Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie
Hi there,
There is an issue with role permissions that is being worked on at the moment.
If you are having trouble with access or permissions on regional forums please post here to get access: https://www.boards.ie/discussion/2058365403/you-do-not-have-permission-for-that#latest
There is an issue with role permissions that is being worked on at the moment.
If you are having trouble with access or permissions on regional forums please post here to get access: https://www.boards.ie/discussion/2058365403/you-do-not-have-permission-for-that#latest
ssri no btter then placebos
-
26-02-2008 1:15pm#1
Comments
-
-
-
-
-
Well why didn't they do a meta-analysis on a trials that were more long-term? SSRI's can take around 6 weeks to 'kick in', I really don't understand why they chose such short term trials to study!
Long term trials with these drugs do not exist in sufficient numbers to do a meta-analysis. You work with what you got! And the original researchers for the drug companies worked within their budget - if they can get what they want in 6 weeks, they won't pursue it any further.I would go so far as to say the researchers have a bias or even an agenda here.
The authors of this meta-analysis? I sincerely doubt it. They almost certainly did have a priori expectations but that does not necessarily bias the data. They selected the studies - the most unbiased sample they could obtain, most likely, as detailed in their methods. They ran their model (which is where the weaknesses of this study really reside, if you're willing to read the the thing) and published their results. The data is what it is. And who honestly wants to piss off Big Pharma??
The biggest problem with this study is that it courts sensationalist media attention.0 -
Advertisement
-
As this study is in fact a meta-analysis and all clinical drug trials need FDA approval, this is a major strength of the study! It uses the FDA trials that were published and also those that were not.
The minimum trial length was 4 weeks, the maximum was 8 eights. Most were 6 weeks long. As a meta-analysis, this was not under the control of this study - it can only attempt to report on the trails done by others. It's not the aim to report on longer term term effects - it is what it is!
The limitations of this kind of data is well known but what else can you do with 'depression' - a poorly defined mental diagnosis? In fact, it's probably the gold-standard for depression assessment.
The this study is in broad agreement with that. They don't say that SSRIs don't work, only that they only have a modest effect that is proportional to the initial severity of the depression.
Why is the lack of peer-review a strength? I see their point about publication bias. But it's not usual practise to rely on unpublished data so heavily, is it? I mean, presumably, we have no access tot his data as doctors trying to examine the trial closely, which always worries me.
The length of time is an issue as they draw the conclusion that "Drug–placebo differences in antidepressant efficacy increase as a function of baseline severity, but are relatively small even for severely depressed patients" in a general sense, without any reference to the timeframe.
The point about the questionnaire is also an issue when the above conclusion is drawn. In particular when psychiatric guidelines state that improvement in depressed patients sholdn't be assesed purely on a symptom tick box basis. I would disagree that depression is neccesarily poorly defined. There's good DSM-IV criteria. I agree that monitoring improvement is difficult, though.
They don't say that SSRIs don't work per se, but they say they're not much better than placebo, which is the next best thing
Having said that, they do also say "we find that the overall effect of new-generation antidepressant medications is below recommended criteria for clinical significance".
I broadly agree with your "it is what it is" stance. I'd have no issue with this study just being "put out there".I just think the conclusions are suggesting that it's more than what it is,0 -
tallaght01 wrote: »Why is the lack of peer-review a strength? I see their point about publication bias. But it's not usual practise to rely on unpublished data so heavily, is it? I mean, presumably, we have no access tot his data as doctors trying to examine the trial closely, which always worries me.
For a meta-analysis, the primary threat to external validity is publication bias. These studies are from drug companies - not independent laboratories. If, say a negative or unimpressive result is found, publication would be (for them) considered low priority. All clinical trials go through a process of review before they are conducted so, in a sense, they have already been peer-reviewed for scientific and ethical purposes. In addition, all human drug trials go through the FDA so they're getting the complete US sample. The fact that they are unpublished in no way means they are bad science. Subsequent journal peer review would largely focus on issues of novelty and clinical impact. Unless, of course, in the unlikely event that a gaping methodological flaw slipped through.
