Contraceptive use in France in 2012 and 2022: a descriptive analysis of two repeated nationwide cross-sectional studies.

Under the terms of the SNDS data use agreement, the complete study data cannot be shared with other investigators ( https://www.snds.gouv.fr ). However, the authors try to share publication related data as much as possible: the data that support the findings of this study are available on reasonable request from the corresponding author ( [email protected] ) and EPI-PHARE research team.

This study was conducted using data from the French National Health Data System (Système National des Données de Santé, SNDS) that includes health claims of 99% of individuals living in France (i.e. over 67 million inhabitants). The SNDS is a comprehensive national health data register that contains anonymised individual data on all reimbursed health expenditures, including hospitalization, birth, mortality, dispensation of prescribed drugs (coded according to the Anatomical Therapeutic Classification-ATC) and medical devices (coded according to the Reimbursed Products and Benefits List-LPP). The SNDS also contains socio-demographic characteristics such as sex, age, area of residence, and social deprivation index. The SNDS has been described in many previous pharmaco-epidemiological studies. 11 , 16 We conducted two cross sectional studies at two time points—January 2012 and January 2022—and compared prevalence measured at both time points. We chose to measure use on a specific month. Assessing contraception use in January 2022, well after the end of the last COVID-19-related lockdown in France (April 2021), we don't expect our results to reflect the decline in contraceptive dispensing observed during the pandemic. 17 We used the term ‘women’ for individuals identified as female through a sex variable in the SNDS database. For most of them, this reflected female-assignation at birth, and could also result from civil status gender transition. At each time point, all women residing in France aged 15–49 years in January of that year, and with at least one health expenditure of any kind recorded in the previous two years, were included in the study. We excluded women with a history of hysterectomy and women with a current pregnancy. We studied contraceptive methods reimbursed under the French National Health Insurance: COC, POP, injectable contraception, contraceptive implant, copper intrauterine device (Cu-IUD), LNG-IUD with 52 mg of levonorgestrel (LNG-IUD 52 mg), LNG-IUD 19.5 mg, LNG-IUD 13.5 mg and female sterilization (including tubal ligation and use of medical device Essure®—withdrawn from the French market in 2017) (ATC and LPP codes available in Supplementary Table S3 ). Use of a contraceptive method was identified through pharmacy dispensations. Since insertions of intrauterine devices or implants are not exhaustively recorded by professionals, we opted to use dispensations as indicators of initiation. Because use at a specific time point is conditioned on a past dispensation (i.e. dispensation of COC in the last three months, dispensation of Cu-IUD in the last five years, etc), we identified methods used at the time points of interest (January 2012 or January 2022) by researching the most recent dispensing of a contraceptive and taking into account maximum duration of use ( Supplementary Table S1 ). Because of a possible time-lag between dispensation and uptake of the method or insertion of the device, we added one month to methods’ duration of use. For example, implants have a maximum duration of use of three years: we identified all implant dispensations starting from 37 months (36 months duration of use + one month) prior to the time point of interest. Discontinuation of a contraceptive method was defined as initiation of another contraceptive method, the start of a pregnancy or maximum duration of use plus one month. Regarding sterilization, for January 2012, we included all women who had had a sterilization from January 2006 to January 2012. For January 2022, we included women already counted in 2012 and alive, and added all women who had a procedure from January 2012 to January 2022. Emergency contraception and barrier methods such as condoms were excluded. A number of COC and POP available on the French market were not reimbursed under the National Healthcare System, in