Big Data: a Case Study of its Use in Dental Public Health

Abstract

Today, in a hyper connected world nearly everyone leaves digital tracks of their activities, life style and behaviours. The intensive development of the Electronic Health Records (EHR) and the wide use of computer systems in health care systems are generating huge amounts of health data. The challenge is the re-use of these data for research purposes. The exploitation of health big data will allow better prevention based on knowledge of patients behaviour, implementation of more efficient diagnostic tools and improvement of care. This should lead to more efficient use of health finance and enhanced access to healthcare. The exploitation of big data is a major issue for public health and dental public health.

Aim: The aim of this abstract is to describe the integration of dental EHR data in a health data warehouse in order to make these data available for research.

Methods: In France, the Health Big Data research team has developed and implemented a biomedical data warehouse (EHOP) - which contains the EHR data of 1.5 million patients seen at Rennes University Hospital The data contained in these databases are semantically poor and can only be efficiently exploited with the use of additional knowledge systems which allow for the description and annotation of the data. These knowledge systems include classifications describing diagnoses (such as ICD-10, SNOMED-CT, DDS: Dental Diagnostic System), interventions and procedures (such as Common Classification of Medical procedures (CCAM) in France or International Classification of Health Interventions (ICHI) at the international level) and drugs knowledge bases (such as Vidal, Thériaque). Dental care data have been integrated in the EHOP warehouse. The Dental Care Centre at Rennes University Hospital has been generating electronic care data since 2008. These structured and unstructured data provide a significant amount of information about dental diagnoses and care. The initial phase of the project was the integration of ICD-10 and SNODDS for diagnoses, CCAM and NGAP for procedures and Theriaque for drugs) and when needed (for all non coded data), the mapping of dental EHR data with these tools allows it to be coded.

Results: The EHR data of the University of Rennes Dental Care Centre, generated since 2008, were integrated into the University Hospital data warehouse (EHOP) during 2016 and thus have been made available to the end-used for research purposes. Research can now be performed on both medical and dental care data. Patients or hospital stays can be retrieved based on structured data, which include diagnoses or treatment provided. Patients’ full medical and dental records have also been made available through full text request tools. It is expected that numerous retrospective studies based on these data will now take place. For example, analysis of the dental or/and medical data will provide information on dental treatments (survival analysis of endodontic treatments, fixed prosthesis versus implants, etc), on possible links between oral health and general health and will allow health economic studies and comparison of actual practice with guidelines.

Conclusion: The availability of big data such as those collected by the University Hospital of Rennes provides exciting new possibilities for research locally, nationally and internationally.