Data From Web-Based Questionnaires Were Valid for Gestational Diabetes and Preeclampsia, but Not Gestational Hypertension

Abstract

Objectives: We aimed to validate Web-based questionnaires for the common pregnancy complications gestational diabetes, gestational hypertension, and preeclampsia.

Study design and setting: We included 1,809 women participating in the PRegnancy and Infant DEvelopment (PRIDE) Study who delivered in 2012-2017, for whom relevant data were complete. Sensitivity, specificity, and positive and negative predictive values of self-reported diagnoses of gestational diabetes, gestational hypertension, and preeclampsia were determined using obstetric records as reference standard. Furthermore, we assessed whether maternal characteristics affected disagreement between questionnaires and obstetric record.

Results: For gestational diabetes and preeclampsia, we observed very few false-positive and false-negative reports, yielding sensitivities of 93% (95% confidence interval [CI] 86-100) and 88% (95% CI 79-98), respectively, and specificities of 100%. Depending on the definition of gestational hypertension, sensitivity and positive predictive values ranged from 62% to 89% and 53% to 64%, respectively. Disagreement on gestational hypertension was associated with prepregnancy overweight and multiparity.

Conclusion: Self-reports of gestational diabetes and preeclampsia in Web-based questionnaires were valid, but the validity of gestational hypertension seemed to be lower because of relatively high numbers of false-positive reports. However, it is questionable whether an appropriate reference standard exists to validate this pregnancy complication.

Keywords: Epidemiologic methods; Gestational diabetes; Gestational hypertension; PRIDE Study; Preeclampsia; Pregnancy.