This application aims to address the unmet critical need to identify, define, and quantify predictors of clinical outcomes in pregnant people, neonates, and infants. Despite the wealth of available data, these frequently excluded populations are grossly understudied, resulting in possibly unnecessary empiric treatment exposures. Major contributors to this problem are underpowered studies of “small” local data sets and practices that are labor intensive and/or siloed. Use of the vast data generated across multiple sites is limited to narrowly focused research studies, limiting the scope of variables that are explored, and the ability to leverage evolving artificial intelligence efforts. The lack of timely, high-quality, forward compatible, and efficient surveillance data prevents the dissemination of key exposures and outcomes that impact pregnant people, neonates, and infants. To address this need, we aim to leverage our institutional expertise at Johns Hopkins in disseminating the infrastructure and architecture necessary to utilize real-world data. We will utilize the OHDSI OMOP CDM (Observational Health Data Sciences and Informatics Observational Medical Outcomes Partnership, Common Data Model). OMOP, a CDM based upon standard clinical terminologies enables extraction, ingestion, and collation of variables of interest into an observational research registry which has the capability for data storage, security, analysis, and transfer among participating sites. Our central hypothesis is that the OMOP CDM will facilitate data collection in a standardized and efficient manner that enables reproducible research in phenotype definitions and analysis, yet is timely and readily computable, scalable, and deployable across institutions and departments for pregnant people, neonates, and infants. Our outstanding multidisciplinary group of data scientists, computer scientists, biostatisticians, epidemiologists, informaticians, physicians, and patient advoc
ates from Johns Hopkins University (JHU), University of Alabama Birmingham (UAB), and the Critical Path Institute (C-Path) CURE Drug Repurposing Collaboratory (CDRC) propose the following three specific aims: Aim 1a. Utilize the OMOP CDM to advance interoperability and data infrastructure for pregnant people and infants. We will use standardized mechanisms developed at Johns Hopkins and the OHDSI community to improve and increase data interoperability, surveillance, and data collection. Aim1b. Develop and revise data schemas to aid investigators from Components A, B and the CDC to study longitudinal mother-baby dyad outcomes by sharing our expertise and code for mother-baby linkage allowing investigation of public health issues that impact pregnant mothers, neonates, and infants. Aim1c. Utilize existing electronic health record data to prepare automated datasets for analysis and reporting where the OMOP CDM will use existing health care vocabularies and ontologies to automate curation of public health datasets for timely reporting of key exposures and outcomes which impact pregnant people, infants, and neonates. Aim 2. Provide a replicable and scalable approach to implement the OMOP CDM via educational materials, technical assistance by disseminating OMOP centric educational materials and providing technical assistance to partners which will result in increased data processing from sites awarded components A and B, and increased awareness by CDC of the data, data sharing, and data transformation processes for public health reporting. Aim 3. Create metrics and processes to evaluate data flow and data quality from clinical and health department sites and identify opportunities for improvement by deploying data quality dashboards which will help generate high quality, high fidelity, timely real word data that can be leveraged for improved surveillance and swift response to emerging threats during pregnancy to pregnant people, neonates, and infants.
