treating a patient’s body or can really look at them isolation from environment (since being mentally healthy is about our subjective experience and how we function within our environment).

    One confounding factor and a rich source of personal biases regarding mental illnesses and mood disorders is that we all have minds and most of us will suffer from some form of depression or mood disorder at some point in time. We also tend to extrapolate our own subjective experiences and to project them onto others (hence the tendency of people to think that depressed people are malingering and should just “get over it” since they themselves can do so when they feel a bit blue) and assume our experience of the world and emotions is exactly equivalent to someone else’s. The reality is that there’s a great deal of neurodiversity, that there’s a huge social component to mental illness and mood disorders, and we still have a great deal to learn before we’re experts regarding the brain (and the experts are generally very aware of this even while they try to find best treatments to alleviate suffering and help people be as functional as possible). Because of the many unknowns and complexities of creating clinical trials – and because psychiatry/psychology/cognitive science has a complex history not always actually tied into biology, rife with ideology and greatly influenced by culture – we should all remain open, curious and wary of making absolutist statements or abusing a highly personal and controversial subject like mental health to promote our own ideologies (particularly since this is where psychiatry has historically been the most problematic and led to, in retrospect, some quite dubious and harmful treatments). Though, interestingly enough, shock therapy – which is often held up as the most inhumane of treatments – may actually (in the much more refined version deep brain stimulation) actually be effective for some types of treatment resistent major depression.

    This is a subject where I’m very biased – not towards a particular treatment but towards proper science-based studies of all forms of psychiatric treatments and a perspective that considers body/brain, mind and environment. As well as an ackowledgment of neurodiversity and the necessity of individualized treatments.

  86. MKandefer says:

    Has SBM done a good overview of evaluating meta-analysis? I know such an effort would require a good understanding of mathematics, but it would help us in the skeptical community.

    Doctor Gorski I have to wonder why the authors couldn’t have used a tiered approach in evaluating the literature, which is to say, give a coefficient that values less those studies that don’t have the raw data published, or available. As another respondent claimed there is 30 years of research behind SSRIs, and while not all studies are close to the gold standard, I believe there are ways of including them in a meta-analysis, but ranking them as less reliable in one’s calculations.

  87. micheleinmichigan says:

    Alison, “Now I have a clear goal for talk therapy: I want her to be nice to me. I told her that when we first met. She was surprised but went along with it.”

    I would never be that brave. Good on you.

  88. David Gorski says:

    If you wanted to show that homeopathy “works,” you’d need to present detailed data on the pharmacodynamics and pharmacokinetics of homeopathic “remedies” AND you would need to explain how it could possibly work without violating known laws of chemistry and physics.

    But, but, but….

    Water has memory, don’t you know? Homeopaths tell us so!

    Actually, it amuses me to no end when Dana Ullman shows up, because, as my friend has said, homeopathy is most akin to primitive sympathetic magic, using the example of Dr. Stephen Strange, Sorcerer Supreme and Master of the Mystic Arts (and one of my favorite comic characters of all time).

  89. micheleinmichigan says:

    “If your mentally ill patient complains that psychotherapy is useless, does that mean that psychotherapy is useless? that the psychotherapist is complete crap?”

    “If the patient is my friend, I concluded that the psychotherapist is crap. If the patient is me, I conclude I (the patient) am incompetent” Good point.

  90. Plonit says:

    Claiming that dilution makes a substance more effective is like claiming that gravity makes objects lift skyward. It is simply impossible in light of the basic laws of physics and chemistry.

    ++++++++++

    Huh? Theoretically that is only true at the point where dilution makes the substance “not present”. Prior to that point, it is certainly possible for less to be more effective without violating any basic laws.

  91. micheleinmichigan says:

    alison “* And I would think to myself, Well, I don’t look depressed to you because I have atypical depression, commonly assiciated with bipolar disorder, and you only see me when I’m sitting chatting with you in your nice sunny office.”

    # daedalus2u “You learn to hide it and deny it. It wasn’t until I was in my 40’s and on meds that worked quite well that I began to appreciate how non-depressed people feel every day.

