(P-069) Improving Accuracy for Prospective Pregnancy Registry Enrollment Targets

Background: To extrapolate from live births to pregnant women to establish enrollment targets, it is necessary to make several assumptions, including assumptions regarding the rates of live birth, 1st trimester exposure, loss to follow-up (LTFU), and exclusion from the analysis population. In previous abstracts, we explored the rates of live birth and LTFU. In this abstract, we explore the rates of 1st trimester exposure as well as exclusions from the analysis population. Since the FDA draft guidance released in May 2019, participants may be enrolled in a registry but excluded from the analysis population for being retrospectively enrolled, exposed to teratogens during pregnancy, or already included in the analysis population for a prior pregnancy.

Objectives: To evaluate the rates of 1st trimester exposure, retrospective enrollment (i.e., enrollment after pregnancy outcome), exposure to teratogens, and subsequent pregnancies observed in prospective pregnancy registries.

Methods: We conducted a search in PubMed with the search term “pregnancy registry” which produced 327 results. Of these, 282 were excluded because the article was not reporting results from a prospective registry (n=207), not focused on drug exposure (n=12), not available in English (n=1), and not the most recent product-specific article from the registry (n=62). Data were extracted from each article and analyzed descriptively.

Results: Among the 45 articles included in the analysis, the proportion of articles with available data was as follows: 1st trimester exposure, 60%; retrospective enrollment, 20%; teratogen exposure, 4%; and subsequent pregnancy, 4%. In most cases, data on teratogen exposure and subsequent pregnancy were simply not reported (96%). For retrospective enrollment, data were largely either not reported (20%) or the article specified that retrospective participants were not eligible for the study (60%). For 1st trimester exposure, 22% of articles did not report the information and the remaining 18% specified that 1st trimester exposure was inclusion criteria for enrollment.

Conclusions: There is a paucity of data in the published literature on rates of retrospective enrollment, exposure to teratogens, and subsequent pregnancies observed in prospective pregnancy registries. For 1st trimester exposure, which is a critical data point, data are missing in nearly one-quarter of articles. When designing a pregnancy registry, these rates should be considered to improve the accuracy of enrollment targets. Registry sponsors would benefit from publishing these metrics to inform the design of future studies.