varying proportions across the period. Therefore, the total prevalence of COC and POP use could not be measured solely from the SNDS data. To account for non-reimbursed oral contraceptives, we used the OpenHealth® sales database, 18 a pharmacy panel for mainland France built on a multi-stratified model taking into account the total sales of pharmacies, including sales data for reimbursed and non-reimbursed medication and medical devices. We used the OpenHealth sales data for 2012 and 2022 to determine the proportion of non-reimbursed oral contraceptives within the total amount of oral contraceptives sold in each of these years, for COC and POP separately ( Supplemental Figures S1 and S2 ). We used the proportion of reimbursed oral contraceptives as a correcting coefficient, and estimated the total number of users, for COC and POP and at each time point: E s t i m a t e d t o t a l n u m b e r o f u s e r s = n u m b e r o f u s e r s o f r e i m b u r s e d O C x ( 1 c o r r e c t i n g c o e f f i c i e n t ) In 2012, of the total amount of purchased COC, 80% were reimbursed and measured through the SNDS database. The remaining 20% were not reimbursed. Likewise, only 27% of POP were reimbursed, yielding correcting coefficients of 0.8 for COC and 0.27 for POP. In 2022, 84% of COC (correcting coefficient = 0.84) and 93% of POP (correcting coefficient = 0.93) were reimbursed. Sales data provided no information on prescribers or characteristics of women purchasing non-reimbursed oral contraceptives. There is previous evidence that non-reimbursed oral contraceptives are more readily prescribed by gynaecologists than general practitioners and that women with higher socioeconomic status are more likely to use them. 19 We also hypothesized that prescription and use of non-reimbursed oral contraceptives varied according to age groups. For these reasons, when analysing contraceptive use according to prescribers, socioeconomic variables or age groups, we used tailored correcting coefficients (see Supplementary Panel S1 ). Characteristics of users included age in January 2012 and in January 2022 (15–19 years, 20–29 years, 30–39 years, 40–49 years), coverage by the universal health insurance (CMU-C) for individuals with an income below the French poverty line (corresponding to below half of the median household income for that year), and the French census-based deprivation index of the municipality of residence 18 used as a proxy for the household socioeconomic category. The French census-based deprivation index is categorized in quintiles, with the fifth quintile representing the most deprived municipalities. We extracted data on type of health care provider prescribing the method (general practitioners, gynaecologists or midwives). For each contraceptive method, number of users and frequency were reported. Frequencies were computed by age groups, by health care providers and by quintiles of social deprivation index. All analyses were performed using SAS Enterprise Guide software, version 7.15. EPI-PHARE has permanent regulatory access to the data from the French National Health Data System (SNDS) via its constitutive bodies ANSM and CNAM, in application of the provisions of the French Decree No. 2016–1871 of December 26, 2016 relating to the processing of personal data called the “National Health Data System”, the French law articles Art. R. 1461-1316 and R. 1461-1417 from the French Public Health Code and the French Data Protection Authority (CNIL) decision CNIL-2016-316.18. In accordance with the permanent regulatory access granted to EPI-PHARE via ANSM and CNAM, this work did not require any specific opinion from the French Ethical and Scientific Committee for Research, Studies and Evaluations (CESREES) nor approval from the CNIL. The study was registered on the study register of EPI-PHARE concerning studies from SNDS data under the reference T-2023-08-407. This research was funded by the French National Health Insurance Fund (Cnam) and the French National Agency for Medicines and Health Products Safety (ANSM) via the Health Product Epidemiology Scientific Interest Group (ANSM-Cnam EPI-PHARE Scientific Interest Group). The funders had no role in considering the study design or in the collection, analysis, interpretation of data, writing of the report, or decision to submit the article for publication.