    Figuring out if what you have is “mild” or “severe” is simply not something that a naive depressed person can figure out on their own, particularly if that depression has existed since childhood. ”

    Yes, one of the kinds of anxiety I have is social anxiety. It can be quite trying. When I shared this with a therapist. They told me that I looked quite relaxed. Sure I look relaxed, I have been working my whole life to not look foolish in front of people. I can not behave how anxious (angry, happy, sad) I feel because I am too afraid to act that way.

    In fact people often remark on how relaxed and laid-back I am in stressful situations. It is a facade, folks.

  92. apteryx says:

    Dr. Amy wrote:

    “Okay, I just finished reviewing the abstracts for all 23 studies. … Of the 17 excluded studies 15 DID show effectiveness of the antidepressant treatment of mild to moderate depression. Two studies showed no difference compared to placebo.”

    Hmm, I thought Dr. Amy had been one of those skeptics who spewed contempt when a non-MD dared to cite a study with only an abstract was publicly available, saying that you MUST read the whole paper to be able to mention it at all, as if many journal editors will let the authors outright lie about the results in the abstract. Well, never mind that. I too just finished reviewing the abstracts, and question Amy’s numbers.

    To begin with, many of the abstracts do not specify the severity of depression. The original blog post said that the meta-analysis was willing to include studies with HAMD scores up in the 30s (severely depressed). Most of the abstracts do not specify severity either with a verbal label or score range, and it’s not possible to confirm just from the abstract that severely depressed patients were not included in the analysis.

    Studies reporting drugs to be superior to placebo specifically for mild to moderate depression include DeMartinis et al. 2007, Gastpar et al. 2006 (which demonstrated equivalence and far lower side effects for hypericum as compared to citalopram), Hollyman et al. 1988, Philipp et al. 1999 (which demonstrated equivalence or superiority of hypericum compared to imipramine)

    Studies reporting one or more drug treatments to be superior to placebo in depression of unknown severity were Boyer et al. 2008, Cohn et al. 1996, D’Amico et al. 1990, DeRubeis et al. 2005 (which studied people with “moderate to severe” depression and did NOT break down efficacy by initial severity), Feiger 1996, Feiger et al. 2006, Lieberman et al. 2008 (which apparently pooled data from two studies because neither initially got independent results), Liebowitz et al. 2008 (which included only patients with HAMD scores at or above 20, not broken down), Mynors-Wallace et al. 1995 (in which 52% “recovered” with amitryptyline vs. 60% with problem-solving treatment, so probably mild to moderate); Rickels and Case 1982; Rickels et al. 1982; Septien-Velez et al. 2007; and Versiani et al. 1990 (which had 25 patients per study group).

    Then there are the remainder:

    Barrett et al. (2001) studied dysthymia and minor depression, and found significant benefit for paroxetine only for dysthymia. The abstract concludes: “For minor depression, the 3 interventions [paroxetine, problem-solving treatment, placebo] were equally effective…”

    Dimidjian et al. (2006) compared behavioral activation, cognitive therapy, and antidepressant medication in people with MDD of unspecified severity. It says “Among more severely depressed patients, behavioral activation was comparable to antidepressant medication, and both significantly outperformed cognitive therapy.” From the abstract, there is no indication whatsoever that antidepressants outperformed a placebo or any other treatment in less severely depressed patients.

    Elkin et al. (1989) compared two types of therapy with imipramine and placebo. They found relatively little difference among groups overall, which they explained as follows: “Significant differences among treatments were present only for the subgroup of patients who were more severely depressed and functionally impaired…. In contrast, there were no significant differences among treatments, including placebo plus clinical manaagement, for the less severely depressed and functionally impaired patients.”

    Liebowitz et al. (2007), no severity measures given, found no difference between desvenlafaxine and placebo on the primary endpoints or most secondary endpoints. The former did significantly improve measures of physical pain, at the expense of significantly greater side effects.

    UK Moclobemide Study Group (1994), with initial HRDS of at least 17, reported that clinical improvement was similar in all 3 treatment groups (using a 50% score reduction as their primary criterion): 53% of patients on moclobemide, 50% on imipramine, and 51% on placebo.

