INTRODUCTION: In 2018, European pregnancy prevention program for oral retinoids, i.e., isotretinoin (acne), acitretin (psoriasis) and alitretinoin (eczema) was updated as part of additional risk minimization measures (RMM) [1, 2] to prevent major congenital anomalies in new-borns whose mothers were exposed to these drugs [3, 4].
OBJECTIVE: To measure the impact of the 2018 RMMs on utilization patterns of oral retinoids, alternative medications, contraceptives, and on rates of pregnancies occurring concurrently with retinoid prescriptions.
METHODS: Retrospective cohort study between Jan-2010 and Dec-2020 in females of childbearing age (12–55 years) derived from six electronic healthcare record data sources in the Netherlands (PHARMO), Spain (BIFAP and VID-Valencia), Italy (Caserta and ARS-Tuscany), and Denmark (Danish National Registries-DNR). Information from data sources was transformed into the ConcePTION Common Data Model [5]. Common analysis scripts were implemented by study sites. Monthly utilization patterns of isotretinoin, alitretinoin and acitretin (incidence/1000 person-months, prevalence/1000 persons, and percentage of discontinuers) were estimated. An interrupted time series analysis was performed to assess immediate changes after the implementation (level change) and over time (trend change). Contraceptive use (percentage) and concurrent pregnancies during retinoid treatment episodes (rates/1000 users) were calculated as well.
RESULTS: The study population comprised 11,570,047 females of childbearing age (12–55 years), 88,992 persons used an oral retinoid at any point during the study period. Monthly incidence and prevalence rates showed that retinoid prescriptions have a strong seasonal pattern with peaks of use in winter months. No significant (p < 0.05) level and trend changes were seen after the implementation of RMMs in August 2018 for incidence, prevalence, and discontinuation rates in any of the study sites. The low level of recording of pregnancy tests did not allow for trend analysis. Contraceptive use could not be measured in ARS and DNR, and was very low in Caserta; in PHARMO, recorded contraceptive use in retinoid users was far below 100%, in VID and BIFAP, recorded contraceptive use increased over time. In BIFAP, there was a significant trend increase after 2018 RMMs in the utilization of contraceptives before the initiation of an oral retinoid treatment, but still far below 100% (15% maximum). Pregnancy rates varied between 0.1 and 0.4 per 1000 retinoid users and occurred in all study sites pre- and post-intervention.
CONCLUSION: Based on the findings, there is very limited measurable impact of the 2018 RMMs among females of childbearing age in the included databases. Moreover, pregnancies still happen during oral retinoid treatment after the implementation.
