This repository is no longer maintained. This project will no longer receive updates, bug fixes, or new features. It is provided "as-is" for reference only.
For the latest version and ongoing updates, please visit the new repository at MII-CALM-QE-repo.
This repository is developed to create "Study Data" for CALM-QE Project.
The purpose of this set of scripts is to determine a cohort dataset specifically for patients whose primary diagnoses are associated with Asthma or Chronic Obstructive Pulmonary Disease (COPD). The scripts identify and extract locally relevant patient data from FHIR server resources available. This obtained cohort dataset determines a primary set to continue a further analysis to determine and quantify patients with secondary conditions, observations, and specific medication records related.
To run this project is necessary to cover the following requirements:
- Connection to a FHIR Server
- Python 3.12
- Docker (optional)
The installation can be orchestrated directly by copying this repository locally and following the Set up instructions, or run it directly with Docker.
1. Install requirements
Install all the required packages:
pip install -r requirements.txt
2. Configure FHIR Server Connection
Before running the scripts, ensure that FHIR server configurations are added in data_extraction/Constants.py file.
You should update the following fields:
- USER_NAME = "Your User Name"
- USER_PASSWORD = "Your Password"
- SERVER_NAME = "Your FHIR Server Base URL"
This section identifies patients with "Asthma" or "COPD" as main diagnosis.
Initially CohortPatientsExecute.py, reads from the input_files folder asthma_copd_codes.json automatically. This json file includes all the ICD-10 codes available related to "Asthma" and "COPD". Modifications to this code list are possible based on unique needs when required.
The usage of this file is determined in Constants.py.
The script outputs all the patients' IDs and corresponding diagnoses in patients_main_diagnosed_asthma_copd.json.
python .\data_extraction\CohortPatientsExecute.py
After the first part is complete, the analysis continues with the fetching, extraction, and count of secondary Conditions, Observations and Medication data available after running ExtractResourcesForCohortExecute.py.
The script generates separate json files for each resource type (e.g., Conditions, Observations, Medications) per patient.
After compiling the script, a metadata.json is generated as part of the outcomes to provide a general and quantitative overview of the items generated.
python .\data_extraction\ExtractResourcesForCohortExecute.py
Instead of setting up and running the scripts manually, you can run the scripts in a container environment. First, define the necessary credentials to connect to a FHIR Server in dockerfile as follows:
- USER_NAME = "Your User Name"
- USER_PASSWORD = "Your Password"
- SERVER_NAME = "Your Fhir Server Base URL"
After making sure the docker is installed, you can run the following commands.
docker build -t fhir-cohort-resources-extraction .
docker run --name calm-qe fhir-cohort-resources-extraction