AI-driven evidence synthesis: data extraction of randomized controlled trials with large language models
Bibliographic record
Abstract
The advancement of large language models (LLMs) presents promising opportunities to enhance evidence synthesis efficiency, particularly in data extraction processes, yet existing prompts for data extraction remain limited, focusing primarily on commonly used items without accommodating diverse extraction needs. This research letter developed structured prompts for LLMs and evaluated their feasibility in extracting data from randomized controlled trials (RCTs). Using Claude (Claude-2) as the platform, we designed comprehensive structured prompts comprising 58 items across six Cochrane Handbook domains and tested them on 10 randomly selected RCTs from published Cochrane reviews. The results demonstrated high accuracy with an overall correct rate of 94.77% (95% CI: 93.66% to 95.73%), with domain-specific performance ranging from 77.97% to 100%. The extraction process proved efficient, requiring only 88 seconds per RCT. These findings substantiate the feasibility and potential value of LLMs in evidence synthesis when guided by structured prompts, marking a significant advancement in systematic review methodology.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.371 | 0.743 |
| Meta-epidemiology (narrow) | 0.006 | 0.004 |
| Meta-epidemiology (broad) | 0.009 | 0.014 |
| Bibliometrics | 0.015 | 0.012 |
| Science and technology studies | 0.002 | 0.002 |
| Scholarly communication | 0.008 | 0.008 |
| Open science | 0.005 | 0.009 |
| Research integrity | 0.004 | 0.006 |
| Insufficient payload (model declined to judge) | 0.018 | 0.005 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; the direct Gemma label and the distilled Codex classifier agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".