EnCoRe MOMS: Engaging Communities to Reduce Morbidity from Maternal Sepsis

Maternal sepsis is the second leading cause of maternal death, major cause of morbidity, and preventable in most cases. Labor, birth, and postpartum are periods of increased sepsis risk. Yet few evidence-based interventions exist. With our extensive community partnerships and community organized leadership advisory board (CoLAB), EnCoRe MoMS: Engaging Communities to Reduce Morbidity from Maternal Sepsis will address three highly related specific aims: (Aim 1) Develop, implement, and evaluate a community-informed maternal sepsis bundle in 4 NYC hospitals; (Aim 2) Develop algorithms to optimize prediction of sepsis around delivery and postpartum; and (Aim 3) Conduct a co-design process and qualitative study to explore the experiences, needs, and perceived solutions for maternal care continuity, sepsis prevention, and promotion of postpartum care. In the UG3 phase we will establish robust community engagement and research infrastructures to: Aim 1a: Design a comprehensive obstetric sepsis bundle that i) applies and optimizes standard evidence-based components of readiness, recognition, response, and reporting ii) incorporates multidisciplinary obstetric provider training, and iii) integrates screening for unmet needs. Aim 2a. Create a rich electronic health records (EHR) database from the Perinatal Research Consortium (PRC). Aim 2b. Collate neighborhood-level datasets characterizing community level factors. Aim 3a. (3a.1) Refine our CoLAB and co-design process; (3a.2) Conduct in-depth individual patient interviews (IDIs) and focus group discussions (FGDs) with community and hospital stakeholders from one site to explore the lived experiences and perspectives of factors relating to care access, quality, outcomes, and solutions for care continuity. In the UH3 phase, we will engage community to implement our maternal sepsis care model, analyze results, and translate findings. Aim 1b. Implement our comprehensive obstetric sepsis bundle and evaluate its effectiveness using process and outcome measures Aim 1c. Define patterns in EHR of provider response to suspected sepsis, pre- vs post-bundle implementation; analyze associations between provider response variation and outcomes Aim 2c. Harmonize patient-level EHR and neighborhood-level datasets and use machine learning models to analyze the individual and joint contributions of patient and neighborhood factors to optimize sepsis risk prediction within the PRC sample Aim 3b. (3b.1) Complete qualitative patient IDIs and stakeholder FGDs for the three additional hospital sites; (3b.2) Co-design an integrative supportive care model, with our community partner co-lead, CoLAB, and results from other aims, that entails maternal sepsis community engagement, care linkages, education, and services. Our resulting model can be scaled to hospitals and communities with differing landscapes and applied to other preventable causes of severe maternal morbidity.