Association for Public Policy Analysis & Management

Literature on the social determinants of health (SDH) has consistently shown that negative social, behavioral, and environmental factors have a substantially larger effect on long-term health outcomes (such as mortality rates and risk of chronic disease incidence) than medical and biological factors. Further, a growing body of literature demonstrates the flip side is true as well, with favorable social conditions positively impacting health outcomes. Due to the nature of SDH (food insecurity, poor housing quality, stress, etc.) these factors disproportionately affect low-income and marginalized populations, contributing to an increase in health disparities.

While researchers and policymakers have begun to intervene on SDH at the individual and community levels, none have examined the role of the household – the socio-physical environment within which individuals spend a majority of their time. Though a multi-sectoral approach for action on the SDH has been called for, obstacles such as competing organizational goals, funding streams, and data collection make integration at a macro-level difficult. However, the household is the point at which different sectors (such as health, education, employment, and housing) interact, and it plays an important role in individuals’ risk exposure, access to resources, and behavior development. Thus, the household exists at a meso-level between the larger social system and individual well-being. The first step in understanding how to intervene at this level is to determine: What does the household socio-physical environment look like, and how can we measure variation across households and over time?

We utilize proprietary data from an innovative household-centered outreach program called NeighborhoodHELP (NHELP), which operates in low-income and immigrant communities throughout Miami-Dade County, Florida. The NHELP data covers a 21-month period (March 2015 through December 2016) and over 500 households. This dataset includes a series of indicators collected during the program intake process, such as household income and employment, pre-existing health conditions, methods of transportation, and housing quality (e.g. leaks, mold, pests). The dataset also includes longitudinal data tracked at each encounter point between NHELP’s outreach workers and the household (number of encounters range from 2 to 15). Longitudinal data includes households’ self-reported “most urgent” social need, services provided by outreach workers, and a “risk score” rating (from 1 to 5) across ten different SDH sectors: health, food security, housing, income, employment, education, transportation, daily activities, technology, and legal support. Our study uses this data to investigate household well-being in terms of urgent social needs and risk scores across the SDH. We also explore variation across households and measure household “churn” in terms of risk and needs over time.

Policymakers interested in building a culture of health or in mitigating social determinants of health often face difficulties in understanding how to implement, measure, and evaluate multi-sectoral interventions. This study focuses on the household as a unique intervention point, and provides actionable information for understanding the household environment, identifying the most at-risk households, and measuring change in risk over time.