Woebot for Postpartum Mood and Anxiety: Randomized Controlled Trial Evaluating Feasibility, Acceptability, and...

Postpartum psychological distress, ranging from transient mood and anxiety disturbances to full-syndrome postpartum depression, is prevalent. Many postpartum individuals lack access to evidence-based interventions due to stigma and insufficient provider availability. The treatment gap is particularly pronounced among historically marginalized groups, including Black, Hispanic/Latina, and low-income mothers. Digital health interventions offer support to address this.

We aimed to evaluate the feasibility, acceptability, and preliminary efficacy of Woebot for Postpartum Mood and Anxiety (W-PPMA), a smartphone-based relational agent delivering cognitive behavioral therapy-informed psychoeducation to postpartum individuals.

This randomized controlled trial recruited participants from the PowerMom study, a digital platform for maternal health research. Eligible individuals (≥16 y, <3 mo post partum) were randomized to W-PPMA or a waitlist control condition. W-PPMA, an investigational digital mental health intervention, features a relational agent delivering cognitive behavioral therapy-based psychoeducation via text-based conversations. Primary outcomes included feasibility (Usage Rating Profile-Intervention Revised, Feasibility Subscale), acceptability (Usage Rating Profile-Intervention Revised, Acceptability Subscale), and satisfaction (Client Satisfaction Questionnaire-8) assessed at the 8-week end of intervention (EOI). The secondary outcome was change in depressive symptoms (Patient Health Questionnaire-8) among those with elevated baseline symptoms. Exploratory outcomes included anxiety (Generalized Anxiety Disorder-7 items), stress (Perceived Stress Scale), perinatal depression (Edinburgh Postnatal Depression Scale), Mother-to-Infant Bonding Scale, and therapeutic alliance (Working Alliance Inventory-Short Revised). This study followed CONSORT (Consolidated Standards of Reporting Trials) guidelines and received Institutional Review Board approval.

Of 267 participants (W-PPMA=144, waitlist=123), mean age was 32.2 (SD 5.1) years; 12.7% (n=34) identified as Hispanic/Latina, 8.6% (n=23) as Black, and 68.5% (n=183) as White. W-PPMA users engaged with the app a median of 9 (IQR 5-24) days across 4 active weeks (IQR 2-7). At EOI, W-PPMA participants reported high feasibility (Usage Rating Profile-Intervention Revised, Feasibility Subscale median 31, IQR 28-34), acceptability (Usage Rating Profile-Intervention Revised, Acceptability Subscale median 30, IQR 28-34), and satisfaction (Client Satisfaction Questionnaire-8 median 26, IQR 24-29). Among those with baseline depressive symptoms (n=97), Patient Health Questionnaire-8 scores improved modestly in the W-PPMA group compared to waitlist (mean change -0.7, SD 3.4, vs 0.0, SD 4.8; Cohen d=-0.16). Exploratory analyses showed a statistically significant between-group difference in Perceived Stress Scale change at EOI (mean difference -1.59, 95% CI -2.4 to -0.73) and a nonsignificant between-group difference in Edinburgh Postnatal Depression Scale change (mean difference -0.51, 95% CI -1.36 to 0.34). Therapeutic alliance (Working Alliance Inventory-Short Revised) was strongest at baseline among Black participants and those from socioeconomically disadvantaged neighborhoods (Area Deprivation Index ≥75). Participants with a high school or General Educational Development reported the highest satisfaction scores, underscoring accessibility across educational backgrounds.

W-PPMA met primary study aims. Preliminary evidence and these findings support W-PPMA, warranting further evaluation in larger trials.