This code accompanies the following paper:

Delos Reyes, R., Capurro, D., & Geard, N. (2024). Modelling patient trajectories in emergency department simulations using retrospective patient cohorts. Computers in Biology and Medicine. https://doi.org/10.1016/j.compbiomed.2024.109147

---

Setting up the environment

1. Download anaconda
2. Run conda env create --name edabm -f edabm.yaml
3. Run conda activate edabm

Getting the required data

1. Download the following datasets (they require credentialed access which can be requested at the provided websites)
   • MIMIC-IV v2.2
   • MIMIC-IV-ED v2.2
2. Store the unzipped datasets inside the data folder
   • data/ed
   • data/hosp
   • data/icu

Preprocessing the data

Open the following Jupyter notebooks in the following order and run all cells:

1. generate_patient_data.ipynb  # To exclude patient records with invalid values
2. generate_event_logs.ipynb  # To convert the records to event logs
3. generate_model_parameters.ipynb  # To generate the parameters needed to run the ED simulation model
4. generate_mci_frequency.ipynb  # To generate modified parameters for experiment 3 (mass casualty incident scenarios)

Running the experiments

1. Run chmod u+x ./run.sh
2. Run ./run.sh

Generating the figures

Open the following Jupyter notebooks and run all cells:

1. plot_results_exp1.ipynb
2. plot_results_exp2.ipynb
3. plot_results_exp3.ipynb
4. plot_results_supplementary.ipynb