placebo test irony

[quarky]

greetings. i'm new; on slow dial-up, and lack the time to review the contents of this site. this is my apology if this matter has been discussed to death already.

perhaps many are aware of how big-pharm skewers double blinds in their favor, in respect to placebos. drugs like prozac beat the placebos partly because participants that show a high affinity for the placebo effect are eliminated prior to testing.
that's the small part of the problem.

in most double blind experiments designed to test the efficiacy of a wanna be drug, the participants are aware of the possibility (50/50) that they only got a sugar pill.

it wouldn't be ethical, but suppose a test was conducted wherein ALL participants believed that they got the 'happy' pill. even if they just got the placebo. wouldn't such a test eliminate the anti-placebo effect? wouldn't a much higher effectiveness be predictable?

perhaps what i'm suggesting seems nit-picky.

yet, playing in reverse, suppose an unscrupulous drug maker had realized this?
a worthless (hopefully harmless) drug that had made it past the testing, could be expected to out preform the placebo two-fold...which would be a major success.

does anyone know what i'm getting at?

imagine the placebo effect without the fore-knowledge that you likely got a placebo!

 

[Capsid]

I don't think you fully understand the use of placebos in clinical trials. The placebo is used to discriminate the real drug effects from those occurring as a result of an identical sham procedure. These placebo effects occur at a low rate that they are useless for real medical interventions.

It would help to read your post if you follow convention and capitalised the first letter of the sentence, please.

 

[quarky]

is it not true that in clinical trials, the participant is forewarned that he may be getting a placebo?
what if they knew that they were not getting a placebo? (unethical, but play along a sec) because they were lied to, and 50% got the placebo anyway. how would the placebo effect look then? pretty hot, i suspect.

better than most antidepressants.
sorry about the missing upper case.

 

[homer]

placebo

Thats the problem . They should not know if they are getting a placebo but in todays PC climate that may no longer be possible . This would be fatal to these tests .

 

[tracer]

It seems like in EITHER a blind test where the recipients are all told that they're receiving the "real drug", OR a blind test where the recipients are all told that they might be receiving a placebo, the placebo effect is being accounted for.


Oh, and instead of saying "sorry about the missing upper case," why don'tcha just use the Shift key?

 

[krazyKemist]

is it not true that in clinical trials, the participant is forewarned that he may be getting a placebo?
what if they knew that they were not getting a placebo? (unethical, but play along a sec) because they were lied to, and 50% got the placebo anyway. how would the placebo effect look then? pretty hot, i suspect.

better than most antidepressants.
sorry about the missing upper case.

The point of the placebo, or control group, is to treat everybody the same, I mean psychologically, except for the drug, so as to see the effect of the drug alone realistically. Especially when the interpretation of results is somewhat subjective, as in, 'are you feeling depressed'.

Telling people they may receive placebo is mandatory for ethical reasons. And in a double-blind, even the doctor who administer the drug must not know whether or not he is giving a placebo, so that he does not give it away. He simply administers a coded lot number, and the information related to this lot will be available when compiling the results, not before, so no cheating allowed.

In correctly made clinical trials, all these steps must be documented and traceable. If they are not, the trial may be rejected. As conducting clinical trials costs millions of dollars, and that if you commit an offense such as doctoring results may not only make it void but put you in a bad position towards drug testing agencies, it would not be a very smart thing to do.

the Kemist

 

[69dodge]

is it not true that in clinical trials, the participant is forewarned that he may be getting a placebo?
what if they knew that they were not getting a placebo? (unethical, but play along a sec) because they were lied to, and 50% got the placebo anyway. how would the placebo effect look then? pretty hot, i suspect.

better than most antidepressants.

The placebo effect can also help people who get the real drug, not just people who get the placebo.

I wouldn't be surprised if both groups did better in your scenario than in the usual scenario, but why would your scenario improve the placebo group more than it improved the drug group?

 

[quarky]

because:

if, in a typical double blind test, wherein 25% of those tested have a positive response to a placebo, and only 50% of the participants recieve the placebo, the implication is an effectiveness of 50% for the placebo.
placebos are not effective if one is told that they are definitely getting a placebo.
in most tests, there is a reasonable doubt regarding whether you got the real pill or the sugar pill. 50% doubt.
suppose that doubt was removed? (unethicly, through lying)
wouldn't one expect a much better showing for the sugar pill vs. the drug being tested?

(geepers. i hope this doesn't sound like i'm accussing big drug companies of unethical behaviour. but if they wanted some, herein lies a goldmine, unless my logic is flawed)

 

[casebro]

Re: no caps-

It is entirely possible that Quarky has some form of physical disability that only allows the use of a minimal number of fingers. Or maybe it's a mental problem. If neither, Quarky has some other form of mental problem. At least he's not using leet. Some people just don't realize that the point of communication is- communication?

 

[ponderingturtle]

is it not true that in clinical trials, the participant is forewarned that he may be getting a placebo?
what if they knew that they were not getting a placebo? (unethical, but play along a sec) because they were lied to, and 50% got the placebo anyway. how would the placebo effect look then? pretty hot, i suspect.

better than most antidepressants.
sorry about the missing upper case.

It is relevant? Both groups get the same level of knowledge about their getting the drug or the placebo. So it is an accurate method of measuring the effect of the drug vs nothing.

You seem to be more interested in getting a maximized "placebo effect" and not caring about getting an accurate measurement of the effect of the drug. Not all placebo effects are real effects in the person, some are from observer bias, hence the double blind test and not a single blind test. You can see improvement most anywhere if you look hard enough.

 

[quarky]

you spelled my name wrong

Re: no caps-

It is entirely possible that Quarky has some form of physical disability that only allows the use of a minimal number of fingers. Or maybe it's a mental problem. If neither, Quarky has some other form of mental problem. At least he's not using leet. Some people just don't realize that the point of communication is- communication?

small q in quarky.

 

[fls]

greetings. i'm new; on slow dial-up, and lack the time to review the contents of this site. this is my apology if this matter has been discussed to death already.

perhaps many are aware of how big-pharm skewers double blinds in their favor, in respect to placebos. drugs like prozac beat the placebos partly because participants that show a high affinity for the placebo effect are eliminated prior to testing.

How are they eliminated?

that's the small part of the problem.

in most double blind experiments designed to test the efficiacy of a wanna be drug, the participants are aware of the possibility (50/50) that they only got a sugar pill.

it wouldn't be ethical, but suppose a test was conducted wherein ALL participants believed that they got the 'happy' pill. even if they just got the placebo. wouldn't such a test eliminate the anti-placebo effect? wouldn't a much higher effectiveness be predictable?

The placebo effect isn't really a specific effect. It's a combination of the natural course of events, subjective perceptions on the part of the participants (both the researchers and the subjects), and the effects of participating in a study and getting extra attention. Believing that you received the real medicine really only has the potential to affect your subjective perceptions, which contributes only a small amount to the end result. There have been studies where people have been told they got the real drug or real treatment. It's not unethical (depending upon the details), but it's also not particularly informative.

perhaps what i'm suggesting seems nit-picky.

yet, playing in reverse, suppose an unscrupulous drug maker had realized this?
a worthless (hopefully harmless) drug that had made it past the testing, could be expected to out preform the placebo two-fold...which would be a major success.

does anyone know what i'm getting at?

imagine the placebo effect without the fore-knowledge that you likely got a placebo!

In addition to not being as pronounced (or exciting) as you seem to think, since what you suggest applies equally to the placebo and active treatment groups, it still wouldn't make a difference to the effect in excess of placebo that is attributed to the treatment.

Linda

 

[Dancing David]

greetings. i'm new; on slow dial-up, and lack the time to review the contents of this site. this is my apology if this matter has been discussed to death already.

perhaps many are aware of how big-pharm skewers double blinds in their favor, in respect to placebos. drugs like prozac beat the placebos partly because participants that show a high affinity for the placebo effect are eliminated prior to testing.

How do they establish that? Do they give a pretrial placebo? Where is your evidence?

that's the small part of the problem.

in most double blind experiments designed to test the efficiacy of a wanna be drug, the participants are aware of the possibility (50/50) that they only got a sugar pill.

There are other double blinding procedures, you can also have a third group taking another medication already in the study. You can also have controls for behavioral variables as well. IE the placebo group should recieve their placebos with the exact same same protocol as the medication groups.

it wouldn't be ethical, but suppose a test was conducted wherein ALL participants believed that they got the 'happy' pill. even if they just got the placebo. wouldn't such a test eliminate the anti-placebo effect? wouldn't a much higher effectiveness be predictable?

