double blind tests

[ungoliant]

when i went to ayurvedic school, the teacher said that double-blind testing is done for 'safety', that you cannot get any valid results unless you use subjects who have something in common.

he said how can fat pills be tested on skinny people? how can those results be valid?

he said double-blind tests don't work.

he said he had participated in them and they are hooey.

he said how can you test a pill for high blood pressure on people who have low blood pressure.

he said double-blind tests are just to test the safety of an item.

please do not respond with put-downs. please give me scientific answers, i need to ponder this.

 

[Lisa Simpson]

he said how can you test a pill for high blood pressure on people who have low blood pressure.
.

Uh...I don't think that a pill for high blood pressure would be given to someone who has low blood pressure. Double blind merely means that neither the doctor nor the patients knows who was given the active drug and who was given the placebo.

 

[ungoliant]

for example, he said if you are gonna test something about sound, you have to find people who hear.

he says double-blind testing has things that do not fit into it's parameters, and it cannot be used to test everything.

 

[Lisa Simpson]

Well, that may be true, however for medicine double blind works exceedingly well.

 

[Hawk one]

he said how can fat pills be tested on skinny people? how can those results be valid?

he said double-blind tests don't work.

he said he had participated in them and they are hooey.

he said how can you test a pill for high blood pressure on people who have low blood pressure.

he said double-blind tests are just to test the safety of an item.

please do not respond with put-downs. please give me scientific answers, i need to ponder this.

The lines I have emphasized are showing a basic misunderstanding of what a double-blind test is.

A double-blind test for a pill that makes you slimmer won't be tested on a fat group, and also on a skinny group.

The pill will be tested on two fat groups, where one group gets the real thing, and the other group gets a placebo. And neither group will know which kind of pill they will get. If there is a significant change in weight loss between those groups, then the medicament is effective.

Same principle with the blood pressure.

 

[ungoliant]

i don't have the words to explain to him, so i came here to ask for your help. i find i understand the concept much better as an abstract in my head and can't find the words to debate this point.


this was all a debate about whether ayurveda can be proven scientifically.

 

[Hawk one]

Do you wish for me (or someone else, of course) to attempt a more detailed explanation (than the post I already made), so that hopefully you'll have a way to express yourself?

 

[ungoliant]

Do you wish for me (or someone else, of course) to attempt a more detailed explanation (than the post I already made), so that hopefully you'll have a way to express yourself?

yes, please.

 

[Hawk one]

OK. I'll do my best, though there probably will be left gaps in the explanation that others might have to fill:

when i went to ayurvedic school, the teacher said that double-blind testing is done for 'safety', that you cannot get any valid results unless you use subjects who have something in common.

The test subjects do in fact have something in common: The illness that the medication is supposed to help treating. That is the one and only variable that we initially need to check upon. That the group is otherwise as varied as possible is deemed a good thing, because that way, we can see that the treatment is effective all over the range of ordinary differences people have.

he said how can fat pills be tested on skinny people? how can those results be valid?

My previous post dealt with this particular claim. I will reword myself in order to (hopefully) make it even clearer:

If overweight is the problem, and someone claims he's made a pill that will deal with this problem, a proper double-blind test will do the following:

It will seek out people that, to some extent or other, are overweight. Young, old, man, woman, it doesn't matter. It's the fact that they are fat that counts.

Now for the next phase. Randomly, they are split into two groups (it doesn't in fact need to be physically split, as long as there are some ways to know which group person A belongs to).

Then one group will be given the fat pill, while the other group receives a placebo. And nobody will know which pill they will get. This is the heart of the double-blind test, because this eliminates the phenomenon known as "observer's bias".

Of course, in the case of the fat pill, we have a rather objective method of checking the effectiveness: Put them on the weight. But even so, a control group will be needed. Perhaps people are eating less, or more, or differently, while taking the pill. Or they are excercising more (or less). If you're part of such an experiment (which naturally will span over at least a couple of months), you will be much more concious of what you eat and how active you are, and you may be changing your behaviour accordingly.

But thanks to double-blind testing, it won't matter if such behaviour emerges. Even if the placebo group starts losing weight, if the fat pill is really more effective, the real group should lose significantly more weight.

he said double-blind tests don't work.

A simple falsehood. Double-blind tests have shown time and again to be the best practical method of testing a medicine's effectiveness.

he said he had participated in them and they are hooey.

Either he went to a rather dubious test, or his basic misunderstanding on how such a test is conducted distorted his experience.

he said how can you test a pill for high blood pressure on people who have low blood pressure.

Again, same as with the fat pills.

he said double-blind tests are just to test the safety of an item.

Nope. There are other kinds of tests that will deal with the safety of an item. At least the short-term effects. Sometimes, a medicine might have a bad effect in the long run, but it can take years and even decades to manifest itself.

please do not respond with put-downs. please give me scientific answers, i need to ponder this.

