Protocol-constrained AI enhances tacrolimus dosing accuracy in kidney transplant care
Benjamin Bizer, Oscar A. Garcia Valencia, Flora Kincses, Jose Arriola-Montenegro, Charat Thongprayoon, Jing Miao, Wisit Cheungpasitporn · 2026
Read Protocol-constrained AI enhances tacrolimus dosing accuracy in kidney transplant care by Benjamin Bizer, Oscar A. Garcia Valencia, Flora Kincses, Jose Arriola-Montenegro, Charat Thongprayoon, Jing Miao, Wisit Cheungpasitporn (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-13233491.txt.
Description
Benjamin Bizer and colleagues published a 2026 study of an artificial intelligence assistant for dosing tacrolimus after kidney transplantation. The drug prevents rejection, but correct doses vary widely between patients and mistakes can harm the graft. The team built TacroDose AI on GPT-4, bound by their center's dosing protocol, and tested it on 300 simulated clinical scenarios.
The first version followed the protocol in about seventy-three percent of cases, with a worrying share of serious errors. After the prompt was tightened, a version called TacroAI 2.0 reached adherence above ninety percent and cut serious errors to one percent. Consistency between repeated answers also improved. The authors frame the design as a template for safe AI prescribing support, pending real-world validation.
Catalog record
| CREATOR | Benjamin Bizer, Oscar A. Garcia Valencia, Flora Kincses, Jose Arriola-Montenegro, Charat Thongprayoon, Jing Miao, Wisit Cheungpasitporn |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 35 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |