Code for: Partial Adoption of the 2018 Fracture-Related Infection Consensus: A Contextual Text Mining Analysis of Diagnostic Terminology and Reporting Practice
Description
This repository contains the code used for the analysis in: “Partial Adoption of the 2018 Fracture-Related Infection Consensus: A Contextual Text Mining Analysis of Diagnostic Terminology and Reporting Practice.” It is a context-aware LLM pipeline for counting fracture-related infection (FRI) terminology in full-text PDFs. It was designed to reduce misleading term matches where a keyword appears in text but is not actually being used to diagnose, classify, or describe infection. The development version of the code is available on GitHub: https://github.com/nickpiccaro/FRI-LLM-Based-Text-Mining. This archived version corresponds to the version used for manuscript submission.
Files
Steps to reproduce
Step-by-step instructions to reproduce the analysis, including data preparation, environment setup, and execution of scripts, are provided in the included README.md file.
Institutions
- University of Colorado Anschutz Medical CampusColorado, Aurora