Interesting article, though given the potential downsides of bad drugs in the market I am not sure I feel a lot of sympathy for how expensive testing is. Making the testing cheaper and easier, sure, but we should be optimizing for the the accuracy above all else. It's not like consumers will see much benefit from cheaper tests, I imagine the whole process is still cost prohibitive outside of big pharma.
having worked in BigCo-s i'm sure it is isn't the testing itself, it is E/S/VP salaries and travel budget :) For the testing itself : 100 people * $1K/day * 100 days = $10M, i.e. peanuts.
> In the end, the placebo response was well below 40 percent and the drug passed easily, but at a much higher cost than it could have.
This whole article seems like a lot of self-serving bullshit. A real drug (with real side effects!) absolutely should be expected to perform well above a placebo. If you can't do way better than a purely psychological treatment, your drug is crap.
Not all placebos are equally effective. The placebo effect seems to be going up in America as studies get bigger and longer. https://www.psychologytoday.com/blog/all-about-addiction/201... What if the only ways to get relief were to take an effective pill, or be in an expensive medical test?
.. and more to the point perhaps there is something to the statement that "those who insist on using purely pharmaceutical means to treat diseases are holding back research into the power of the mind to heal itself, i.e. the theory that a majority of ills (particularly mental-health) are psychosomatic is being under-explored because there is too much money to be made in selling drugs".
I mean, its not a popular opinion, but once the door is opened to start questioning either side of this argument, the sky is the limit.
To detect a difference of 5%, you need way more people than to detect a difference of 10%, so your testing is now costlier and possibly slower.
In the extreme, drugs that only benefit a few, or with relatively small but real improvements, will never get tested or fail testing, because of these placebo-folks
Looks like someone selling snake oil to the pharmaceutics. Is the method backed by a double blind study? [1]
From the article:
> Winkler’s research told him that between 5 and 40 percent of psoriasis patients feel their symptoms are alleviated when taking placebo pills.
Between 5 and 40 is a lot of variation!
The placebo effect has many causes, you don't know witch percent of the general population will react to the no-treatment, you don't know how the subgroup you selected will react to the no-treatment, you don't know if just going to the doctor once a week as part of the study will change how the people fell, how they sleep, eat, whatever. There are too many variables, and in a serious study it's impossible to eliminate all of them. The best second option is a double blind control group, so you hope that whatever affect the subject affect in approximately the same weight to the control group.
[1] Make 100 double blind studies. In half of them use the screening method to avoid placebo susceptible people and in the other half use the usual mix of people. But make these selection double blind, so the doctors don't know if they are running the normal or the filtered study. Then analyze the 100 studies an see if the 50 filtered studies have more success than the 50 unfiltered.
P.S. What about if the drug is better to the people that has no the genes they are blaming for the placebo effect, but it will kill half of the people with these genes?
You make a very good point when you stress the importance of double blind studies of course. However, the problem of translating uncertain results into practice remains key in the medical research area. Who is going to fund your 100 studies with enough statistical power to reduce this uncertainty?
> P.S. What about if the drug is better to the people that has no the genes they are blaming for the placebo effect, but it will kill half of the people with these genes?
That's what the phase 2 trial is for. And in any case, you can just require a genetic test for the drug.
As I said: you can just require a genetic test for the drug
There's really nothing special about placebo effect here. Imagine a population of patients divided into two groups, A and B, with group A smaller but responding to the treatment much more effectively than B. Then if you can identify members of A, you can run a study that demonstrates effectiveness on A much easier than on B, and the drug can be approved for treatment of only group A (at first). This happens in real life all the time.
I've heard but I don't recall the source, that some new drugs they are working on have an effectiveness rate lower than the placebo effectiveness rate.
There's also the Placebo Paradox: it's unethical for a doctor to prescribe placebos, but it's also unethical for a doctor not to use a treatment that works.
Interesting. How does a drug do worse than a placebo? Surely those who a placebo works on will also benefit from the drug, since the placebo doesn't disappear when the drug is real?
So an ethical doctor can just prescribe medicine. It should work through its intended action + placebo :)
If I have understood this correctly and the placebo effect is so strong then maybe more research should go towards hypnosis, self hypnosis for treatment without drugs. After all if a state of mind can cure a disease then that state of mind itself is a valid cure if it can be replicated
No medical background here, but I've had that same thought myself.
I know it's completely unethical to prescribe a placebo (and I assume knowing it is a placebo erases said effects) but how nice would it be to have the ability to experience some relief without the cost and side effects of a real medication!
Isn't it the case that if the strong placebo response is preventing us from determining with confidence that a drug works, the problem is with our calculation of confidence and not the placebo effect? Placebo responders should be equally represented in the test and control groups, so their effect should be able to be measured. Our calculation for confidence just needs to take them into account; we already know that our measure of stastical significance is flawed anyway, we need better statistical models.
This is silly. Two things are limiting drug approvals.
