Place-Based Influences on Mortality among People Experiencing Homelessness

PROJECT SUMMARY/ABSTRACT Research Challenge: More than 582,000 people in the US experience homelessness each night, and a staggering 1.25 million people enter the shelter system each year. People experiencing homelessness endure extreme adversity, drastically increasing their risk of dying young. Although the response to the homelessness crisis primarily occurs at the local level, local data on homeless mortality is largely unavailable. The limited existing data show substantial variation in homeless mortality rates by place, suggesting that place-based factors may impact mortality risk. However, methodological differences in how existing data are collected and compiled make valid cross-location comparisons difficult. As a result, at present there is no way to rigorously test how local place-based factors—such as policies, service availability, and economic conditions—relate to homeless mortality, which impedes the ability to identify promising local policy solutions or interventions. Objective & Specific Aims: The overall objective of the proposed research is to identify policy-sensitive drivers of homeless mortality. This objective will be achieved through the following specific aims: (1) calculate homeless mortality rates for each US state and Continuum of Care (local planning bodies that coordinate homeless services for a specific city, county, or small group of counties), (2) document frontline providers’ perceived place-based risk and protective factors of homeless mortality, and (3) investigate the association of place-based factors with homeless mortality. Approach: The proposed study will use a sequential mixed methods design to provide insight about homeless mortality and its place-based predictors. In Aim 1, restricted-access US Census Bureau data on people experiencing homelessness will be linked to Social Security Administration mortality records to estimate area- level homeless mortality rates. In Aim 2, 48 key informant interviews will be conducted with frontline providers of homeless and health care services to identify perceived place-based predictors of homeless mortality. In Aim 3, using homeless mortality estimates from Aim 1 and potential predictors identified in Aim 2, multi-level modeling will be used to test the association between place-based factors and homeless mortality. Impact: This project will yield the first comparable state-level and Continuum of Care-level estimates of homeless mortality as well as identify place-based correlates of mortality. This proposal, which crosscuts the missions of NIA, NICHD, and NIMHD, would have positive impact by identifying potential policy avenues that could bring to bear improved health and reduced mortality among people experiencing homelessness.