DataGauge: A Practical Process for Systematically Designing and Implementing Quality Assessments of Repurposed...

Area of Science:

• Health Informatics
• Data Science
• Clinical Data Management

Background:

• Repurposing clinical data for secondary uses presents significant data quality challenges.
• Existing methods lack a systematic approach for implementing data quality assessments in secondary data use.
• Ensuring data validity is crucial regardless of the analytical methods employed.

Purpose of the Study:

• To present DataGauge, a systematic process for designing and implementing data quality assessments for repurposed clinical data.
• To provide a framework for evaluating data quality tailored to specific secondary data use cases.
• To bridge the gap between general data quality theory and practical electronic health record data quality guidelines.

Main Methods:

• DataGauge involves five key steps: defining information needs, developing a Data Needs Model (DNM), creating goal-specific DQ assessment requirements using the DNM and DQ theory, extracting specified data, and evaluating against requirements.
• The process integrates general data quality theory with data quality assessment methods.
• A theory-based guidance framework links the DNM to DQ testing methods for each DQ dimension.

Main Results:

• DataGauge provides a systematic process for data quality assessment of repurposed clinical data.
• It facilitates the integration and practical application of existing Electronic Health Record-specific data quality assessment guidelines.
• The developed framework aids in designing data quality assessments by connecting the DNM to DQ testing methods.

Conclusions:

• DataGauge establishes a systematic process for data quality assessment, adaptable to various clinical datasets and secondary uses.
• It lays the foundation for future research in data quality theory integration, requirements portability, and assessment tool development.
• This systematic approach enhances the reliability and validity of results derived from repurposed clinical data.