Not really, you will just have a giant placebo group then.

perhaps what i'm suggesting seems nit-picky.

The issue is thatthen you have only a single group and no experimental variable.

yet, playing in reverse, suppose an unscrupulous drug maker had realized this?
a worthless (hopefully harmless) drug that had made it past the testing, could be expected to out preform the placebo two-fold...which would be a major success.

does anyone know what i'm getting at?

imagine the placebo effect without the fore-knowledge that you likely got a placebo!

The effect should be the same.

 

[quarky]

Thanks for your response, Linda. Out of respect for you, I'm using upper-case, where appropriate.
Perhaps I am mistaken in my understanding of the placebo effect.
I think not, and hope to express the idea more elegantly.

Firstly, I am mostly concerned with anti-depressants; clearly a different category than, say, anti-biotics.
Nobody wants the damn sugar pill when they have a raging infection.
Is it safe to say that the placebo effect is stronger in relation to psycho-somatic maladies and their inherrant drug treatments, than it is in general?

As I have read, in the history of prozac's launching, there was a filtering system that rejected people that were overly prone to placebo effects. I can't provide a link, but if it is true, than the results would have given the drug an unfair advantage over the usual background placebo affectiveness. As it was, the placebos were batting around 300, and the s.s.r.i. hit the low 400's. Not super compelling, for the drug...especially weighing in the side effects. Nevertheless, millions have been helped with prozac. Millions have been made, as well.
Suppose, in stead, the placebo effects were maximized?
Suppose you were testing a placebo, as if it was a real drug? Imagine that millions of dollars were at stake. Its not hard to believe that the placebo's batting average would go up...perhaps neck to neck with prozac; minus the side affects.

I see that i have resorted to sports metaphors in the above.
(That happens when I use upper-case)

I'd like to return with a hypothetical drug test.
In it, all participants are lied to. They are all depressed; they are all given a sugar pill; they are all told that they got the new 'happy pill', and the test is only for side affects.

 

[drkitten]

As I have read, in the history of prozac's launching, there was a filtering system that rejected people that were overly prone to placebo effects. I can't provide a link, but if it is true, than the results would have given the drug an unfair advantage over the usual background placebo affectiveness. As it was, the placebos were batting around 300, and the s.s.r.i. hit the low 400's.

How would this have given the drug an advantage over the placebo? Assuming that the filter was applied to both groups (the control and the experimental group), then the batting average would have been raised for both groups equally.

The only way you could make a case for comparative advantage is if you have some reason to believe that people who are particuarly succeptible to the placebo effect are particularly resistant to the effects of SSRI's (and vice versa). But that assumption can just as easily go the other way -- perhaps people who are strongly succeptible to the placebo effect are particularly sensitive to SSRIs and would have responded even more strongly.

Absent some evidence,.... I just don't see this argument flying in any way.

Its not hard to believe that the placebo's batting average would go up...perhaps neck to neck with prozac; minus the side affects.

Unless the drug's batting average went up as well, as it almost certainly would have.

I'd like to return with a hypothetical drug test.
In it, all participants are lied to. They are all depressed; they are all given a sugar pill; they are all told that they got the new 'happy pill', and the test is only for side affects.

This doesn't test anything, as there is no control group against which to test. You have "no experimental variable."

 

[fls]

Is it safe to say that the placebo effect is stronger in relation to psycho-somatic maladies and their inherrant drug treatments, than it is in general?

When sugar pill is compared to no treatment, any differences are limited to some subjective perceptions and pain. Specifically, sugar pills have no effect on depression.

http://content.nejm.org/cgi/content/abstract/344/21/1594

As I have read, in the history of prozac's launching, there was a filtering system that rejected people that were overly prone to placebo effects. I can't provide a link, but if it is true, than the results would have given the drug an unfair advantage over the usual background placebo affectiveness. As it was, the placebos were batting around 300, and the s.s.r.i. hit the low 400's. Not super compelling, for the drug...especially weighing in the side effects. Nevertheless, millions have been helped with prozac. Millions have been made, as well.

It is not unusual, when performing a clinical study, to try and minimize the amount of 'noise' and to maximize your ability to find an effect, should one exist. This is good research practice. Is that what you're talking about?

