Data mining application in customer relationship management for hospital inpatients

Area of Science:

• Healthcare Management
• Data Mining Applications
• Customer Relationship Management

Background:

• Hospitals increasingly seek to identify and retain loyal patients.
• Understanding patient loyalty is crucial for effective customer relationship management (CRM).
• Data mining offers advanced tools for analyzing complex patient data.

Purpose of the Study:

• To discover and model the medical service usage patterns of loyal hospital inpatients.
• To propose a data mining application for CRM in hospital settings.
• To identify key factors influencing patient loyalty.

Main Methods:

• Applied the Recency, Frequency, Monetary (RFM) model to 14,072 discharged patients.
• Utilized cluster analysis for patient segmentation.
• Employed decision tree analysis to model loyal customer usage patterns.

Main Results:

• Identified a loyal customer segment characterized by high medical service utilization and expenses.
• Key predictors of loyalty include length of stay, treatment certainty, surgery, number of treatments, room type, and discharge department.
• A patient in internal medicine, without surgery, staying over 13.5 days had a 70% probability of being a loyal customer.

Conclusions:

• Data mining effectively identifies loyal patients and models their behavior.
• Combining RFM model with data mining provides actionable insights for hospital CRM.
• Integrating segmentation, targeting, positioning (STP) with data mining enhances CRM strategies.