JMIR Publications

ABSTRACT

Background:

The huge amount of clinical, administrative and demographic data recorded and maintained by hospitals can be consistently aggregated into Health Data Warehouses (HDWs) with a uniform data model. In 2017, Rouen University Hospital (RUH) initiated the design of a Semantic Health Data Warehouse (SHDW) enabling both semantic description and retrieval of health information.

Objective:

Our objectives were: first, to present a proof of concept of this SHDW, based on the data of 250,000 patients from RUH and second, to assess its ability to assist health professionals to select patients in a clinical trials context.

Methods:

The SHDW relies on three distinct semantic layers: (a) a Terminology and Ontology (T&O) portal, (b) a Semantic Annotator and (c) a Semantic Search Engine and a Not Only SQL (NoSQL) layer to enhance data access performances. The system adopts an entity-centered vision which contrasts with the usually patient-centered vision adopted by existing systems such as Informatics for Integrating Biology and the Bedside (i2b2). This vision notably provides generic search capabilities able to express data requirements in terms of the whole set of interconnected conceptual entities that compose health information. We assessed the ability of the system to assist the search for 95 inclusion and exclusion criteria originating from five randomly chosen Clinical Trials from RUH.

Results:

The system succeeded in fully automating 39.19% of the criteria and was efficiently used as a pre-screening tool for 72.97% of them.

Conclusions:

The semantic aspect of the system combined with its generic entity-centered vision enables the processing of a large range of clinical questions. However, an important part of health information remains in Clinical Narratives and we are currently investigating novel approaches (deep learning) to enhance the semantic annotation of those unstructured data.