Key Takeaways
Extended Takeaways
- The single prescreening question, “Overall, are you feeling happy or sad?” should be used only as a triage step, not as a substitute for the Edinburgh Postpartum Depression Scale, because sensitivity was 65.7% despite specificity of 92.6%.
- Universal Edinburgh Postpartum Depression Scale administration after postpartum week 2 appears clinically important, since sensitivity of the prescreener dropped from 88.5% in weeks 1 and 2 to 45.0% in weeks 3 through 6, meaning some participants with positive Edinburgh Postpartum Depression Scale scores still reported feeling “happy.”
- The app generated substantial clinical workload that should be planned for in implementation: 8,049 mental health screening assessments led to 475 nurse follow-up calls, all intended for same-day outreach using a structured protocol.
- Among actively monitored pregnancies, 13.6% (n=202) had symptoms indicative of perinatal depression, which is close to the nearly 13% prevalence cited for US women and suggests the cohort reflects routine postpartum depression risk seen in practice.
- Escalation beyond monitoring was uncommon but clinically meaningful: app-triggered follow-up resulted in 4 psychiatric hospitalizations, 3 emergency department referrals, 3 well-checks, 58 clinic referrals, and 11 lactation referrals.
- Attrition is a major implementation challenge for digital postpartum depression care, with 30.2% (n=697) lost to follow-up and 5.6% (n =129) withdrawing, leaving 64.2% (n=1,481) monitored through the full 6-week period.
