Development of an empathetic chatbot to monitor perinatal mental health longitudinally - 1 ABSTRACT 2 One in five pre- or postnatal women will develop a perinatal mood disorder (PMD). While temporary and 3 treatable, 50-65% of mothers with PMD symptoms never receive a diagnosis or treatment due to inadequate 4 screening practices. PMDs remain undetected because mothers do not trust clinicians due to fear of judgment, 5 stigma, or even loss of parental rights. Untreated PMDs increase risk for maternal mortality, maternal suicide, 6 preterm birth, low birth weight, and adverse childhood experiences. There is an urgent unmet need for 7 comprehensive, real-time perinatal mental health monitoring to connect women to effective treatment. To meet 8 this need, BetterHope LLC (dba Poisera) is developing a technology platform to systemically monitor, screen, 9 refer, and treat PMDs for women from early pregnancy through 36 months postpartum. The platform includes a 10 mobile health (mHealth) app, where a chatbot conversationally engages mothers in daily reflection exercises 11 about pregnancy, motherhood, and mental health. Each week, the chatbot prompts mothers to complete a 12 state-of-the-art, validated adaptive diagnostic screening for PMDs. When a woman screens positive for PMD 13 symptoms, the Poisera platform connects the woman to a Psychiatric Perinatal Access Program (PPAP), 14 where a behavioral health clinician (BHC) will assess risk, deliver a brief therapeutic intervention, and 15 coordinate long-term treatment. BHCs update her risk profile and treatment information in Poisera’s 16 dashboard, where the patient’s clinician can track treatment progress. PPAPs exist in 30 states. PPAPs are 17 proven to improve PMD symptoms and guarantee continuity of care. Using a chatbot to longitudinally monitor 18 and screen for PMDs is an innovative approach because chatbots promote honest disclosure and increase 19 adherence. Further, the Poisera platform transforms the current, inadequate clinical practice for PMD 20 screening and treatment to one where clinicians are confident their patients receive effective treatment at PMD 21 onset because the information is seamlessly integrated into their clinical workflow. The long-term goal of this 22 research is to eliminate adverse health consequences from PMDs. In this Phase I SBIR, we will demonstrate 23 the Poisera platform meets the needs of both pregnant women and clinicians. In Aim I, we will demonstrate 24 that pregnant women find the Poisera app and chatbot usable. In Aim 2, we will demonstrate obstetric 25 clinicians find the Poisera dashboard usable in their clinical workflows. In Aim 3, we will conduct a two-arm 26 feasibility randomized controlled trial (RCT) to measure screening adherence and determine dropout behavior. 27 These outcomes will inform the design of our Phase II RCT, where we aim to demonstrate the Poisera platform 28 cost-effectively improves PMD outcomes for women from pregnancy to 36 months postpartum. Demonstrating 29 effective outcomes on these measures will provide us with the validation needed to sell the Poisera platform to 30 health systems and health insurers, a $2.8 billion annual market size. 31
