Opportunities and challenges for biomarker discovery using electronic health record data

Area of Science:

• Biomedical Research
• Health Informatics
• Genomics and Precision Medicine

Background:

• Electronic health records (EHRs) are increasingly utilized for biomedical research, particularly for identifying biomarkers in complex conditions.
• The primary design of EHRs for clinical care, not research, presents inherent challenges for data extraction and analysis.
• Despite limitations, EHRs hold substantial promise for advancing disease etiology understanding and informing precision medicine.

Purpose of the Study:

• To review the current applications of EHR data in phenotyping and molecular biomarker identification.
• To critically analyze the challenges and limitations associated with using EHRs for research.
• To propose strategies for overcoming these obstacles and maximizing the potential of EHRs for biomarker discovery.

Main Methods:

• Literature review of studies employing EHR data for phenotyping and biomarker discovery.
• Analysis of methodologies used for extracting and interpreting EHR data for research purposes.
• Examination of existing challenges in EHR data quality, standardization, and accessibility.

Main Results:

• EHR data is valuable for identifying patient states and molecular biomarkers, especially in complex diseases.
• Significant challenges exist, including data heterogeneity, missing information, and the need for robust phenotyping algorithms.
• Various strategies, such as data standardization, advanced analytics, and collaborative efforts, are being developed to mitigate these issues.

Conclusions:

• Leveraging EHR data for biomarker discovery is crucial for advancing precision medicine and understanding disease.
• Addressing data quality, interoperability, and analytical challenges is essential for realizing the full potential of EHRs in research.
• Future research should focus on developing and implementing standardized methods for EHR data utilization in biomarker identification.