COVID-19 TestNorm: A tool to normalize COVID-19 testing names to LOINC codes

Area of Science:

• Biomedical Informatics
• Public Health Research
• Health Data Management

Background:

• Electronic Health Records (EHRs) are vital for COVID-19 research.
• Data normalization across institutions presents a significant challenge for EHR data.
• Standardized terminology for COVID-19 diagnostic tests was published after clinical use.

Purpose of the Study:

• To develop an automated tool for normalizing COVID-19 diagnostic test names.
• To map local test names to standard LOINC codes for research consistency.
• To facilitate the secondary use of EHR data for COVID-19 research.

Main Methods:

• Developed a rule-based tool named COVID-19 TestNorm.
• Utilized 568 COVID-19 test names from 8 healthcare systems for development and evaluation.
• Assessed accuracy on an independent test set.

Main Results:

• COVID-19 TestNorm achieved 97.4% accuracy in normalizing test names.
• The tool successfully mapped local test names to standard LOINC codes.
• Demonstrated high performance in a multi-institutional dataset.

Conclusions:

• COVID-19 TestNorm effectively automates COVID-19 diagnostic test normalization.
• The tool supports improved data quality and consistency for COVID-19 research.
• An open-source package and web application are available to support EHR data utilization.