All women in France aged 15–49 years with at least one health expenditure recorded in the previous two years, without hysterectomy or current pregnancy, were included in the study: 14.2 million women in January 2012 and 14.6 in January 2022 ( Table 1 ). Overall prevalence of contraception use has remained stable, concerning 6,664,325 women or 47% of included women in 2012 and 6,728,680 or 46% in 2022. Table 1 Utilization of contraceptive methods among women aged 15–49 years in France in January 2012 and January 2022. 2012 a 2022 a n % (95% CI) of all women n = 14,172,538 % (95% CI) of contraception users n = 6,664,325 n % (95% CI) of all women n = 14,616,875 % (95% CI) of contraception users n = 6,728,680 No method b 7,508,213 53.0 (53.0–53.0) 7,888,194 54.0 (53.9–54.0) Oral contraceptives  COC 3,602,803 25.4 (25.4–25.4) 54.1 (54–54.1) 2,370,205 16.2 (16.2–16.2) 35.2 (35.2–35.3)  POP 666,030 4.7 (4.7–4.7) 10.0 (10.0–10.0) 1,293,073 8.8 (8.8–8.9) 19.2 (19.2–19.2) LARC  All LNG-IUD 1,132,081 8.0 (8.0–8.0) 17.0 (17.0–17.0) 1,007,094 6.9 (6.9–6.9) 15.0 (14.9–15)  LNG-IUD 52 mg 1,132,081 8.0 (8.0–8.0) 17.0 (17.0–17.0) 677,167 4.6 (4.6–4.6) 10.1 (10–10.1)  LNG-IUD 19.5 mg 0 0.0 0.0 274,101 1.9 (1.9–1.9) 4.1 (4.1–4.1)  LNG-IUD 13.5 mg 0 0.0 0.0 55,826 0.4 (0.4–0.4) 0.8 (0.8–0.8)  Cu-IUD 791,231 5.6 (5.6–5.6) 11.9 (11.8–11.9) 1,428,837 9.8 (9.8–9.8) 21.2 (21.2–21.3)  Implant 315,893 2.2 (2.2–2.2) 4.7 (4.7–4.8) 412,717 2.8 (2.8–2.8) 6.1 (6.1–6.2) Injectable 2248 0.02 (0.02–0.02) 0.03 (0.03–0.04) 1996 0.01 (0.01–0.01) 0.03 (0.03–0.03) Sterilization 154,040 1.1 (1.1–1.1) 2.3 (2.3–2.3) 214,759 1.5 (1.5–1.5) 3.2 (3.2–3.2) COC: combined oral contraceptives; POP: progestogen-only pill; LNG-IUD: levonorgestrel-releasing intra-uterine device; LARC: long-acting reversible contraceptives. Differences between the two time-points in total distribution and for each method separately are all statistically significant, with p-value <0.0001. a Number in italics are corrected to include estimated numbers of users of non-reimbursed oral contraceptives. b Refers to women who are either not taking any form of contraception (including women trying to conceive), or who are taking non-reimbursed contraception (other than oral contraception) that we are unable to trace (patch, ring, condoms). Utilization of contraceptive methods among women aged 15–49 years in France in January 2012 and January 2022. COC: combined oral contraceptives; POP: progestogen-only pill; LNG-IUD: levonorgestrel-releasing intra-uterine device; LARC: long-acting reversible contraceptives. Differences between the two time-points in total distribution and for each method separately are all statistically significant, with p-value <0.0001. Number in italics are corrected to include estimated numbers of users of non-reimbursed oral contraceptives. Refers to women who are either not taking any form of contraception (including women trying to conceive), or who are taking non-reimbursed contraception (other than oral contraception) that we are unable to trace (patch, ring, condoms). COC was the most used method in 2012, with 3,602,803 users, or 54% of contraceptive users, and decreased by a third to 2,370,205 or 35% in 2022. Despite this, COC remained the most popular method in 2022. In an opposite dynamic, POP and Cu-IUD use has doubled: from 10% of contraceptive users (n = 666,030) in 2012 to 19% (n = 1,293,073) in 2022 for POP, from 12% (n = 791,231) to 21% of contraceptive users (n = 1,428,837) for Cu-IUD. Taken together, all LNG-IUD reduced slightly in 2022, and remained less widely used than Cu-IUD, despite the two low-dose LNG-IUD newly available. Implant use was stable, as was the use of injectable contraceptives. In 2012, 2% of women using contraception, or 154,040 women, had undergone sterilization, reaching 3% in 2022, or 214,759 women. Supplementary Figure S1 displays the shifts that occurred within COC use: the proportion of 3–4Gs has more than halved (from 50% in 2012 to 15% in 2022, from 27.2 million dispensations during the year 2012 to 5.2 million dispensations during the year 2022) in favour of 2Gs, the latter making up 84% of purchased COC in 2022, or 27.1 million dispensations during the year 2022. Fig. 1 displays stark differences in