Development of a time-oriented data warehouse based on a medical information event model

Area of Science:

• Medical Informatics
• Health Data Management
• Clinical Data Warehousing

Background:

• Medical information systems often involve complex data structures, hindering efficient data retrieval and secondary use.
• The timing of clinical decisions and treatments is critical, necessitating efficient access to patient information.
• Existing systems, like PC-ORDERING2000, utilized extensive database tables (approx. 600), leading to complexity.

Purpose of the Study:

• To design a new, simplified medical information event model.
• To develop a time-oriented data warehouse based on this novel model.
• To enhance the efficiency and flexibility of clinical data management and secondary use.

Main Methods:

• A unique and simple medical information event model was designed as the basic data unit.
• A time-oriented data warehouse was developed utilizing this event model.
• Clinical data was migrated from an old order entry system to the new system.

Main Results:

• The number of data structures was reduced from approximately 600 to a single, simple event model.
• A flexible system was created, enabling easy secondary use of patient clinical data.
• Data retrieval efficiency was significantly heightened, with response times shortened by a factor of 1:600.

Conclusions:

• The developed medical information event model and time-oriented data warehouse offer a simplified and flexible approach to managing clinical data.
• The system demonstrably improves data retrieval efficiency and response times.
• This approach facilitates easier secondary use of patient data, benefiting research and clinical decision-making.