![[dr. bob]](/dr-bob-sm.gif)
Dr. Bob is Robert Hsiung, MD,
bob@dr-bob.orgShown: posts 106 to 130 of 166. Go back in thread:
Posted by SLS on July 15, 2006, at 7:27:24
In reply to Re: couldn't have said it better myself, posted by cecilia on July 15, 2006, at 7:02:43
> You definitely have a more optimistic outlook than me, Scott! You see the glass as half full, the people who did achieve good results. I see the glass as half empty, all the people who went through the different phases of the study and still didn't get any better. And as others have pointed out, since there was no placebo arm we have no idea what percentage of the people with good results were placebo responders. It seems to me that a study with no placebo arm is going to have a much higher percentage of placebo responders than one without, simply because everyone knows for sure they are getting a real drug and have higher expectations. Cecilia
I predict that the percentage of placebo responders in the STAR*D study will be much lower than you will see for any standard clinical antidepressant trial. This is because most of the participants were true MDD sufferers with chronic and recurrent courses. The selection of this patient population included standardized inclusion criteria and rating scales. That the majority of patients achieved complete remission by their third drug trial paints a very optimistic picture for the average sufferer of depression.
- Scott
Posted by Klavot on July 15, 2006, at 7:32:31
In reply to Re: couldn't have said it better myself, posted by SLS on July 15, 2006, at 7:27:24
The actual paper of the controversial Zoloft / St John's Wort study can be found at http://jama.ama-assn.org. Restricted membership to JAMA is free, and I would strongly encourage anybody interested in psychotropic medication to read this paper. Taken in its entirety, it gives a rather more positive depiction of Zoloft than the frequently pandered claim that Zoloft has no super-placebo efficacy. A few pertinent points:
1. Some critics have said that the placebo success rate in this study was unusually high, that in other similar studies (similar in terms of duration, patient profile etc.) the placebo remission rate is often much lower. So perhaps the placebo success rate in this particular study is a statistical aberration.
2. The average dose of Zoloft used was relatively low (75 mg / day during the acute phase and 89 mg / day during the maintenance phase) especially considering that the mean 17-point HAM-D score for the Zoloft arm was a relatively high 22.5.
3. Patients with a HAM-D suicide score of more than 2 were excluded for ethical reasons. The role of low serotonergic activity in suicidal ideation is well correlated. Had these patients been included in the study, this would likely have improved the outcome of the Zoloft arm relative to placebo.
4. While Zoloft did not differ to a statistically significant extent from placebo in producing full remission, the case of strictly partial remission is another story. Zoloft produced strictly partial remission in some 26 % of patients compared to the 13 % placebo rate.
5. Zoloft had an overall higher response rate than placebo (46 % Zoloft vs 36 % Placebo).
6. Zoloft gave a greater improvement in mean HAM-D, GAF, BDI, SDS and CGI scores than placebo, admittedly not all of them statistically significant.
7. 22 % of placebo patients reported "forgetfulness", compared with 12 % Zoloft patients. Thus it may be that, on average, Zoloft actually improves memory (this has certainly been my personal experience).
8. As far as I can tell, all the authors of the study have financial interests in Pfizer.
Vitamin Z has vindicated itself in my book!
Klavot
Posted by SLS on July 15, 2006, at 7:54:01
In reply to Re: couldn't have said it better myself, posted by SLS on July 15, 2006, at 7:15:39
> It is my contention that the problem with clinical trials of antidepressants is precisely that the inclusion criteria are not rigid enough. Too many people with mild to moderate psychogenic depressions are allowed to participate. That is why the rate of placebo response is so high, and perhaps why the rate of response for the active compound is so low.
- Scott
Posted by Klavot on July 15, 2006, at 13:36:28
In reply to Re: couldn't have said it better myself, posted by Klavot on July 15, 2006, at 7:32:31
Sorry, I misinterpreted the table. Points 4 and 5 above / below should be
4. While Zoloft did not differ to a statistically significant extent from placebo in producing full remission, the case of strictly partial remission is another story. Zoloft produced strictly partial remission in some 23.9 % of patients compared to the 11.2 % placebo rate.
5. Zoloft had an overall higher response rate than placebo (48.6 % Zoloft vs 43.1 % Placebo).
(Both points are still valid!)
