A Python-based exploratory data analysis (EDA) project for MedLab Diagnostics in Kumasi, Ghana — uncovering patient behavior, test utilization trends, and operational inefficiencies to drive data-backed strategic decisions.
# MedLab Diagnostics - EDA & Strategic Insight Project
## Project Overview
**MedLab Diagnostics**, a biomedical facility in Kumasi, Ghana, needed actionable insights to improve operational efficiency and patient retention. Using Python-based exploratory data analysis (EDA), this project uncovered usage trends, patient behavior, and time-based bottlenecks that were slowing growth.
> The goal? Transform raw healthcare data into crystal-clear insights that spark better decisions.
---
## Key Objectives
- Identify **high-performing** and **underutilized** diagnostic tests.
- Compare **new vs returning patient trends**.
- Pinpoint **peak vs off-peak congestion** to optimize staffing.
- Recommend **revenue-boosting strategies** using data.
---
## Tools & Tech Stack
- **Python** – pandas, matplotlib, seaborn, plotly
- **Jupyter Notebook** – end-to-end analysis and visual storytelling
- **Excel** – initial data source
---
## Deliverables
- Cleaned dataset and reproducible Python scripts
- EDA visualizations highlighting trends and inefficiencies
- Business Case Presentation (PDF with recommendations)
---
## Key Findings
- Underutilized tests present revenue growth opportunities.
- Returning patient rates were **lower than expected**—signaling a retention gap.
- Peak hours revealed operational congestion needing smarter scheduling.
---
## Recommendations
1. Introduce a **digital appointment scheduler** to flatten peak congestion.
2. Launch **patient loyalty programs** to improve retention.
3. Bundle underutilized tests into promotional packages.
4. Leverage insights to align **staffing with traffic patterns**.
---
## About Me
**Frank Nketiah** – Data Analyst | Evidence-Based Strategist | Future Agribusiness Leader
Building bridges between data and decisions in healthcare, public planning, and agribusiness.
Connect with me on LinkedIn
Email: franknketiah@gmail.com
---