method use across age groups. In women below 30 years, contraceptive utilization was dominated by COC in 2012. In the 20–29 years age group, we observed a greater diversity of contraceptive methods in 2022, with increase of POP, Cu-IUD and implants. Fig. 1 Utilization of contraceptive methods among women using contraception aged 15–49 years in France in January 2012 and January 2022, by age groups . COC: combined oral contraceptives; POP: progestogen-only pill; LNG-IUS: levonorgestrel-releasing intra-uterine system. Percentages were corrected to include estimated number of users of non-reimbursed oral contraceptives. Utilization of contraceptive methods among women using contraception aged 15–49 years in France in January 2012 and January 2022, by age groups . COC: combined oral contraceptives; POP: progestogen-only pill; LNG-IUS: levonorgestrel-releasing intra-uterine system. Percentages were corrected to include estimated number of users of non-reimbursed oral contraceptives. In the 30–39 years age group, the reduction in COC was mirrored by a rise in POP and Cu-IUD use. In the 40–49 years age group, COC use decreased while POP use surged (from 9.5%–24% of users, or from 190,809 to 498,123 users) and sterilization reached 9% of users (179,884 women). IUD use remained stable, concerning 50% of women aged 40–49 years, or 1,051,066 women, comprising an increase of Cu-IUD (from 19% to 23% of users, or 384,669 to 477,603 users). Between 2012 and 2022, the proportion of prescriptions by general practitioners remained stable at 50%–48% (or from 3,224,949 to 3,100,737 prescriptions), while that of gynaecologists decreased from 48% to 36% (or from 3,102,884 to 2,280,331 prescriptions) and that of midwives increased from 0.3%–13%, or from 19,118 to 854,889 prescriptions ( Supplementary Figure S3 ). Over the period, use of hormonal methods decreased from 86% to 76% of women using contraception (or from 5,719,054 to 5,085,084 users) ( Fig. 2 ), with non-hormonal methods (Cu-IUD and sterilization) increasing from 14% to 24% (or from 945,271 to 1,643, 596 users). In 2022, nearly one in four of the women with prescribed contraceptives used Cu-IUD. Fig. 2 Proportions of hormonal versus non-hormonal method use in women aged 15–49 years, in January 2012 and January 2022 . Proportions of hormonal versus non-hormonal method use in women aged 15–49 years, in January 2012 and January 2022 . Table 2 shows important variations in the prescribing practices of different types of healthcare providers. Oral contraceptives remained the first prescribed methods for general practitioners in January 2022, with COC and POP together accounting for 71% of general practitioners’ prescriptions (or 2,224,720 prescriptions). In contrast, gynaecologists and midwives alike appeared as strong prescribers of intrauterine devices. Table 2 Distribution of utilization of contraceptive methods among women aged 15–49 years in France in January 2012 and January 2022, according to prescribing health care professional (%). GP Gynecologist Midwife Other HCPs 2012 N = 3,224,949 2022 N = 3,100,737 2012 N = 3,102,884 2022 N = 2,280,331 2012 N = 19,118 2022 N = 854,889 2012 N = 119,679 2022 N = 177,252 % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) Oral contraceptives  COC 68.0 (67.9–68.1) 48.3 (48.2–48.4) 44.5 (44.4–44.6) 25.8 (25.7–25.9) 25.8 (25.2–26.4) 20.9 (20.8–21.0) 51.7 (51.4–52.0) 36.3 (36.1–36.5)  POP 11.2 (11.2–11.2) 23.4 (23.4–23.4) 8.1 (8.1–8.1) 16.6 (16.6–16.6) 45.4 (44.7–46.1) 15.9 (15.8–16.0) 13.1 (12.9–13.3) 22.3 (22.1–22.5) LARC  LNG-IUD 52 mg 9.0 (9.0–9.0) 6.4 (6.4–6.4) 25.4 (25.4–25.4) 16.2 (16.2–16.2) 10.6 (10.2–11.0) 9.4 (9.3–9.5) 17.1 (16.9–17.3) 9.8 (9.7–9.9)  LNG-IUSD19.5 mg 0.0 2.0 (2.0–2.0) 0.0 6.7 (6.7–6.7) 0.0 5.8 (5.8–5.8) 0.0 3.9 (3.8–4.0)  LNG-IUD 13.5 mg 0.0 0.5 (0.5–0.5) 0.0 1.2 (1.2–1.2) 0.0 1.1 (1.1–1.1) 0.0 0.9 (0.9–0.9)  Cu-IUD 6.3 (6.3–6.3) 12.5 (12.5–12.5) 17.9 (17.9–17.9) 28.4 (28.3–28.5) 14.3 (13.8–14.8) 39 (38.9–39.1) 12.5 (12.3–12.7) 20.5 (20.3–20.7)  Implant 5.4 (5.4–5.4) 6.7 (6.7–6.7) 4.0 (4.0–4.0) 5.1 (5.1–5.1) 3.9 (3.6–4.2) 7.9 (7.8–8.0) 5.6 (5.5–5.7) 6.3 (6.2–6.4) Injectable 0.06 (0.06–0.06) 0.04 (0.04–0.04) 0.01 (0.01–0.01) 0.01 (0.01–0.01) 0.02 (0–0.04) 0.03 (0.03–0.03) 0.03 (0.02–0.04) 