Improving Diagnostic Test Ordering Accuracy Through EMR Optimization: Impact of an Integrated Order Panel for BCR::ABL1 Assays in a Tertiary Academic Center

This repository contains analysis code and figure outputs supporting a manuscript on improving BCR::ABL1 test ordering accuracy through a standardized EMR order panel / order set (clinical decision support–enabled ordering workflow). The goal of sharing this code is reproducibility: readers can inspect the full analytic approach and re-run the pipeline on their own appropriately structured dataset.

Repository structure

.
├── figures/
│   ├── Fig1.tiff
│   ├── Fig2.tiff
│   ├── Fig3.tiff
│   ├── Fig4_cyto_over_time.tiff
│   └── Fig5_panel_utilization.tiff
├── scripts/
│   ├── bcr_abl_analysis.ipynb
│   └── bcr_abl_hash_salt_deid.ipynb
└── .gitignore

What you need to run this

• Python 3.10+ (3.11+ recommended)
• JupyterLab or Jupyter Notebook

Because environments vary by institution, the most reliable way to reproduce results is to create a clean environment and install dependencies as needed while running the notebooks.
Additionally, data should be consistently formatted, considering variations in electronic medical records (EMR) and data across different institutions.

Citation

To be updated after publication.

Questions

For questions or issues, please open a GitHub Issue and the output of session_info.show().