> The actual paper of the controversial Zoloft / St John's Wort study can be found at http://jama.ama-assn.org. Restricted membership to JAMA is free, and I would strongly encourage anybody interested in psychotropic medication to read this paper. Taken in its entirety, it gives a rather more positive depiction of Zoloft than the frequently pandered claim that Zoloft has no super-placebo efficacy. A few pertinent points:
>
> 1. Some critics have said that the placebo success rate in this study was unusually high, that in other similar studies (similar in terms of duration, patient profile etc.) the placebo remission rate is often much lower. So perhaps the placebo success rate in this particular study is a statistical aberration.
>
> 2. The average dose of Zoloft used was relatively low (75 mg / day during the acute phase and 89 mg / day during the maintenance phase) especially considering that the mean 17-point HAM-D score for the Zoloft arm was a relatively high 22.5.
>
> 3. Patients with a HAM-D suicide score of more than 2 were excluded for ethical reasons. The role of low serotonergic activity in suicidal ideation is well correlated. Had these patients been included in the study, this would likely have improved the outcome of the Zoloft arm relative to placebo.
>
> 4. While Zoloft did not differ to a statistically significant extent from placebo in producing full remission, the case of strictly partial remission is another story. Zoloft produced strictly partial remission in some 26 % of patients compared to the 13 % placebo rate.
>
> 5. Zoloft had an overall higher response rate than placebo (46 % Zoloft vs 36 % Placebo).
>
> 6. Zoloft gave a greater improvement in mean HAM-D, GAF, BDI, SDS and CGI scores than placebo, admittedly not all of them statistically significant.
>
> 7. 22 % of placebo patients reported "forgetfulness", compared with 12 % Zoloft patients. Thus it may be that, on average, Zoloft actually improves memory (this has certainly been my personal experience).
>
> 8. As far as I can tell, all the authors of the study have financial interests in Pfizer.
>
> Vitamin Z has vindicated itself in my book!
>
> Klavot
>
Posted by ttee on July 15, 2006, at 16:14:56
In reply to Re: couldn't have said it better myself, posted by SLS on July 14, 2006, at 22:55:22
The link with the STAR*D study information was interesting. Particularly, that all the "thought leaders" and psychiatrists involved are the same psychiatrists that get grants from the drug companies to give talks (sell) their drugs. These names are all in the top ten in receiving drug company money in psychiatry today. I understand that the drug companies didn't pay for STAR*D directly, but indirectly, they pay each of these scientists handsomely. As such, it is not surprising that the STAR*D data is very vague and does not discredit any patent meds, nor does credit any patent or non-patent ones either. One can say that no drug company was harmed during the course of the study; therefore, no drug company money to any of the scientists is in jeopardy. They way the system works; it is not possible to do any type of study these days without the drug companies having some level of influence. The best that could be done is to hire scientists for a NIH funded study that never had, or promise to never receive any sort of funding, grant, etc. from any drug company. None of the scientist thought leaders in the STAR*D study could qualify for that.
I hope I am wrong and may be at some point they will report something that could be used clinicaly from the STAR*D study. As it stands now, there is nothing except what we already knew and that for TRD, you just have to try everything and anything and hope for the best.
> > Off course Mirtazapine and Nortriptyline did not fare well in the Star*D study, as they are both off patent and the scientist don't have to worry about a drug company cutting off their funding.
>
> You might want to read this:
>
> http://www.nimh.nih.gov/healthinformation/stard_qa_general.cfm
>
>
> - Scott
Posted by linkadge on July 15, 2006, at 21:31:12
In reply to Re: couldn't have said it better myself, posted by SLS on July 15, 2006, at 5:58:14
I don't think the evidence is overwhealming at all. I'd call it underwhealming.
Linkadge
Posted by linkadge on July 15, 2006, at 21:43:43
In reply to Re: couldn't have said it better myself, posted by cecilia on July 15, 2006, at 7:02:43
>It seems to me that a study with no placebo arm >is going to have a much higher percentage of >placebo responders than one without, simply >because everyone knows for sure they are getting >a real drug and have higher expectations.
Bingo. It was also noted before, that dispite the 'best possable treatment', a significant portion of those with depression will not achieve adequate results.
In addition, there aren't too many studies that follow these individuals to see if their treatments actually continue to work down the road. This is usually the ultimate test. Its easy to get a response with all the bells a whistles, but when the support leaves, and all you've got is a drug, relapse rates are high. Same thing goes with hospital treatment. Who knows what percentage of persons are getting better due to the "treatment" effect. I relapsed within 2 weeks of leaving the hospital on a high dose of zoloft. I wasn't quite sure if I got better because of the zoloft, or the fact that the hospital cafeteria served really good bacon.
