The Digital Press

Automated Meta-analysis (AMA) uses AI to speed up combining scientific studies, but current tools mainly focus on data extraction, not full synthesis. This review analyzed 54 studies and found only limited progress toward end-to-end automation.

Why it matters: AMA can improve research efficiency and reproducibility but currently lacks comprehensive automation across all meta-analysis steps.

The big picture: Most AMA work targets early data processing; advanced synthesis and bias evaluation remain underdeveloped.

Stunning stat: Only 2% of studies explored preliminary full-process automation, highlighting a major gap in AMA capabilities.

The stakes: Without full automation and robust methods, AMA's potential for scalable, domain-wide evidence synthesis is restricted.