0.05 (0.04–0.06) Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 COC: combined oral contraceptives; POP: progestogen-only pill; LNG-IUS: levonorgestrel-releasing intra-uterine device; HCPs: Health care professionals. Percentages were corrected to include estimated number of users of non-reimbursed oral contraceptives. Due to missing data in health care professional variable, total numbers of dispensations are smaller than numbers reported in Table 1 . Distribution of utilization of contraceptive methods among women aged 15–49 years in France in January 2012 and January 2022, according to prescribing health care professional (%). COC: combined oral contraceptives; POP: progestogen-only pill; LNG-IUS: levonorgestrel-releasing intra-uterine device; HCPs: Health care professionals. Percentages were corrected to include estimated number of users of non-reimbursed oral contraceptives. Due to missing data in health care professional variable, total numbers of dispensations are smaller than numbers reported in Table 1 . Percentages of women benefitting from universal health insurance for low-income individuals are presented in Supplementary Table S3 , showing a smaller proportion of low-income women in gynaecologists and midwives’ patient base, compared to general practitioners. In Table 3 , distribution of use of each method is presented according to social deprivation index. Social disparities existed in the use of POP in 2012 only and in injectable contraceptives across the period (near 30% of injectable contraception users lived in the most deprived areas at both time points, 488 users in January 2012 and 491 users in January 2022). Social disparities were observed for all IUDs, with a higher share of IUD users living in the least deprived areas. The opposite pattern was observed for implants and sterilization, with more users living in the most deprived areas, largely stable over the period. Table 3 Utilization of contraceptive methods among women aged 15–49 years in France in January 2012 and January 2022, according to social deprivation index of place of residence. COC % POP % Injectable % LNG-IUD 52 mg % LNG-IUD 19.5 mg % LNG-IUD 13.5 mg% Cu-IUD % Implant % Sterilization % 2012 N = 3,474,975 2022 N = 2,293,283 2012 N = 630,389 2022 N = 1,247,776 2012 N = 1660 2022 N = 1600 2012 N = 1,102,140 2022 N = 656,655 2022 N = 266,250 2022 N = 52,881 2012 N = 763,418 2022 N = 1,380,946 2012 N = 301,723 2022 N = 390,167 2012 N = 148,699 2022 N = 207,483 Q1 19.6 20.0 17.3 18.9 13.4 13.8 20.0 19.8 21.7 27.2 21.3 23.1 15.0 15.0 13.9 12.0 Q2 21.1 20.7 20.1 20.8 15.8 18.6 22.0 21.9 23.7 23.7 21.8 23.1 19.8 19.3 20.0 19.2 Q3 20.3 20.1 20.5 20.0 20.5 18.1 21.1 20.8 21.3 20.1 21.1 20.9 20.9 20.8 21.1 21.2 Q4 20.1 19.9 21.0 20.4 20.8 18.9 20.4 20.5 18.9 16.5 19.5 18.2 22.0 21.6 23.1 23.8 Q5 18.9 19.2 21.1 19.9 29.4 30.7 16.5 17.0 14.4 12.6 16.4 14.7 22.4 23.3 21.9 23.9 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 COC: combined oral contraceptives; POP: progestogen-only pill; LNG-IUD: levonorgestrel-releasing intra-uterine device. Q1: less deprived quintile; Q5: most deprived quintile. For COC and POP, percentages were corrected to include estimated number of users of non-reimbursed oral contraceptive. Due to missing data in social deprivation index variable, total numbers of dispensations are smaller than numbers reported in Table 1 . Differences between the two time-points are all statistically significant with p-value <0.0001, except for Injectable (p-value = 0.08). Utilization of contraceptive methods among women aged 15–49 years in France in January 2012 and January 2022, according to social deprivation index of place of residence. COC: combined oral contraceptives; POP: progestogen-only pill; LNG-IUD: levonorgestrel-releasing intra-uterine device. Q1: less deprived quintile; Q5: most deprived quintile. For COC and POP, percentages were corrected to include estimated number of users of non-reimbursed oral contraceptive. Due to missing data in social deprivation index variable, total numbers of dispensations are smaller than numbers reported in Table 1 . Differences between the two time-points are all statistically significant with p-value <0.0001, except for Injectable (p-value = 0.08).

