Graphical modeling and query language for hospitals

Area of Science:

• Health Informatics
• Information Systems in Healthcare
• Clinical Process Modeling

Background:

• Limited evidence exists for health information technologies (HIT) driving healthcare cost savings.
• Complexity in hospital business process ownership hinders HIT impact.
• Need for a method enabling direct ad-hoc data retrieval by clinicians from hospital databases.

Purpose of the Study:

• To develop a business model-based method for hospitals.
• To enable doctors to directly retrieve ad-hoc information from diverse hospital databases.
• To address the complexity of business process ownership in healthcare settings.

Main Methods:

• Developed MedMod, a domain-specific process modeling language, as a profile on UML Class diagrams.
• Created the Process Query Language (PQL), based on MedMod, for querying runtime hospital process data.
• Demonstrated MedMod semantics through informal explanations and examples.

Main Results:

• MedMod overcomes limitations in existing process modeling languages for clinical processes.
• PQL offers enhanced usability and efficiency through familiar process views and simple filtering.
• PQL provides dynamic query construction, reducing learning curves and error rates.
• PQL achieves O(n) time complexity for data retrieval regarding dataset size.

Conclusions:

• Future development includes user-friendly graphical editors for MedMod and PQL.
• Usability evaluation with physicians using real hospital data is planned.
• An efficient implementation of the PQL query language will be developed.