A Data Quality Assessment Guideline for Electronic Health Record Data Reuse

Area of Science:

• Health Informatics
• Clinical Research Methodology
• Data Science

Background:

• Electronic Health Records (EHR) are crucial for clinical research, but ensuring data quality is challenging.
• Existing methods for EHR data quality assessment are often inconsistent or insufficient.
• A dynamic, evidence-based guideline is needed to standardize EHR data quality assessment and reporting.

Purpose of the Study:

• To formulate, develop, and conduct an initial expert review of the 3x3 Data Quality Assessment (DQA) guideline.
• To create a dynamic, evidence-based framework for assessing and reporting EHR data quality in clinical research.
• To establish best practices for EHR data quality assessment.

Main Methods:

• The 3x3 DQA guideline was developed using a triangulation approach, combining literature review, quantitative EHR data completeness analysis, and qualitative interviews with researchers.
• A panel of EHR data quality experts reviewed the developed guideline.
• The framework operationalizes data quality constructs (complete, correct, current) across EHR data dimensions (patients, variables, time).

Main Results:

• The 3x3 DQA guideline framework integrates data quality constructs with EHR data dimensions, resulting in nine operationalized constructs.
• Each operationalized construct corresponds to a specific methodological recommendation for EHR data quality assessment.
• Initial expert review indicated a positive reception of the framework, with suggestions for improvement.

Conclusions:

• The initial version of 3x3 DQA provides a foundation for guideline-based best practices in EHR data quality assessment and reporting.
• Future development will focus on enhancing clarity regarding the guideline's application, improving the feasibility of recommendations, and refining the process for selecting relevant constructs and recommendations for specific datasets and studies.