And you do have access to this information - the FDA has to release all results under the Freedom of Public Information act - that's where this group got the results from!tallaght01 wrote: »The length of time is an issue as they draw the conclusion that "Drug–placebo differences in antidepressant efficacy increase as a function of baseline severity, but are relatively small even for severely depressed patients" in a general sense, without any reference to the timeframe.
I agree, but that's their sloppy interpretation - doesn't make the result bad science - just a bad discussion and interpretation. Tbh, I never read the discussion until I've had time to fully digest the results for myself. Sometimes I don't even bother reading the discussion. This paper obviously has courted controversy (intentionally, no doubt!) with how they wrote the paper. Sometimes you kind of have to look behind the bluff and bluster.tallaght01 wrote: »They don't say that SSRIs don't work per se, but they say they're not much better than placebo, which is the next best thing Having said that, they do also say "we find that the overall effect of new-generation antidepressant medications is below recommended criteria for clinical significance".
I broadly agree with your "it is what it is" stance. I'd have no issue with this study just being "put out there".I just think the conclusions are suggesting that it's more than what it is,
And the attitude taken to 'clinical vs. statistical' significance is also imperfect here - the drugs basically have to >3 better than placebo before they are considered of benefit, which needless to say can be unreasonable. They have certainly drawn conclusions you and I don't agree with but their results are able to speak for themselves.0 -
For a meta-analysis, the primary threat to external validity is publication bias. These studies are from drug companies - not independent laboratories. If, say a negative or unimpressive result is found, publication would be (for them) considered low priority. All clinical trials go through a process of review before they are conducted so, in a sense, they have already been peer-reviewed for scientific and ethical purposes. In addition, all human drug trials go through the FDA so they're getting the complete US sample. The fact that they are unpublished in no way means they are bad science. Subsequent journal peer review would largely focus on issues of novelty and clinical impact. Unless, of course, in the unlikely event that a gaping methodological flaw slipped through.
And you do have access to this information - the FDA has to release all results under the Freedom of Public Information act - that's where this group got the results from!
I agree, but that's their sloppy interpretation - doesn't make the result bad science - just a bad discussion and interpretation. Tbh, I never read the discussion until I've had time to fully digest the results for myself. Sometimes I don't even bother reading the discussion. This paper obviously has courted controversy (intentionally, no doubt!) with how they wrote the paper. Sometimes you kind of have to look behind the bluff and bluster.
And the attitude taken to 'clinical vs. statistical' significance is also imperfect here - the drugs basically have to >3 better than placebo before they are considered of benefit, which needless to say can be unreasonable. They have certainly drawn conclusions you and I don't agree with but there results should be able to speak for themselves.
With regard to the FDA approval, I know we can apply for them under the FOI rules. i know these guys did that. But it's a big hassle for everyone who wants to read a study to do that in order to check out the results of these trials, and their methodology.
I never thought the results were bad science because they weren't in a journal. I just like to see the trials were talking about before I alter prescribing practise. I'm sure you'll agree there's been plenty of peer-reviewed stuff over the years that you wouldn't use for anything but to light the fire. It was the lack of easy accesibility that was my problem.
I agree with you about the placebo Vs drug significance ratio. But they claim NICE used that as their cut off for clinical significance!
I think they did court controversy intentionally. Otherwise there's no way I'd have ever read this study
My favourite bit was in the editor's summary:
"The researchers conclude that there is little reason to prescribe new-generation antidepressant medications to any but the most severely depressed patients unless alternative treatments have been ineffective."
I think that's a bit of a leap. They haven't evaluated the alternative treatments, and there's a body of evidence to say that these drugs are effective in mild to severe depression (well, according to NICE there is).