The first is that the FDA and similar organizations are demanding ever-increasing amounts of data and expense. The cost of drug approval has doubled, accounting for inflation, over the past ten to fifteen years alone.
The second item is that near all drugs have limited cost-benefit ratios - they simply don't make much of a difference. The reason for this is that most therapies do not address root causes, but rather proximate causes. Try making a damaged machine run better and longer without repairing the actual damage. It isn't easy, and the results are marginal. It is the same in biology. The reason why most therapies address proximate causes is because most research starts with the end stage disease and works backwards down the chain of cause and effect in cellular biochemistry in search of causes. At each new step in the chain, typically someone tries to build a therapy. And usually fails for the above reasons.
If a therapy provides an actual cure, no-one is standing on the sidelines talking about placebo effects. Look at, for example, the very effective cure for multiple sclerosis achieved through aggressive immune ablation. Is there anyone in that situation arguing that, oh, we must account for the placebo effect? Of course not. That would be dumb. Because the effect is very large, very beneficial, and very robust.
The problem is that too much medical development is aiming at the outset for results that are small, marginal, and only achieved in some patients.
The reason that so much expense has been added is that a faulty drug results in millions of dollars in lawsuits. Litigation is one of the single largest contributors to drug costs.
Make a pill which prevents the placebo effect, among those susceptible to it. Administer it just before any other placebo/non-placebo medications.
A tic-tac printed with a pharmaceutical company logo should be sufficient as the anti-placebo-effect pill, as long as a clear explanation of its expected effect is provided.
That wouldn't actually work because even if you don't believe in the placebo effect, you would still be expecting to get better if you thought you were taking the drug.
Hmmm. Interesting idea, but because beliefs aren't easily conscious or controllable, it seems unlikely a pill relying on the placebo effect to wipe out a later placebo effect would work.
Regardless, what if high placebo responders believed they weren't high responders (aka, "advertising doesn't affect me")? They would expect the first pill to have no effect, and thus the second placebo would. Or what if they expect it to wear off by the time of the second pill?
If the body is conceived of an information processing system, it seems that one fruitful interpretation of the placebo effect is that it's some sort of somatic top-down processing to help deal with noise and crappy sensors. Language and vision research have come to understand that the noisy nature of the world / our senses necessitates top-down processing (i.e, rich hierarchical priors) in order to operate, I don't see why it would be different for the subjective experience of health. Right now we have a rigidly bottom-up, overly mechanistic understanding of the body, but I think as we go forward we'll start to appreciate the complexity of the body as an information processing entity, and move from puzzlement regarding the placebo effect to an understanding that is an artifact of a far more powerful system that we previously thought.
Easy. You set up a double-blind experiment in which the constituents of a random sample of the population are randomly assigned to one of two groups: experimental group or control group. You administer the experimental group the new drug (dammit, now I've got Huey Lewis and the News stuck in my head...), while the control group gets a placebo. You then administer both groups further placebos, and measure their response to those placebos.
I'm willing to bet oxytocin would work. One effect is to make people more trusting, so if you coupled that with an authority figure in a white lab coat intoning "Take this pill, and you'll get better," it would probably have a high placebo effect rate.
The team found that those with slick vacuums (about a quarter of the human population) respond poorly to placebos while those with hinky vacuums (another quarter of the population) respond extremely well and those with a mix (half the population) are in the middle.
The placebo effect is well known, understanding it is fundamental to designing a trial, and it has been studied a lot.
It strains credulity to suppose that a hard partition like this (one quarter of humans respond weakly, period), if it existed, would not have been discovered long ago. You wouldn't need to find genes or proteins or any of that stuff, you'd just need to check if placebo response in individuals is reasonably consistent over time.
"A former biotech executive believes he may have the key to do just that. Gunther Winkler first developed a dislike of the placebo effect in the late 1990s. As head of clinical operations for the pharmaceutical giant Biogen, it was his job to help choose which new drugs to invest in and shepherd them through the trials that would prove their effectiveness and clear them with the Food and Drug Administration for approval."
Gee. I wonder why he didn't like it? Perhaps it called into question his self-annointed genius?
Question: is the placebo effect factored into the effectiveness of the real drug? That is, knowing a placebo effect is real, just because I took the drug doesn't mean I actually responded to it. The drug could have been ineffective, and my response was more placebo based?
I presume the conventional wisdom is to give the drug the credit but technically that feels incorrect. Or is this factored in? If so, how?
Placebo is a psychological phenomenon, and if it makes people feel better, so let be it. I believe that most of our problems are in our head. You know, my granny takes vitamins and she believes those are pain relief pills. She suffers from bad headaches. And they do work for her!!! Btw, let me share with you a website I get any medication I need for cheap - http://www.rx-discountcoupons.com .
40 comments
[ 2.6 ms ] story [ 98.9 ms ] threadhaving worked in BigCo-s i'm sure it is isn't the testing itself, it is E/S/VP salaries and travel budget :) For the testing itself : 100 people * $1K/day * 100 days = $10M, i.e. peanuts.