I've tried my best giving answers describing the method of double-blind tests. I am not a scientist, but the principle of double-blind testing should be understood even by a layman as myself. (Of course, expressing this principle is another issue, and I may have misunderstood aspects of it.)

With regards to ayurveda. I have no knowledge about ayurveda, and won't insult anyone's intelligence (including mine) by taking a peek-a-boo at Google to search on some fast info on this. Someone that's got the real knowledge (or at least know where to find proper info.) should be able to tell you more about this.

 

[Beth]

when i went to ayurvedic school, the teacher said that double-blind testing is done for 'safety', that you cannot get any valid results unless you use subjects who have something in common.

he said how can fat pills be tested on skinny people? how can those results be valid?

True. Weight loss pills are tested on subjects who are trying to lose weight.

he said double-blind tests don't work.

False.

he said he had participated in them and they are hooey.

he said how can you test a pill for high blood pressure on people who have low blood pressure.

True. You test all medications on people with the condition the pill is hoping to alleviate.

he said double-blind tests are just to test the safety of an item.

False.


Your instructor seems clueless regarding what double-blind means or why it is used.

 

[McCragge]

Doesn't double blind mean that both the doctor and the patient are unaware of which pill (the real or the fake) they are recieving...thus the term "double" blind?

Just wanted to clear that up for my own knowledge

Thanks

McCragge

 

[brodski]

when i went to ayurvedic school, the teacher said that double-blind testing is done for 'safety', that you cannot get any valid results unless you use subjects who have something in common.

he said how can fat pills be tested on skinny people? how can those results be valid?

he said double-blind tests don't work.

he said he had participated in them and they are hooey.

he said how can you test a pill for high blood pressure on people who have low blood pressure.

he said double-blind tests are just to test the safety of an item.

please do not respond with put-downs. please give me scientific answers, i need to ponder this.

the best way to understand double blind placebo controlled studies is to work from first principles.

You want to test if treatment X works, so you give it to a paitents,
do they get better? if they do, all well and good, but would they have got better anyway? you'll never know, ideally you would run the test twice on the same patient in the same conditions, once giving them the treatment, once not giving them the treatment.
Unfortunately this is impossible.
SO instead you get another patient as similar to the first patient as possible and you give them no treatment while you give the first patient the treatment you want to test. If the treated patient got better, well this indicates that the treatment is effective.

However, there are so many variables which can affect recovery that having a sample size of 1 for each group is meaningless, so you replicate the experiment with as many paitents as practical in each group.

The two groups should be as similar as possible, to get as close to repeating the experiment on exactly the same patient as possible.


OK, so some people get better others don't. But how do you judge who got better and who didn't in an objective manner?
The researchers have a vested interest in seeing the treated group get better, and so are likely to interpret those patients results as getting better, even if there has been no significant improvement.
The patients in the treated group are likely to be under some pressure to say they are feeling better , where there has in fact been no improvement.

How do you stop both the researchers and the patients supplying misleading data (usually unwittingly), simple, you don't let either group know who is receiving real treatment and who is not, and you don't let the researchers know either.
That data is held by a third party who only releases this data once the trial has finished, and who does not have any involvement in assessing the patients during the trial (usually they have no involvement with the trial at all, other than to hold the active group/ placebo group data)


We can then count who got better in each group whether they believed they where receiving treatment or not. If the treated group did significantly better than the control group, then we have evidence that the treatment is effective, if the treatment group did not do any better than the placebo group, then it indicates that the treatment is not effective.

Double blind trials underline an important fact about real medicine, it works whether you believe in it or not.

 

[MRWiffen]

Safety for medication is usually not done double blind. The standard method (IIRC I don't work in a clinical lab) they bring a group of healthy, non-smoking men into a clinical lab (ads for this can be found in newspapers). They give all the people there a standard dosage regimen. They then take blood tests of the entire group and see if there is any changes to the blood that are unexpected or hazardous. I think this is clinical phase I, it has to pass this before it can move on to the double blind efficacy testing. I wouldn't be surprised to find out your instructor signed up for this testing and thought that all clinical testing was done double blind instead of just the eficacy testing.

 

[MRC_Hans]

Doesn't double blind mean that both the doctor and the patient are unaware of which pill (the real or the fake) they are recieving...thus the term "double" blind?

Just wanted to clear that up for my own knowledge

Thanks

McCragge

Correct. Because even the best observer will, to some extent, be biased. And even if he is not, the double blind procedure ensures that he is not suspected of bias.

As for ayurvedic and testing, the rationalist answer vill have to be:

Does it claim to objectively change anything, in the real world?

If yes, then it is testable.

Hans

 

[Mojo]

The lines I have emphasized are showing a basic misunderstanding of what a double-blind test is.

A double-blind test for a pill that makes you slimmer won't be tested on a fat group, and also on a skinny group.

