Placebos work, even when patients are in the know, study finds

[Emet]

Science-Base Medicine article:

Placebo effects without deception? Well, not exactly…
Dec 27 2010

http://www.sciencebasedmedicine.org/?p=9339

 

[fls]

In some cultures, the 'doctor' prescribes a meaningless and illogical ceremony, such as bringing squash to the base of a certain tree at a specific time, for 10 days in a row.
The 'doctor' pretends that the gesture is gravely significant in its own right, knowing that its purpose was to distract the mental rut the patient had gotten into.

Even taking a placebo, in a test, is something that would slightly alter or upset the normal routine. A symbolic gesture can be legitimate and sincere, and quite possibly beneficial, though one would expect that the 'doctor' would study the patient before prescribing the healing woo, and fine-tune the prescription for the individual.

I also suspect ritualized distraction is one of the benefits of placebo use. If we stopped fooling around with this silly mind-body stuff and actually investigated how we can make expectation and distraction work for us as psychological tools, I think we'd make much more progress.

Was it you who asked about whether the placebo effect (i.e. expectation) is watered down in clinical studies because some people guess/expect that their 'pill' has no effect? This would be relevant to the results of this study and I think it was mentioned in one of the commentaries which has been linked.

Linda

 

[Badly Shaved Monkey]

Now covered at SBM where I have posted to summarise the situation as follows;

Maximal placebo effect offers little real benefit to patients.

 

[Uncayimmy]

I also suspect ritualized distraction is one of the benefits of placebo use. If we stopped fooling around with this silly mind-body stuff and actually investigated how we can make expectation and distraction work for us as psychological tools, I think we'd make much more progress.

Was it you who asked about whether the placebo effect (i.e. expectation) is watered down in clinical studies because some people guess/expect that their 'pill' has no effect? This would be relevant to the results of this study and I think it was mentioned in one of the commentaries which has been linked.

I recently came across a study looking at expectancy when it came to nausea meds after chemotherapy (I believe that was it). Results were better when patients were led to believe the medicine would work well. The conclusion was for doctors to work on leveraging expectancy in cases like this.

I also recently came across a study where doctors and subjects were asked about whether they could tell who was on the medication or placebo. A very high percentage knew. Side-effects are a big clue. I would suppose with some meds (like anti-depressants) if a patient forgets, which happens a lot, they feel a withdrawal effect of sorts.

This, as I understand it, is at the center of the debate over anti-depressants in general. The statistical significance was there, but only in severe cases was it clinically significant. If a statistically significant number can detect that they are on the real med, that could explain at least some of the results in something as subjective as depression.

Fascinating stuff, really. Years ago some researchers argued that anti-depressants should include no-treatment groups as well as sham treatment groups. Unfortunately, I don't think that's happened in but a few studies.

 

[quarky]

I also suspect ritualized distraction is one of the benefits of placebo use. If we stopped fooling around with this silly mind-body stuff and actually investigated how we can make expectation and distraction work for us as psychological tools, I think we'd make much more progress.

Was it you who asked about whether the placebo effect (i.e. expectation) is watered down in clinical studies because some people guess/expect that their 'pill' has no effect? This would be relevant to the results of this study and I think it was mentioned in one of the commentaries which has been linked.

Linda

It was me. And i believe it was you that straightened me out with mention of the nocebo effect.
My contention was that a double blind test wherein 1/2 got a placebo and 1/2 got a drug, but all were told they got the drug (unethical, i suppose) that the positive response to the placebo would be higher than normal, because the doubt would be removed.

Or, (also unethical, but this is hypothetical) if a placebo was marketed and hyped, it might register a higher % of perceived benefit than usual.
In the case of certain anti-depressants, the difference is slight enough, and the profits large enough, that corruption of the science must be a potent temptation.

 

[rjh01]

<snip>
I also recently came across a study where doctors and subjects were asked about whether they could tell who was on the medication or placebo. A very high percentage knew. Side-effects are a big clue. I would suppose with some meds (like anti-depressants) if a patient forgets, which happens a lot, they feel a withdrawal effect of sorts.
<snip>

This is why I think people in the placebo group should not always receive only sugar pills. The pills they should receive should produce similar side effects to the real pills. To confuse things still tell all participants that one side effect is ###, but then only some in the placebo group will actually have this side effect. Then it would be hard to tell who is in which group.

