COVID-19 TestNorm

COVID-19 TestNorm normalizes local COVID-19 diagnostic test names in electronic health records (EHRs) by mapping them to Logical Observation Identifiers Names and Codes (LOINC) to enable standardized data integration and interoperability.

Key Features:

• Rule-Based Normalization: Uses a rule-based approach to automatically map diverse local COVID-19 test nomenclature in EHRs to standardized LOINC codes.
• Evaluated Accuracy: Demonstrated 97.4% accuracy on an independent test set using 568 test names collected from eight different healthcare systems.

Scientific Applications:

• EHR Data Harmonization: Standardizes diagnostic test records across institutions to support pooled analyses of clinical data.
• Epidemiological Studies: Enables consistent identification and comparison of COVID-19 test results for large-scale population research.
• Clinical Trials: Facilitates reliable aggregation of diagnostic test data for multicenter trial analyses.
• Public Health Surveillance: Supports interoperable reporting of COVID-19 testing information across health systems.

Methodology:

Applies a rule-based mapping system to convert local COVID-19 diagnostic test names to LOINC and was evaluated on 568 test names from eight healthcare systems, achieving 97.4% accuracy on an independent test set.