Electronic Health Record Data Quality and Performance Assessments: Scoping Review.

Area of Science:

• Medical Informatics
• Health Data Science
• Clinical Research Methodology

Background:

• Electronic Health Records (EHRs) offer vast potential for medical research via accessible databases.
• Realizing this potential is hindered by challenges in EHR data quality (DQ) and performance assessment.

Purpose of the Study:

• To review and standardize best practices for EHR data quality and performance assessments.
• To establish a replicable standard for researchers in the field.

Main Methods:

• Systematic PubMed search for original research articles on EHR DQ and performance.
• Data collected until May 7, 2023.

Main Results:

• 26 articles were reviewed, with common limitations including reporting inconsistencies, poor replicability, and limited generalizability.
• Key DQ indicators were completeness, conformance, and plausibility; correctness/accuracy was primary for performance.
• Artificial intelligence (AI) techniques show promise in improving EHR dataset quality and performance.

Conclusions:

• Standardization and incentivization of EHR DQ and performance assessments are necessary.
• AI-based techniques can enhance EHR data quality and performance, unlocking their full potential for medical research and practice.