ckakalou/Healie_SDoHScreener

SDoH Screener Demo (Python + Streamlit)

This is a quick demo app that renders the Social Determinants of Health (SDoH) screener from a JSON config, validates responses with a JSON Schema, and computes key flags and scores (PHQ-2, HITS, Physical Activity).

1) Set up (macOS)

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

2) Run the app

streamlit run app.py

The browser will open automatically (usually at http://localhost:8501).

3) What this demo does

• —Renders form sections from sdoh_screener.json
• —Applies basic skip logic (employment sub-questions, unmet-needs follow-ups)
• —Validates the submitted payload with sdoh_screener_response_schema.json
• —Computes and displays:
• —PHQ-2 total and screen result
• —HITS total and screen result (Spanish cutoff toggle)
• —Physical-activity weekly minutes and need flag (based on age)
• —Other flags: housing instability, transportation need, food insecurity, medical cost barrier
• —Lets you download the response payload and evaluation results as JSON

4) Files

• —app.py – Streamlit app
• —sdoh_screener.json – Form configuration
• —sdoh_screener_response_schema.json – JSON Schema for responses
• —requirements.txt – Dependencies

5) Notes

• —This is a demo; production apps should persist responses and add authentication/consent.
• —You can customize the UI easily by editing app.py.