Guideline-augmented large language models for contraindication screening in interventional spine care
Jack Lott, Nicholas Dietrich · 2026
Read Guideline-augmented large language models for contraindication screening in interventional spine care by Jack Lott, Nicholas Dietrich (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/pmc-13543940.txt.
Description
Jack Lott and Nicholas Dietrich tested whether chatbots, given the rulebook, can screen spine procedure cases for safety risks, in a 2026 study. Interventional spine care has clear lists of absolute and relative reasons not to proceed. The team built three hundred eighteen scenarios from the field's safety guidelines, spanning twelve procedure types, and asked three leading models to class each case. Each model worked once unaided and once with guideline text fetched on demand.
Unaided accuracy sat near eighty-six to eighty-seven percent. With the rulebook attached, accuracy climbed to between ninety-two and ninety-eight percent. Every model caught nearly all absolute dangers in both modes. The softer, relative risks were the weak spot unaided, and they gained the most from retrieval. The authors back guideline-fed models for pre-procedure screening, with human review.
Catalog record
| CREATOR | Jack Lott, Nicholas Dietrich |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 23 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |