Key Points: 1. Traditionally, assessing publication bias has been recommended for meta-analyses and as part of judgments concerning certainty of evidence (e.g., Grading of Recommendations, Assessment, Development, and Evaluations [GRADE]). 2. Funnel-plot based tests attempting to assess publication bias (and related concepts such as reporting bias and small-study bias) are underpowered and, furthermore, allow for multiple interpretations, and therefore publication bias cannot be ascertained. 3. At a minimum, "risk of publication bias" should be used instead of "publication bias," and the inclusion and interpretation of this item in meta-analyses and in GRADE should be critically reviewed.
© Copyright 2024 Sports Medicine. Springer. All rights reserved.