Complex analyses on clinical information systems using restricted natural language querying to resolve time-event...

Area of Science:

• Clinical Informatics
• Biomedical Data Analysis
• Health Data Science

Background:

• Complex research questions in clinical studies often involve intricate time-event dependencies.
• Existing data analysis methods can be time-consuming and inefficient for such complex queries.

Purpose of the Study:

• To present a generic framework for complex clinical data analysis using cascaded queries.
• To extend the Clinical Data Analytics Language (CliniDAL) for enhanced research capabilities.

Main Methods:

• A cascaded query model with up to five levels was developed to resolve time-event dependencies.
• The model simulates scientific experiments by defining subjects, time spans, control groups, variables, and outputs.
• Flexibility is offered by merging or nesting queries based on research complexity.

Main Results:

• The CliniDAL solution efficiently answers complex queries with time-event dependencies within hours.
• Manual analysis, which would typically take many days, is significantly accelerated.

Conclusions:

• The proposed CliniDAL framework offers high usability and efficiency for complex clinical data analysis.
• It provides a powerful tool for clinicians to conduct sophisticated research.