Estimating Inter-Organizational Network Effects on the Efficiency of Implementing Core Public Health Services



Background: The Affordable Care Act created new incentives for hospitals, insurers, public health agencies, and others to contribute to disease prevention and health promotion activities, potentially changing inter-organizational relationships and expanding implementation of strategies that improve population health. This study uses data from the 1998-2014 National Longitudinal Survey of Public Health Systems to examine: (1) the extent and nature of change in inter-organizational contributions to public health activities; (2) whether network changes attenuate or exacerbate disparities in public health implementation across communities; and (3) how network changes affect preventable mortality and resource use.

Methods: We follow a cohort of more than 350 U.S. metropolitan communities over time using survey data collected initially in 1998 and again in 2006, 2012 and 2014. Local public health officials report on the availability of 20 guideline-recommended public health activities in the community, the organizations that contribute to each activity, and the perceived effectiveness of each activity. We construct network-analytic measures of inter-organizational connectedness in implementing activities (density, degree and betweenness centrality), with a focus on hospitals, public health agencies, and community-based organizations. We link survey data with outcome measures that include county-level cause-specific mortality rates and measures of public health agency expenditures. Fixed-effects models with instrumental-variables are used to estimate changes in preventable mortality and expenditures attributable to changes in network structure, controlling for observable and unmeasured confounders.

Findings: During 2012-14, hospitals increased their implementation of public health activities by 20.1 % in the average U.S. community, compared with an increase of 8.7 % by local public health agencies, an increase of 6.3 % by community-based nonprofit organizations, and a reduction of 6.6 % by state agencies. Disparities in implementation between the top and bottom 20 % of U.S. communities increased by 30.2 % during the full 1998-2014 period, with more than two-thirds of this disparity attributable to changes in inter-organizational network structure. Increases network density and centrality were associated with statistically-significant reductions in infant mortality and deaths due to cardiovascular disease, diabetes, and cancer, and also reductions in public health resource use.

Implications for D&I Research: Highly-connected and integrated inter-organizational structures support improved implementation of recommended public health services, improved outcomes, and lower resource use.

Project: Health Reform Implementation and Comprehensive Population Health SystemsNALSYS Resources and Results
Resource: Download
Authors: Hogg RA, Mays GP, Butler JS