I assume it's not because sugar could be causative too.
September 29, 2017 11:15 AM Subscribe
According to an aside in this article, modern ethics boards frown on useless “sugar pill” placebos in medical studies. Really? Why? Tell me more about this!
This doesn't really jive with my understanding of the way you'd want a study to be designed, and my googling mostly comes up with debates about doctors prescribing placebos rather than medical studies providing them. I'm so curious! Does anyone have any additional information about this?
This doesn't really jive with my understanding of the way you'd want a study to be designed, and my googling mostly comes up with debates about doctors prescribing placebos rather than medical studies providing them. I'm so curious! Does anyone have any additional information about this?
But what if what you want to test is the current standard of care?
posted by clawsoon at 11:19 AM on September 29, 2017
posted by clawsoon at 11:19 AM on September 29, 2017
It's not ethical to withhold effective treatment, so we don't test against nothing anymore--we test new things against what participants would be receiving anyhow if there were no trial.
As a side to the ethical issue, it also may incentivize participation: if you are a cancer patient for example, you would not really be likely to enroll in a clinical trial that may or may not treat you at all. But if you know that you're going to get at least the standard of care, you're probably more likely to participate.
posted by sockermom at 11:27 AM on September 29, 2017 [10 favorites]
As a side to the ethical issue, it also may incentivize participation: if you are a cancer patient for example, you would not really be likely to enroll in a clinical trial that may or may not treat you at all. But if you know that you're going to get at least the standard of care, you're probably more likely to participate.
posted by sockermom at 11:27 AM on September 29, 2017 [10 favorites]
The book Placebo: A Reader goes into this in detail in the third section, on ethics. You could ask your public library to interlibrary loan it for you if you'd like.
posted by sockermom at 11:36 AM on September 29, 2017 [3 favorites]
posted by sockermom at 11:36 AM on September 29, 2017 [3 favorites]
But what if what you want to test is the current standard of care?
That happens, but again: the current standard of care is something that is widely believed to work, and to deny someone that standard treatment is kind of a big deal. Like, there have been a lot of interesting sham surgery studies that suggest that fake surgery works about as well as some standard surgeries. But yikes, what if instead you had found that fake surgery wasn't nearly as good as real surgery and people died or lived in pain for years or something?
Basically it's really hard to ethically test that kind of stuff! Is it better to continue to provide a treatment of dubious usefulness to a whole lot of people, or to withhold treatment from a smaller group of people without their knowledge? It's not an easy question.
The more common issue, though, is that it's generally A LOT EASIER to demonstrate that your treatment is better than a placebo than it is to demonstrate that your treatment is better than other treatments that already work, so if you want to "prove" that your treatment exists (and is therefore marketable) you might be tempted to test against a placebo rather than another treatment.
So, two reasons:
1) withholding treatment is ethically dubious
2) placebo studies are too easy
posted by mskyle at 11:37 AM on September 29, 2017 [6 favorites]
That happens, but again: the current standard of care is something that is widely believed to work, and to deny someone that standard treatment is kind of a big deal. Like, there have been a lot of interesting sham surgery studies that suggest that fake surgery works about as well as some standard surgeries. But yikes, what if instead you had found that fake surgery wasn't nearly as good as real surgery and people died or lived in pain for years or something?
Basically it's really hard to ethically test that kind of stuff! Is it better to continue to provide a treatment of dubious usefulness to a whole lot of people, or to withhold treatment from a smaller group of people without their knowledge? It's not an easy question.
The more common issue, though, is that it's generally A LOT EASIER to demonstrate that your treatment is better than a placebo than it is to demonstrate that your treatment is better than other treatments that already work, so if you want to "prove" that your treatment exists (and is therefore marketable) you might be tempted to test against a placebo rather than another treatment.
