LAUSR

Are French national health insurance data reliable for studying the use of medical contraception? Health insurance data produce a measurement of contraceptive use consistent with population-based survey data, which affords new opportunities for studying contraception. Medical contraception is a major public health issue as most women of reproductive age use it. It is usually studied through population-based surveys. However, such surveys are conducted only every 10 years, and analyses are limited by their sample size. French national health insurance data provide comprehensive and time-continuous information on each reimbursed contraceptive. However, because these data have been collected for a different purpose (reimbursement), their relevance for measuring the use of contraceptives needs to be assessed. Two sources were analysed. First, a cross-sectional cohort was extracted from the health insurance database, which includes all health reimbursements (such as those for medical contraceptives) and covers 98% of the French population, including 14 million women aged 15–49. Secondly, we used the last French survey on contraception, a cross-sectional study including 4,508 women aged 15–49 interviewed by phone. From both sources, we selected all women aged 15–49 living in metropolitan France. We identified the last medical contraceptive purchased by each woman between 2014 and 2019. The woman was then classified as currently using this contraceptive if the recommended duration of use for this contraceptive was still ongoing on 31 December 2019. Prevalences were compared to those observed in the population based survey. Among the 14.3 million women aged 15–49 living in metropolitan France covered by the health insurance, 26.0% were using the pill, 17.4% an IUD (7.6% hormonal IUD; 9.9% copper IUD), and 3.1% an etonogestrel implant. These proportions are very close to and not statistically different from those observed in the population-based survey (26.2% for the pill, 18.4% IUD, and 3.1% implant). Contraceptive use varied widely with women’s age. At ages 20–24, the most widely used contraceptive was the pill (42.2%), and very few long-acting contraceptives were used (7.6% IUD; 4.9% implant). At ages 30–34, the pill was less frequently used (21.6%) and IUD more frequently used (copper IUD: 15.79%; hormonal IUD: 7.06%). It cannot be ruled out that some contraceptives were purchased but never used and that a few women stopped using the contraceptive before the end of its recommended duration. Wider implications of the findings: To our knowledge, this study is the first to estimate prevalence for Copper IUD and for hormonal IUD in France. Using the national health insurance database, it is now possible to monitor the use of each type of medical contraceptive over time in a reliable population-based approach. Not applicable