Patient Healthcare Analytics Dashboard

Project Overview

Developed an interactive Power BI dashboard at Georgia State University that analyzed 25GB of healthcare datasets to identify improvement opportunities in patient care and emergency department operations. The project involved comprehensive exploratory data analysis using Python to uncover trends among patients and their respective diseases.

The dashboard visualized healthcare trends across multiple demographic dimensions including city, area, age, gender, race, and occupation, resulting in a 17% improvement in operational efficiency for ER departments.

Key Features

• Multi-Dimensional Analysis: Analyzed healthcare trends by city, area, age, gender, race, and occupation
• Interactive Visualizations: Created dynamic Power BI dashboards for stakeholder exploration
• Predictive Insights: Identified patterns that enabled proactive healthcare resource allocation
• Real-Time Updates: Dashboard updated with latest patient data for current insights
• KPI Tracking: Monitored key performance indicators for emergency department efficiency

Technologies Used

• Python: Data processing and exploratory data analysis
• Pandas: Data manipulation and cleaning
• Matplotlib: Statistical visualizations
• Scikit-learn: Machine learning models for pattern identification
• Power BI: Interactive dashboard development
• Google Colab: Collaborative data analysis environment

Research Methodology

The project followed a systematic approach:

1. Data Collection: Aggregated 25GB of healthcare records from various sources
2. Data Cleaning: Processed and cleaned raw data for analysis
3. Exploratory Analysis: Identified patterns and trends using statistical methods
4. Visualization: Created interactive dashboards for stakeholder communication
5. Validation: Tested findings with clinical stakeholders

Challenges & Solutions

• Data Volume: Handled large datasets through efficient data processing techniques and sampling strategies
• Data Privacy: Ensured HIPAA compliance through proper data anonymization
• Data Quality: Addressed missing values and inconsistencies in healthcare records
• Stakeholder Communication: Translated complex analytical findings into actionable insights for non-technical users

Business Impact

The healthcare analytics dashboard delivered significant value:

• 17% improvement in emergency department operational efficiency
• Better resource allocation based on patient flow predictions
• Improved patient care through identification of high-risk populations
• Data-driven decision making for hospital administrators
• Foundation for predictive healthcare analytics initiatives

My Role

As Graduate Research Assistant, I conducted all phases of the data analysis project including data collection, cleaning, exploratory analysis, and dashboard development. I worked directly with clinical stakeholders to understand their needs and presented findings to university administration.