It is general practice to compare a new treatment against a sham treatment (placebo), and then use those results to compare efficacy of the new treatment (call it B) to an existing treatment (call it A), at least when existing treatments are available.

However, it seems reasonable and more ethical to use A and B together in a study, rather than giving one group a sham treatment. In other words, instead of a placebo group and a treatment group, why aren't there two treatment groups, one with A and one with B? Statistical noise such as regression to the mean, natural progression of the disease, etc would all show up in both groups, so if B gave better results than A, we should be able to say directly from that result that B works better than A, and we would be doing so in a more direct way than comparing the effect sizes of B vs placebo to the effect sizes of A vs placebo.

  • 6
    $\begingroup$ While more ethical and reasonable, and often / usually legally required, comparing a drug to another drug can also has one potential draw-back: assume that drug A) improves survival from 1% to 10%, and drug B), which would be tested after A), would improve from 10% to 11%. In this case it might be practically impossible without a large group of patients to statistically show an improvement over A) - as the potentially less sophisticated, drug A) would have already been released $\endgroup$ – tsttst Dec 6 '17 at 20:57
  • $\begingroup$ I fail to see the point. $\endgroup$ – Daniel Goldman Dec 7 '17 at 0:15
  • 1
    $\begingroup$ @DonBranson I see nothing wrong with them knowing they may be given the 'placebo', as long as they don't know which one they are ('placebo'/other drug vs the drug being tested) $\endgroup$ – Riker Dec 7 '17 at 2:59
  • 4
    $\begingroup$ Ben Goldacre talks about this is in his book "Bad Science" for a fair while. He makes the point that good science now-a-days compares against the current market leader, while bad science compares against a placebo. It's relatively easy to do better than placebo. It's also worth noting that just because a drug doesn't do outright better than the market leader doesn't mean it isn't useful. For example, it may be applicable to people with certain allergies or have a less intense dosing regime. It may be able to be used in a combination therapy while the market leader can't. $\endgroup$ – SGR Dec 7 '17 at 9:39
  • 2
    $\begingroup$ When you want to check if your light bulb is broken or not, do you flick the light switch and see if the room gets illuminated; or do you measure the difference in light intensity between one room and another? The two tests are completely different. The first tries to confirm that the drug works; the second compares the drug's working to another drug. You're essentially suggesting that a driving exam (testing aptitude) can be done by having students race (testing who's best) and giving a driving license to whoever wins the race (as they are clearly better at driving than the others). $\endgroup$ – Flater Dec 7 '17 at 9:43

You are correct. For serious diseases, it would be unethical to withhold the existing treatment (Treatment A) for a placebo. Treatment A is called "standard of care", and new treatments are compared to this.

  • 1
    $\begingroup$ I can't ask this as a main question, but I guess in comments it's fine. How frequently is an A vs B as opposed to placebo vs B method used, or can you at least provide an example clinical trial? $\endgroup$ – Daniel Goldman Dec 6 '17 at 20:26
  • 4
    $\begingroup$ @DanielGoldman This review seems to be an extensive discussion and provides examples. $\endgroup$ – canadianer Dec 6 '17 at 20:51
  • $\begingroup$ Could you perhaps expand on your answer a bit, and maybe include something from what tsttst said? I could just take your answer and the comments and write my own answer, but I think if you expanded on yours, I could tag yours as accepted. $\endgroup$ – Daniel Goldman Dec 6 '17 at 23:47
  • $\begingroup$ @canadianer Doesn't seem to provide figures for rates of A vs B vs Placebo vs B, just a comprehensive answer for the question here. $\endgroup$ – SGR Dec 7 '17 at 9:45

It's worth pointing out that the purpose of the placebo isn't really to test against no treatment, it's to test against the belief of receiving an effective treatment, for both the subject and the researcher. Loooong answer incoming...

Firstly, it's a lovely issue with medical trials that if you get a statistically valid comparison between a group who don't believe they're taking an effective drug and a control group that know they aren't taking anything, you'll find the placebo group has a better result. Part of it will be psychosomatic, part of it will be biased reporting of symptoms, part of it might be slight changes to routine during the trial. The end result, though, is that a placebo treatment does tend to have a slight effect in the direction intended.

IIRC, there were some less ethical studies at one point that demonstrated that if you tell a subject that their placebo drug has negative side effects, they may actually experience those effects even though the drug has no ability to cause them. Certainly there are studies in of the negative effects during studies not intended to cause them, documenting such things as withdrawal effects from being on a placebo course. In the case of negative results the term "nocebo" tends to be more appropriate.

Secondly, a placebo is important for the researchers as well. The core requirement of a double blind test is that neither the subject nor the researcher knows which treatment the subject is on. You might be able to pull off a single blind test for a comparative test but I would personally say it would be ethically problematic for a researcher to be administering an active treatment without knowing which it was. You can dodge the problem with a placebo by making the inactive treatment identical in administration to the active one.

If your test isn't double blind, there's a real risk of researchers having expectation bias skewing the data collecting and results. Unchecked this can result in tests seeming to show a treatment as being far more effective than it actually is, or as having far less serious side effects than it actually causes.

Thirdly, building on those two points and more directly answering the question, you tend to need a placebo group because that can be considered a standard reference.

So, in the first case, even assuming a blind trial, there is going to be some placebo effect going on, but having only two active groups means you can't filter out that effect. How much of the new treatment's effect was due to placebo? Would the old treatment's group have gotten better or worse results if they knew they were on a "proven" treatment? Would the new treatment's group have gotten better or worse results if they didn't think there was a chance of them being on the old treatment? You can't easily and reliably answer these questions without control groups, or if you can the results still won't be accepted as readily.

