A Pragmatic Trial of a Multilevel Community-Centered Intervention to Reduce Pregnancy Related and Associated Morbidity and Mortality and Disparities: A Study Protocol

Introduction: Maternal morbidity and mortality in the US disproportionately affect some women, including the Medicaid-insured, rural, and Black, Hispanic, and Native American populations. The drivers of disparities are complex and there is an urgent need for multilevel interventions focused on pregnancy-related and associated morbidity and mortality (PRAMM). This study will be the first large-scale test of Community Health Worker (CHW) inclusive home visiting and provider/practice/system level improvement strategies that support empowered patients, pro-active providers, and integrates community and clinical care designed to reduce PRAMM and disparities among women who are disproportionately affected. Methods and analysis: The primary study outcome will be rates of PRAMM - a composite outcome including pregnancy-associated and pregnancy-related morbidity, severe maternal morbidity, and mortality - and disparities among women disproportionately affected. A quasi-experimental, stepped wedge design will be used. Participants will be Medicaid-insured women in three Michigan counties observed during pregnancy, at birth, and up to one year postpartum, who give birth between 2021-2028 (>101,000 births). Individual health outcomes will be assessed at three steps of the stepped-wedge design. Analyses will use a statewide linked data system including all Medicaid birth and death records, Medicaid claims, and other program data. In the context of Michigan extending pregnancy-related Medicaid, the study will evaluate the impact of the proposed multilevel intervention vs. usual care from early pregnancy through 12 months postpartum. The study was developed with full community engagement and participation including shared leadership with academic and community-based Principal Investigators representing intervention communities. Ethics and dissemination: An Institutional Review Board determined the study exempt from the human subject's regulations, for the following reasons. Two of the three intervention components will not interact with human subjects to collect data - we will only rely on secondary research using data or biospecimens not collected specifically for this study, and the data will be provided without identifiable information by someone without any role in this research study except providing the data. The third intervention component, medical provider surveys, is a benign behavioral intervention where information will be recorded using methods that prevent subjects' identities from being readily ascertained. Results will be published in peer-reviewed journals following EQUATOR guidelines.