 

[Percent]

Now covered at SBM where I have posted to summarise the situation as follows;

Maximal placebo effect offers little real benefit to patients.

Well, certainly in this case:

My female cousin is recovering from major abdominal surgery after a ruptured appendix. She has been having serious problems for quite some time and was diagnosed over the last two months with 1) food poisoning, 2) flu, and 3) ovarian cyst. When she had the “ovarian cyst”, a few weeks ago, she couldn’t even manage to get into the car. Her husband had to call an ambulance. But, dutifully, she went home from the hospital with her “cyst”. She immediately returned to work and to caring for her one-year old son.

Of course that “cyst” was her appendix rupturing and she was filled with “cottage-cheese” infection when surgery was finally done last Thursday. She wasn’t even able to sit up in bed for holiday visitors, 3 days later. She will be in the hospital for some time.

I’ve got to think that double-blind expectancy and deception played a large part in her ability to cope for so long. I mentioned “placebo” to my mother on the phone and she said “but they didn’t give her any pills for the cyst!”

I recently learned that my brother, who practices dermatology, sometimes uses placebo as his first line of tx for children with plantar warts. If nothing else, he says he likes to get the parents to wait a while for them to resolve. He hates to inflict pain on kids.

I also suspect ritualized distraction is one of the benefits of placebo use. If we stopped fooling around with this silly mind-body stuff and actually investigated how we can make expectation and distraction work for us as psychological tools, I think we'd make much more progress.
Linda

The mind-body nonsense is a favorite of an acquaintance at my gym. He professes to be a psychiatrist but I have had my doubts about that. He is a proponent of acupuncture and consistently tells me that I, as a psychologist, should understand the influence of the mind (and Christ) on the body. He commonly has questionable methods to heal ills, including a host of supplements and pushing SPECT scans by Daniel Amen. Unfortunately, this acquaintance was rushed to emergency surgery last July with a baseball-sized malignant brain tumor.

I recently came across a study looking at expectancy when it came to nausea meds after chemotherapy (I believe that was it). Results were better when patients were led to believe the medicine would work well. The conclusion was for doctors to work on leveraging expectancy in cases like this.

This seems plausible. No disrespect to people’s endurance during chemo therapy, but who watches Top Gear USA? Rutledge’s speedboat seasickness, in the race to Key West from Miami, went away when he saw that he was beating Tanner’s Lotus. I’ll be using all of this with a dear friend, a car nut, who was just diagnosed with lymphoma.

Fascinating stuff, really. Years ago some researchers argued that anti-depressants should include no-treatment groups as well as sham treatment groups. Unfortunately, I don't think that's happened in but a few studies.

Yeah, I’d love to see more research as well. I can find studies on dopamine that include control groups separate from placebo treatment showing an expectancy effect within the dopamanergic system but there aren’t as many studies to be found with transmitters targeted by common anti-depressasnts.

Anne

 

[Uncayimmy]

This is why I think people in the placebo group should not always receive only sugar pills. The pills they should receive should produce similar side effects to the real pills. To confuse things still tell all participants that one side effect is ###, but then only some in the placebo group will actually have this side effect. Then it would be hard to tell who is in which group.

How do you determine what side-effects a medication has without using inert placebos during the testing phase?

And since the goal is to test the medication against no treatment, why would you introduce a substance that has an effect?

 

[Alan]

I think I can answer the second part. The point of an active placebo is to avoid people taking a placebo from realising it is a placebo due to the lack of ANY effect. It gives the person taking it a side-effect that the actual medication might have, without actually treating anything. It is a step to improve blinding because it makes the placebo more similar to what is being tested. It's not to compare treatment vs no treatment, but treatment without that active ingredient that is being tested for vs. treatment with that ingredient. Everything else about the treatment should be the same...

