Feasibility of an ADAPTive intervention to improve food security and Maternal-Child Health - Project Summary/Abstract The US is facing a maternal and infant health crisis with stark disparities according to race, ethnicity, and socioeconomic status. Food insecurity (FI) affects 30% of pregnancies and is a significant contributor to these health disparities. FI leads to worse health in pregnant people and their children, including an increased risk of gestational diabetes, pre-term birth, and future cardiometabolic chronic conditions. Interventions used to address FI in clinical care settings include 1) referring patients to the Special Supplemental Nutrition Program for Women, Infants and Children (WIC) or providing food via 2) produce vouchers and 3) medically-tailored meals. Although there is evidence for each, patients differ in the support they need to reduce FI. Rather than using a single FI intervention, adaptively allocating interventions could be a more effective, equitable, and efficient approach to improve food security for the large, vulnerable population of pregnant patients with FI. Our long-term goal is to develop an effective adaptive intervention that health systems can use to address FI and improve metabolic health in pregnant patients and their offspring. The objectives of this pilot study are to determine 1) the feasibility of recruitment and retention and 2) the potential effectiveness of the interventions in anticipation of a large, definitive Sequential Multiple Assignment Randomized Trial (SMART). SMART uses experimental design principles to build optimal adaptive interventions. In this pilot study, pregnant patients with FI will be randomized to 1 of 2 first-stage FI interventions: 1) an electronic health record (EHR)-based WIC referral or 2) EHR WIC referral + care navigation. Participants who do not have an improvement in FI will be re-randomized to 1 of 2 second-stage interventions: 1) produce vouchers or 2) medically-tailored meals. In Aim 1, we will determine the feasibility of recruitment and which first-stage intervention is potentially more effective in reducing FI. In Aim 2, we will evaluate the feasibility of re-randomization, data collection, and retention and assess which FI intervention may be the best next step for those who do not respond to the initial intervention. In Aim 3, we will explore heterogeneity in response to the first and second stage interventions by conducting semi-structured interviews with participants. The proposal complements NIDDK’s priority to “understand the impact of metabolic dysfunction on the intrauterine environment and subsequent metabolic health of mother and offspring.” This study is the first to evaluate an adaptive FI intervention in pregnant patients. Given health systems’ focus of integrating FI interventions in clinical practice, an efficacious adaptive intervention could be broadly disseminated. Adaptive interventions can reduce the burden and costs for patients and providers by providing no more treatment than is needed for patients who respond to a less intensive approach and redirecting the saved resources to those who need a more intensive approach. Our team of established investigators has successfully used the methods proposed and has the full support of the health system where the study will occur.