The prevalence of contraceptive use in France has remained stable over the last decade. Combined oral contraception, used by the majority of contraceptive users in 2012, has decreased by a third, while remaining the most used method in 2022. Use of Cu-IUD and POP have doubled. Within COC, the proportion of 3–4Gs has more than halved in favour of 2Gs. Contraception has become more diversified at all ages. The changes have been most marked in women over 30 years, with the increase of POP and Cu-IUD. The proportion of prescriptions dispensed by gynaecologists has decreased, with midwives taking over. While consistent with previous work showing a reduction in oral contraceptive use and an increase in IUD use, 13 our study allowed for more nuance, disaggregating COC from POP and identifying opposite dynamics within oral contraceptive use. We were also able to be more precise with the evolution of IUD use, identifying an increase in Cu-IUD and stability in the use of LNG-IUD. Comparable trends of COC and POP use in 2010–2019 were documented in Denmark, 4 although with an increase in use of LNG-IUD unseen in our data. In the United Kingdom, contraceptive methods used were roughly comparable to ours at the start of the period, with dominant prevalence of COC and comparable prevalence of POP and Cu-IUD. 20 A decrease in COC prescriptions coupled with an increase in POP prescriptions during the COVID-19 pandemic was later reported. 21 A decrease in oral contraceptive use was documented in other European countries, in favour of condoms in Germany and Switzerland 3 and in favour of IUDs in Belgium. 5 While our results displayed high Cu-IUD use, there are large variations in prevalence of Cu-IUD use across Europe. 22 Our findings document a decrease in hormonal method use. The decrease in COC use in favour of Cu-IUD potentially reduces the cardiovascular and thromboembolic risk associated with COC. Indeed, previous work in France found that the pulmonary embolism hospitalization incidence rate in women aged 15–49 years decreased by 11% between 2012 and 2013, following a decrease in use of 3-4G COC by half. 23 This decrease in pulmonary embolism was in line with the documented association of 3-4G COC and venous thromboembolism morbidity. Moreover, because of change in prescribing practices, hormonal methods used in 2022 have lower cardiovascular risk compared to methods used in 2012: increased POP use, and the shift from 3-4G to 2G COC, mean potentially lower cardiovascular risk for users. Parallel to concerns relative to vascular risks associated with COC use, concerns about the effects of hormonal contraceptives on mental health may have fuelled hormonal hesitancy in women 6 , 24 , 25 and encouraged them to seek out non-hormonal options. We described important changes in contraceptive use in women above 30 years old, with large decrease of COC and strong increase in Cu-IUD. These changes suggest that recommendations to limit COC use among women with cardiovascular risk factors, 14 more frequent in women above 30 years old, or with tobacco smoking, have been heard by healthcare providers to some extent. It may also be that these age groups have been more prompt in consulting midwives, possibly becoming familiar with this health care option during pregnancies. Midwives, with a long-term practice of postnatal Cu-IUD prescription and solid training in IUD insertion, have been instrumental in the increase of Cu-IUD use. Unlike Cu-IUD use, LNG-IUD use has not increased, in spite of the two additional LNG-IUD available. The pill crisis may have fuelled a more general hormonal hesitancy that may have reached hormonal IUD use. Moreover, since research has shown an association between LNG-IUD 52 mg and subsequent use of anti-depressants, 24 concern over mental health might also participate in the decrease in LNG-IUD 52 mg. Cu-IUD and LNG-IUD use may differ in other respects: previous research using the SNDS database 26 has shown that women using LNG-IUD tended to be older and have more frequent gynaecological history than Cu-IUD users, possibly using LNG-IUD for their non-contraceptive benefits. Few previous works have explored the use of POP, 20 , 21 and this is the first estimate of POP use in France, its evolution and the socioeconomic profile of its users. Before the decrease in COC use and the shifts described here, POP was prescribed mostly in the postpartum period, and to women with obesity or other cardiovascular risk factors, which could explain the social disparities in POP use observed in 2012. As prescribers and women alike turned away from COC, much of COC use may have been transferred to POP, recommended by healthcare providers as a low-risk oral contraception option concerning vascular risk. Reimbursement of desogestrel-based POP and the arrival in 2020 of a drospirenone-based POP, not reimbursed to date, may have additionally contributed to its prescription on a wider social spectrum. Disaggregating Cu-IUD and LNG-IUD, we were able to show that IUD increase in women with higher socioeconomic status existed only in Cu-IUD. Following the pill crisis, women with higher socioeconomic status may have been more aware of the risks associated with 3-4G COC and more willing to transition to hormone-free methods. We found striking social disparities in the use of injectable contraceptives: this is as expected, as in France, injectable contraceptives are mostly used by immigrant women in the first years after their arrival, a very economically vulnerable group. The observed drop in COC use surely has meant questioning the central place of COC as an obvious go-to method. It might have functioned as an opportunity for contraception counselling, opening the range of choices for both providers and users, a chance to tailor contraception use to individual preferences and needs of women. And indeed, the large increase in Cu-IUD use seems to echo women's demand for effective and hormone-free contraceptive options. 