Suppose, in stead, the placebo effects were maximized?
Suppose you were testing a placebo, as if it was a real drug? Imagine that millions of dollars were at stake. Its not hard to believe that the placebo's batting average would go up...perhaps neck to neck with prozac; minus the side affects.

You misunderstand what is meant by treatment effect and what is meant by placebo effect. The treatment effect is that which is seen in addition to the effect that was seen in the placebo group. 'Boosting' the effect seen in the placebo group does not somehow decrease the overall effect seen in the treatment group (since that group would also be 'boosted'). Also, there is no evidence that you can boost the effect seen in the placebo group, as sugar pills don't seem to have an effect on depression.

I see that i have resorted to sports metaphors in the above.
(That happens when I use upper-case)

I'd like to return with a hypothetical drug test.
In it, all participants are lied to. They are all depressed; they are all given a sugar pill; they are all told that they got the new 'happy pill', and the test is only for side affects.

What good would that do?

Linda

 

[quarky]

Actually, sugar pills do have some effect on depression; quantifiable, in fact.
So does St. John's wort.

What I'm suggesting is that sugar pills would score better in a test that was funded by a sugar pill company.
Eliminating the most prone to placebo effects doesn't clean up the test at all...it skewers the scores in favor of the non-placebo. A more normal cross section of participants would, if anything, give the placebo more points than if the participants are pre-anti-woogie.

(gosh...did i just write "pre-anti-woogie?"

 

[fls]

Actually, sugar pills do have some effect on depression; quantifiable, in fact.

No. People with depression who are in a placebo group show improvement, but it makes no difference whether or not you also receive a sugar pill. These findings are reported in the link I provided.

I describe in this post some of the reasons why participation in a clinical study gives a false impression of benefit.

What I'm suggesting is that sugar pills would score better in a test that was funded by a sugar pill company.
Eliminating the most prone to placebo effects doesn't clean up the test at all...it skewers the scores in favor of the non-placebo. A more normal cross section of participants would, if anything, give the placebo more points than if the participants are pre-anti-woogie.

(gosh...did i just write "pre-anti-woogie?"

It cannot skewer the results unless you think that being less prone to placebo effects somehow influences your randomization procedure. Explain how you think that could happen.

Linda

 

[drkitten]

Eliminating the most prone to placebo effects doesn't clean up the test at all...it skewers the scores in favor of the non-placebo.

How? At least three people have called you on this unsupported assertion.

I don't think you understand what the placebo effect is. The placebo effect is not how well you respond to sugar pills. It's how well you respond to treatment, in general, independent of the contents of the pills. The placebo effect is exactly the same whether you are handed a sugar pill or a "real" drug -- and the "treatment effect" is the difference that remains when you subtract the response of the control group from the experimental group.

To use your numbers -- let's say that there were 1000 people in each group, and that 200 people were excluded because they would have shown a positive response to ANYTHING they were handed. With 1000 people in each group, 400 of the experimental group and 300 of the control group showed improvement.

If we added the other people in (100 to each group), we would have seen 500/1100 and 400/1100 positive responses, instead. The treatment effect is exactly the same -- 100 more people did better in the experimental group than the control group.

 

[quarky]

have placebos been tested?

How? At least three people have called you on this unsupported assertion.

I don't think you understand what the placebo effect is. The placebo effect is not how well you respond to sugar pills. It's how well you respond to treatment, in general, independent of the contents of the pills. The placebo effect is exactly the same whether you are handed a sugar pill or a "real" drug -- and the "treatment effect" is the difference that remains when you subtract the response of the control group from the experimental group.

To use your numbers -- let's say that there were 1000 people in each group, and that 200 people were excluded because they would have shown a positive response to ANYTHING they were handed. With 1000 people in each group, 400 of the experimental group and 300 of the control group showed improvement.

If we added the other people in (100 to each group), we would have seen 500/1100 and 400/1100 positive responses, instead. The treatment effect is exactly the same -- 100 more people did better in the experimental group than the control group.

Again, I would suggest that it does matter, whether you are handed a placebo, or a real drug. Most participants should half doubt they got a real drug. These hopefull gamblers get the placebo effect from the real drug, in their minds. They put no faith in the placebo. They put their faith (and upwards of 30% effectiveness) in the 50/50 equation presented in the double blind.

Pure drug sucess implies half the people in the test remain unhappy, or sneezing...the one's on the fake drug. But it doesn't pan out that way