Development and feasibility of a novel mhealth resource for food insecurity: Mixed methods cohort study

Background: Pediatric clinical practice guidelines recommend identifying and addressing food insecurity (FI) as part of routine care. However, methods for health systems to connect families experiencing FI to community food resources are lacking. Confidential SMS text messaging can increase equity in resource delivery, is user-friendly, is aligned with caregiver preferences, and is feasible for health systems to implement. Despite the promise of this approach, SMS text messaging has not been widely tested in pediatric settings. 

Objective: This paper details (1) the process of developing a novel, mobile health intervention to help families access local food resources and (2) results on reach, engagement, usability, and acceptability of the intervention following a 1-year pilot. 

Methods: We designed and evaluated an automated SMS text messaging system that delivers geographically tailored food resource information to families with FI after hospital discharge at a single US children’s hospital. English- and Spanish-speaking caregivers of hospitalized children with a positive FI screen documented during clinical care were included. Caregivers received a food resource text message 1 and 4 days postdischarge. In addition, 2 subsequent text messages asked about reach and engagement. We used system-reported (primary) and caregiver-reported (secondary) measures of reach and engagement and caregiver-reported resource connection as a preliminary measure of effectiveness. We assessed usability (Simplified System Usability Scale [SUS]; >75 indicates good usability), acceptability, and caregiver preferences for resource provision through semistructured interviews among a subset of caregivers (20 English-speaking and 11 Spanish-speaking caregivers). 

Results: Of 194 patients with a positive FI screen during the study period, 187 (96%) spoke English or Spanish and were included in the cohort. Primary, system-reported measures indicated that the food resource message successfully reached 175 (94%) participants; of these, 102 (58%) engaged with the text messages in some way, with 65 (37%) clicking the link and 92 (53%) responding to a text message. Among the subset of text message respondents (n=92), 88 (96%) reported receiving the resource message, 83 (90%) read the message, and 42 (46%) used the information to search for food resources. Among the subset of interviewed caregivers (n=31), the median SUS score was 86.1 (IQR 66.7‐91.7); 97% (30/31) of caregivers felt the intervention was acceptable. Caregivers preferred receiving food resource information via text message rather than paper handouts because it felt more accessible. 

Conclusions: Providing automated, geographically tailored food resource information via text message to families with FI after hospital discharge was feasible, the information was usable, and the delivery mode was acceptable to families, with SMS text messaging preferred over paper handouts. SMS text messaging offers a promising low-intensity approach to social resource provision for health systems. Future research should assess effectiveness and strategies to increase uptake in clinical care contexts.