As our resident sciencey-type, 2scoops, would you alter prescribing based on this trial? I guess that's what I'm getting at. I won't be altering mine, but I'd be interested to see if you think there's a practical application of this data.0 -
tallaght01 wrote: »With regard to the FDA approval, I know we can apply for them under the FOI rules. i know these guys did that. But it's a big hassle for everyone who wants to read a study to do that in order to check out the results of these trials, and their methodology.
Well then, you should be thankful these guys went to the bother for you!tallaght01 wrote: »would you alter prescribing based on this trial? I guess that's what I'm getting at. I won't be altering mine, but I'd be interested to see if you think there's a practical application of this data.
Well, ignoring the interpretation and editors comments, I certainly wouldn't stop prescribing ADs if I were currently doing so. They work. The data from here say that they work, regardless of the spin they put on it. Even if the analysis was wholly negative (which it isn't), there's more than enough conflicting evidence to not stop prescribing. We've mentioned the limitations of this specific analysis above as well.
The value of these data lie in the fact that they show that, if look in as impartial a manner as we can, the effect of ADs in the short term is modest. This means maybe we should investigate the phenomenon further and try to develop strategies that are more effective in the short term after acute diagnosis. Also, coupled with what we already know about ADs, it would mean that failure to see an effect in patients in the short term with these meds should not be taken as completely discouraging.
Their results aren't nearly as challenging to the currently accepted practice as they would have us believe - I think they're actually in broad agreement with them. This new wave of headline-seeking MD-PhD - they're probably the type that would probably sell their wedding photos to Hello! magazine. tut-tut! 0 -
-
Advertisement
-
Well I don't think the methodology is particularly bad here, tbh. It has the limitations you'd expect from any meta-analysis. I think PLoS Med is a suitable home for the information but obviously I would have preferred if it was a little more conservative in tone. Even without the outlandish claims they make the PLoS Med paper is certainly novel and a little exciting, even in its most conservative interpretation. As to why bad papers get published, I can suggest a few reasons but they're really only my subjective opinions:
Primarily, there are some bad reviewers out there. Really bad. Some of the reviews I've read of my own stuff or as 2nd reviewer on somebody else's paper or grant has been shockingly bad. They can let through some bad stuff [more likely though, they hold up good stuff!]. That's why they try and send it out to ~3 reviewers at once. PLoS Med is a good journal and should have good reviewers on their books (although with a bias towards MDs) but sometimes people just disagree with me!
Obviously, the editor here was right behind them and endorsed their views! Somehow though, I can't see the same paper getting into JAMA as written.
Secondly, if the science is good enough, new, interesting and important, like above, picking holes in the discussion is hardly worth the time (it takes hours spread over a couple of days to properly review a paper - all unpaid). That said, I still wouldn't have let this go through with such an overstatement. In general, though, I think if the data can speak for itself then it doesn't really matter what the investigators write in the discussion.
The assumption is that no one with a bit of common sense is going to make an important decision without understanding the data. That is not always true - I mean, I'm not going to do anything crazy, you're not because we can read between the lines and people who are not 'it-getters' generally aren't ever in a position to do anything stupid - but when journalist hacks get involved, studies like these can gain nationwide exposure and not the good kind. Joe Public or someone with more influence may now get a bee in their bonnet over ADs when there's really no need to.0 -
The assumption is that no one with a bit of common sense is going to make an important decision without understanding the data. That is not always true - I mean, I'm not going to do anything crazy, you're not because we can read between the lines and people who are not 'it-getters' generally aren't ever in a position to do anything stupid - but when journalist hacks get involved, studies like these can gain nationwide exposure and not the good kind. Joe Public or someone with more influence may now get a bee in their bonnet over ADs when there's really no need to.
Ohhhh the painful MMR memories that brings back
Still dealing with the fallout of that!!!0 -
tallaght01 wrote: »Ohhhh the painful MMR memories that brings back
Still dealing with the fallout of that!!!
Ha, quite! Felt most keenly in paeds, no doubt 0 -
+1ronbyrne2005 wrote: »Havent read the article/study but would be weary of it.