It is easy enough to ramp people up to a high dose of antidepressants to get some initial responses, but with the growing information bank regarding AD poop out we could add to the statistics. I read some statistic stating that ~30% of AD responces end up pooping. So maybe only 35% of people actually get any long term benifit.
Linkadge
Posted by linkadge on July 15, 2006, at 21:58:11
In reply to Re: couldn't have said it better myself, posted by SLS on July 15, 2006, at 7:15:39
"That is why the rate of placebo response is so high, and perhaps why the rate of response for the active compound is so low."
Well, seeing as these drugs were approved based on the fact that they could undepress regular, non genetically chemically imballenced mice coupled with the fact that these drugs have never been proven by any strech of the imagination to correct any genuine depression related chemical imballences, I would have to disagree with the above comment.
Studies *have* shown that even in those with the most marked depression related disturbances in brain function, recovery in any form, is associated with resolution of such peturbations. Ie. there is just as much evidence to suggest that depression causes chemical imballences as there is to suggest that chemical imballances cause depression.
Unfortunately, a response to an antidperssant does not, in any way, prove the existance of a "genuine chemical imballence". Like I said above, mice respond to prozac. Regular mice who have been depressed by chronic stress. Mice with no known gentic biochemical abnormalities.
Placebo responce has never been shown to be restricted to those with "situational depression". Even in those with strong genetic dispositions to depression often have spontainious recovery. In addition, there has been no evidence to suggest that situational depression does not respond to antidepressants. Infact, there is ample evidence to suggest that even situational depression responds to antidepressants. Just the same way as ritalin will improve concentration in people without ADHD, that is why it is abused in universities.
Linkadge
Posted by linkadge on July 15, 2006, at 22:09:11
In reply to Re: couldn't have said it better myself, posted by SLS on July 15, 2006, at 7:27:24
"I predict that the percentage of placebo responders in the STAR*D study will be much lower than you will see for any standard clinical antidepressant trial"
Why should this be so? Most clinical studies seem to show similar or worse statistics. I don't know of too many trials that show better statistics.
But, I would interprate the study this way. The study basically showed that if you have failed to respond to two antidepressnats, you are highly unlikely to respond to a different intervention.
I would personally see this as suggesting that those who responded to the first two drugs were likely placebo responders, and by the third drug, most placebo responders had been filtered out.
This is where the placebo arm would have been *very* usefull information.
Could you imagine, if reponse to the third pacebo had dropped similarly to response to the third active drug? That would confirm that each additional stage was filtering out more placebo responders.
Thats where I think these results are really meaningless.
Linkadge
Posted by linkadge on July 15, 2006, at 22:31:49
In reply to Re: couldn't have said it better myself, posted by Klavot on July 15, 2006, at 7:32:31
>1. Some critics have said that the placebo >success rate in this study was unusually high, >that in other similar studies (similar in terms >of duration, patient profile etc.) the placebo >remission rate is often much lower. So perhaps >the placebo success rate in this particular >study is a statistical aberration.
Of course, it could be a statiscial abberation. Though the drug responce could have just as likely been a statiscal aberation.
>2. The average dose of Zoloft used was >relatively low (75 mg / day during the acute >phase and 89 mg / day during the maintenance >phase) especially considering that the mean 17->point HAM-D score for the Zoloft arm was a >relatively high 22.5.
The dose of SJW could have been increased too, though higher doses often do not correlate to better responces. The placebo dose could have been increased, that can work too!
>3. Patients with a HAM-D suicide score of more >than 2 were excluded for ethical reasons. The >role of low serotonergic activity in suicidal >ideation is well correlated. Had these patients >been included in the study, this would likely >have improved the outcome of the Zoloft arm >relative to placebo.
The >role of low serotonergic activity in suicidal >ideation is well correlated.
In some ways that is a perpetuated myth that people base a lot of assumptions on. It essentially is not true. The only real way to get an accurate idea of serotonergic functioning is post mortem. Those with suicidal "ideation" are not yet dead. In addition, no consistant findings regarding typtophan depletion tests, tryptophan hydroxylase genes, or serotonin transporter genes have ever been made in suicidal patients. Altered serotonergic binding has been noted, but that is not necessarily due to "low serotoniergic function". Some have suggested too that "low serotonin" is really just a marker of impulsivity, rather than depression at all. Then there is the recent discovery that serotonergic drugs can often make people suicidal. It all throws a kink in the theories.
