Social Vulnerability of Older Adults and the Risk of Medical Hospitalization
Funded Grant
Overview
Affiliation
View All
Overview
description
Project Summary/Abstract Decades of research establish that social determinants of older adults affect their health, yet, we lack ways to implement this knowledge. We fall short without a coherent synthesis: What basic social determinants should an investigator include in their aging cohort study? How can a Medicare accountable care organization include social determinants to identify at-risk patients in the population for which they are accountable? Lacking practical ways to incorporate social risk factors, risk models often exclude social risk factors, and in doing so, risk creating biased estimates. For practical applications in research, population health, and policy settings it is useful to identify a subset for social determinants that predict risk efficiently. Prior attempts have fallen short by not accounting for the unique social determinant of older adults. To address this gap, we propose a study to develop a social vulnerability index – a summary measure reflective of the social risk factors of older adults. We will test the social vulnerability index’s ability to predict medical hospitalization, a common health event among older adults, and assess the index’s ability to improve traditional risk models. The central hypothesis is that social risk factors – specifically aging-associated social risk factors – will significantly improve the predictive model discrimination of traditional risk models. This hypothesis will be tested using the nationally- representative sample of older adults in the Health and Retirement Study Medicare-linked cohort. We have two specific aims: 1) Develop a social vulnerability index to predict medical hospitalization and 2) Assess the ability of the social vulnerability index to improve upon traditional comorbidity-based prediction models. This project is innovative and significant because it will develop a summary measure of social vulnerability that can be used to improve prediction. This will, in turn, will improve the quality of risk models used in research, population health, and health policy.