    Wichers et al. (2009) looked at 83 depressed patients (no severity given) and 22 healthy controls, who received imipramine or placebo for six weeks. They report improvement in “Stress-Sensitivity and Reward Experience,” but don’t say anything about their results from using the Hamilton scale, which suggests that they did not see any benefit versus placebo.

    Now, Dr. Amy claims that 15 of these 17 studies showed antidepressants to be effective for mild to moderate depression. This means that she must have been able to determine this to be the case not only for every single study in the second group above (DeRubeis et al. and Liebowitz et al. included) but for four of the last six studies I just individually summarized. However, THREE of those – Barrett et al., Elkin et al., and the UK Moclobemide Study Group – all appear to state the exact opposite (while two others seem to have found no efficacy, in depression of unknown severity, with standard primary outcome measures, and the sixth hints that they did not find benefit in less depressed patients).

    How frankly untrustworthy does one of your bloggers have to be, in terms of stating facts about data and literature that are just not true, before you recognize her as a drag on your team?

  93. David Gorski says:

    Doctor Gorski I have to wonder why the authors couldn’t have used a tiered approach in evaluating the literature, which is to say, give a coefficient that values less those studies that don’t have the raw data published, or available. As another respondent claimed there is 30 years of research behind SSRIs, and while not all studies are close to the gold standard, I believe there are ways of including them in a meta-analysis, but ranking them as less reliable in one’s calculations.

    I think that can be done, but it introduces another problem. Somehow investigators would have to assign weights to each study, and that would be a process rife with possibilities for manipulation and accusations of manipulating ratings of various studies. It’s something that would have to be done very, very carefully, and even then would probably cast doubt on the methodology.

  94. “It is a facade, folks.”

    Yep.

    I encountered two basic ways that people determined how sick I was.

    Psychologists would ask me to fill out a depression scale. (After that first disastrous encounter with depression scales I learned to refuse to fill them out.)

    The other was used by doctors and was more informal. The doctor would lay an imaginary stopwatch on the desk and see how many seconds it took to break my defences down until I cried. (I actually have no idea if they were consciously doing this, but after several repetitions with different doctors it certainly felt like it.)

    Subjectively I felt like the stopwatch test was more accurate. Unfortunately, you can’t use the stopwatch test for research.

    I’ve often wondered if the depression studied with scales and inventories is the same as the depression that people in doctors’ offices report. Research subjects are often college students who are recruited with ads asking if they’ve had sleep and appetite changes accompanied with feeling down for two weeks, the rationale being that depression is defined as six weeks but most people who have been depressed for two weeks will stay depressed for six weeks. And then their depression is rated on tools that capture some people but not others.

    What does a student who has been unusually sad and sleepless for two weeks and whose depression is captured nicely on an inventory have in common with an adult who has been struggling with hopelessness for so long that it is their “normal” and “usually”?

    The student with a “mild” depression score might be a very different person from the adult with a “mild” depression score. As a very biased, extremely subjective, non-researcher I am simply skeptical about depression scored on inventories. An individual’s scores going up and down from their own baseline is one thing. That’s fine. But is it actually possible to say that your “mild” depressives are the same as my “mild” depressives, if yours are college students and mine are adults highly invested in their facades?

    Maybe the tools used for these groupings are different from the tools that I’m familiar with and they will always compare apples to apples. I know I’m not the first person to ask this question. But if we’re reviewing studies that distinguish between “mild,” “moderate” and “major,” could we have an overview of the tools used to make these groupings? It would help me understand the research better.

  95. Fifi says:

    As the recounting of personal experiences here illustrates, it’s very hard to generalize about treatments for mental illness and mood disorders. Even if two people are diagnosed with moderate depression and score relatively the same on various tests (which themselves rely upon subjective reporting), they may both respond differently to different types of therapy depending upon a variety of factors (personality, both family and larger culture, contributing environmental/social factors, individual neurobiology, and on and on the list goes). It’s also seems important to recognize that anxiety disorders, while often associated with depression, aren’t actually depression and there are different parts of the brain involved. What works beautifully for one person, may be useless or actually quite negative or even damaging for another – this is true of medications, as well as different forms of talk and behavioral therapies – which is why proper diagnosis of each individual is so important and needs to consider not only neurobiology and general biological health, but also relationships and environment (including life events).