6 However, it seems that structural determinants remain largely at play in shaping contraceptive use. Social disparities have deepened in the use of IUDs and implants, in spite of a widening range of contraceptive options, with newly available reimbursed methods. Also, to some extent, the channelling of women towards specific contraceptive options according to their social background could be partly explained by the healthcare providers they access. 27 Women whose contraception was prescribed by general practitioners tended to have lower income and they received oral contraceptives more frequently and IUDs less frequently. On the contrary, women whose contraception was prescribed by gynaecologists and midwives—frequent prescribers of Cu-IUD and LNG-IUD—tended to have higher income. However, as of 2022, consulting a midwife for contraception prescription was a fairly new option in France. Women with higher education and more resources generally may have been more aware of this new healthcare option. Therefore, we can expect that in time, more low-income women will seize the opportunity of consulting midwives for their contraception, as a low-barrier sexual health specialist, possibly bringing new shifts in contraceptive use. Although consultation fees of all medical professionals are covered by the National Health Insurance (for up to 65% of the fee), in 2023 75% of gynaecologists charged additional consultation fees, 28 not subject to reimbursement, while midwives had no additional consultation fees. But beyond the financial aspect, additional barriers exist in accessing gynaecologists, driven by the growing scarcity of gynaecologists nationwide (e.g. difficulty obtaining an appointment, distance to a practitioner). This study has some limitations. We chose not to use information on parity. The SNDS data is available only from 2006, meaning that information on parity can only include births that occurred from 2006. Information on parity status is therefore partial, producing misclassification for women who gave birth before that year. Without exhaustive information on LARC removal, we chose to define LARC discontinuation as initiation of another contraceptive method, the start of a pregnancy or maximum duration of use. This may have led to an overestimation of LARC prevalence, in cases of LARC removals not followed by initiation of another method or the start of a pregnancy. Emergency contraception and non-hormonal methods such as condoms and other barrier methods, although very partially reimbursed in France, were excluded from our study due to lack of exhaustive data and their inherently occasional use. Women classified as “No method” may, however, have used these forms of contraception. Similarly, women classified as “No method” might have relied on their partner's vasectomy for birth control. Recently published work has shown that the number of vasectomies remain low in France, despite a fast increase in recent years, with up to 30,000 procedures carried out in 2022. 29 This study presents notable strengths. This large real-life study provides an overview of the evolution of contraception use in the last decade in a country with high use of prescribed contraception. We presented previously unavailable results on prescribed contraception use in the early 2010s using real-world data, allowing us to map out the evolution of use over the decade. We provided novel and detailed results on the evolution of use of different types of oral contraceptives and intrauterine devices. Using sales data allowed us to estimate prescribed contraceptive method use exhaustively. Moreover, tailoring correcting coefficients according to age, providers and social deprivation of the areas of residence made it possible to assess the evolution of contraceptive use according to these major determinants. Lastly, we have included all women who use prescribed contraception, regardless of the indication (whether they use it for birth control or other potential benefits). Indeed, besides high contraceptive efficacy, COC and LNG-IUD 52 mg may have health benefits regarding menstrual cycle disorders and endometriosis. 15 , 30 Therefore, non-contraceptive indications are likely to affect prescription rates. This differs from national contraceptive use surveys in France, where contraceptive use is collected only in individuals reporting recent heterosexual intercourse. Our study has shown that the use of hormonal contraception among French women has changed considerably since 2012. The use of COC has decreased across all age groups with a larger decrease with increasing age. Second-generation COC has largely replaced use of third- and fourth-generation COCs. In addition, women are increasingly using POP and IUDs, highly effective contraceptive methods with lower cardiovascular risk.

