The relationship between virtual antenatal care and pregnancy outcomes in a diverse UK inner-city population; A group-based trajectory modelling approach using routine health records

Abstract

Background: The COVID-19 pandemic resulted in major reconfiguration of maternity services, particularly an increase in virtual antenatal care. We explored associations between virtual antenatal care trajectories and pregnancy outcomes.

Methods: Pregnancy and birth outcome data were obtained from a multiethnic and socioeconomically deprived UK inner-city population before and during the pandemic (with and without lockdown). Data were collected using a health record data linkage from the Born in South London, (eLIXIR-BiSL) cohort. Antenatal care was characterised by the number of outpatient contacts during six gestational windows: 0-14⁺⁶, 15-20⁺⁶, 21-27⁺⁶, 28-32⁺⁶, 33-36⁺⁶ and ≥37 weeks' gestation. In each window, the proportion of virtual antenatal care was grouped into quartiles, and group-based trajectory modelling was used to extract virtual antenatal care trajectories. Associations between these trajectories and pregnancy outcomes were explored using adjusted multinominal logistic regression.

Results: The analysis included 34,114 mother-child dyads (Oct-2018-Jul-2023). Group-based trajectory modelling suggested four trajectories of virtual antenatal care contacts: Low and stable virtual care throughout pregnancy (Trajectory-0; n=27,751 pregnancies, 81.3%); High 1st trimester virtual care (Trajectory-1; n=832, 2.4%); High 2nd trimester virtual care (Trajectory-2; n=2,410, 7.1%); and high 3rd trimester virtual care (Trajectory-3; n=3,121, 9.2%). Following adjustment, compared with the low and stable group (Trajectory-0), high 2nd trimester virtual care was associated with less gestational hypertension (adjusted relative risk ratio 0.84, 95% confidence interval [0.74-0.96]) and assisted vaginal birth (0.87 [0.76-1.00]), and more premature births (<37 weeks, 1.21 [1.02-1.44]), labour induction (1.13, 1.02-1.25), breech presentation (1.92, 1.02-3.62) and postpartum hemorrhage (1.14, 1.00-1.30). Similarly, compared to the low and stable group (Trajectory-0), high 3rd trimester virtual care had less gestational hypertension (0.84 [0.73, 0.96]), more premature births (<37 weeks, 1.35, 1.16-1.58) and elective (1.54, 1.38-1.72) or emergency (1.21, 1.01-1.34) Caesarean sections, and neonatal intensive care admissions (1.28, 1.09-1.50); fewer 3^rd /4^th degree vaginal tears (0.82, 0.75-0.90); and less early infant skin-to-skin contact (0.82, 0.73-0.92) and breastfeeding (0.90, 0.81-0.99).

Conclusion: A higher proportion of virtual care contacts in antenatal care in the second or third trimesters was associated with a greater risk of adverse pregnancy outcomes.

Keywords: Trajectories; antenatal care; birth outcomes; electronic health records; virtual care.