Real-World Data Analytics Demo (SQL + Python, OMOP-CDM)

This repository demonstrates how to analyze synthetic electronic medical records (EMR) using the OMOP Common Data Model (CDM) framework.

---

Key Features

• SQL for cohort building
  • Example queries for patient selection, condition prevalence, and comorbidity co-occurrence.
• Python for predictive modeling
  • Logistic regression applied to respiratory outcomes.
• OMOP-CDM structure
  • person table (demographics: age, sex, race).
  • condition_occurrence table (SNOMED-coded conditions).
• Clinical terminologies
  • Uses OMOP condition_concept_ids mapped to SNOMED CT, compatible with ICD.
• Visualization
  • Co-occurrence heatmaps, cohort characterization plots, feature importance analysis.

---

Relevance to RWE

This demo showcases the core workflow used in RWE analytics:

• Cohort definition
• Comorbidity analysis
• Predictive modeling
• Visualization & interpretation

Even though data are synthetic, the methods are directly transferable to EMR and claims data.

---

Quick start

git clone https://github.com/ivkarla/rwd-sql-python-demo
cd rwd-sql-python-demo
conda create -n rwd-demo python=3.10
conda activate rwd-demo
pip install -e .