Beyond Missingness: Systematizing Methods for Comprehensive Data Fitness Assessment in Clinical Research.

Area of Science:

• Health Informatics
• Clinical Data Science
• Research Methodology

Background:

• Secondary use of clinical data accelerates research but is hindered by data quality issues.
• Current data quality approaches are often network-specific or lack methods for ensuring fitness for analysis.
• Existing tools offer basic checks but do not fully capture study-specific data fitness requirements.

Purpose of the Study:

• To introduce a systematic model for study-specific data quality assessment (SSDQA).
• To guide the design and implementation of improved SSDQA processes.
• To enhance the consistency, completeness, and reproducibility of clinical data quality assessments.

Main Methods:

• Developed a model integrating theoretical data quality principles with practical clinical data considerations.
• Incorporated metadata for consistent annotation and reporting of quality assessment results.
• Proposed regularization of check application using a standard set of options.

Main Results:

• The SSDQA model provides a consistent framework for specifying data quality assessment checks.
• It builds upon current practices to offer more complete, sound, and reproducible assessments.
• Facilitates multidisciplinary collaboration to identify data quality issues.

Conclusions:

• The proposed SSDQA model addresses limitations in current data quality assessment practices.
• It promotes more robust and reliable clinical research through systematic data quality evaluation.
• Ultimately supports better study design and informed adoption of research findings.