A population-based cohort to evaluate drug safety during pregnancy (PREGVAL): rationale, design, and baseline...

The main aim of the PREGVAL cohort is to evaluate and monitor the in-utero safety of medication administered to pregnant women, as usually this population is excluded from clinical trials assessing drugs. It is a population-based cohort, comprised by pregnant women and their offspring in the Valencia Region (Spain), between July 1st 2009 and December 31st 2021, which will be followed from 6 months pre-conception until death, loss of coverage or end of study, with over 520.000 pregnancies for the study period. The data source will be the Valencia Integrated Database (VID), which links records at the individual level using a unique personal identification code, with exhaustive data on comorbidities (diabetes, hypertension, etc.) and other clinical variables, sociodemographic variables (age, socio-economic status, risk of social exclusion) and other potential confounders (smoking, alcohol intake). It also provides accurate data on timing of exposure for most livebirths and all of the stillbirths occurred in the region, and partially for spontaneous abortions and elective terminations (due to the availability of gestational age by ultrasound or last menstrual period for these outcomes allowing the calculation of the date of conception) and allows the identification of medications prescription and dispensation data with a very high detail, avoiding the problem of potential exposure misclassification present in many of the existent registries/cohorts. Regarding outcomes, there is exhaustive and accurate data on pregnancy complications, spontaneous abortions, terminations (in this case, those occurred within the public health system), birth weight, preterm birth, congenital anomalies, and perinatal mortality; as well as the availability of post-natal long-term follow-up. These features make the PREGVAL study one of the largest and most exhaustive cohorts to assess real-world in-utero safety of medications to date.