>4. While Zoloft did not differ to a >statistically significant extent from placebo in >producing full remission, the case of strictly >partial remission is another story. Zoloft >produced strictly partial remission in some 26 % >of patients compared to the 13 % placebo rate.
Yeah, that was due to their crafty "second measures" if I recall. Essentially, once they saw that the drug didn't perform better than placebo, they intoroduced a set of second measures which of course showed the drug was better than placebo. But, you can't do that. You can't just introduce new measures after the study has been started.
>6. Zoloft gave a greater improvement in mean HAM->D, GAF, BDI, SDS and CGI scores than placebo, >admittedly not all of them statistically >significant.>7. 22 % of placebo patients >reported "forgetfulness", compared with 12 % >Zoloft patients. Thus it may be that, on >average, Zoloft actually improves memory (this >has certainly been my personal experience).
Serotonergic drugs may improve acute memory recall. They can act like stimulants for many patients. Memory problems have been associated with long term SSRI use, so who knows how this affect pans out.
>8. As far as I can tell, all the authors of the >study have financial interests in Pfizer.
No comment.
It is easy in hindsight to speculate why a drug failed.
Linkadge
Posted by linkadge on July 15, 2006, at 22:45:02
In reply to Re: couldn't have said it better myself, posted by SLS on July 15, 2006, at 7:54:01
Thanks for that link. It really helped to summarize what I have been saying.
"In fact, more than half of all recent clinical trials of commonly used antidepressants failed to show statistical superiority for the drug over placebo."
Bingo.
"This is not necessarily because of the ineffectiveness of the antidepressant, but rather because of an increased response to placebo."
Oh I see, so instead of making drugs that actually work, we'll just try and find ways to dampen the placebo effect.
"Factors that may contribute to these findings remain elusive. Using data from U.S. Food and Drug Administration (FDA) Summary Basis of Approval (SBA) reports and from studies conducted by our group, we reviewed methodological factors used in clinical trials of antidepressants."
"The 2 most notable factors affecting positive trials are (1) the inclusion of patients with more severe depression, and (2) the use of a flexible-dose design; these may yield results identifying true antidepressant-placebo differences. Severely ill patients with depression respond well to antidepressants but poorly to placebo."
Any references to support that claim?
I've heard the contrary, that severely depressed people don't respond well to either antidepressants or placebo.>Flexible dosing paradoxically reduces the >response to placebo without augmenting the >response to the antidepressant. All of these >findings suggest that the use of placebo is >mandatory when assessing new antidepressants.
Thats exactly what I have been saying. The use of placebo is "mandatory". It can absolutely bring contest to any trial.
Its easy to suggest why the drugs often do worse than placebo, but harder to actually change the outcomes.
Linkadge
Posted by linkadge on July 15, 2006, at 22:59:21
In reply to Re: Zoloft / SJW errata, posted by Klavot on July 15, 2006, at 13:36:28
5. Zoloft had an overall higher response rate than placebo (48.6 % Zoloft vs 43.1 % Placebo).
Yeah, on "second measures". Though I don't know if we're talking abbout the same study.
http://www.eurekalert.org/pub_releases/2002-04/du-sjw040402.php
Davidson said something stupid. He said that taking SJW was like playing russian Roulette with your health. Though, this study suggests that taking zoloft is more of a risk.
He also said. "It is very important to treat it early and effectively,". But the study doesn't offer any concrete answers on how one should go about doing that.
The problem is that news stations only broadcasted that "SJW is no more effective than placebo", instead of saying that "SJW and Zoloft are no more effective than placebo"
Linakdge
Posted by Klavot on July 16, 2006, at 3:11:35
In reply to Re: couldn't have said it better myself, posted by linkadge on July 15, 2006, at 22:31:49
> >2. The average dose of Zoloft used was >relatively low (75 mg / day during the acute >phase and 89 mg / day during the maintenance >phase) especially considering that the mean 17->point HAM-D score for the Zoloft arm was a >relatively high 22.5.
