Nudges for Networks: Community-Level Effects of a Health Plan Initiative to Address Social Determinants
Abstract
Research Objective
Health plans are among the many stakeholders seeking to improve health and reduce costs by screening and intervening on social determinants of health (SDOH). Humana Health Plan launched its Bold Goal Initiative in 2015 to test a novel approach for SDOH intervention not only at the individual patient level, but also at the community level through multi-sector partnerships. We examined the community-level effects of this initiative using data from a national, longitudinal survey of community health networks collected during 2012-2018.
Study Design
The National Longitudinal Survey of Public Health Systems was used to measure networks of community organizations that work together in implementing health improvement activities in a national sample of U.S. communities. Changes in the network characteristics of Bold Goal communities were compared with a matched, synthetic comparison group of similar U.S. communities using difference-in-difference analysis focusing on the years before (2012-2014) and after (2016-2018) implementation. Data from each wave of the survey were linked with county-level community characteristics from the Area Health Resource File, mortality data from the CDC Compressed Mortality File, and Medicare spending data from the Dartmouth Atlas of Healthcare. A new surrogate-index methodology proposed by Athey, Chetty and colleagues was used with the full panel of network survey data from 1998-2018 to estimate the possible long-term effects of Bold Goal on mortality and medical spending over 10 years, using network measures as short-term statistical surrogate variables.
Population Studied
The Bold Goal initiative was initially implemented in seven U.S. metropolitan areas with large concentrations of Humana members and employees: Baton Rouge, LA; Broward County, FL; Knoxville, TN; Louisville, KY; New Orleans, LA; San Antonio, TX; and Tampa Bay, FL. The synthetic control group was constructed from a national sample of 280 U.S. communities having at least 100,000 residents that were included in the National Longitudinal Survey of Public Health Systems.
Principal Findings
Cross-sector network density increased by an average of 18.4 percentage-points in Bold Goal Communities compared to 1.8 percentage points in the comparison group (p<0.01). Results were driven by large increases in network participation rates among employers, hospitals, health insurers, physicians, schools, and faith-based organizations in the Bold Goal communities (all p<0.05). Surrogate index results indicated that the program could generate a 5.1 percentage point reduction in all-cause mortality and a 6.4 percentage-point reduction in Medicare spending per person in affected communities if network effects are maintained over 10 years.
Conclusions
A health plan-sponsored initiative can stimulate substantial changes in community networks with the potential to benefit not only enrolled plan members but also the public at large.
Implications for Policy or Practice
The expanding array of SDOH screening and referral initiatives hold considerable health improvement potential, but only if community capacity to address unmet social needs can be strengthened commensurately. Results show that health plan engagement in community-level SDOH initiatives are likely feasible and effective in strengthening community capacity and improving outcomes by expanding community networks. As such, insurers and other sponsors should consider expanded engagement in these initiatives as part of their SDOH strategies.
Primary Funding Source
The Robert Wood Johnson Foundation
