[Performing pharmacoepidemiological studies using the French health insurance data warehouse (SNDS): How to translate guidelines into practice]. / Réalisation d'études pharmacoépidémiologiques à partir du système national des données de santé (SNDS) : application des recommandations dans la pratique.

The French health insurance data warehouse named SNDS is one of the largest medico-administrative in the world allowing for powerful pharmacoepidemiological studies, based on real-life data collected prospectively. In addition to the absolute necessity of a strong pharmacological rationale, recommendations have been thought to improve the quality of pharmacoepidemiological studies. These guidelines emphasize the importance of an accurate definition of the study population, outcome and exposure, especially for studies performed on medico-administrative databases. Compliance with certain guidelines, particularly those concerning the identification of a specific population or an outcome and the definition of risk periods or exposure periods, may be difficult when performing studies on the SNDS because of its structure and the nature of the data recorded. The objective of this article is to provide advice for the conduct of pharmacoepidemiological studies according to the recommendationswhen using the SNDS, given its specificities. The performing of reliable studies from this rich but complex data warehouse requires the expertise of researchers with deep knowledge both in the SNDS and in pharmacological reasoning.

The Characteristics of Care Provided to Population(s) in Precarious Situations in 2015. A Preliminary Study on the Universal Health Cover in France.

Background: The French Universal Health Cover (CMU) aims to compensate for inequalities between precarious and non-precarious populations, enabling the former to access to free healthcare. These measures rely on the principle that precarious populations' health improves if healthcare is free. We designed a study to examine whether CMU fails to compensate for inequalities in reimbursed drugs prescriptions in precarious populations. Material and method: This retrospective pharmaco-epidemiological study compared the Defined Daily Dose relative to different reimbursed drugs prescribed by general practitioners (GPs) to precarious and non-precarious patients in France in 2015. Data were analysed using Mann-Whitney tests. Findings: 6 out of 20 molecules were significantly under-reimbursed in precarious populations. 2 were over-reimbursed. The 12 remaining molecules did not differ between groups. Interpretation: The under-reimbursement of atorvastatin, rosuvastatin, tamsulosine and timolol reflects well-documented epidemiological differences between these populations. In contrast, the equal reimbursement of amoxicillin, pyostacine, ivermectin, salbutamol and tiopropium is likely an effect of lack of compensation for inequalities. Precarious patients are more affected by diseases that these molecules target (e.g., chronic bronchitis, bacterial pneumonia, cutaneous infections). This could also be the case for the equal and under-reimbursement of insulin glargine and metformin (targeting diabetes), respectively, although this has to be considered with caution. In conclusion, the French free healthcare cover does not fail to compensate for all but only for some selective inequalities in access to reimbursed drugs prescriptions. These results are discussed with respect to the interaction of the doctor-patient relationship and the holistic nature of primary care, potentially triggering burnout and empathy decrease and negatively impacting the quality of care in precarious populations.