Breast Cancer Clinical Data Analysis

This repository contains exploratory data analysis and visualization of a breast cancer clinical dataset, focusing on staging patterns, treatment distributions, and data completeness.

Project Objectives

• Summarize baseline clinical characteristics
• Explore staging and treatment patterns
• Assess missing data structure and completeness
• Identify limitations affecting downstream analysis
• Provide recommendations for data improvement

Clinical Characteristics & Data Quality Overview

AJCC Stage Distribution

• Most patients were classified as Stage III, followed by Stage IV
• Very few cases were recorded as Stage I or II
• A substantial number of records were missing staging information

Interpretation:

• This distribution suggests a late-stage presentation pattern, commonly observed in resource-limited settings, with significant implications for treatment planning and outcomes analysis.

Chemotherapy Regimen Frequency

• AC/EC-based regimens were most commonly administered among documented cases
• Taxane-based therapies were the second most frequent
• The largest proportion of entries were missing regimen data
• Hormonal and “Other” treatments were rarely recorded

Implications:

• High missingness limits reliable treatment-effect analyses. Improving regimen documentation should be prioritized in future data collection.

Missing Data Overview

• Percentage of Missingness per Variable
• Missingness ranged from <5% to nearly 100%
• Several clinically relevant fields (e.g., imaging results, adjuvant therapy details) had >70% missingness
• Demographic variables were relatively complete

Missingness Distribution Across Observations • Overall missingness was approximately 49.5% • Missingness was not random, with blocks of consistently unreported variables

Next Steps

• Improve data-collection protocols to reduce missing information
• Emphasize complete documentation of clinical and treatment variables
• Avoid listwise deletion due to high data loss