Cancer patient flows discovery in DRG databases

Area of Science:

• Health Services Research
• Data Science in Healthcare
• Oncology Network Management

Background:

• Cancer care in France is transitioning towards regional networks to optimize expertise, services, and resource distribution.
• Existing health information systems and data-mining tools offer potential for understanding patient flow dynamics.

Purpose of the Study:

• To analyze cancer patient flows within a French region using a national healthcare administrative database.
• To demonstrate the utility of Formal Concept Analysis (FCA) for visualizing and understanding complex patient pathways.

Main Methods:

• Utilized one year of data from the French Diagnosis Related Groups (DRGs) based administrative system.
• Applied Formal Concept Analysis (FCA), an unsupervised conceptual clustering technique.
• Constructed Iceberg Lattices to represent and visualize cancer patient flows in the Lorraine region.

Main Results:

• Successfully generated visual representations of cancer patient flows using FCA.
• The Iceberg Lattices provided an understandable overview of patient movement and service utilization patterns.
• Identified key insights into the distribution and coordination of cancer care services within the region.

Conclusions:

• Formal Concept Analysis is an effective method for describing and visualizing cancer patient flows in regional healthcare networks.
• This approach can support improved knowledge of patient journeys, aiding in the coordination of expertise, services, and resource allocation.
• The findings contribute to the ongoing evolution of cancer care delivery through data-driven network design.