Data structuring of electronic health records: a systematic review

The medical field has experienced a series of transformations with the adoption of new technologies. One of the aspects that experienced significant changes is how a patient’s information is stored. Electronic health records have brought a series of advantages but still present many issues. One of them is the degree of structuring for contained information. More structuring brings a greater richness of information. On the other hand, it contains more challenging and complex content when most of the information is stored in free text (unstructured information). In this sense, many studies focused on structuring the information contained in free text have emerged. This work aims to review the studies focused on the structuring of unstructured health record information, seeking to answer key questions to propose new studies in the field on topics such as the form in which information is structured, the main techniques used, and how data acquisition for development and evaluation is done. To answer these questions, a wide systematic review of the field was conducted since the emergence of BERT networks. In addition to answering those questions, this systematic review identified the main challenges, such as difficulty in data acquisition, problems with natural language processing, and the specific challenges of the studies that process non-English languages, finalizing with a general view of the state of the art in the field and its future opportunities.