POSA412 Using French National Healthcare Databases (SNDS) to Estimate Economic Burden of Diseases: A Literature Review

Abstract

The French national medico-administrative database called SNDS (“Système National des Données de Santé”) contains hospital records, primary and secondary care for 66 million people. All dispensed and reimbursed drug prescriptions are collected and these data are individualized and anonymized. The General Beneficiaries Sample (EGB) is a representative sample (1/97) of the SNDS. We aim to provide an overview of cost of illness assessment from medico-administrative databases, focusing on data extracted from SNDS and EGB. A literature review of French and English language publications from 2011 to 2021 was conducted in PubMed, searching keywords related to the SNDS database studies. Only human studies were selected. All publications including a cost analysis conducted through these databases were included. We excluded studies without any cost analysis and studies with cost data extracted from other databases or retrieved from PMSI (hospital database) only. Overall, 43 articles were selected after screening 121 publications, which most of them were published in clinical journals (54%). The EGB database was used more frequently to estimate costs than the SNDS (63% vs 37%). Cardiovascular diseases (16%), cancers (16%) and diabetes (16%) were the three pathologies most studied. Almost half of the studies considered costs over only a 1-year period or less (46.5%). Cost analyses were performed from the French health Insurance System (47%) or societal (35%) perspective, rarely both (9%), when the information was provided. Direct medical (hospitalizations and ambulatory care) or non-medical costs (transportation) were retrieved from French administrative databases, and 8 publications also considered some indirect costs (sick leaves, daily allowances, disability pensions). Discounting was not applied in most cases (79%). The SNDS is an exhaustive medical-administrative database which can provide real world cost data to assess burden of diseases, but also to provide inputs for medico-economic and budget impact models.