A question about testing a new medicine.
When testing a new medicine for safety and effectiveness it is common to divide the test subjects into two groups. One group gets the experimental medicine, the other group is a control, who receive placebos.
Does this mean that the control group doesn’t actually receive treatment for their condition? Doesn’t that potentially risk their lives, or their health?
The people in the control group do not get actual medicine, but everything else about the treatment will (must) remain exactly the same. For example, if you’re testing a new anti-depressant, some get the real med, some get the placebo, but they’d all still be in therapy. If you’re testing new arthritis med, the people in the control group would continue to do go for their daily walk (assuming they do currently).
So, a minor nitpick between medicine and treatment, but it’s important. The only thing that should be different between the two groups is the medicine they’re taking.
They’re going to be monitored very closely. Either by the people treating them or, in a double blind study, whoever those people report to.
I would assume that if someone appears to be going (unexpectedly) down hill, they’d be removed from the study.
Also, keep in mind, these meds are being tested to determine safety and efficacy. It’s entirely possible, especially early in the study, for the people that are getting the real med to not get any benefits from it. Subjecting them to the same risks as someone getting the placebo.
Also, you have to take into consideration that the people in the study aren’t there involuntarily. Every one of them signed up for it and all (well, most) understand that they may be getting a placebo and what the risks are for that.
But what if, for example, they are testing a cancer treatment. Existing treatments for that particular form of cancer has a 30% success rate. They hope the new drug will have a 50% success rate.
So, some get the new drug that is maybe 50% effective, and have a good chance of surviving, while others get a placebo that is 0% effective, and die. Is that how it works?
That’s my understanding. But, like I said, both groups will still be doing the same thing and they’ll be getting monitored. So, for example, if the group getting the real drug is also getting radiation, the control group will get radiation as well. And, if they’re expecting the tumor(s) to shrink, but they grow, those people will likely be pulled from the study and get traditional/proven treatments.
Drug trials, IIRC, are pretty closely regulated, you can probably find a lot of guidelines online.
Yes, it is, and there have been many, many drug trials that ended early because it was so obvious that the actual drug worked well (or, for that matter, was too toxic, or didn’t work either). This has also happened with AIDS meds, and meds from other categories as well.
All of this is done with informed consent of the patient, or their guardian if they cannot consent.
I believe that, in the initial stages, where there is not enough proof of effectiveness, if a new drug is being tested for something that already has an effective treatment, study subjects would be limited to people for whom the usual treatment failed, or whose disease had progressed beyond the established medicine’s ability to be beneficial.
IOW, for someone with cancer, no study would be enrolling subjects for a new medicine of unknown effectiveness if there’s already a known, effective, medicine available.
But if that medicine leads to 50% of patients being cancer free in 5 years, then you’ll have a bunch of people for whom it failed. You’ll also have some number of people who are terminally ill, and have exhausted all of the existing treatments. That is where a completely new drug might start. A study of whether it prolongs the lives of terminal patients, or whether it has beneficial effects for people who don’t benefit from the existing treatments.
Later, when it’s known to be effective, it might be compared to the existing treatment rather than a non-treatment control, if providing no treatment would be unethical.
If there is a clear positive [or negative] result to using DrugX versus the placebo during testing, the trial is likely to be suspended on ethical grounds - prolonging it exposes one group or the other to avoidable harm.
I had a friend in a trial for a glial blastoma. He was in stage 4 when he entered the trial, which meant that even if the medication worked, he might not be getting it soon enough. That was done with informed consent. Basically, he was out of options. He was going to die in n months with or without standard treatment. If he was not in the placebo group, he might get a few extra months. Even if he were miraculously cured, the drug would not reverse damage done by the cancer. He was contributing to scientific knowledge, though.
He was also told that if there was such obvious gain by the non-control group, they could halt the study, and treat the placebo group with the medication, and treatment would be free, as part of the program.
He died about a year after the program started. We think he was in the non-control group, because his life-expectancy had been 3 months upon entering it. He also was in a lot less discomfort while in the trial.
He was allowed to take pain medication, and something else that was supposed to help him with mental clarity. I’m not sure what it was. He was also on an antidepressant. IIRC, any medication he was on going into the trial, he stayed on, other than something that was a direct cancer therapy.