I am guessing with the first question. If the side-effects of a treatment aren't known because it's early on, does it matter if the active placebo's effect is increased sweating when the actual treatment's side-effects don't include that?

 

[Alan]

I'll pick an extreme fictional example.

Let's say we are researching whether Cthulex helps people stop smoking. It has the well-known side-effect of causing people using it to suffer excruciating pain.

The control group gets given sugar pills.

.............

That would not be a good experimental design. There would be far too little suffering, so that blinding would have failed.

The control group should be given some medication that has no effect on smoking cessation, but still causes the person to suffer excruciating pain.

And I would probably be found guilty of committing first degree mad science.

 

[CapelDodger]

The results didn’t surprise me when I saw them yesterday.

As fls reported, the demand characteristics all but guaranteed sweet returns from the participants who received the placebo.

fls posted what I expected to conclude, and your contribution has saved me a little time.

Irritable Bowel Syndrome as an exemplar hoists a red flag right from the start, apart from anything else. The rest is silence, with noise attached.

 

[Uncayimmy]

I think I can answer the second part. The point of an active placebo is to avoid people taking a placebo from realising it is a placebo due to the lack of ANY effect. It gives the person taking it a side-effect that the actual medication might have, without actually treating anything. It is a step to improve blinding because it makes the placebo more similar to what is being tested. It's not to compare treatment vs no treatment, but treatment without that active ingredient that is being tested for vs. treatment with that ingredient. Everything else about the treatment should be the same...

Have you ever seen that used anywhere outside of pain trials, where, as I understand it, it's an unusual practice as it is?

I am guessing with the first question. If the side-effects of a treatment aren't known because it's early on, does it matter if the active placebo's effect is increased sweating when the actual treatment's side-effects don't include that?

How would you figure it out?

 

[rjh01]

How do you determine what side-effects a medication has without using inert placebos during the testing phase?

And since the goal is to test the medication against no treatment, why would you introduce a substance that has an effect?

My answer to the first question.
At least some side effects of a drug would be known. There are several stages of testing a drug before a double blind test. One such stage is to give the drug to healthy people. They would tell the researchers about side effects.

That assumes it is a new drug. If it is a drug that has been used (or even tested for other treatments) before then the side effects would be well known.

Alan has done a good job of answering the second question.


Remember if the participants can easily work out who has the sugar pill and who has the real pill then it is no longer a good double blind test.

 

[blutoski]

I recently came across a study looking at expectancy when it came to nausea meds after chemotherapy (I believe that was it). Results were better when patients were led to believe the medicine would work well. The conclusion was for doctors to work on leveraging expectancy in cases like this.

Yeah, that's a boilerplate conclusion, though.

I also recently came across a study where doctors and subjects were asked about whether they could tell who was on the medication or placebo. A very high percentage knew. Side-effects are a big clue. I would suppose with some meds (like anti-depressants) if a patient forgets, which happens a lot, they feel a withdrawal effect of sorts.

The word is 'confounding'. Protocols take that into consideration as much as possible. There is a set of statistical and procedural strategies intended to detect and prevent confounding. For example, additional active ingredients can be added to placebos to match side effects. Medications can be dispensed in single doses to prevent sharing. Subjects will not be started until they've finished a tapered 'washout' period with the previous medication to mitigate withdrawl (if required) or the protocol can stipulate only naiive subjects can participate.

This, as I understand it, is at the center of the debate over anti-depressants in general. The statistical significance was there, but only in severe cases was it clinically significant. If a statistically significant number can detect that they are on the real med, that could explain at least some of the results in something as subjective as depression.

Fascinating stuff, really. Years ago some researchers argued that anti-depressants should include no-treatment groups as well as sham treatment groups. Unfortunately, I don't think that's happened in but a few studies.

I was in a conversation about this over the weekend. The main challenge is ethical, especially if we're dealing with severe depression. "Ah, the nontreatment group had eleven suicides, but the active drug group had none. We'll send a letter to their kids thanking them for advancing science."