>
> The dose of SJW could have been increased too, though higher doses often do not correlate to better responces. The placebo dose could have been increased, that can work too!Personally I am not interested in the SJW aspect of the study (since I've never much subscribed to alternative modalities). Rather, I am concerned over the apparent sub-placebo efficacy of Zoloft. The fact remains - and there is no way to maneuver around this - that the dose of Zoloft used was relatively low.
> >3. Patients with a HAM-D suicide score of more >than 2 were excluded for ethical reasons. The >role of low serotonergic activity in suicidal >ideation is well correlated. Had these patients >been included in the study, this would likely >have improved the outcome of the Zoloft arm >relative to placebo.
>
> The >role of low serotonergic activity in suicidal >ideation is well correlated.
>
> In some ways that is a perpetuated myth that people base a lot of assumptions on. It essentially is not true. The only real way to get an accurate idea of serotonergic functioning is post mortem. Those with suicidal "ideation" are not yet dead. In addition, no consistant findings regarding typtophan depletion tests, tryptophan hydroxylase genes, or serotonin transporter genes have ever been made in suicidal patients. Altered serotonergic binding has been noted, but that is not necessarily due to "low serotoniergic function". Some have suggested too that "low serotonin" is really just a marker of impulsivity, rather than depression at all. Then there is the recent discovery that serotonergic drugs can often make people suicidal. It all throws a kink in the theories.
>I read somewhere - though I cannot back this up with any references - that biopsies of the brains of suicide victims tend to show low serotonin levels.
Anyway, this particular study was skewed in favour of nonsuicidal patients, and as SLS points out in his cogent analysis of placebo efficacy, had those patients been included, the results of the Zoloft arm may well have been different.
> >4. While Zoloft did not differ to a >statistically significant extent from placebo in >producing full remission, the case of strictly >partial remission is another story. Zoloft >produced strictly partial remission in some 26 % >of patients compared to the 13 % placebo rate.
>
> Yeah, that was due to their crafty "second measures" if I recall. Essentially, once they saw that the drug didn't perform better than placebo, they intoroduced a set of second measures which of course showed the drug was better than placebo. But, you can't do that. You can't just introduce new measures after the study has been started.
>Since there are baseline measurements for these secondary measures, I am not sure that these measurements were only introduced afterwards. Your claim that they decided afterwards to include these measurements to save face is mere speculation. And again, the fact remains that Zoloft did outperform placebo on these measurements, irrespective of any possible dubious motivations for including these measurements.
> Serotonergic drugs may improve acute memory recall. They can act like stimulants for many patients. Memory problems have been associated with long term SSRI use, so who knows how this affect pans out.
>> >8. As far as I can tell, all the authors of the >study have financial interests in Pfizer.
>
> No comment.It is interesting how, when a study supports the efficacy of a drug, suddenly issues of collusion between researchers and pharmaceutical companies are raised, while when a study such as this one finds no statistically significant super-placebo efficacy in terms of full remission, then the industry involvement of the authors is a suddenly a non-issue.
> It is easy in hindsight to speculate why a drug failed.
>The drug did not "fail". The whole point that I am making is that the coverage given of this study in the popular media has been highly distorted. Furthermore, the points raised above are based on fact rather than speculation.
> Linkadge
>
>
>
>
>
>
Posted by Klavot on July 16, 2006, at 3:19:06
In reply to Re: couldn't have said it better myself » SLS, posted by linkadge on July 15, 2006, at 22:45:02
> >Flexible dosing paradoxically reduces the >response to placebo without augmenting the >response to the antidepressant. All of these >findings suggest that the use of placebo is >mandatory when assessing new antidepressants.
>
> Thats exactly what I have been saying. The use of placebo is "mandatory". It can absolutely bring contest to any trial.
>Placebo should be mandatory when testing absolute efficacy of a drug, I agree. But as others have pointed out already, the STAR*D study was designed to *compare* algorithm success rates, not to test whether the drugs actually work. Besides, from the "real-world" setting of the study, placebo would have been impractical. The study was conducted on patients actively seeking treatment rather than patients consenting to participate in research.
Posted by SLS on July 16, 2006, at 5:49:59
In reply to Re: couldn't have said it better myself, posted by linkadge on July 15, 2006, at 21:58:11
> "That is why the rate of placebo response is so high, and perhaps why the rate of response for the active compound is so low."
>
> Well, seeing as these drugs were approved based on the fact that they could undepress regular, non genetically chemically imballenced mice coupled with the fact that these drugs have never been proven by any strech of the imagination to correct any genuine depression related chemical imballences, I would have to disagree with the above comment.Mice are a model to screen for drugs, nothing more. You are disagreeing with human results. Studies using people chosen with more severe depression have lower placebo response rates.