    An interesting article about the neurobiology of depression…
    http://www.sciencedaily.com/releases/2009/12/091208132724.htm

  96. micheleinmichigan says:

    # daedalus2uon 05 Feb 2010 at 11:31 am

    I looked at the authors of the included studies and noticed that of the 3 SSRI studies, two of them have authors that were also authors of this meta-analysis. They only got patient level data from one outside group for the SSRI.

    Ummm, to a non-research person, that sounds very weird. It sounds like they are just repackaging their own study’s for re-release with a new and exciting spin on them.

    I know someone said up thread that they come up with the criteria before hand, but isn’t there a point when it just seems too thin?

  97. BillyJoe says:

    Dana Ullman said:

    “So, either Prozac works and homeopathy works for moderate or severe depression OR neither works. What is it going to be?

    ecam.oxfordjournals.org/cgi/content/abstract/nep114 ”

    Maybe you could tell us:

    The trial compared homeopathic drops to placebo drops identical in appearance to the homeopathic drops; and fluoxetine capsules to a placebo capsules identical in appearance to the fluoxetine capsules.
    So, in your analysis of this trial…
    1) Did the homeopathic drops work any better than the placebo drops?
    2) Did the paroxetine capsules work any better than the placebo capsules?
    Based on these results, what, if anything, can you conclude about homeopathic drops vs fluoxetine capsules in the treatment of moderate to severe depression?

    Please also comment on the following:
    “The research pharmacist randomly delivered homeopathy and placebo or fluoxetine and placebo, according to a randomized assignment sequence to either homeopathy or fluoxetine group….Only the senior author and the pharmacist had access to the code of the randomized sequence during the study. ”
    The pharmacist delivered the treatment and he also had access to the code. Is that correct? If so, what possible effect do you think this could have on the outcome?

    BJ

  98. Reviewer 3 says:

    Hi Dr Gorski,

    You wrote “However, the most rigorous meta-analysis methods do require the use of the original data if it is obtainable. Investigators calculate the common estimate and its confidence interval using these studies. To do this most accurately it’s best to have the original data from all the studies if they can be obtained.”

    I’m not sure this is right. Meta-analyses using summary data from trials should produce pretty similar results to meta-analyses using individual patient data. The reason for doing a patient-level analysis is if you want to do an analysis that takes into account time (eg time-to-first event analysis) or you want to include baseline variables into the analysis, not to improve the accuracy of the analysis.

    In this paper, the authors have divided the cohort into subgroups according to their baseline severity of illness. An important issue with this is that it breaks the randomization of the original trials, which may introduce confounding. For each subgroup, it is incorrect to assume that the placebo and treatment groups will have the same baseline characteristics and that any differences are due to randomization (ie chance). eg it is possible that the treatment group might be older/younger, or have different proportions of men and women etc than the placebo group which may contribute to the differences between the treatment groups. This was not addressed at all in the paper.

    Dr Tuteur,

    I also think the conclusion is too strong. My reason (that you may wish to comment on) is that the study addressed a question of whether treatment effect for depression varied by baseline severity and found it did. The study didn’t ask the question: are antidepressants effective in mild depression? So the authors shouldn’t try to answer that question, because the study wasn’t designed to answer it, and was underpowered to detect any effect in the subgroup with mild depression.

  99. Reviewer 3 says:

    @ Daedalus2u

    “I looked at the authors of the included studies and noticed that of the 3 SSRI studies, two of them have authors that were also authors of this meta-analysis. They only got patient level data from one outside group for the SSRI.”

    I’m surprised that ALL the groups who supplied patient data don’t have an author. It would be fairly standard to write a protocol, send it to interested authors, invite them to join the collaboration, provide feedback on the protocol and then supply the data after the protocol was finalised. People generally need some incentive to give away data for free, especially to rival research groups.

Comments are closed.