The current ethical standard is to compare a new drug to the existing effective treatment, or at least to some effective baseline (IIRC, we were testing prospective HIV meds against AZT for years). The baseline was often tested against nontreatment (because it was a pioneer drug) and really we only want to know if the new drug is more effective than the current treatment. If it's better than nontreatment but worse than the current treatment, it's probably still a flop.

 

[Uncayimmy]

Yeah, that's a boilerplate conclusion, though.

What does that mean?

The word is 'confounding'. Protocols take that into consideration as much as possible. There is a set of statistical and procedural strategies intended to detect and prevent confounding. For example, additional active ingredients can be added to placebos to match side effects. Medications can be dispensed in single doses to prevent sharing. Subjects will not be started until they've finished a tapered 'washout' period with the previous medication to mitigate withdrawl (if required) or the protocol can stipulate only naiive subjects can participate.

What do you mean the word is confounding?

Can you cite some studies adding active ingredients to placebos? I've read probably 500 to 1,000 abstracts/studies in my time, and I've never seen it done. I've seen references to something similar being done with pain meds.

I was in a conversation about this over the weekend. The main challenge is ethical, especially if we're dealing with severe depression. "Ah, the nontreatment group had eleven suicides, but the active drug group had none. We'll send a letter to their kids thanking them for advancing science."

What is the ethical issues? Some people are getting placebos while others do not participate in the study. What is unethical about tracking the no-treatment group? Obviously if you are tracking them, they are seeing doctors. If they seem suicidal, which can happen at any time to any group, the doctor can deal with it.

The current ethical standard is to compare a new drug to the existing effective treatment, or at least to some effective baseline (IIRC, we were testing prospective HIV meds against AZT for years). The baseline was often tested against nontreatment (because it was a pioneer drug) and really we only want to know if the new drug is more effective than the current treatment. If it's better than nontreatment but worse than the current treatment, it's probably still a flop.

It takes a while for a "gold standard" to be put in place. There are still numerous placebo controlled studies out there even in areas where treatments already exist.

 

[blutoski]

How do you determine what side-effects a medication has without using inert placebos during the testing phase?

It depends. Many drugs are already out there, but now being tested for a new condition. eg: an antidepressant may be explored as a mood stablizer for bipolar. Or it may have anecdotal evidence that it's an antipsychotic that's now being formally tested.

If it's a completely new drug, hopefully we've seen some animal testing, and some of they pharmacology will already make sense due to its mechanism of action.

The main point of of Phase I Trials is to look for showstopper side effects rather than quantifying effectiveness. Phase I trials are often done on healthy patients who do not even have the target condition.

By Phase II, short-term side effects should be mostly identified and can be simulated in an active placebo if confounding is a risk.

And since the goal is to test the medication against no treatment, why would you introduce a substance that has an effect?

I'm not sure that's "the goal". The goal is to test medication against a control. That can be nontreatment, but it can also be a currently accepted treatment if there is an ethical conflict with nontreatment.

Another type of nontreatment control is 'waiting list'.

 

[Uncayimmy]

By Phase II, short-term side effects should be mostly identified and can be simulated in an active placebo if confounding is a risk.

How do you know what they are? Take a look at the side-effects for placebos in drug trials. We need to run statistical analysis to find out which ones are areal. How can you do that without a control?

I'm not sure that's "the goal". The goal is to test medication against a control. That can be nontreatment, but it can also be a currently accepted treatment if there is an ethical conflict with nontreatment.

We're talking about placebo controlled studies, so obviously we're talking about looking at no treatment. You're talking about a comparison study, which means the whole issue of placebos is off the table. Both groups are getting medications with side effects.

If you are going to test with an active placebo, you have to be certain that whatever you use is known not to have an effect on what you're testing.

 

[Squeegee Beckenheim]

There has been a small study like this before, which also came to the conclusion that placebos can be more effective than no treatment. I can't post links yet, but if you google for "An Exploration of Neurotic Patients' Responses to Placebo When Its Inert Content Is Disclosed" you should find it.