- Scott
Posted by SLS on July 16, 2006, at 6:13:38
In reply to Re: couldn't have said it better myself, posted by linkadge on July 15, 2006, at 21:43:43
> >It seems to me that a study with no placebo arm >is going to have a much higher percentage of >placebo responders than one without, simply >because everyone knows for sure they are getting >a real drug and have higher expectations.
> Bingo.
You guys are speculating. I think you are both wrong. I think the degree of placebo response is dependent on the selection criteria. I have already cited a paper regarding this issue.
> In addition, there aren't too many studies that follow these individuals to see if their treatments actually continue to work down the road.How long do you think they should be followed for? Is the STAR*D design sufficient?
This is usually the ultimate test. Its easy to get a response with all the bells a whistles, but when the support leaves, and all you've got is a drug, relapse rates are high.
Is this speculation? What are the statistics?
Certainly psychotherapy will enhance one's chances of getting well and remaining well. There are numbers on this. However, I have not seen the numbers on relapse rates once supportive psychotherapy is withdrawn. Let's see them.
> I relapsed within 2 weeks of leaving the hospital on a high dose of zoloft.
That would be considered a placebo effect if you began to "respond" during the first two weeks according to some researchers. You might not really have been responding to the drug. Placebo responders do not remain in remission for very long compared to true responders. You can look at the work of Frederick Quitkin regarding this.
> I wasn't quite sure if I got better because of the zoloft, or the fact that the hospital cafeteria served really good bacon.
It was likely neither.
> It is easy enough to ramp people up to a high dose of antidepressants to get some initial responses, but with the growing information bank regarding AD poop out we could add to the statistics.What exactly is that bank? Where can we find these statistics?
> I read some statistic stating that ~30% of AD responces end up pooping. So maybe only 35% of people actually get any long term benifit.
Cite it. Let us see if they separate out placebo responders first. If they don't, the statistics are worthless.
- Scott
Posted by SLS on July 16, 2006, at 6:41:21
In reply to Re: couldn't have said it better myself, posted by linkadge on July 15, 2006, at 22:09:11
> "I predict that the percentage of placebo responders in the STAR*D study will be much lower than you will see for any standard clinical antidepressant trial"
>
> Why should this be so? Most clinical studies seem to show similar or worse statistics. I don't know of too many trials that show better statistics.I believe I've already addressed this issue in previous posts. I have seen other abstracts on Medline and elsewhere reporting that clinical trials of antidepressants using rigid criteria allowing only the more severe cases of MDD demonstrate lower placebo response rates and greater superiority for putative antidepressants.
The following paper describes the failure of trials with high placebo response rates to demonstrate superiority of antidepressants, but not those with low placebo response rates. It is my contention that high placebo response rates are indicative of poorly designed inclusion criteria that allow people who do not have true MDD to enter the studies. I feel that this is more of a problem today than it was 20 years ago. I believe this is due to the increase in monetary pressure to produce subjects for clinical trials. This is born out by the steadily increasing rate of placebo response observed in these studies over the years. It is indeed ironic that the drug companies, in their haste and zeal, have cost themselves dearly in time and money by applying this pressure.
- Scott
Posted by SLS on July 16, 2006, at 7:23:10
In reply to Re: couldn't have said it better myself » SLS, posted by linkadge on July 15, 2006, at 22:45:02
> > I wasn't quite sure if I got better because of the zoloft, or the fact that the hospital cafeteria served really good bacon.
>
> It was likely neither.
That was a silly thing for me to say. I apologize.
- Scott
Posted by SLS on July 16, 2006, at 7:27:45
In reply to Re: couldn't have said it better myself, posted by SLS on July 16, 2006, at 5:49:59
> > "That is why the rate of placebo response is so high, and perhaps why the rate of response for the active compound is so low."
> >
> > Well, seeing as these drugs were approved based on the fact that they could undepress regular, non genetically chemically imballenced mice coupled with the fact that these drugs have never been proven by any strech of the imagination to correct any genuine depression related chemical imballences, I would have to disagree with the above comment.
>
> Mice are a model to screen for drugs, nothing more. You are disagreeing with human results. Studies using people chosen with more severe depression have lower placebo response rates.
