Facilitating harmonized data quality assessments. A data quality framework for observational health research data...

Area of Science:

• Health Research
• Data Science
• Biostatistics

Background:

• Lack of standardized data quality handling and reporting in health research.
• Need for harmonized data quality assessments in observational studies.

Purpose of the Study:

• Introduce a comprehensive data quality framework for observational health research.
• Provide supporting software implementations for data quality assessments.
• Facilitate harmonized and reproducible data quality evaluations.

Main Methods:

• Developed a data quality framework based on existing evaluations and literature reviews.
• Implemented data quality indicator computations using R functions.
• Illustrated the framework with data from the Study of Health in Pomerania (SHIP).

Main Results:

• The framework includes 34 data quality indicators covering integrity, completeness, consistency, and accuracy.
• R functions generate data quality metrics and reports.
• A dedicated website provides guidance on the framework and tools.

Conclusions:

• This is the first data quality framework for observational health research linking a formal concept with R implementations.
• The framework and tools enhance transparency and reproducibility in research.
• Applicable for ongoing study monitoring, pre-analysis, and beyond research contexts.