There is a principle in designing randomized placebo-controlled trials called clinical equipoise. It basically means that you can only use placebos if there is no better alternative.
So, if there is a treatment that is known to be effective, you cannot deny that treatment to anyone, even to test a treatment that might be better. You’d usually have to test “current treatment only” vs. “current treatment + new treatment” and not have an untreated control group - the “current treatment only” group would act as your control.
In an extreme edge case, with a condition that has severe negative outcomes, and a current treatment that only has marginal benefits, and if the current treatment would interfere with the proposed new treatment, and there’s sufficient prior reason to think the proposed new treatment would likely be significantly better, an Institutional Review Board might authorize trials where the current treatment is withheld from the experimental group, but there’d have to be a very stringent review, very stringent oversight, and very detailed and explicit informed consent from the participants.
I bungled that a bit. In the case of a condition with a known treatment, you would still use a placebo, but you would use it in addition to the current standard of care.
So, your control group would receive the current standard of care + placebo, and your experimental group would receive the current standard of care + the experimental treatment.
Clinical equipoise demands that you never offer one group in a trial an option that you know beforehand is better than an option you’re offering another group.
Which is wrong.
No, it isn’t.
@gdave has it right.
People, it’s literally the second sentence in the Wikipedia article:
In some medical studies, where it may be unethical not to treat patients who present with symptoms, controls may be given a standard treatment, rather than no treatment at all.[2] An alternative is to select controls from a wider population, provided that this population is well-defined and that those presenting with symptoms at the clinic are representative of those in the wider population.[5] Another method to reduce ethical concerns would be to test early-onset symptoms, with enough time later to offer real treatments to the control subjects, and let those subjects know the first treatments are “experimental” and might not be as effective as later treatments, again with the understanding there would be ample time to try other remedies.
You mean the second sentence that says " In comparative experiments, members of a control group receive a standard treatment, a placebo, or no treatment at all. " Which (getting a placebo) is exactly what I and nearwildheaven said?
Or are you talking about the part that you quoted which brings up using placebos in place of actual drugs as an ethics issue and suggests using non-effective treatments in the control group for those early enough in the onset of the illness that there is (or should be) plenty of time to pull them out of the trial and get them moved over to a proven course of action if/when the need arises? It goes on to say that (known) effective treatments should be used for the control group in cases where it would be unethical to take a do-nothing approach.
Different tests are going to have different approaches, both for strictly scientific rigor and ethical issues. But I’m not sure where you’re getting the idea that placebos aren’t used for control groups.
My dad once tested a drug as treatment for an always-fatal disease that at that time had no known treatment. The drug was hugely effective. Many of the study participants taking the drug showed signs of improvement, something that has never been seen for this disease.
I remember him wrestling with when to halt the study. On the one hand, he wanted to give the drug to everyone, of course. On the other hand, he wanted to make sure he had solid enough evidence to convince all the other doctors in the world treating all the other people who had this disease that it was worth offering this treatment to their patients.
He ended the study quite early, and relied partly on the historical data about the disease.
But it was a really hard ethical question, and that’s what he spoke about at dinner for several days.
Joey_P’s response is correct, but I think starting it by saying “That’s my understanding” in reply to Peter_Morris’s hypothetical is a bit confusing, because in that hypothetical, the control group seems to only be receiving a placebo, which is wrong.
If existing treatments have a 30% success rate, clinical equipoise demands that both the experimental and control groups receive existing treatments. The control group would receive “existing treatments + placebo.” The experimental group would receive “existing treatments + new drug.” No one would receive only a placebo.
In the real world, there are often a variety of treatment options, so it’s likely you would have multiple control and experimental groups getting different combinations, but no group would ever receive care that would be worse than the standard of care.
In the case of a drug that is expected to have a 50% success rate versus a current standard of care with a 30% success rate, if the existing treatments interfered with the new drug or vice versa, such that the only way to test it would be to actually withhold existing treatments, I seriously doubt any IRB would approve such a test.
What a lot of people overlook is that the experimental group is also risking their life and health, possibly to a greater degree than the control group. For some reason, people have this idea that “experimental” means “better than anything we had before”. It doesn’t. Sometimes something experimental is better, and sometimes it’s worse. We don’t know which. And that’s why we need to do the experiment.
