Efficacy of a Large Language Model Data Extraction System in Evidence Reviews for Emerging Infectious Diseases: A Randomized Crossover Trial
Masahiro Ishikane, Yuki Kataoka, Yasushi Tsujimoto, Yuki Moriyama, Yukimasa Matsuzawa, Norio Ohmagari · 2026
Read Efficacy of a Large Language Model Data Extraction System in Evidence Reviews for Emerging Infectious Diseases: A Randomized Crossover Trial by Masahiro Ishikane, Yuki Kataoka, Yasushi Tsujimoto, Yuki Moriyama, Yukimasa Matsuzawa, Norio Ohmagari (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13393321.txt.
Description
Masahiro Ishikane and five colleagues ran a randomised crossover trial testing whether a large language model speeds the tedious data extraction behind evidence reviews for emerging infections. At Japan's national reference centre, five seasoned reviewers pulled set items from mpox articles twice over, once helped by OpenAI's o3 model producing set summaries and once reading PDFs by hand. Twenty task-level checks were completed.
Assisted extraction averaged about eight minutes faster per article, near a quarter saving, with accuracy at one hundred percent either way and no adverse events. The time gap did not reach the bar for statistical proof, and its confidence interval crossed zero, so the authors stay cautious. Still, they argue the practical value for rapid evidence gathering during outbreaks looks real as tools and prompting improve.
Catalog record
| CREATOR | Masahiro Ishikane, Yuki Kataoka, Yasushi Tsujimoto, Yuki Moriyama, Yukimasa Matsuzawa, Norio Ohmagari |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 24828 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |