Dynamic meta-analysis: a method of using global evidence for local decision making.

Area of Science:

• Evidence synthesis
• Meta-analysis
• Agroecology
• Conservation biology

Background:

• Meta-analysis typically generates global generalizations from available evidence.
• Global generalizations may lack relevance for local decision-making.
• Dynamic meta-analysis can assess the local relevance of global evidence.

Purpose of the Study:

• To introduce and demonstrate dynamic meta-analysis as a method to bridge the gap between global evidence and local decision-making.
• To showcase how interactive meta-analysis can improve the applicability of synthesized evidence.

Main Methods:

• Developed Metadataset (www.metadataset.com) as a proof-of-concept for dynamic meta-analysis.
• Utilized dynamic methods for subgroup analysis, meta-regression, and recalibration to filter and weight evidence.
• Applied dynamic meta-analysis to an agroecology example to illustrate its application.

Main Results:

• Dynamic meta-analysis can yield different conclusions for distinct subsets of global evidence.
• This approach allows evidence users to participate in decisions traditionally made by systematic reviewers, such as study inclusion and data handling.
• Metadataset enables interactive filtering, weighting, and recalculation of meta-analysis results.

Conclusions:

• Dynamic meta-analysis addresses the challenge of applying global evidence to local decisions.
• This method is applicable across various scientific disciplines, including agroecology and conservation biology.
• Future work should focus on standardizing metadata, developing software packages for dynamic meta-analyses, and establishing protocols for responsible evidence use.