Data mining applications in the context of casemix

Area of Science:

• Health economics
• Health informatics
• Data science

Background:

• Singapore implemented casemix-based funding in public hospitals in 1999.
• This funding model aims for equitable, rational healthcare financing and improved cost efficiency.
• Concerns exist regarding the "quicker and sicker" syndrome, potentially leading to premature patient discharges.

Purpose of the Study:

• To explore data mining applications within the casemix framework.
• To analyze hospital readmission data to identify patterns related to the "quicker and sicker" syndrome.
• To demonstrate how data mining can enhance understanding of complex healthcare data.

Main Methods:

• Utilizing readmission data as a case study for data mining.
• Applying data mining techniques to detect systematic patterns in patient readmissions.
• Comparing data mining's capabilities with traditional statistical methods for complex data analysis.

Main Results:

• Data mining can effectively analyze complex, non-linear relationships in large datasets.
• It offers a methodology to uncover hidden patterns within healthcare data, such as readmission trends.
• Analysis of readmission data can provide insights into potential issues arising from casemix-based funding.

Conclusions:

• Data mining is a valuable tool for complementing traditional statistical analysis in healthcare.
• It can help identify and understand potential negative consequences of funding models like casemix.
• Further application of data mining can support quality improvement and cost-efficiency initiatives in hospitals.