It won’t happen because the culture is pretty set. We’re taught from early in graduate school that when a study doesn’t work, you dust yourself off and move on. You don’t write it up; there’s no point. Writing an article and getting it published is about a year or more worth of effort. If I write up everything that doesn’t work, I’ll do nothing else. Hell, if you have a 10% success rate in the lab, you’re a hall of famer!
That being said, if there were a push for a central repository for all clinical trial data, that would be a good thing, and I would have little problem with that.
Even journals that focus on basic research don’t like to publish negative data; it’s less interesting to the community as a whole. Every lab I’ve been in has been delighted when we’ve actually gotten a phenotype or had an effect in a treatment group, because it’s very hard to publish otherwise. We do experiments all the time that don’t work for one reason or another, but no one wants to read about the genes that aren’t upregulated in your mouse model of multiple sclerosis. It’s horribly inefficient, as labs probably end up investigating the same interesting hypotheses that don’t generate anything over and over independently because there’s no body of literature for the negative results. Oh, well. I guess that’s what meetings are for.
I’m not sure how deliberate the conspiracy is or whether they need to try to suppress the studies, though. The company doesn’t want to publish negative data, the journals don’t want to publish negative data, and the scientific community doesn’t want to read about negative data, so I’m not sure there’s a practical way to make anyone publish it.
/edit: What FiveYearLurker said.
If the results helped the drug companies you can bet they’d be published as fact. As the current study results cast a negative light on the drugs, well, obviously more testing needs to be done. I’m sure cigarette companies had absolutely no idea their products led to cancer until countless court battles and 50 years of studies. :rolleyes:
Are any of these studies actually claiming, as the thread title implies, that antidepressants Don’t Work? Is Prozac just a placebo? That seems pretty damn unlikely after all this time.
But that just throws ice-water on the much ballyhooed Peer Reviewed Journal Cite. If we can assume that they were less likely to publish contrary data, then citing what they did publish is citing nothing more than their bias.
Wrong. The real conspiracy is that drug companies are allowed to advertise prescription drugs to consumers. Who know shit all about the drugs, about proscribing drugs, and are just influenced by the happy people on TV who are in their demographic.
If i’m a doctor or a scientist interested in the usefulness of antidepressants, and i read 35 studies that indicate that they are effective under particular circumstances, or with particular patients, then why wouldn’t i want to know about the 35 other circumstances or patients where they didn’t work? Not wanting to know about the negative results seems like little more than wilful ignorance.
This is what I’d like to know…speaking as someone who’s been on a variety of antidepressants, having finally settled on Effexor because of the immense changes I saw in myself. In addition to alleviating depression, other things that improve with the slow reuptake of seratonin (I think that’s what I mean to say) like improved digestion, better sleep patterns and less physical pain have all happened with Effexor. If it worked solely on a placebo basis, then other drugs that I’ve tried (celexa, lexapro) would have given me the same benefits, no?
But seriously, I don’t see the big deal. Antidepressants, unlike things like cholesterol drugs, are drugs that you can pretty much tell immediately if they are working for you. If you take them for two months and you’re still depressed, that one isn’t working for you, try another. I’ve taken them in the past, one type worked, one type didn’t- on the one that didn’t, I told the doctor I wasn’t seeing benefits and he gave me some of the kind I had success with. And hell, if 99% of the studies showed they didn’t work, but you took some and they worked for you, what do you care about the studies if you’re feeling better?
Because the stigma of mental illness is already quite strong, in the U.S. at least, and people are constantly being told “just think happy thoughts” or “man up, nancy” or that they crave attention or are being dramatic or whathaveyou. For me, at least, having a pill that helped, sort of illustrated that true clinical depression is a chemical imbalance, with a chemical solution (and therapy of course). Also, if pills are proven to work only on a placebo level, they may no longer be covered by insurance some time in the future.
Just pulling this right out of my ass, but I’m guessing what we call “depression” is not one thing, but a thousand different things, with a thousand different causes. We should not expect any one chemical to have a broad range of effectiveness, because that fundamentally implies a single cause.
But the money is in broad spectrum applications, hence broad spectrum sales.
No, because the claims of a given paper still have to be internally justified, and reviewed. Moreover, peer review is not held to be some panacea; publication is not a gold standard of truth, it is merely a point in an ongoing process. “Scientists” do not automatically assume everything they read in a peer reviewed journal is true (well, maybe shit ones do); they apply their critical appraisal skills to assess for themselves whether they believe the paper supports its claims. If physicians are limiting their appraisal of what to prescribe to published claims, and ignoring other available data, then they are delinquent in their duty to their patients.
There’s a difference between “not wanting to know about” something, and not thinking it’s interesting enough to publish in a top journal with limited space. The assumption amongst several people here appears to be that if something isn’t published, it didn’t happen. Not so.
Journals aren’t mere catalogues of study results; they’re records of novel results. There are, of course, boring successes and interesting failures. However, since a positive result pretty much always hints at an underlying cause, it is more frequently of interest since it leads to more avenues of research for the community at large. By contrast, negative results (bread and butter of the scientific community though they may be) are less likely to be interesting simply because the space of Things Which Do Not Work is vaster than that of Things That Work by many orders of magnitude. Journals are supposed to act as a filter on current scientific work; that’s why they’re valuable. But pretending that they are the be-all and end-all of scientific endeavour is damaging in the extreme.
Crudely put, one might think of a given scientific field as an optimisation problem. We’re trying to find a local (even global) minimum, but we can’t see very far. We putter around, mostly, not getting very much better results (negative findings). When someone does stumble across a valley in our solution space, we want to know about it; such a finding is objectively more interesting than our previous putterings, essential though they were. It’s important that it’s highlighted, but no-one should believe that it occurred in a vacuum.
This is not to say that trials should be permitted to be re-run and re-run until a positive result is obtained; quite the opposite. Data should invariably be public (indeed, the only ethical approach IMO is to treat it as belonging to the patients in the trial). It should not be possible for companies to hide their bad results completely (or indulge in any of the more subtle means of distorting results). But focusing on journal publication as the sine qua non of academic endeavour is just flawed. The whole point is that the data is available for you to make up your own mind.
This is also true. If only there were some medical testing available to actually prove the chemical imbalances of depression (low endorphin levels, norepi re-uptake rates, seratonin reuptake) to prove that just because Smallville isn’t on this week, it doesn’t mean you have depression, but that the girl down the street who pretends to be happy all day and then goes home and drinks herself into a coma IS.
Please note, the article does not say “anti-depressants found to be ineffective”; it says that negative results are less likely to be published. The negative trials may have had wildly divergent outcomes specified for study. For example, one trial may (hypothetically!) have discovered that Paxil is no good for bipolar disorders, while another shows it to be effective for common or garden depression. The latter finding might have been published, the former not. Drugs are trialled for wildly different things these days - many will be aware that Viagra is used for pulmonary hypertension as well as erectile dysfunction. To say there has been a “negative finding” is meaningless unless you know what was actually measured.
Disclaimer: I am most definitely not a medical researcher. I’m just illustrating how a “negative outcome” for a given trial does not mean that the drug involved has no effect for all measured outcomes. Indeed, outcome modification is one of the more insidious ways in which trial results can be manipulated. My point is that not all trials measure the same thing.