AIXIAL Review #4
A partnership to match real-life data to SNDS databases
Since 2019, France has created a health data platform, the Health Data Hub (HDH), whose objective is to promote research by sharing data compiled from various sources . These data include those from the French databases of the national health data system (SNDS).
In pharmacoepidemiology, claim databases, like those of the SNDS, are important, if not essential, data sources. With the expansion of their scope to all data reimbursed by health insurance, SNDS databases are exceptionally rich and exhaustive. For a long time not very accessible and underexploited, they have gradually opened up, under conditions, to research pursuing a purpose of public interest. They are used in particular by the health authorities and the “Assurance Maladie” itself. The one-stop-shop that constitutes the HDH, the simplification of procedures and the provision of data, should also facilitate its use by epidemiologists and manufacturers.
The place occupied by the bases of the SNDS is now such that, with regard to real-life studies on health products (drugs, medical devices), the first question that comes to mind when attempting to meet health authority expectations has become in France: should recourse to the SNDS databases provide an answer? It is only by default that more restrictive, less exhaustive and more expensive RWE studies have to be undertaken.
Originally designed to provide public authorities with a tool for managing health expenditure, the SNDS databases are however not very medicalized and therefore limited. If the data concerning the consumption of healthcare reimbursed by the community are exhaustive, precise and irreplaceable, the identification of pathologies, their stage, lines of treatment, their severity, etc., is much more problematic. In particular, the pathologies are not always directly identifiable. While there are sometimes tracer health products that can be used as proxies, for example, patients who have been reimbursed for anti-diabetic agents will be considered diabetics. In addition, the development of substitution algorithms for medical criteria partially compensates for this lack of medical data. These can, as in the case of the PROGRESS study in which we are participating , give rise to scientific publications because they are intended to be widely shared with the research community.
Despite the interest of database analysis, it often remains essential to collect real world data, (clinical, paraclinical and quality of life data) from both caregivers and patients, within the appropriate framework of research questions, using rigourous epidemiological methodology rigorously, in particular cohorts, which constitutes the core business of AIXIAL.
Sometimes, however, none of these methods can fully answer the questions asked and we can see the benefit of mutually enriching the data provided by the SNDS databases and those of cohorts by matching them. This is now being facilitated by HDH.
With its RWE department, AIXIAL is already developing an international offer intended to carry out non interventional studies (including on database) for public interest purposes, and is working with researchers experienced in health insurance databases, CLINITYX, for its part, has developed solid experience in matching RWE data with that of the SNDS, in particular with the French Society of Cardiology, and offers its services to contribute to such matching in association with the AIXIAL teams.
In order to carry out the RWE studies best able to meet the needs of pharma and medical device industries and those of health authorities, AIXIAL and CLINITYX have henceforth joined forces and are opening a consultation, with the help of methodologists likely to guide these industries to reach and carry out, together, useful studies matching non interventional studies, realized by AIXIAL, with the SNDS databases, thanks to the contribution of CLINITYX.
— Written by Jacques Massol, MD PhD, prof of Therapeutics, AIXIAL consultant
 https://solidarites-sante.gouv.fr/IMG/pdf/181012_-_rapport_health_data_hub.pdf (last visited, 2020, 23-11)
 The databases including those of SNIIRAM / EGB for ambulatory care, PMSI for hospital care (on 4 fields: MCO, SSR, HAD and psychiatry) and CépiDc for vital status and causes of death. They have the advantage of being chained
 Access after validation of the request involving, according to requests, the Ethical and Scientific Committee for Research, Studies and Evaluations in the field of Health (CESREES) and the National Commission for Informatics and Freedoms (CNIL). The processing of SNDS data is governed by the European regulation on the protection of personal data (RGPD and subject to the authorization of the CNIL
 A Scientific Interest Group (GIS), associating the CNAM and the ANSM: Epi-PHARE was created in 2018 to meet the growing demand for studies based on the use of complex and very large-scale data from the System. national health data (SNDS), in a reactive manner. These studies make it possible in particular to identify more early and more precisely the risks associated with health products and to inform public authorities’ decisions in terms of health safety
 M. Nasser, S. Larrieu, S. Si-Mohamed, K. Ahmad, L. Boussel, M. Brevet, D. Revel, F. Thivolet-Bejui, J. Traclet, M. Lederlin, S. Marque, J. Massol (Aixial), D. Maucort-Boulch, S. Zeghmar, E. Hachulla, S. Jouneau, V. Cottin, Développement d’algorithmes pour identifier les patients atteints de pneumopathie interstitielle diffuse de forme progressive (hors fibrose pulmonaire idiopathique) (PID-FP) en France (étude PROGRESS), Revue des Maladies Respiratoires Actualités, Volume 12, Issue 1, 2020, Page 150, ISSN 1877-1203