The pill will be tested on two fat groups, where one group gets the real thing, and the other group gets a placebo. And neither group will know which kind of pill they will get. If there is a significant change in weight loss between those groups, then the medicament is effective.

Unless, of course, the placebo group is the one that loses significantly more weight. In that case we can conclude that the remedy is worse than useless.

 

[Capsid]

Clinical trials have 3 phases. Phase I is a safety study in a small group of about 20-40 healthy volunteers to establish that the drug is not toxic or causes any adverse events. The drug would have gone through animal toxicity testing before this along with other pre-clinical studies to determine a drug effect. Phase 2 is an efficacy study done in a small group (~20- 40) of the target group. Phase 3 is a large efficacy study using enough subjects so that appropriate statistical analysis may be performed. Each phase is a check before proceeding to the next phase.

Usually the phases are done double blind although single blind studies are conducted if appropriate.

One way of blinding is to issue codes numbers for the drug and placebo and on the day of treatment a random code number is selected (by computer or automated telephone system for example), so no one knows who gets what until the end of the study. The drug and placebo are visually indistinguishable too, so that the doctor can not tell what he is giving to the trial participant.

There are inlcusion and exclusion criteria for the patients enrolled into the study depending on what the target disease is. So for example, a drug for hepatitis B treatment would not recruit people who are alcoholics as the effect of possible cirrhosis might be a confounding factor.

HTH

 

[Mojo]

the best way to understand double blind placebo controlled studies is to work from first principles.

<snip>

Double blind trials underline an important fact about real medicine, it works whether you believe in it or not.

Nominated.

Happy birthday!

 

[brodski]

Nominated.

Happy birthday!

well thanks,

and thanks!

 

[kevin]

double-blinds are done for one reason: to determine if there is a difference between two things, without undue influence of the tester's or testee's own bias.

In medicine a double-blind is done to determine if a pill/regime/etc... is different than the placebo. If the patients on the real drug have more of a beneficial effects than those on the placebo then it is considered "better". If those on the drug do worse than the placebo the test will probably be ended early. If those on the drug do the same as on the placebo the drug probably won't go any further in testing.

In audio a double-blind is done to see if a component produces an audible effect. It will not answer subjective questions such as which is "better". So much audio crap is sold by dishonest companies that double-blinds are important first step in weeding the liars out. First find out if a piece of equipment has an audible difference -- then argue about which result is "better."

Double-blind means both the tester, and the testee are unaware of which test is being performed -- the placebo or the drug, system a or system b. Why do this?

First consider the case where the testee knows they are getting a drug that is supposed to cure them. When asked about effects, their hopes at being cured may make them over-emphasize feelings of improvement, while de-emphasizing side-effects. This can skew the results for a drug that really does lower blood-pressure (improvement) but makes the heart race (probably bad for heart patients) and causes heart attacks. Patients may not report the heart racing side-effect if they think they're on a beneficial drug.

What about when the tester knows they are giving a drug vs. giving a placebo? This is a more subtle situation and is more psychologial based. The tester may treat a patient getting the real drug differently from one getting the placebo and therefore skew the results. For example, if the tester knows the patient is on the real drug and the patient reports no improvement the tester may then "push" the patient ("are you sure?") to change their answers.

Double-blind eliminates (or at least reduces) many of these psychological effects and focuses the study on the real effects of a drug, or audio component.

I've seen several believers in alt medicines will claim that their particular product is taking advantage of the "placebo" effect. Remember -- the placebo is the same as doing nothing and that is what their medicines are doing.

I've seen unscrupulous audio salesman demoing systems for customers take advantage of knowing which system (the expensive or the cheaper one) is being demoed. They'll make the more expensive one slightly louder because peopel tend to equate louder with better.

 

[Jon.]

I've seen several believers in alt medicines will claim that their particular product is taking advantage of the "placebo" effect. Remember -- the placebo is the same as doing nothing and that is what their medicines are doing.

Actually, it's not really true to say that placebo is the same as doing nothing. Placebo is the effect caused by having treatment of any kind, whether it is physiologically effective or not. Anti-depressants are a perfect example. For someone who is depressed, the fact that they are being given medication that is designed to make them feel less depressed may give them hope, which is a good first step out of depression. Independent of the pharmacological effect of the drug, the hope it represents can help to alleviate depression. (Please note that I'm not saying anti-depressants have not pharmacological or physiological effect - they definitely do - but that the placebo effect comes into play as well.)

As for alt medicines taking advantage of the placebo effect, this may well be true for some kinds. IIRC, the famous study of homeopathy recently published in The Lancet showed that h'pathy had no effect above placebo. It would be interesting to know what the placebo effect would be of a substance that the users knew had no pharmacological or physiological effect, and that it was placebo only. That would make a good psychological study.

The whole purpose of DBT of drugs is to separate the placebo effect from the pharmacological effect.