-------------------------------------------------
Just one more from Google:
Severity of Depression and Response to Antidepressants and Placebo: An Analysis of the Food and Drug Administration Database.Articles
Journal of Clinical Psychopharmacology. 22(1):40-45, February 2002.
Khan, Arif MD *+; Leventhal, Robyn M. BA *; Khan, Shirin R. *; Brown, Walter A. MD ++Abstract:
Some studies suggest that more severely ill patients with depression respond well to antidepressants and poorly to placebo, whereas those who are mildly ill respond equally well to antidepressants and placebo. This notion has implications for the design of clinical trials. To further assess and substantiate these putative predictors of antidepressant and placebo response, we assessed the Food and Drug Administration database of 45 phase II and III antidepressant clinical trials. The frequency of statistically significant differences between antidepressants and placebo was higher in the trials that included patients with more severe depression. In the antidepressant-treated groups, the magnitude of symptom reduction was signif-icantly related to mean initial Hamilton Rating Scale for Depression (HAM-D) score; the higher the mean initial HAM-D score, the larger the change. With placebo treatment, however, the higher the mean initial HAM-D score, the smaller the change. Early discontinuation was more frequent among patients whose mean initial HAM-D scores were higher. These data may help inform the design of future antidepressant clinical trials.
- Scott
Posted by linkadge on July 16, 2006, at 18:08:54
In reply to Re: couldn't have said it better myself, posted by Klavot on July 16, 2006, at 3:11:35
>Personally I am not interested in the SJW aspect >of the study (since I've never much subscribed >to alternative modalities). Rather, I am >concerned over the apparent sub-placebo efficacy >of Zoloft. The fact remains - and there is no >way to maneuver around this - that the dose of >Zoloft used was relatively low.
Nobody was stopping them from using higher doses. Higher doses do not always equate to higher efficacy.
>I read somewhere - though I cannot back this up
>with any references - that biopsies of the >brains of suicide victims tend to show low >serotonin levels.Some studies show this, some studies don't. There have been studies that show completely normal levels of serotonin biproducts in the suicide victom. There have been inferences too that low serotonin is simply a marker of impulsivity.
>Anyway, this particular study was skewed in >favour of nonsuicidal patients, and as SLS >points out in his cogent analysis of placebo >efficacy, had those patients been included, the >results of the Zoloft arm may well have been >different.
May have. Its easy in hindsight to speculate, but designing a clinical trial to show these drugs are significantly better than placebo is hard.
>Since there are baseline measurements for these >secondary measures, I am not sure that these >measurements were only introduced afterwards. >Your claim that they decided afterwards to >include these measurements to save face is mere >speculation. And again, the fact remains that >Zoloft did outperform placebo on these >measurements, irrespective of any possible >dubious motivations for including these >measurements.
It was my understanding, (I will look for a reference), that the areas in which zoloft performed better than placebo, were selected after the study was done. Ie, we set out to show it was better on these measures, which it wasn't, so we will select the areas which it did perform better on, and pertend as if those measures are the ones that matter.
>It is interesting how, when a study supports the >efficacy of a drug, suddenly issues of collusion >between researchers and pharmaceutical companies >are raised, while when a study such as this one >finds no statistically significant super-placebo >efficacy in terms of full remission, then the >industry involvement of the authors is a >suddenly a non-issue.For starters, the study (at least the one I am referring to) did not support the efficacy of zoloft. The industry involvement is always an issue, thats not the way these trials should be conducted.
>The drug did not "fail". The whole point that I >am making is that the coverage given of this >study in the popular media has been highly >distorted. Furthermore, the points raised above >are based on fact rather than speculation.
I think we are talking about a different study.
This is the one I was thinking of, and yes sertaline did fail in this one.
http://www.herbalgram.org/ogdenpress/herbalgram/articleview.asp?a=2351&p=Y
Linkadge
Posted by linkadge on July 16, 2006, at 18:11:35
In reply to Re: couldn't have said it better myself, posted by Klavot on July 16, 2006, at 3:19:06
>But as others have pointed out already, the >STAR*D study was designed to *compare* algorithm >success rates, not to test whether the drugs >actually work.
Oh, I see....to determine wheather red pills work better than blue.
Linkadge
Posted by linkadge on July 16, 2006, at 18:13:09
In reply to Re: couldn't have said it better myself, posted by SLS on July 16, 2006, at 5:49:59
>Studies using people chosen with more severe >depression have lower placebo response rates.
