Streamlining electronic medical record data extraction and validation in digital hospitals: A systematic review to...

Extracting and curating data from large clinical information systems is challenging, and the optimal methodology is often unclear. This review was to systematically investigate and appraise the research literature to assess existing methods used by healthcare organizations to extract data from the electronic medical record (EMR). The Observational Medical Outcomes Partnership (OMOP) common data model (CDM) is used as a comparator for the various methods of data extraction. Our specific research question was: what lessons can be learned from healthcare organizations' experiences with data extraction from EMRs using OMOP CDM as a standardized use case?

We searched PubMed, Web of Science, Embase, the snowballing citation, and potentially relevant gray literature via Google Scholar for EMR data extraction and validation with OMOP CDM as the standardized use case for studies published between June 2017 and December 2022. A total of 316 candidate articles were examined, but only nine met the inclusion criteria. Two authors screened and assessed articles based on predetermined criteria to examine prevalent techniques and challenges through thematic synthesis and data analysis.

Among all the included articles, the most frequently discussed challenges in EMR data extraction and validation are the lack of a standardized process, data structure, and skilled personnel. Five of nine studies scored above 70% in the article quality assessment process. Three studies used Observational Health Data Sciences and Informatics's suite, and two utilized Staged Optimization of Curation, Regularization, and Annotation of clinical text alongside the semantic transformation framework.

The study revealed the importance of standardizing a uniform approach, consistent processes, and tools for EMR data extraction and validation. The identified methods and techniques could streamline the EMR data extraction processes. Our future work will empirically evaluate these methods in collaboration with real-world healthcare organizations.