Prior research shows persistently greater burden of illness and poorer access to care among racial minorities and low-income individuals; older adults with chronic illness are no exception. Although evidence on the social determinants of health has grown, many questions remain about the complex interplay of individual, organizational, and environmental factors in health disparities generally, and in healthcare access and outcomes in particular. This developmental study will assess the feasibility of using a state-of-the-art systems science method ? Agent-Based Modeling and Simulation (ABMS) ? to examine and address the complex array of factors contributing to racial and socio-economic disparities in healthcare access and outcomes among chronically ill older adults. We will capitalize on an existing dataset that includes three years of health care claims on a cohort of 278,820 New York City-dwelling Medicare beneficiaries ages 65 and older with diabetes, heart failure and/or hypertension linked with data on neighborhood characteristics (e.g. supply of primary care and federally qualified health centers, neighborhood walkability, and transit access). Applying ABMS to the existing data could establish the groundwork for developing a decision-support tool to guide interventions in health disparities; however, it is essential to first identify the most influential factors in each level of analysis and how they interact. Focusing on the three levels of analysis needed to build a prototype agent-based model, the specific aims are to assess the feasibility of: (1) developing an individual-level agent model with the available data on demographics (e.g. race, age, low income), diagnoses, patterns of outpatient healthcare use and outcomes (e.g. hospitalizations for ambulatory care-sensitive conditions, emergency department visits); (2) modeling the organization of health services across the care continuum by combining data on sequencing and intensity of service delivery and health care supply; and (3) developing a neighborhood-level model that includes socio-demographic composition, walkability, public transit, safety, amenities and social services as environmental determinants of healthcare access and outcomes. This study will determine whether agent-based modeling is feasible for capturing the multiple dimensions of phenomena that influence healthcare access and outcomes among vulnerable populations. This work would lay a foundation for a larger study that includes simulation of promising interventions using an expanded dataset that could include both Medicare and Medicaid claims for services delivered since the enactment of the Affordable Care Act. The long-term goal of this research is to develop a model that serves as a decision support tool to guide policy and practice interventions at the provider and/or community level to improve the health of underserved populations of older adults with chronic illness and, in turn, reduce health disparities.
Among the growing numbers of older adults with chronic illness, racial minorities and low-income individuals frequently have poorer access to care and worse health outcomes overall. This developmental study will assess the feasibility of using an advanced method ? Agent-Based Modeling and Simulation (ABMS) ? to investigate the complex factors contributing to racial and socio-economic disparities in the health of older adults with chronic illness. The long-term goal of this research to develop a decision support tool that can be used by policymakers and health care administrators to guide interventions to improve the health of underserved populations of older adults with chronic illness and, in turn, reduce health disparities.
