Key points
- A placebo is an inactive treatment designed to look like the real drug. Comparing against it shows what the drug adds beyond expectation, natural recovery, and chance.
- Randomization decides who gets the drug or placebo, so the groups are comparable. Blinding keeps participants and investigators from knowing who got which.
- Placebo controls are generally not used when an effective treatment exists and withholding it would risk serious or irreversible harm.
- Alternatives include comparing against an active drug, adding the new drug or placebo on top of standard care, or comparing against historical data.
- Uncontrolled or historically controlled results are weaker evidence, because many diseases improve or fluctuate on their own.
If patients improve after taking a new drug, did the drug cause the improvement? Not necessarily. People recover on their own, symptoms fluctuate, and simply expecting to feel better can change how people feel and what they report. A placebo-controlled trial is designed to separate the drug's real effect from all of these.
How it works
In a placebo-controlled trial:
- Participants are randomly assigned to receive either the experimental drug or a placebo.
- The placebo looks identical to the real drug, for example a pill of the same size and color with no active ingredient, or a saline injection.
- Participants, and usually investigators, are blinded, meaning they do not know who is receiving which treatment.
- Both groups are followed in the same way, and their outcomes are compared at the end.
Because the only systematic difference between the groups is the drug itself, a difference in outcomes can be attributed to the drug.
Why each element matters
| Element | What it protects against |
|---|---|
| Placebo | Improvement from natural recovery, expectation, extra medical attention, and fluctuating symptoms, all of which affect both groups equally |
| Randomization | Doctors or patients choosing who gets the drug in ways that make the groups different, for example giving it to healthier patients |
| Blinding | Patients reporting, or investigators assessing, outcomes differently because they know which treatment was given |
Trials in which both participants and investigators are unaware of assignments are called double-blind.
The placebo effect
The placebo effect refers to real improvements, especially in symptoms such as pain, mood, or fatigue, that occur in people receiving an inactive treatment. Placebo responses can be substantial in conditions measured by patient-reported symptoms, such as pain and depression. They are one reason uncontrolled studies in these areas are particularly hard to interpret.
Placebo is one of several control types
The International Council for Harmonisation's guideline on control groups (ICH E10) describes several kinds of comparison. FDA regulations (21 CFR 314.126) recognize the same types as possible bases for an adequate and well-controlled study:
| Control type | How it works | Typical use |
|---|---|---|
| Placebo concurrent control | New drug compared with an inactive look-alike | When no effective treatment exists, or when the new drug or placebo can be added to standard care |
| No-treatment concurrent control | New drug compared with no treatment at all | When blinding is impossible |
| Dose-response concurrent control | Different doses of the new drug compared with each other | To show that more drug produces more effect |
| Active concurrent control | New drug compared with an existing effective treatment | When withholding treatment would be unethical |
| External (historical) control | New drug compared with patients treated elsewhere or earlier | Rare, serious diseases where randomized controls are impractical |
Add-on designs are a common compromise. Every participant receives the current standard treatment, and in addition is randomized to the new drug or a placebo. This tests whether the new drug adds benefit without withholding effective care.
When placebo is not appropriate
The ethics of clinical research, set out internationally in the World Medical Association's Declaration of Helsinki, limit the use of placebos. In general, a placebo control is acceptable when no proven treatment exists, or when there are compelling scientific reasons to use one and participants would not face additional risk of serious or irreversible harm.
In practice, this means a new cancer drug is rarely compared against placebo alone when an effective treatment exists. It is more likely to be compared against the current standard of care, or added on top of it.
Why controls matter when reading results
Results without a concurrent control group are much harder to interpret:
- Single-arm trials give everyone the drug. Any improvement could reflect the disease's natural course.
- Historical comparisons measure against patients treated at another time or place. Differences in patient selection, supportive care, and measurement methods can create apparent benefits that are not real.
Single-arm studies can still be valuable, particularly when a drug produces effects that rarely happen on their own, such as tumor shrinkage in an aggressive cancer. In general, though, a randomized, placebo-controlled or active-controlled result is stronger evidence than an uncontrolled one.
What to look for
- Was the trial randomized?
- What was the comparator: placebo, standard care, or nothing?
- Was it blinded, and could side effects have revealed which patients got the drug?
- Was the endpoint objective (such as death) or subjective (such as a symptom score)? Blinding matters more for subjective endpoints.
Sources
- International Council for Harmonisation. E10: Choice of Control Group and Related Issues in Clinical Trials.Regulator
- 21 CFR 314.126: Adequate and well-controlled studies.Statute / regulation
- World Medical Association. Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Participants.Regulator
- National Library of Medicine. ClinicalTrials.gov: Learn About Clinical Studies.Registry
This article is for education. It is not medical advice and not investment advice. Figures are drawn from the sources above and describe historical averages, not the prospects of any particular drug. See our editorial standards and disclaimer.