Validation of The Umbrella Collaboration for Tertiary Evidence Synthesis in Geriatrics: Mixed Methods Study.

Background: The synthesis of evidence in health care is essential for informed decision-making and policy development. This study aims to validate The Umbrella Collaboration (TU), an innovative, semiautomated tertiary evidence synthesis methodology, by comparing it with traditional umbrella reviews (TURs), which are currently the gold standard.

Objective: The primary objective of this study is to evaluate whether TU, an artificial intelligence-assisted, software-driven system for tertiary evidence synthesis, can achieve effectiveness comparable to that of TURs, while offering a more timely, efficient, and comprehensive approach.

Methods: This comparative study evaluated TU against TURs across 8 matched projects in geriatrics. For each selected TUR, a parallel TU project was conducted using the same research question. Outcomes of interest (OoIs), effect sizes, certainty ratings, and execution times were systematically compared. Effect sizes were assessed both quantitatively, by transforming TUR metrics to Cohen d and correlating them with TU's RTU metric, and qualitatively, through categorical classifications (trivial, small, moderate, and large). Certainty levels were compared by mapping Grading of Recommendations Assessment, Development, and Evaluation (GRADE) ratings and TU's sentiment analysis scores onto a common 0-1 scale. Execution time was measured precisely in TU, while TUR durations were estimated from literature benchmarks. Statistical analyses included chi-square tests and Spearman correlations.

Results: Eight TURs in geriatrics were matched with parallel projects using TU. TU replicated 73 of the 86 (85%) OoIs identified by TURs and reported an additional 337 OoIs, representing a 4.77-fold increase in outcome identification. In the comparison of effect size classifications, full concordance was observed in 24 of the 48 (50%) cases, and consistent concordance (full plus 1-level deviation) in 45 of the 48 (94%) cases, with a moderate strength of association (Cramér V=0.339). The correlation of transformed certainty values between TU and GRADE yielded a statistically significant Spearman coefficient (ρ=0.446; P=.02). The average execution time per TU project was 4 hours and 46 minutes, compared with estimated durations of 6-12 months for TURs.

Conclusions: The TU demonstrated high concordance with TURs, replicating 73 of the 86 (85%) outcomes identified by TURs and identifying nearly 5 times as many additional outcomes. The experimental effect size metric (RTU) showed moderate agreement with conventional measures, and the certainty ratings derived from sentiment analysis correlated acceptably with GRADE-based assessments. While further validation is needed, TU appears to be a valid and efficient approach for tertiary evidence synthesis, offering a scalable and time-efficient alternative when rapid results are required.