MZ, NR and LP, conceived and planned the study. LP, RF and KH-M performed data management. LP, NR and RF had access to raw data. LP performed statistical analyses. NR and RF verified the data. LP drafted the manuscript. All authors reviewed the manuscript. MZ supervised the study and had final responsibility to submit for publication. All authors approved the final manuscript.

A decrease in oral contraceptive use in several European countries in the 2010s has been documented. 1 , 2 , 3 , 4 , 5 Across Europe, women had voiced concern about hormonal contraception affecting physical and mental health. 6 , 7 Despite a historically high use of oral contraceptives in France, 8 , 9 heightened public and media attention was drawn to the increased risk of venous thromboembolism, 10 pulmonary embolism and other cardiovascular events 11 , 12 associated with combined oral contraceptives (COC) use, triggering what was called a national pill crisis of large magnitude. 13 Studies have found that risk varies according to the type of progestogen, with third and fourth generation (3G and 4G) COC (containing desogestrel, gestodene, norgestimate, norgestrel, dropirenone, dienogest, nomegestrol as progestogen) and cyproterone acetate having a more elevated cardiovascular risk compared to second generation (2G) COC containing levonorgestrel. 10 , 11 The European Medicines Agency (EMA) issued guidelines in 2014 affirming that second generation COC was to be preferred. 14 While the pill crisis affected all social groups, with a general decrease in COC use, uptake of intrauterine devices (IUD) was more likely in women with higher socioeconomic status. 13 Under the French National Health Insurance, medical contraceptive methods are reimbursed for women of all ages up to 65% of their price (100% for individuals with low income benefitting from universal health insurance coverage) with optional private health insurance covering the remaining amount. All physicians are able to prescribe all contraceptive methods but general practitioners and gynaecologists are the main contraception prescribers. Until 2013, midwives could prescribe all contraceptives but only in the perinatal period. Starting in 2013, they are able to prescribe contraception throughout the reproductive life, while referring women with pathologies to gynaecologists. Contraceptives are largely prescribed in private practice. In most cases, long-acting reversible contraceptives (LARCs) are inserted by the prescriber. French medical guidelines highlight non-contraceptive benefits of contraceptive methods 15 : COC and levonorgestrel-releasing intrauterine device (LNG-IUD) 52 mg might be used for the treatment of menstrual cycle disorders and endometriosis. Over the period 2012–2022, four major developments have been observed. First, in 2013, coverage by the French National Healthcare system of third generation COC was ended, although they remained available in pharmacies. Second, from 2014, desogestrel-based progestogen-only pill (POP) started to be reimbursed, adding to levonorgestrel-based POP as a reimbursed option. Third, two lower-dose LNG-IUD entered the French market and were reimbursed in 2013 and 2018. And fourth, in 2013 the professional competences of midwives have been extended to include the prescription of contraceptives for women in good reproductive health throughout their lives (see Supplementary Chart S1 ). While these elements suggest important changes in contraceptive use in the past decade, little is known of the recent evolution in contraceptive use. Our main objective was to describe the trend in the use of contraceptive methods from 2012 to 2022 in France, in different age groups. We also aimed to describe changes in healthcare providers’ prescribing practices and in the socioeconomic profiles of women using different contraceptive methods.

The authors declare that they have no competing interests, no relationships, activities or interest to disclose that are related to the content of the manuscript.