Trust in publication

Don’t assume peer review and publication is enough.

Peer review (assessment of a study by other researchers working in the same field) and publication, even in prestigious medical journals, are poor indicators of reliability.


Peer review does not guarantee that published studies are reliable. Medical journals rely on peer review to ensure the quality of the research they publish. However, peer review is highly variable, inconsistent, and may not be systematic. For the most part it is done by volunteers, few of whom have formal training, and who commonly do not detect major errors.

Similarly, just because research is published in a medical journal, even one with a good reputation, this does not guarantee that the research is reliable. Sometimes, research that has been peer reviewed and published is so untrustworthy that it is later withdrawn. Many published studies are too small to have reliable results, and small studies are more likely to report extreme results than large studies. Although only a small proportion of published papers are withdrawn, many more are corrected or proven wrong by more reliable fair comparisons.

Published, peer-reviewed studies and published information written for busy decision makers sometimes contain misleading information on the effects of things we do for our health (health actions). Research reports commonly emphasise results that suggest benefits, while ignoring other findings and they often have a high risk of bias, which can result in overestimating or underestimating the effects and cost-effectiveness of health actions. News reports about published comparisons of health actions often do not consider the reliability of the studies, may exaggerate benefits, and can be misleading.


Perhaps the most well-known example of published research that was later withdrawn, was a small study published in a prestigious medical journal called The Lancet. The article about the study suggested that measles, mumps, and rubella (MMR) vaccination might cause autism. Publication of that paper led to a decrease in vaccinated children, outbreaks of measles, serious illness, and at least four deaths that could have been prevented.

Remember: Always consider whether a published study of the effects of a health action is a fair comparison and whether the results of the comparison are reliable.

Back to Top