Does this mean that there is a chance that there are long term side effects that are unknown, un-investigated? Are patients the subjects of long term study?Long term trials with these drugs do not exist in sufficient numbers to do a meta-analysis.
I agreeThe biggest problem with this study is that it courts sensationalist media attention.I would go so far as to say the researchers have a bias or even an agenda here.They selected the studies - the most unbiased sample they could obtain, most likely, as detailed in their methods.
I haven’t read it - What method did they use to select the studies – meta- analysis or not the selection process itself can engender bias.And who honestly wants to piss off Big Pharma??
I agree he is finishedtallaght01 wrote: »I agree with you about the placebo Vs drug significance ratio.
Perhaps its an increased placebo effect...tallaght01 wrote: »"The researchers conclude that there is little reason to prescribe new-generation antidepressant medications to any but the most severely depressed patients unless alternative treatments have been ineffective."
I think that's a bit of a leap. They haven't evaluated the alternative treatments,
exercise has been found to be effective in mild depression has it not and I thought I read somwhere years ago just as effective as prozac....tallaght01 wrote: »and there's a body of evidence to say that these drugs are effective in mild to severe depression (well, according to NICE there is).
who is right then??
thats a pity considering you are dealing with paeds and there is less proof in their regard and scant data in regard to side effects of anti depressants short or long term...tallaght01 wrote: »would you alter prescribing based on this trial? I guess that's what I'm getting at. I won't be altering minetallaght01 wrote: »Ohhhh the painful MMR memories that brings back
Still dealing with the fallout of that!!!
Unbelieveable given there are never any side effects and that measles is a murderer with no benefit to childrenAnd who honestly wants to piss off Big Pharma??
Remember Wakefield?
Who would dare now question their god like authority? Especially when they pay for everything including university0 -
Being wary is one thing but be wary for the right reasons - at least read the thing!+1Does this mean that there is a chance that there are long term side effects that are unknown, un-investigated? Are patients the subjects of long term study?
No - it just means there aren't a sufficient number of studies to make a meta-analysis viable. A meta-analysis is technique that, in simple terms, uses the results from several studies as a single data points. Therefore, you need lots of comparable studies available before you would have enough statistical power to reveal any overarching trends. For example, this paper used data from 47 separate clinical trials.I haven’t read it - What method did they use to select the studies – meta- analysis or not the selection process itself can engender bias.
Every single clinical drug trial in the US needs FDA approval. Whether they get published after that is at the discretion of the investigators. The FDA must release all their info to the public. These investigators accessed the all the data on these drugs from the US, whether it was published or not. So many questions... why not read it yourself??I agree he is finished
Hardly. But he's not going to get a job off them, that's for sure.who is right then??
Me, obviously! 0 -
-
-
And the attitude taken to 'clinical vs. statistical' significance is also imperfect here - the drugs basically have to >3 better than placebo before they are considered of benefit, which needless to say can be unreasonable. They have certainly drawn conclusions you and I don't agree with but their results are able to speak for themselves.
Could you please explain why you think a level of clinical significance >3 could be unreasonable. Surely N.I.C.E. has set this standard to ensure the benefits of taking the medication outweight the costs in terms of money and side effects. A treatment can be statistically significant but have such a marginal clinical benefit as to be irrelevant, hence the need for clinical significance.
