804: eFGR prevalence, prognosis and outcomes in a UK population dataset

Abstract

Extremely early onset fetal growth restriction (eFGR; FGR < 3rd centile diagnosed prior to 28 weeks’ gestation, delivered prior to 33 weeks’ gestation) is a rare, but devastating pregnancy complication. Relatively little is known about its prevalence and prognosis, complicating counselling and management. In this study we have examined eFGR using a large national dataset combined with detailed fetal medicine measurements. Scottish national health service data (333,213 births from 2012-2017) was used to retrospectively determine prevalence and survival in eFGR. Prognostic information was derived from a single fetal medicine database of eFGR cases (n=183). Potential prognostic indicators of fetal death in-utero (FDIU) including presenting characteristics and longitudinal growth velocity were examined using logistic and mixed level regression. eFGR prevalence is 3 per 1000 pregnancies. Survival was >75% for those born after 29 weeks, and at a birthweight above 800g. Using the detailed eFGR database, sonographic factors predictive of FDIU at diagnosis (adjusted for gestation) included estimated fetal weight (EFW) (25g increment: aOR 0.71 (95% confidence interval 0.62-0.81)), umbilical artery end-diastolic flow (present compared to absent/reversed: 13.4 (3.70-48.6)) and mean uterine artery resistance index (0.1 increment: 2.05 (1.23-3.43)). A multivariate prognostic model using these factors correctly identified 43/45 cases of FDIU, with a positive likelihood ratio (LR) of 4.4 but a negative LR of 0.09. Growth velocity in pregnancies ending in FDIU was significantly slower (Figure 1), with a 50g/week increase in EFW between 24-26 weeks’ gestation associated with a 53% reduction in FDIU. This is the largest single study to date of eFGR prognosis prediction, and the first to determine eFGR prevalence in a population dataset. We have identified factors predictive of FDIU and proposed a potential model to use in clinical care which, due to its strong negative LR, would allow clinicians to reassure those with a normal result that pregnancy is likely to result in a live birth.