New EHR-based multimorbidity index for diverse populations across the lifespan: development, validation, and application

点击翻译按钮获取中文摘要

Project Summary / Abstract Virtually all U.S. adults will develop multimorbidity (coexistence of multiple chronic conditions) by late adulthood. The sequelae are substantial: vulnerability to acute illness, disease exacerbation, hospitalization, disability, poor health-related quality of life, and mortality. Despite this, there are no viable, patient-centered measures for multimorbidity in the electronic health record (EHR) that include a comprehensive inventory of conditions based on their impacts on physical functioning in community-dwelling adults and are thus broadly applicable for the general population. The absence of such tools impedes systematic efforts to develop effective interventions for patients with multimorbidity. To bridge these gaps, this proposal aims to develop and validate a robust, clinically relevant, readily-available EHR-based multimorbidity-weighted index (eMWI) that accurately ascertains disease presence using EHR data and is applicable for diverse populations across the lifespan. The central hypothesis is that a comprehensive multimorbidity index that weights conditions based on their impacts on physical functioning can more precisely quantify multimorbidity and provide a better model fit to predict key health outcomes than prior measures. This hypothesis is strongly supported by our preliminary results using large national surveys and survey-linked claims data, in which we rigorously developed and validated a comprehensive set of 91 chronic conditions weighted by their average impacts on physical functioning over the disease life course, thus incorporating illness burden and physical functioning into a clinically meaningful measure applicable for the general population. As a transformative step for multimorbidity measurement in patient care, population health, and research using EHR data, the team aims to 1) improve multimorbidity measurement by more accurately ascertaining disease cases, and merging these with validated physical functioning disease weights to create a new patient-centered eMWI applicable to diverse populations; 2) assess the validity of eMWI via its association with key clinical outcomes: multimorbidity progression, hospitalization, and mortality; and 3) test the applicability of eMWI to national population health and policy by applying it to evaluate the risk of severe and fatal COVID-19 among vaccinated vs. unvaccinated adults based on their multimorbidity. This study uses large, diverse EHR data from 6 California health systems (>6 million adults) with unique data linkages to census data, and the largest, most nationally-representative National COVID Cohort Collaborative (N3C) dataset (>5 million COVID cases). The results will yield a new, validated, patient-centered multimorbidity index for EHR data – the eMWI – to help guide clinical decisions, population- health management, policy, and research for diverse populations. The team anticipates that eMWI can directly impact future practice and outcomes in which multimorbidity and functional status impact everyday treatment decisions and outcomes. eMWI will be readily available in standardized code for other EHR data. Overall, these results will improve the quality of care and health outcomes for diverse, aging adults with multimorbidity.