So, two reasons:
1) withholding treatment is ethically dubious
2) placebo studies are too easy
posted by mskyle at 11:37 AM on September 29, 2017 [6 favorites]
I recently read the 2016 book Suggestible You, by Erik Vance, which talks about the science of human suggestibility, including the placebo effect. Per that book, we are starting to research placebos more than ever (there having been a giant dearth of research into the effect itself outside of psychology labs), as well as look for genetic markers indicating strong vs weak susceptibility to placebo (I had no idea that was a thing, but evidently it is). Such genetic testing on the pool of test subjects could reduce the placebo-susceptible population of subjects, which could produce better (and way cheaper) results about how a drug performs compared to placebo in double-blind testing.
The book does talk about how in the US it's considered unmedical and unethical to treat with placebo, but plenty of other countries do it, and the book also talks about how there's a whole lot of medical ritual that is using a similar effect of creating the expectation of healing, which for some damn reason actually makes your brain do a lot of the healing work on its own.
It's a fascinating book, talking science where a million other books would be talking woo nonsense. Great read.
posted by Sunburnt at 11:38 AM on September 29, 2017 [1 favorite]
The book does talk about how in the US it's considered unmedical and unethical to treat with placebo, but plenty of other countries do it, and the book also talks about how there's a whole lot of medical ritual that is using a similar effect of creating the expectation of healing, which for some damn reason actually makes your brain do a lot of the healing work on its own.
It's a fascinating book, talking science where a million other books would be talking woo nonsense. Great read.
posted by Sunburnt at 11:38 AM on September 29, 2017 [1 favorite]
Response by poster: To add a little context, the "placebo" in this typhoid vaccine test was a meningitis vaccine (instead of a typhoid vaccine). All participants did get the current standard of care, which was antibiotics, if/after they were infected, so this wasn't adding any additional treatment for the typhoid itself. What I'm curious about was why they gave a meningitis vaccine, rather than saline or whatever neutral medium.
posted by mosst at 11:41 AM on September 29, 2017 [1 favorite]
posted by mosst at 11:41 AM on September 29, 2017 [1 favorite]
One reason for giving a different active product is to maintain blinding. Many vaccines have at least mild side effects, and the complete absence of side effects would give the participant and study staff a better chance at guessing that the participant received placebo. I have worked on clinical trials where a great deal of effort was put into designing control treatments that mimicked the side effects of the experimental treatment.
posted by esoterrica at 11:41 AM on September 29, 2017 [5 favorites]
posted by esoterrica at 11:41 AM on September 29, 2017 [5 favorites]
Response by poster: That sounds like a logical research design choice, but this specified that it was ethics boards making that call. If it's about maintaining similar side effects, which does make sense, why would that decision fall under ethics rather than generic study design/efficacy/anti-bias best practices?
This may also just be poor science journalism, maybe I should go read the study now
posted by mosst at 11:50 AM on September 29, 2017
This may also just be poor science journalism, maybe I should go read the study now
posted by mosst at 11:50 AM on September 29, 2017
Ethics and good research design can be strongly overlapping; if a study is not well-designed but executed anyway, then participants may be exposed to an experimental treatment (risk) for no good reason (i.e., because valid inferences about the experimental treatment can't be made). A substantial part of research ethics is assessment of risk-benefit ratio.
posted by mean square error at 12:17 PM on September 29, 2017 [4 favorites]
posted by mean square error at 12:17 PM on September 29, 2017 [4 favorites]
Placebos are still occasionally used when there's no accepted standard of care to compare to. In this study I suspect the point was to make sure the effects were specific to the typhoid vaccine and not just related to a general immune system boost from having received a vaccine of any kind. Most people in that age group would not have already received meningococcal vaccination and researchers could be reasonably sure that it would be useless against typhoid.
posted by The Elusive Architeuthis at 12:53 PM on September 29, 2017 [3 favorites]
posted by The Elusive Architeuthis at 12:53 PM on September 29, 2017 [3 favorites]
« Older Where (PNW) should I head to see big rocks... | Partial hospitalization for depression Newer »
This thread is closed to new comments.
posted by Jorus at 11:18 AM on September 29, 2017 [14 favorites]