(A-292) Validation of Self-Reported Registrations on Maternal Health and Mother-Infant Outcomes: Insights from the BELpREG cohort

Background: The BELpREG registry in Belgium collects real-world data on medication use in pregnancy and mother-infant outcomes self-reported by women during and after their pregnancy.

Objectives: To validate the quality of maternal health and mother-infant outcome data reported by women in the BELpREG pregnancy registry by assessing its consistency with medical records.

Methods: This validation study compared self-reported outcomes related to maternal health, delivery outcomes, and infant outcomes with medical records’ data from gynaecologists, general practitioners, and midwives. The study included BELpREG participants (i.e., pregnant women aged 18 or older and receiving healthcare in Belgium) who delivered before April 2024, provided informed consent, and had medical record data available from at least one healthcare provider. Continuous variables such as preconceptional maternal BMI, gestational age at delivery, and infant metrics at birth were analyzed using intraclass correlation coefficients (ICC) with 95% Confidence Intervals (CIs). Dichotomous and categorical variables, including pregnancy complications (i.e., hypertensive disorders, gestational diabetes), delivery characteristics (i.e., mode of delivery, onset of delivery, preterm birth), and mother and infant hospital admissions were assessed using sensitivity, specificity, and Cohen’s Kappa.

Results: This study included 175 unique participants with available medical record data. Data from medical records were mostly returned by gynaecologists (40%), midwives (25.7%), or both (13.7%). High agreement was found between the BELpREG registry data and medical records for most studied variables. For example, the sensitivity and specificity for hypertensive disorders were 1.00 (95% CI: 0.72-1.00) and 0.96 (95% CI: 0.92-0.99), respectively, and for gestational diabetes, both were 1.00 (95% CI: 0.59-1.00 and 0.98-1.00). ICCs for continuous variables were high, with gestational age at delivery at 0.99 (95% CI: 0.99-0.99), birth weight at 0.97 (95% CI: 0.69-0.98), head circumference at 0.86 (95% CI: 0.80-0.90), and the Apgar score at 5 minutes at 0.88 (95% CI: 0.84-0.92). However, some outcomes showed moderate agreement, such as preconceptional maternal BMI with an ICC of 0.75 (95% CI: 0.67-0.82), and complications related to preterm birth, with a sensitivity of 0.69 (95% CI: 0.39-0.91). Mode of delivery had a Cohen’s Kappa of 0.88, and the onset of delivery had a Cohen’s Kappa of 0.90. The low prevalence of hospital admissions, delivery and infant complications limited the ability to validate these outcomes effectively.

Conclusions: These findings demonstrate the value and relevance of the self-reported BELpREG data, substantiating its use in future drug safety assessments.