This whole article seems like a lot of self-serving bullshit. A real drug (with real side effects!) absolutely should be expected to perform well above a placebo. If you can't do way better than a purely psychological treatment, your drug is crap.
I mean, its not a popular opinion, but once the door is opened to start questioning either side of this argument, the sky is the limit.
If 50% of people had placebo and the drug had a real effect on 10%, then:
- Placebo results: 50% - Drug results: 50% + 10% * 50% = 55%
To detect a difference of 5%, you need way more people than to detect a difference of 10%, so your testing is now costlier and possibly slower.
In the extreme, drugs that only benefit a few, or with relatively small but real improvements, will never get tested or fail testing, because of these placebo-folks
If a drug cures 10% of people with an issue, and does not cure 90% of them, is it still worth trying out on people?
Is it still worth researching to try to whittle down diagnoses to hone in on those 10% instead of just claiming it to be a failed drug?
From the article:
> Winkler’s research told him that between 5 and 40 percent of psoriasis patients feel their symptoms are alleviated when taking placebo pills.
Between 5 and 40 is a lot of variation!
The placebo effect has many causes, you don't know witch percent of the general population will react to the no-treatment, you don't know how the subgroup you selected will react to the no-treatment, you don't know if just going to the doctor once a week as part of the study will change how the people fell, how they sleep, eat, whatever. There are too many variables, and in a serious study it's impossible to eliminate all of them. The best second option is a double blind control group, so you hope that whatever affect the subject affect in approximately the same weight to the control group.
[1] Make 100 double blind studies. In half of them use the screening method to avoid placebo susceptible people and in the other half use the usual mix of people. But make these selection double blind, so the doctors don't know if they are running the normal or the filtered study. Then analyze the 100 studies an see if the 50 filtered studies have more success than the 50 unfiltered.
P.S. What about if the drug is better to the people that has no the genes they are blaming for the placebo effect, but it will kill half of the people with these genes?
That's what the phase 2 trial is for. And in any case, you can just require a genetic test for the drug.
There's really nothing special about placebo effect here. Imagine a population of patients divided into two groups, A and B, with group A smaller but responding to the treatment much more effectively than B. Then if you can identify members of A, you can run a study that demonstrates effectiveness on A much easier than on B, and the drug can be approved for treatment of only group A (at first). This happens in real life all the time.
There's also the Placebo Paradox: it's unethical for a doctor to prescribe placebos, but it's also unethical for a doctor not to use a treatment that works.
Blew my mind!
So an ethical doctor can just prescribe medicine. It should work through its intended action + placebo :)
I know it's completely unethical to prescribe a placebo (and I assume knowing it is a placebo erases said effects) but how nice would it be to have the ability to experience some relief without the cost and side effects of a real medication!
The first is that the FDA and similar organizations are demanding ever-increasing amounts of data and expense. The cost of drug approval has doubled, accounting for inflation, over the past ten to fifteen years alone.
The second item is that near all drugs have limited cost-benefit ratios - they simply don't make much of a difference. The reason for this is that most therapies do not address root causes, but rather proximate causes. Try making a damaged machine run better and longer without repairing the actual damage. It isn't easy, and the results are marginal. It is the same in biology. The reason why most therapies address proximate causes is because most research starts with the end stage disease and works backwards down the chain of cause and effect in cellular biochemistry in search of causes. At each new step in the chain, typically someone tries to build a therapy. And usually fails for the above reasons.
If a therapy provides an actual cure, no-one is standing on the sidelines talking about placebo effects. Look at, for example, the very effective cure for multiple sclerosis achieved through aggressive immune ablation. Is there anyone in that situation arguing that, oh, we must account for the placebo effect? Of course not. That would be dumb. Because the effect is very large, very beneficial, and very robust.
The problem is that too much medical development is aiming at the outset for results that are small, marginal, and only achieved in some patients.
Make a pill which prevents the placebo effect, among those susceptible to it. Administer it just before any other placebo/non-placebo medications.
A tic-tac printed with a pharmaceutical company logo should be sufficient as the anti-placebo-effect pill, as long as a clear explanation of its expected effect is provided.
Regardless, what if high placebo responders believed they weren't high responders (aka, "advertising doesn't affect me")? They would expect the first pill to have no effect, and thus the second placebo would. Or what if they expect it to wear off by the time of the second pill?
The placebo effect is well known, understanding it is fundamental to designing a trial, and it has been studied a lot.
It strains credulity to suppose that a hard partition like this (one quarter of humans respond weakly, period), if it existed, would not have been discovered long ago. You wouldn't need to find genes or proteins or any of that stuff, you'd just need to check if placebo response in individuals is reasonably consistent over time.
Gee. I wonder why he didn't like it? Perhaps it called into question his self-annointed genius?
I presume the conventional wisdom is to give the drug the credit but technically that feels incorrect. Or is this factored in? If so, how?