Constructing a Pre-Emptive System Based on a Multidimentional Matrix and Autocompletion to Improve Diagnostic Coding...

Area of Science:

• Health Services Research
• Medical Informatics
• Health Economics

Background:

• Short-stay hospitalizations (Medicine, Surgery, Obstetrics) are funded via service charges (T2A).
• Accurate clinical coding, reflecting patient severity, is crucial for appropriate hospital reimbursement.
• Current coding practices may lack optimization, potentially impacting funding accuracy.

Purpose of the Study:

• To propose an autocompletion process and multidimensional matrix to improve clinical information expression.
• To optimize the accuracy of clinical coding for hospital funding.
• To assist physicians, even those unfamiliar with encoding rules, in refining diagnosis codes.

Main Methods:

• Development of an autocompletion process for clinical information.
• Utilization of a multidimensional matrix for coding optimization.
• Integration with optimized knowledge bases of diagnosis codes for semantic refinement.

Main Results:

• Physicians can start with a general concept, refining it through semantic proximity.
• The system aids in associating rough concepts with specific diagnosis codes.
• Improved information expression leads to more optimized clinical coding.

Conclusions:

• The proposed system enhances the accuracy of clinical coding for hospital funding.
• It empowers physicians to optimize coding without deep knowledge of encoding rules.
• This approach ensures better alignment between patient severity and hospital reimbursement.