For example take a product used to re-grow hair on the top of the head. Scientific research could show that the hormones in regain statistically significantly increase the rate of hair growth when rubbed into the scalp. You could then do a controlled trial with 100 men who are as bald as an egg and find it makes no difference at all. It would not be clinically significant for this patient group.Well, ignoring the interpretation and editors comments, I certainly wouldn't stop prescribing ADs if I were currently doing so. They work. The data from here say that they work, regardless of the spin they put on it. Even if the analysis was wholly negative (which it isn't), there's more than enough conflicting evidence to not stop prescribing. We've mentioned the limitations of this specific analysis above as well.D
You say the data say AD's work and you seem very sure about it. At best the data tells us that SSRIs have a marginally statistically significant benefit compared to placebo for all patient groups but a clinically significant (>3) benefit is only found among those with very severe depression. The methodology and data are very clear.The value of these data lie in the fact that they show that, if look in as impartial a manner as we can, the effect of ADs in the short term is modest. This means maybe we should investigate the phenomenon further and try to develop strategies that are more effective in the short term after acute diagnosis. Also, coupled with what we already know about ADs, it would mean that failure to see an effect in patients in the short term with these meds should not be taken as completely discouraging. D
The pharmaceutical companies choose to run trials of only 4-6 weeks. This evidence was taken at face value by the medical community and was the basis for the huge number of prescriptions for SSRIs written over the last 20 years. What it shows more than anything else is that medical professionals got into bed with the pharmaceutical industry and were prepared to trust them to do objective scientific research into products from which the pharmaceutical companies stood to make billions of dollars. Sooner or later there was going to be a scandal. Now it has happened and the medical community has egg on its face.
Do you have any evidence to suggest that if someone has been on the recommended dose of an SSRI for six weeks, and has not benefited at all, that they will do so at a later time? Clinical experience would suggest otherwise. The best psychiatrists in the country will switch medications after six weeks if there has been no response.Their results aren't nearly as challenging to the currently accepted practice as they would have us believe - I think they're actually in broad agreement with them. This new wave of headline-seeking MD-PhD - they're probably the type that would probably sell their wedding photos to Hello! magazine. tut-tut!
Currently accepted practice is to prescribe antidepressants to those with moderate and severe depression. In reality GPs have been prescribing them for all types of depression, mild, moderate and severe. We all know anecdotally that there are GPs out there prescribing SSRIs for everything from back pain to migraine head-ache. This study would suggest that only those with severe depression should be prescibed SSRIs. Those with mild and moderate depression can be treated with evidence based psychological therapies such as CBT if available and a medication other than an SSRI. Should CBT fail, medication on its own could then be tried.
Your views seem to differ from those of Professor Patrick McKweon, the leading expert on mood disorders in Ireland. At one of his lectures last year he was asked about the appropriate prescribing of medication and he said it should be reserved for those with severe clinical depression. CBT has been shown to work for those with mild and moderate depression, but not those with severe depression. Often CBT is recommended with medication. It all depends on how depressed the person is.
These drugs were never meant to be prescribed to such a widespread extent in the community. For all the money that has been spent on SSRI medication we could have trained an army of psychotherapist to deal with mild and moderate depression as well as built better social support services and outreach programmes.
In case there is any confusion, I fully support the use of medication for severe mental illness and believe there is no substitute for it. But if the science says that SSRIs are largely ineffective for mild and moderate depression then they shouldn't be prescribed in these circumstances. Other antidepressants, particularly the MAOIs, have been shown to be very effective in the treatment of non-biological atypical depression and chronic low mood. I think this study should cause doctors to question how they are prescribing medication.0 -
-
-
Advertisement
-
-
-
-
You seem to have conflated the issues of current best practice and currently practiced practice.
I agree. There is a huge difference between the two. You said you didn't think this study would pose a challenge to "currently accepted practice". If you meant best practice then fair enough. I don't think this study will hugely change what is currently best practice, but it will have serious implications for what has been seen as acceptable by the medical profession and GPs in particular. In the late 1980's, there was concern among psychiatrists that GPs were not picking up enough depression in the community. Increased emphasis was placed on diagnosis and treatment to reduce suicide rates. With the arrival of the SSRIs (their low side effect profile and relative harmlessness in overdose), doctors felt much more comfortable prescribing antidepressants and this led to an explosion in the number of prescriptions written. Now the pendulum is swinging the other way with psychiatrists and the Irish Council for General Practicioners urging much greater caution. This study will add to that momentum, with greatly increased demand for alternatives to SSRI antidepressants for mild and moderate depression.tallaght01 wrote:There's evidence for their effectiveness for moderate to severe depression, though.