Yeah, there are situations where, with the current best treatments, you’d only have six months to live. Experimental treatments look pretty appealing, at that point. But there’s still the possibility that the experimental treatment might cut that down to three months.
Oddly, the word “experiment” has vastly different connotations than the word “experimental”, here. Ask people if they want to receive an experimental treatment, and most will say yes. Ask them if they want doctors to experiment on them, and most will say no. Even though both questions mean exactly the same thing.
It’s also probably worth pointing out that medical interventions typically go through multiple phases of clinical research.
Preclinical research is conducted to see if there’s good reason to think there’s anything to research.
Phase 0 trials are conducted on a handful of volunteers, basically just to see how to administer the treatment (to make sure an oral drug is actually digested and becomes bioavailable rather than simply passing through, for example).
Phase I trials are conducted on a larger number of people, mainly to try to work out dosages.
Phase II trials are where you start to see RCT (Randomized placebo-Controlled Trials), but not necessarily. They’re for initial assessments for safety and efficacy.
Phase III trials are the big time. This is where you recruit large numbers of subjects, and conduct large scale RCTs to figure out how well the treatment actually works.
Phase IV trials are basically just ongoing monitoring of actual usage “in the wild.” In a sense, all medical interventions are experimental - individuals and disease progression between individuals vary far too much to ever be able to say with 100% confidence how a given treatment will work.
So, by Phase II, and certainly by Phase III, when you actually use placebos, the researchers should have some idea of how effective a treatment will be, and how dangerous.
As Chronos rightly points out, a treatment may not just wind up being useless, it might be worse. In the infamous recent example, hydroxychloroquine for treating COVID-19, some preliminary studies showed some effects, but the Phase II and Phase III RCT trials pretty much all showed either no clinical effects, or worse clinical outcomes for those receiving HCQ.
I’m assuming that this is just for issues where life or severe illness is on the line. If they’re testing a vaccine for the common cold I don’t think there would be anything unethical with letting the control group get nothing more than a sugar pill and monitoring them for symptoms as they go about their daily lives.
Or to make it even more similar, if there’s a new drug that could make atopic dermatitis disappear forever (cure) in 7 days, you could ethically have the control group take a do-nothing/placebo approach even though there’s multiple other meds they could be taking.
Since there’s no actual clinically-proven prophylactic treatment for the common cold, clinical equipoise wouldn’t be violated by giving the control group a placebo. That would be true for any medical issue where there currently is no effective intervention.
Maybe, but an IRB would have to take a very close look at the study. Again, what you would normally do is have the control group of “current standard of care + placebo” and the experimental group as “current standard of care + new drug”. If you have good reason to believe the new drug and current standard of care would interfere with each other, an IRB might approve a pure placebo vs. new drug trial, but it would be ethically tricky, since you’d be knowingly inflicting pain and suffering (even if minor and short-term) on the placebo group. And even with the best informed consent in the world, I suspect you’d have a hard time actually keeping sufficient numbers of people the control group from self-treating when the discomfort just gets to be too much for them.
That’s a real edge case, though. In general, you have to abide by clinical equipoise. Certainly, in a situation like that outlined by the OP, you’re never going to withhold a treatment with a 30% success rate for a terminal disease to test a new treatment that you think might have a 50% success rate.
A couple more points.
The double-blinded randomized placebo-controlled trial is the gold standard. But it’s not the only way to conduct a trial. You can conduct a trial with current standard of care as the control arm and NewWonderDrug as the experimental arm. In that case, clinical equipoise is violated for the experimental group, since they’re not receiving current standard of care.
Also, to re-emphasize, the entire point of an RCT is to see if your new treatment actually works. It might work. It might have no effect. It might make things worse. It might work, but not as well as the current standard of care. It might work better than the current standard of care in some cases, but be worse in other cases. It might work better at treating the targeted condition but have worse side-effects. And so on. Even in a pure placebo-vs.-experimental trial, the experimental group may not be getting the better end of the deal.
You can violate clinical equipoise. There are real-world situations where it makes sense to do so. But you need to acknowledge that you are doing it and convince your Institutional Review Board that it’s justified.
Everything I ever heard about clinical trials suggests that gdave has it exactly right.