Perinatal depression, defined as a depressive episode during pregnancy or in the first year postpartum, is the most common complication of childbirth, affecting up to 20% of peripartum women globally. An enormous gap exists in supporting women experiencing perinatal depression. Mobile health (mHealth) interventions such as interactive SMS text messaging with healthcare workers (HCWs) have been proposed as resource-efficient, accessible adjuncts to in-person care. Realizing the full public health potential of mHealth for mental health will require strategic use of automation and empiric definition of interventions that dynamically adapt to user needs. This K18 mentored career enhancement award aims to support Dr. Keshet Ronen, an epidemiologist with multidisciplinary training, to become an expert in analysis of mHealth communication and development of adaptive mHealth interventions. Building on Dr. Ronen?s established expertise in development and evaluation of mHealth interventions using SMS and social media, and supported by a team of experienced mentors, the following research and training aims are proposed. Training plan: Through didactic coursework, individual mentor meetings, seminars, and conferences, Dr. Ronen seeks to accomplish the following Career Enhancement Goals. (1) Gain proficiency and experience using natural language processing and machine learning methods to analyze mHealth communication. (2) Gain proficiency and experience in design of an adaptive mHealth intervention. (3) Deepen understanding of mHealth intervention design for mental health. (4) Enhance skills in international study implementation and clinical trial design. The proposed training plan will augment Dr. Ronen?s prior training and experience and allow her to complete the proposed Research plan: Mobile WACh is a unique interactive SMS messaging platform that has been shown to improve maternal-child health and whose impact on perinatal depression is currently being evaluated in a randomized controlled trial in Kenya, Mobile WACh Neo. We propose to leverage a dataset of >100,000 SMS messages with >3000 peripartum women in previous and ongoing Mobile WACh studies to (1) develop a predictive model that can detect client SMS indicating elevated depression symptoms, and (2) identify HCW message characteristics associated with improvements in depression symptoms. (3) Models from Aims 1-2 will be implemented in the Mobile WACh software to develop an adaptive version of Mobile WACh Neo that flags concerning messages and guides HCWs on SMS composition. A pilot study of Adaptive Mobile WACh Neo will be conducted, nested within the Mobile WACh Neo randomized controlled trial in Kenya (R01HD098105), to evaluate its acceptability and preliminary impact on time taken for HCWs to respond to client messages. Collectively, these activities will enable Dr. Ronen to become a leader in the study of adaptive mHealth interventions to support maternal mental health in resource-limited settings.
Depression is the most common complication of pregnancy, but it is undertreated due to lack of mental healthcare resources and barriers to in-person care. Short message service (SMS) text messaging is a promising tool to support mental health in the high-risk peripartum period. This project aims to use natural language processing and machine learning to develop an adaptive SMS intervention that supports healthcare workers to deliver effective SMS messages to support perinatal depression.
