Discovering medical resource utilization in total knee arthroplasty (TKA) using rule-based method.

Area of Science:

• Orthopedic Surgery
• Health Services Research
• Data Science

Background:

• Total knee arthroplasty (TKA) effectively treats advanced knee arthritis but faces challenges in analyzing complex medical data.
• High surgical volume in TKA is linked to reduced morbidity and better economic outcomes, yet data complexity hinders identifying key resource utilization factors.

Purpose of the Study:

• To develop a robust method for identifying critical features affecting healthcare resource utilization in TKA.
• To enhance the accuracy of medical data analysis for improved healthcare management and policy development.

Main Methods:

• Expert-guided data screening to filter relevant patient records.
• A two-stage feature selection process involving Analysis of Variance (ANOVA) and an Integrated Feature Selection Approach (IFSA).
• Application of Rough Set Theory (RST) for data discretization and rule generation, with accuracy comparison against other methods.

Main Results:

• The proposed integrated model successfully identified important features from complex medical datasets.
• Validation across three datasets demonstrated the model's accuracy in feature selection and rule generation.
• The approach effectively addresses the challenge of irrelevant features in medical data analysis.

Conclusions:

• The developed feature selection and rule generation model offers a valuable tool for analyzing TKA data.
• Findings can inform the National Health Insurance Bureau (NHI) in establishing TKA treatment standards.
• This research provides a framework for more accurate healthcare resource utilization analysis in TKA.