Reliability of Self-Reported Maternal Health and Mother-Infant Outcome Data in Web-Based Questionnaires.

OBJECTIVES

The BELpREG registry in Belgium collects self-reported real-world data on perinatal medication use and mother-infant outcomes. Given the importance of self-reported data for pregnancy safety studies, this study aims to assess the reliability of maternal health and mother-infant outcome data reported in BELpREG.

METHODS

We compared self-reported maternal health and mother-infant outcomes in the BELpREG registry with medical record data from gynaecologists, general practitioners, and midwives. Participants included pregnant women enrolled in BELpREG, aged 18 years or older, who delivered between November 2022 and April 2024 and had medical record data provided by at least one healthcare professional (HCP). Comparisons for continuous variables such as pre-pregnancy body mass index (BMI), gestational age at delivery, and infant metrics at birth were analysed using intraclass correlation coefficients (ICCs). For dichotomous and categorical variables, including pregnancy complications, delivery characteristics, and hospital admissions, we used Cohen's Kappa and observed proportions of positive and negative agreement.

RESULTS

This study included 175 participants with available medical record data. Medical records were mostly returned by gynaecologists (40%) and midwives (25.7%). Reliability was high for pregnancy complications, onset, and mode of delivery. Gestational age showed almost perfect concordance (ICC 0.99). Infant outcomes, including birth metrics and Apgar scores, were highly reliable (ICC > 0.90), with almost perfect categorisation into birth weight percentiles (Cohen's Kappa 0.97). Self-reported maternal BMI exhibited moderate reliability as a continuous measure (ICC 0.75), but almost perfect agreement when categorised (Cohen's Kappa 0.89). The low prevalence of hospital admissions, delivery complications, infant complications, postpartum complications, and birth defects limited the assessment of these outcomes.

CONCLUSIONS

The BELpREG registry demonstrated high agreement between self-reported and HCP-reported data for most maternal health and mother-infant outcomes, supporting its use in future pregnancy safety studies.