Mynd Healthcare

Home / Evidence & Research Literacy

When Evidence Changes

Research conclusions can change as new findings emerge, errors are corrected and the strengths and limits of existing evidence become clearer.

#An update can change more than the sample size

Cochrane describes a review update as a new edition that may include new data, methods or analyses. More studies can change an estimate, make it more precise, show whether it applies in different settings or add outcomes not previously reported. Revised methods can also change how much confidence is placed in the findings.

Compare the old and new question, eligibility criteria, outcomes and analyses. A changed result does not by itself establish that earlier researchers acted improperly. It may reflect information that was previously unavailable, a different question or a correction that needs to be explained.

#Read what the publication notice actually changes

ICMJE recommends notices that identify corrections and versions that describe and date their changes. A correction may affect a detail or an important result. COPE describes retraction as correcting the literature and warning about unreliable findings; honest error as well as misconduct can be involved. The label alone is not a verdict about intent.

An expression of concern alerts readers to reliability questions. COPE says it can be an interim notice or a final outcome if the issue cannot be resolved. Read the notice, article and any later version together. Do not assume that every warning will soon become a retraction or that an old downloaded PDF is the current record.

#A current review needs a current evidence check

The date of publication is not the same as the date of the search for evidence. Cochrane recommends assessing whether the question remains important and whether new data or methods could meaningfully affect the findings before deciding to update. A living review involves continuing searches and rapid incorporation of evidence, not merely a page with a recent date.

When reading an update, check the search date, newly included studies, reassessed bias and certainty, and the explanation of what changed. New evidence should be integrated into the relevant body of work rather than used to replace it just because it is the newest paper.

#Changed recommendations need their own explanation

A recommendation can change because evidence about benefits or harms changed, but also because resources, feasibility or the weight given to outcomes changed. NICE asks guideline committees to explain how they moved from evidence to a recommendation. Strength of a recommendation and certainty in an effect estimate are related but different ideas.

Ask whether the update concerns the same population, comparison and outcome, and which part of the reasoning changed. Keep new uncertainty visible. This educational page does not promise continuous surveillance of every cited study or advise anyone to stop or change their treatment without clinical guidance.

Source note

The sections above were checked against the linked sources. No clinical review has been performed. This is general research education, not a clinical guideline.