It is my understaning that severely depressed also respond poorly to active treaments, unless you have data to suggest otherwise.
Linkadge
Posted by linkadge on July 16, 2006, at 18:44:59
In reply to Re: couldn't have said it better myself » linkadge, posted by SLS on July 16, 2006, at 6:13:38
>You guys are speculating. I think you are both >wrong. I think the degree of placebo response is >dependent on the selection criteria. I have >already cited a paper regarding this issue.
Everybody here has been speculating. You sited a paper that *suggests* this *could* be a reason why placebo responces is high. To proove that this is true is much harder.
>How long do you think they should be followed >for? Is the STAR*D design sufficient?I'd say that a drug needs to work for a full year to be considered effective, but hey, thats just me.
>Is this speculation? What are the statistics?
Thats a good question. It is my understanding that depression has a high relapse rate.
>Certainly psychotherapy will enhance one's >chances of getting well and remaining well. >There are numbers on this. However, I have not >seen the numbers on relapse rates once >supportive psychotherapy is withdrawn. Let's see >them.
I'd have to look. I guess my best counterargument at this point would be the lack of data showing that antdiepressants work long term.
>That would be considered a placebo effect if you >began to "respond" during the first two weeks >according to some researchers. You might not >really have been responding to the drug. Placebo >responders do not remain in remission for very >long compared to true responders. You can look >at the work of Frederick Quitkin regarding this.
People can relapse too if psychosocial stressors rise above the power of the AD. Perhaps this was the case with me leaving the hospital, and return to a high level of psychosocial stress.
>It was likely neither.
My money is on the bacon.
> It is easy enough to ramp people up to a high dose of antidepressants to get some initial responses, but with the growing information bank regarding AD poop out we could add to the statistics.
>What exactly is that bank? Where can we find >these statistics?
The bank I am referring to, is the increase in awareness that antidepressants can loose their effectiveness over time. Simply typing in "antidepressnant + poop out" in a search engine will return dozens of results. Doctors have been forced to confront the issue, such as in the article below, but I don't know if studies have been conducted. Poop out is a well known phenomina.
>Cite it. Let us see if they separate out placebo >responders first. If they don't, the statistics >are worthless.
Good point! I would imagine that studies about the issue would be in the works. The problem with things such as this is funding. Drug companies in general, do not want to pay for a study to determine the poop out rate of their drug. In addition, they have no requirement to proove that their drug works long term.
Linkadge
Posted by linkadge on July 16, 2006, at 18:57:42
In reply to Re: couldn't have said it better myself » linkadge, posted by SLS on July 16, 2006, at 6:41:21
>I believe I've already addressed this issue in >previous posts. I have seen other abstracts on >Medline and elsewhere reporting that clinical >trials of antidepressants using rigid criteria >allowing only the more severe cases of MDD >demonstrate lower placebo response rates and >greater superiority for putative antidepressants.
Oh of course you'll find a few studies that suggest this.
>The following paper describes the failure of >trials with high placebo response rates to >demonstrate superiority of antidepressants, but >not those with low placebo response rates.
Clearly
>It is my contention that high placebo response >rates are indicative of poorly designed >inclusion criteria that allow people who do not >have true MDD to enter the studies.
When you're faced with the fact that overall the antidepressant effect vs. placebo is weak, it becomes natural to want to try and come up with reasons why.
>I feel that
>this is more of a problem today than it was 20 >years ago. I believe this is due to the increase >in monetary pressure to produce subjects for >clinical trials. This is born out by the >steadily increasing rate of placebo response >observed in these studies over the years.Do you have any data to support this? There are certainly other possable reasons why placebo response may have increased over the years.
Show me data to suggest that increased sevarity of depression has a higher ratio of active treatment to placebo response.
Linkadge
Posted by linkadge on July 16, 2006, at 18:58:25
In reply to Re: couldn't have said it better myself - Apology » linkadge, posted by SLS on July 16, 2006, at 7:23:10
Go forward in thread:
Psycho-Babble Medication | Extras | FAQ
Dr. Bob is Robert Hsiung, MD,
bob@dr-bob.org
Script revised: February 4, 2008
URL: http://www.dr-bob.org/cgi-bin/pb/mget.pl
Copyright 2006-17 Robert Hsiung.
Owned and operated by Dr. Bob LLC and not the University of Chicago.