Design, Implementation, and Feasibility–Acceptability of a Text Message–Based Intervention to Prevent Perceived Insufficient Milk Supply

To examine the feasibility and acceptability of a semiautomated text system designed to prevent perceived insufficient milk (PIM) and its consequences. Design and implementation considerations of the text system are also addressed. Within this randomized, controlled trial, participants were assigned to either a PIM-specific text-message support intervention (MILK messages) or a control group receiving general perinatal text-based support. Participants were recruited via online advertising and at prenatal clinics in Pittsburgh, Pennsylvania. We randomized 250 nulliparous women who were between 13 and 24 weeks of pregnancy and who intended to breastfeed. MILK messaging was developed through rigorous pilot research using ecological momentary assessment, focus groups, and consultation with experts in breastfeeding and social marketing. Message content and delivery was grounded in principles of behavioral economics, health communication, and breastfeeding self-efficacy theory. Women in both groups received text messages three–seven times per week from 25 weeks of pregnancy to 8-weeks postpartum. Intervention participants could text keywords for more detailed information and request “help” from an on-call study lactation consultant. We conducted feedback interviews with intervention participants at 8-weeks postpartum. Recruitment occurred for 15 months, with an approach-to-consent ratio of 46%. The sample racial–ethnic composition reflected the recruitment catchment area. Ten women stopped MILK messages and 3 opted to reduce message frequency. There were 45 texts for individualized help (56% on weekends or after-hours). The most commonly texted keywords occurred in Weeks 5–6 and addressed breastfeeding and maternal sleep (n = 106), breastfeeding “schedules” (n = 56), and milk supply (n = 56). Women reported that MILK messages provided reassurance and timely anticipatory guidance. Suggested improvements included extending messaging past 8 weeks, tailoring content based on preexisting breastfeeding knowledge, and providing messaging for partners. Solutions were developed to address unanticipated issues, including message consolidation for delayed birth notification and flags–notifications for certain participant-texted content. Prenatal and postpartum evidence-based breastfeeding support delivered via semiautomated text message is a feasible and acceptable intervention for first-time mothers. Refinements to this and future text systems might be achieved through greater personalization of messaging (e.g., application of machine-learning principles).