This project explores pulmonary embolism (PE) patient data using a blend of statistical and geospatial analytics. The goal is to uncover patterns related to mortality, readmission rates, gender, smoking status, and rural versus non-rural healthcare disparities.
The google collab notebook applies advanced visualization techniques and geospatial mapping to understand how various clinical outcomes differ across patient groups. This project not only demonstrates technical proficiency in Python and data visualization but also reflects a strong analytical mindset focused on real-world healthcare impact.
🔍 Key Analyses and Visualizations:
Rural vs. Non-Rural Comparisons:
-
ICU mortality rate differences
-
30-day readmission patterns
-
Gender and smoking prevalence comparison
Categorical Outcome Visualization:
-
Used Seaborn bar plots and count plots to visualize subgroup comparisons.
-
Cleaned and standardized data to enable accurate comparisons across demographics.
Geospatial Analysis:
-
Plotted patient data using GeoPandas and Contextily for base maps.
-
Mapped distribution of PE cases to highlight regional healthcare patterns.
-
Analyzed rurality using RUCA codes for location-based disparities.
Pairwise Feature Comparisons (Suggested for future integration):
Pair plots or correlation heatmaps can add further insight into clinical feature relationships.
💡 What Makes This Project Unique
-
Combines traditional EDA with geospatial healthcare mapping, offering a broader perspective.
-
Highlights disparities in patient outcomes based on location—a critical lens for public health.
-
Designed for clarity, reproducibility, and interpretability, making it ideal for clinical teams or public health stakeholders.
🚀 Potential Applications & Extensions:
-
Can be expanded to include predictive modeling for risk stratification.
-
Ideal for integration into hospital dashboards or decision-support tools.
-
Could be adapted for other conditions to assess urban-rural care inequities.