Sufficient data quality in large Electronic Health Record (EHR) databases is important to producing valid analytical findings. Existing data quality assessment tools currently lack comprehensive rules for assessing laboratory results recorded in EHR databases. Our study and resulting database (called ThemisUnits) addressed this gap using real world data using multiple international EHR datasets. ThemisUnits database was developed within the Observational Data Science and Informatics (OHDSI) consortium as part of the Themis initiative which aims to arrive at stricter data model specifications that would promote higher semantic interoperability within the OHDSI Common Data Model (CDM; www.ohdsi.org/data-standardization). We obtained data from 13 OMOP datasets. The database contains 437 distinct measurements and 513 measurement-unit pairs. It can serve as a knowledge base for data quality assessment of laboratory data.
Learning Objective 1: Understand data quality assesment of laboratory data
Vojtech Huser (Presenter)
Doyeop Kim, AJOU
Ajit Londhe, J&J