Of course, the effects being tested are highly subjective, is easily influenced by the expectations of the researchers, the study is very, very small and, like in the study in this thread, the doctors told the patients that the placebo had helped others with similar conditions, but it's interesting nonetheless. What I found particularly interesting was that they found the placebo more effective than placebos and even drugs in other trials scored the same way, although they do offer a possible explanation for this.

It is a matter of investment. You make an investment, i.e. time and effort to take pills. Therefor you have a vested interest in thinking (or is that stating?) that it works. Of course what would be of interest is if there was some objective way to measure how much of an improvement there was. I predict that would say none.

Actually, you can test the effectiveness of placebo treatment with tests which objectively measure how much of an improvement there is - by measuring the size of stomach ulcers. They work. A quick google should turn up some papers, or at least their abstracts.

 

[blutoski]

What does that mean?

What I meant was that I keep seeing this as a thowaway recommendation whenever there's a benefit in a placebo group. One of the purposes of a control group is to identify imagined effects. If the effects are influencable by expectation, this should really be categorized as a probable imagined effect, and manipulating expectations should not be advised.

What do you mean the word is confounding?

Confounding^WP is the name for patients reverse-engineering the blinding. I put this in the wiki tag so you can review the entry on wikipedia.

Can you cite some studies adding active ingredients to placebos? I've read probably 500 to 1,000 abstracts/studies in my time, and I've never seen it done. I've seen references to something similar being done with pain meds.

Good question. Active placebo^WPs are probably not highlighted in the protocols as such. It's actually unusual for there to be a detail about placebos at all. Some people consider it an important deficiency.

What is the ethical issues? Some people are getting placebos while others do not participate in the study. What is unethical about tracking the no-treatment group? Obviously if you are tracking them, they are seeing doctors. If they seem suicidal, which can happen at any time to any group, the doctor can deal with it.

If we're talking about (in my example) severe depression and they're feeling suicidal, the standard treatment would almost certainly be for them to be on breakthrough meds (that's why they're candidates for a clinical trial for an antidepressant medication in the first place!). Putting them in a nontreatment group contrary to best treatment guideline when there's elevated risk of suicide is totally unethical. It's not even my opinion - this is the ruling by ethics boards throughout the West.

It takes a while for a "gold standard" to be put in place. There are still numerous placebo controlled studies out there even in areas where treatments already exist.

Yes, that's why I brought up those two examples: severe depression and HIV. Standard guidelines are pretty clear, and standard treatments do exist. Witholding treatment would be unethical.

A pharmaceutical company based in the US got into big trouble over an HIV ARV clinical trial because their protocol was rejected domestically so they went and did the trial in Africa with a nontreatment group. Their defense was that the 'standard treatment' there was nontreatment, which is a grey area. But here in the West, treatment against the standard is strongly enforced if the condition carries risk of mortality or serious morbidity.

There are going to be exceptions, of course.

 

[blutoski]

How do you know what they are? Take a look at the side-effects for placebos in drug trials. We need to run statistical analysis to find out which ones are areal. How can you do that without a control?

I was saying that by Phase III we should know, because the previous Phase 0 and Phase I were done with real patients for the purpose of identifying side effects. There would be multiple controls, depending on the condition.

Also many of the drugs have been around for decades but are being repurposed. Their side effects have probably been established long ago. There may even be published established standard placebos for them from previous trials. Even unpublished, they're a phonecall or email away.

We're talking about placebo controlled studies, so obviously we're talking about looking at no treatment. You're talking about a comparison study, which means the whole issue of placebos is off the table. Both groups are getting medications with side effects.

OK: I misinterpreted your paragraph. You said:

And since the goal is to test the medication against no treatment, why would you introduce a substance that has an effect?

I thought you meant the goal of investigating drugs in clinical trials. Obviously, the goal for clinical trials is to determine the efficacy and side effects of the proposed drug.

To address the question again: yes, the goal of a placebo is to create a control that does not affect the target outcomes (symptomatic or therapeutic) but if it can be identified as a control and this defeats blinding, there's no point.

If you are going to test with an active placebo, you have to be certain that whatever you use is known not to have an effect on what you're testing.

Yes, that would be best.