This study casts serious doubt on the benefit on SSRIs for moderate depression. Only those at the very severe end of the depressed scale [i.e. actually completely bed-ridden with the disease] are likely to get a clinically significant benefit from SSRIs.
The evidence which shows SSRIs are effective for moderate depression is based on published trials. What is valuable about this study is that it looks at all the clinical trials, published and unpublished, submitted to the FDA for four of the most commonly prescibed SSRIs before these drugs went onto the market. Bearing in mind this assumes these trials were carried out without any bias on the part of the pharmaceutical companies whatsoever. These trials were designed, funded and executed by the pharmaceutical companies themselves. In fairness to these researchers they are taking the data presented by the pharmaceutical companies completely at face value and yet they still did not find a clinically significant benefit for mild and moderate depression.
Obviously these results will have to be replicated elsewhere, taking into account research into SSRIs post their release onto the market.The biggest problem with this study is that it courts sensationalist media attention.
This is hardly the researchers fault. The researchers are hardly to blame if the data tells a story which the media want to jump upon and sensationalise.Tallaght01 wrote:I think they did court controversy intentionally. Otherwise there's no way I'd have ever read this study
My favourite bit was in the editor's summary:
"The researchers conclude that there is little reason to prescribe new-generation antidepressant medications to any but the most severely depressed patients unless alternative treatments have been ineffective."
I think that's a bit of a leap. They haven't evaluated the alternative treatments, and there's a body of evidence to say that these drugs are effective in mild to severe depression (well, according to NICE there is).
As our resident sciencey-type, 2scoops, would you alter prescribing based on this trial? I guess that's what I'm getting at. I won't be altering mine, but I'd be interested to see if you think there's a practical application of this data.
I don't see how this conclusion courts controversy intentionally. It seems perfectly reasonable based on the evidence they have presented.
You say they haven't evaluated alternative treatments. That would be completely outside the scope of this study. It is an established scientific fact that alternative treatments to SSRIs do exist for mild to moderate depression (including pharmalogical and psychological therapies). These have been reviewed extensively elsewhere. This review of psychotherapy was published just a few months ago: http://www.psychotherapy-ireland.com/2007/09/20/the-effectiveness-of-psychotherapy-a-review-of-research-by-prof-alan-carr/ The executive summary is available here: http://www.psychotherapy-ireland.com/wp-content/uploads/2007/09/executive-summary.pdf
The alternative would not necessarily have to be psychotherapy but maybe another class of antidepressants such as the MAOIs, or a combination of MAOIs and psychotherapy. MAOIs have been around for 50 years and have been proven particularly effective for the non-endogenous, atypical depression and chronic low mood.I understand the reason why clinical significance of a treatment is set at 3. I don't unequivocally disagree with it. But it is still an arbitrary number and it is self-evident that it will bias against treatments with a small effect size:
Sorry 2 scoops I just spotted this. I doubt very much it is an arbitrary number, as if N.I.C.E. pulled this number out of a hat. I'm sure a good deal of thought went into setting this number at three to ensure the benefits of the treatment would be significant enough to outweight the costs. It is indeed "self-evident that it will bias against treatments with a small effect size", in other words it does exactly what it is supposed to do.
My own two cents on this study is that it is excellent and adds real value by opening up the debate on how we treat the different types of depression most effectively. I think the researchers involved in this study were well aware that doctors, scientists etc around the world would go through this study with a fine tooth comb looking for any flaw and they seem to have covered their bases very well. I also think doctors gave up on the older tried and trusted medications much to quickly without properly evaluating the evidence for the newer medications. The tricyclics, MAOIs and lithium revolutionised the treatment of mental health and emptied psychiatric hospitals in the 60's/ 70's /80's/. This led doctors, understandably, to believe anything was possible. They put way too much trust in research carried out by the pharmaceutical industry and if this study can be replicated then it will be clear they were misled. I think we could find doctors falling back on the older medications again.0 -
-
-
-
-
-
Advertisement
-
Advertisement