Comparison of Depression Diagnosis and Treatments during Pregnancy from Electronic Health Data with Self-Reported Information
Recommended Citation
Loree A, Cao J, Goodman S, Ehresmann K, Seburg E, Rossom R, Avalos L, Negriff S, Daida Y, Lipkind H, Vazquez Benitez G, Palmsten K. Comparison of Depression Diagnosis and Treatments during Pregnancy from Electronic Health Data with Self-Reported Information. Pharmacoepidemiol Drug Saf 2025; 34.
Document Type
Conference Proceeding
Publication Date
8-19-2025
Publication Title
Pharmacoepidemiol Drug Saf
Keywords
antidepressant agent, adult, Caucasian, conference abstract, controlled study, depression, diagnosis, drug therapy, electronic health record, female, health care personnel, health care system, health insurance, human, ICD-10, major clinical study, male, pregnancy, prevalence, psychotherapy, self report, therapy
Abstract
Background: Automated electronic health data (EHD; electronic health records or administrative claims data) are often used to study the effects of depression treatments during pregnancy. There are limited data comparing depression-related information from EHD with self-report. Objectives: To compare agreement on pre-pregnancy depression, and depression, antidepressants, and psychotherapy during pregnancy using information from EHD versus patient survey. Methods: We used data from a survey of people with a new episode of depression during pregnancy. Using EHD from 5 US healthcare systems (2022-2023), we identified patients whowere 2-4 months postpartum, had a depression diagnosis during pregnancy; had no depression diagnosis, antidepressants, or psychotherapy in the 6 months before pregnancy; and had health insurance 6 months before pregnancy through delivery. Participants completed the survey online or on paper, reporting depression symptoms, whether a health care provider diagnosed them with depression before and/or during pregnancy, antidepressant use, and psychotherapy received during pregnancy. We linked survey data with participant's EHD on ICD-10 codes for depression diagnoses, antidepressant fills, and procedure codes for psychotherapy visits during pregnancy and up to 3 years before pregnancy. We calculated prevalence according to data source and assessed agreement between the two sources using percent agreement and Cohen's kappa. Results: Of the 339 patients who participated (26% of invited), 41% were non-Hispanic white, and 18% had public insurance. Although all patients had depression diagnoses during pregnancy according to EHD (per survey inclusion criteria), 71% had depression during pregnancy according to survey. According to EHD, 57% had pre-pregnancy depression; 74% had pre-pregnancy depression according to survey. Percent agreement was 60%, Kappa was 0.14 (95% CI: 0.04, 0.24), and 61% of those who had pre-pregnancy depression according to survey had a pre-pregnancy depression diagnosis in EHD. Antidepressant fills during pregnancy occurred in 31% according to EHD and 20% according to survey. Percent agreement was 78%, Kappa was 0.44 (95% CI: 0.33-0.55), and 73% of those who reported antidepressant use during pregnancy had antidepressant fills in EHD. Psychotherapy was received by 45% of participants according to EHD and by 49% according to survey. Percent agreement was 66%, Kappa was 0.33 (95% CI: 0.21-0.45), and 61% of those who reported psychotherapy had EHD documentation of psychotherapy. Conclusions: Agreement on depression-related information between EHD and self-report was low, particularly so for pre-pregnancy depression. Both sources alone have limitations and approaches to overcome misclassification may be needed.
Volume
34