In the second case the issue is that of comparability. Even with decent correction for researcher bias you can't really get rid of the issue that the results are a comparison between two treatments. To compare to a third you'd have to try and infer it against the existing treatment's results against placebo. It'd go something like "Treatment A resulted in 10 more successful cases than Treatment B; Treatment B resulted in 20 more successful cases than placebo; Treatment C resulted in 40 more successful cases than placebo, so we thin Treatment C would be more effective than Treatment A when compared with no treatment at all". How much trust would you place in that sort of reasoning? And that's assuming you have comparisons against placebo for the other treatments... if every treatment has been measured against one or more other active treatments in studies you'll instead have some horrible network of relative efficacy to work though. Common reference points are your friend.

That's for the simple case... now imagine the more complex case. Your two groups are being tracked for a dozen or more data points, the new treatment does better than the old in some but worse in others. You have lists of side effects from both groups and there's only partial overlap there. Of course you have no control to help you remove biases and placebo effects. The resulting paper is going to end up as a straight comparison between the two treatments and a pretty high standard of objectivity and clean structuring is going to be needed to avoid a reviewer concluding that the results are subjective and unreliable. Trying to go from that paper to compare with other treatments is similarly harder.

It's easier to convince someone that your treatment is safer and more effective if you have clean studies to back it up. That's not to say comparison trials aren't useful in some cases, particularly where there's a strong ethical reason for not withholding a known treatment, but I personally would prefer to have an experiment against an inactive control group before trying to compare two active processes.

I'm also reasonably sure that tests against placebo are done with the subject's expectations being managed appropriately. Ethically the subject has to be informed that there is a chance they will be in the control group and a chance the treatment won't have the expected results. To give informed consent to the trial the subject has to understand that the treatment might not be as good as the older treatments. Thus while there is the ethical issue of effectively withholding treatment, it's being done with the consent and co-operation of the patient, hopefully with the intention of improving medicine as a whole. Having a test protocol that doesn't feature informed consent is risking some serious consequences down the line due to the much bigger ethical no-no that is forced human experimentation.

Even finding a universal cancer cure won't save a researcher from a major breach of ethics in the process of finding it.

  • 3
    $\begingroup$ Also, be aware of the existence of counterintuitive non-transitivity in cases where you are trying to say A>B & B>C => A>C in a statistical context. Look up Nontransitive dice and Simpson's Paradox for examples. $\endgroup$ – Dannie Dec 7 '17 at 19:22
  • $\begingroup$ @user6697063 yeah, Simpson's is really vicious if not accounted for. Non-trans dice is an interesting one, the fallacy is obvious only once you're looking for one. An obvious degenerative for Wiki's example is "1/3 A > B > C; 1/3 B > C > A; 1/3 C > A > B;" to give a 2/3 chance for each of A>B, B>C and C>A $\endgroup$ – Kaithar Dec 8 '17 at 14:51

Drug companies often do exactly what you describe. That's why there are active control drug trials where there is no placebo - one group receives the experimental drug, and the other receives the "standard" treatment.

I worked on the technology side of drug trial administration for a Contract Research Organization (CRO) for awhile, and many of our clients (which included large, recognizable pharmaceutical companies) were doing this.


Approval pathways for investigational treatments for a particular indication involve a development program in which multiple studies at different phases are conducted in order to establish statistical evidence of efficacy and safety. A well-designed program proceeds from animal studies of the mechanism of action and toxicology, to large-scale pivotal (adequate and well-controlled) studies that furnish the bulk of evidence to support regulatory filing and approval; and if approved, are followed by post-marketing studies that continue to collect safety information and how the treatment is utilized in less structured clinical settings and patient populations.

As such, early-phase clinical studies expose a limited number of subjects and may not be controlled (whether with placebo or an active comparator or standard of care). The study population may not even have the condition being treated. The goal here is to show the treatment is safe, or safe enough to expose a larger population to it.

Mid-phase studies are intended to begin gathering preliminary evidence of efficacy as well as to explore avenues by which efficacy can be established (e.g., endpoint selection, population selection, dose selection) in order to support the design of a pivotal trial. These studies have medium-sized sample sizes and should employ a type of treatment control, whether with placebo, placebo on top of standard of care, or with active comparator, depending on numerous factors, such as whether there exists approved therapy for the indication or if it is an orphan indication; if it is ethical to use placebo alone; and so forth. This is the essence of your question and to address it fully is beyond the scope of this reply. It is the responsibility of IRBs to ensure that the trial is designed ethically and the rights of participants protected. However, because in a mid-phase trial the goal is not to formally conclude efficacy but merely establish preliminary evidence of such, less importance is placed on comparing the experimental therapy to existing available treatments. In these studies, it is generally acceptable to use placebo control (unless doing so is clearly unethical) and implement a protocol-defined rescue therapy option.

For a late-stage or pivotal trial intended to support NDA, there are clear statistical requirements for ensuring that the evidence for a treatment effect is in fact valid and genuine. If a previously approved therapy is available, such a trial will generally be required to compare the experimental treatment(s) against it, in order to answer the question of whether the experimental therapy offers a superior benefit/risk profile.

The takeaway is that there are many considerations that delineate how trials may be designed: what is the indication; who will be exposed; are there other available therapies; does the route of administration permit placebo; and so forth. As a result, scientists and statisticians have developed numerous and often clever approaches to designing controlled clinical trials in such a way that participants are not unduly harmed, even when placebo controls are used.

  • 1
    $\begingroup$ "such a trial will generally be required to compare the experimental treatment(s) against it" <--- that's an interesting piece of information... who actually has that requirement? Are we talking legal, FDA approval or medical association setting that requirement? $\endgroup$ – Kaithar Dec 8 '17 at 15:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.