Harnessing Artificial Intelligence with Community Engagement to Devise Precision Public Health Strategies to Reduce Vaccine Hesitancy among Pregnant and Lactating Refugee Women

Harnessing Artificial Intelligence with Community Engagement to Devise Precision Public Health Strategies to Reduce Vaccine Hesitancy among Pregnant and Lactating Refugee Women Harnessing Artificial Intelligence with Community Engagement to Devise Precision Public Health Strategies to Reduce Vaccine Hesitancy among Pregnant and Lactating Refugee Women The health of pregnant and lactating (P/L) refugee women are endangered due to their extremely low COVID-19 vaccination rates. To provide efficient and effective vaccination outreach to P/L refugee women, precision population health strategies applied through a racial equity lens are crucial. This community-engaged population health research incorporates the use of Artificial Intelligence (AI) and machine learning techniques to identify patterns in pregnant and lactating refugee womens COVID-19 vaccine-related behaviors. It contextualizes the AI findings with in-depth interviews with pregnant and recently pregnant refugee women and through iterative community stakeholder engagement. Together, we will combine these mixed methodologies to yield precision population health strategies to aid in the prevention of COVID-19 among pregnant and lactating refugee women. The research will also demonstrate the power of the integrative multidisciplinary approach to precision population health, which will open a new direction in this field. This approach can be used by other healthcare systems to create precision population health strategies. Objectives for this Study: Objective 1: Discover risk factors of vaccination inequities and perform risk stratification via AI data mining. We will analyze longitudinal health, syndemic and vaccination data, and SDoH of ~6,000 P/L female patients including ~900 refugees. Using machine-learning methods, we will model the relationships between vaccine uptake and COVID-19 syndemics conditional on both refugee and P/L status. Using these models, we will discover risk factors for vaccine hesitancy and predict outcomes of potential interventions. We will cluster patients into subgroups, estimate individual-level and subgroup-level risks, and identify PDs in each subgroup. Objective 2: Discover barriers and facilitators to COVID-19 vaccination via semi-structured interviews. We will sample PD refugee women (identified in Objective 1) who were either vaccinated during pregnancy (n=15) or while lactating (n=15), and vaccine-hesitant P/L refugee women (n=15) receiving care in one healthcare system. We will conduct semi-structured interviews with these women to elucidate their concerns about COVID-19 vaccination and factors influencing their decision-making to become vaccinated. Objective 3: Work with community stakeholders to iteratively triangulate discoveries from AI models with qualitative data to contextualize the findings and inform strategies to decrease P/L refugee women's vaccine hesitancy.