Analyse de trajectoires de soins à partir de bases de données médico-administratives : apport d'un enrichissement par des connaissances biomédicales issues du Web des données

Abstract : Reusing healthcare administrative databases for public health research is relevant and opens new perspectives. In pharmacoepidemiology, it allows to study large scale diseases as well as care consumption for a population. Nevertheless, reusing these information systems that were initially designed for accounting purposes and whose interoperability is limited raises new challenges in terms of representation, integration, exploration and analysis. This thesis deals with the joint use of healthcare administrative databases and biomedical knowledge for the study of patient care trajectories. This includes both (1) exploration and identification through queries of relevant care pathways in voluminous flows, and (2) analysis of retained trajectories. Semantic Web technologies and biomedical ontologies from the Linked Data allowed to identify care trajectories containing a drug interaction or a potential contraindication between a prescribed drug and the patient’s state of health. In addition, we have developed the R queryMed package to enable public health researchers to carry out such studies by overcoming the difficulties of using Semantic Web technologies and ontologies. After identifying potentially interesting trajectories, knowledge from biomedical nomenclatures and ontologies has also enriched existing methods of analysing care trajectories to better take into account the complexity of data. This resulted notably in the integration of semantic similarities between medical concepts. Semantic Web technologies have also been used to explore obtained results.
Document type :
Theses
Complete list of metadatas

Cited literature [163 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-02137442
Contributor : Abes Star <>
Submitted on : Thursday, May 23, 2019 - 9:33:38 AM
Last modification on : Tuesday, January 14, 2020 - 3:52:39 AM

File

RIVAULTYann.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02137442, version 1

Citation

Yann Rivault. Analyse de trajectoires de soins à partir de bases de données médico-administratives : apport d'un enrichissement par des connaissances biomédicales issues du Web des données. Santé publique et épidémiologie. Université Rennes 1, 2019. Français. ⟨NNT : 2019REN1B003⟩. ⟨tel-02137442⟩

Share

Metrics

Record views

111

Files downloads

112