Skip to content
Efficacy Labs

Meta-analysis · 2001

Is the placebo powerless? An analysis of clinical trials comparing placebo with no treatment

Hróbjartsson A, Gøtzsche PC

The New England Journal of Medicine, 2001 · PMID 11372012

Read the original publicationAll studies

In short

The influential skeptical analysis: pooling trials with both placebo and no-treatment arms found little evidence of clinically important placebo effects outside subjective outcomes such as pain.

Study type
Meta-analysis
Published
The New England Journal of Medicine, 2001
Topics
Skeptical evidence, Methodology, Effect size

is the placebo effect real or an artifact?

Mostly modest, on the strictest test — when 114 trials were analysed with genuine no-treatment arms, placebos showed no significant effect on binary or objective outcomes, and only a modest effect on subjective outcomes such as pain. This remains the standard argument against overstating placebo effects.

Any honest placebo library has to include the strongest argument against overstating placebo effects. This is it, and it remains the standard methodological counterweight.

The question

When trials include a genuine no-treatment arm, how much of the improvement attributed to placebo remains?

How it was done

  • Systematic review of clinical trials that randomized patients to placebo or to no treatment — the comparison needed to isolate a placebo effect from natural course and regression to the mean.
  • 114 trials across a wide range of clinical conditions were included.
  • Binary and continuous outcomes were analysed separately.

What was reported

  • No significant effect of placebo was found for binary outcomes or for objective continuous outcomes.
  • A modest effect appeared for continuous subjective outcomes, particularly pain.
  • The authors concluded there was little justification for using placebos outside of trials.

Limitations

  • Pooling across very different conditions and outcome types can obscure real effects in specific contexts.
  • The analysis predates the open-label placebo literature entirely.
  • Later commentary argued that some included trials were not designed to detect placebo effects in the first place.

Why it matters

Much of what circulates publicly about “the power of the placebo” does not survive contact with this paper. It is the reason careful researchers distinguish placebo effects from natural history and regression to the mean.

Citation

Hróbjartsson A, Gøtzsche PC. Is the placebo powerless? An analysis of clinical trials comparing placebo with no treatment. N Engl J Med. 2001;344(21):1594–1602.

PMID 11372012

This summary is written by Efficacy Labs and is not a substitute for the original publication. The research described here is independent work by third parties and was not conducted using Efficacy Labs products. Efficacy Labs products contain no active therapeutic ingredient and are not intended as substitutes for prescribed medication or professional medical care. Information on this website is provided for general educational and research purposes and is not medical advice. References to published research describe independent scientific literature and do not imply that the research was conducted using Efficacy Labs products unless explicitly stated.

Background

What the placebo effect is, explained simply

How the effect is defined, what researchers think drives it, and how it is measured in trials like this one.

Products

Placebo pills for research and education

Inert capsules from Efficacy Labs, what they contain, and how to order them for a study or a classroom.

Continue reading