“One key application of mediation analysis is in clinical trials, where it allows us to answer the “how” question – how does a given treatment work? Critically, it also allows us to drill down further and make evidence-based improvements if we find a treatment didn’t work as well as we had hoped. Regardless of discipline or study type, AGREMA will help ensure mediation analyses are conducted and reported in a robust and replicable manner, improving the information available to practitioners and researchers, and making it more straightforward to synthesise information across studies.”
Dr Kimberly Goldsmith, Biostatistician and Clinical Trials specialist from King’s IoPPN
21 September 2021
New guidelines to improve reporting standards of studies that aim to understand causal mechanisms
Researchers have developed a new set of guidelines for reporting mediation analyses in health research.
A new guideline has been developed to help scientists publish their research accurately and transparently. Published in JAMA, the AGReMA Statement (A Guideline for Reporting Mediation Analyses) provides recommendations for researchers who want to describe mediation analysis in their paper. Mediation analysis is primarily used to understand how an intervention works or why it does not.
The checklists can be found on the AGReMA website, and the explanation and elaboration paper in JAMA, explains the importance of each item and how the checklist can be used by authors, peer reviewers and journal editors.
Hopin Lee, NHMRC Postdoctoral Research Fellow at NDORMS and lead author explained, “The use of mediation analysis has grown rapidly over the past 10 years across a wide range of disciplines. Our research has shown that their reporting has been poor and inconsistent, which makes it difficult to understand how the research was conducted and how reliable the findings are. We hope that the AGReMA statement will fix some of these issues.”
Gary Collins, director of the UK EQUATOR centre noted, “AGReMA was developed through a rigorous evidence- and consensus-based process using the EQUATOR methodological framework. We think it will be a useful tool for many researchers conducting mediation analyses of trials and observational studies.”
Developed by an international team of methodologists, statisticians, clinical trialists, epidemiologists, psychologists, applied clinical researchers, clinicians, implementation scientists, evidence synthesis experts, representatives from the EQUATOR Network, and journal editors, the consensus-based checklist provides detailed steps to help researchers present clear and transparent reports that will raise the standard of reporting of mediation analyses.
“AGReMA is not tied to a particular disease condition or subspecialty of medicine,” said Hopin. “Our working group will liaise with journal editors and funding agencies to increase awareness and encourage its use. Our hope is that it will be endorsed by journal editors, peer reviewers, and authors and improve the accuracy, completeness and consistency in reporting mediation analyses.”
The development of AGReMA was supported by the Berkeley Initiative for Transparency in the Social Sciences, the Center for Effective Global Action, and the Laura and John Arnold Foundation.
A guideline for reporting mediation analyses of 1 randomized trials and observational studies: The AGReMA Statement (DOI:10.1001/jama.2021.14075) was published in JAMA.
For more information, please contact Patrick O’Brien (Senior Media Officer)
In this story
Professor of Medical Statistics and Complex Intervention Methodology