New Jersey’s Bottom-Up Data Partnership for Population Health

Concern with the social drivers of health — housing, education, income, healthy food access, criminal justice, and more — mushroomed over the last decades and grew further in the wake of the COVID-19 pandemic, economic upheaval, and a racial justice reckoning. Yet, traditional “silos” in government operate as barriers to progress on the social drivers, especially when progress demands working across sectors. Integrated public administrative data — from Medicaid claims, cash benefits, vital records, homeless services, and other important domains — can help support “smart government” with data-driven, cross-sector solutions. However, the fragmentation of these systems, federal and state privacy and research subject protection regulations, bureaucratic culture that eschews data sharing, and myriad technical considerations pose major challenges to maximizing the effective use of these data.

Despite these obstacles, some states have linked administrative data to produce actionable evidence for policymakers and administrators while preserving privacy and confidentiality safeguards. States like Indiana, Rhode Island, and South Carolina have used integrated data to help officials address some of the most pressing health problems facing their residents. Indiana’s initiative laid the groundwork for actionable COVID research and enhanced vaccine distribution. Rhode Island developed data linkages and models to foster a cross-agency response to the state’s opioid overdose and addiction crisis. South Carolina employed integrated administrative data to enable mental health practitioners to better serve the state’s rural residents. These and other examples of successful use of administrative data are encouraging.

However, there has been little exploration of the governance models that underpin successful data integration projects that promote work to drive creative policy solutions and overcome bureaucratic inertia. In this blog, we describe one promising project governance model undergirding an ambitious initiative in New Jersey called the Integrated Population Health Data (iPHD) Project.

Innovative Governance

The iPHD is a state-initiated university partnership that aims to improve population health and the cost-effectiveness of state programs. The initiative features a formal governance structure with four pivotal characteristics.

• It relies on a bottom-up process whereby, within parameters established by a governing board, external researchers, rather than state officials, identify specific study questions and research designs. In addition to setting broad research priorities, the governing board assesses the merits of proposals and decides whether to approve them. Designated in the enabling legislation, the board consists of four cabinet-level officials, two Rutgers representatives, and four public members appointed by the governor, senate president, and speaker of the assembly. The statute calls for public members to be selected to represent consumer, privacy, security, and research perspectives.
• It is intended to be ongoing. Rather than focus on one-shot requests by researchers or public officials, the iPHD project established a periodic process for proposal submission for the indefinite future.
• It rests on a full-throated university partnership which assigns responsibility for implementation to the Rutgers Center for State Health Policy at the state’s flagship public university.
• State policymakers expect the iPHD to be fiscally self-sustaining with user fees from approved projects and grant funds generated by the Rutgers Center.

Enabling legislation. The legislation that established the iPHD was derived from a proposal by the New Jersey Good Care Collaborative, organized by the Camden Coalition of Healthcare Providers, which included care providers and other stakeholders from across the state.   The collaborative stressed the benefits integrating physical and behavioral health services and addressing medical and social complexity. Emphasizing that “good care requires good data,” the collaborative released a brief that outlined the model, attracting bipartisan support. The state legislature, with Democrats in control of both houses, passed authorizing legislation in 2016 and Republican Governor Chris Christie promptly signed it. The law went to great detail in specifying the structure, functions, and authorities of iPHD governance.

A strong state-university partnership. The New Jersey model builds on a long-standing state-university partnership with the Rutgers Center through which state agencies facilitate or contract for analytic work deemed important to the state, including evaluations of the New Jersey Medicaid 1115 demonstration waivers and investigator-initiated research on services for high-need populations and health equity questions. The iPHD legislation assigns administrative responsibility for the program to the state Department of Health, but calls on the Rutgers Center to receive, integrate, and manage pertinent administrative data and to support the board and the research application process.

Initial priorities. The iPHD governing board has identified four broad initial priorities.

• Informing state’s approach to addressing the opioid epidemic
• Improving maternal and infant health
• Assuring access to physical and behavioral health services and addressing social determinants of health
• Supporting the state’s response to COVID-19 and other public health emergencies

The board has thus far approved integrating five data sets from the state Department of Health – birth records, death records, all-payer hospital billing data, emergency medical services records, and COVID-19 surveillance data. It anticipates adding additional data sources in future years, possibly including Medicaid claims and social services program records. The law delegates responsibility for determining whether it is legally permissible to make data available through the program to the State Office of Attorney General, and while agencies housing data sets are not required to contribute them, the model seeks to promote a culture of data sharing.

Potential and Challenges

The iPHD is an ambitious initiative that institutionalizes an enduring state-university data partnership and seeks to encourage a culture of evidence-based policymaking in state government. The first data in the iPHD are from the Department of Health, which has a legacy of sharing data with researchers. A key test of the initiative’s ultimate efficacy will be the degree to which other agencies adopt a similar posture. While the initiative holds great potential, two additional challenges remain.

Lengthy start-up period. First, the time needed to make this model operational greatly exceeded the initial expectations of policymakers. After more than 25 governing board meetings, the project is just now poised to be operational. The initial absence of start-up funding slowed implementation, and the pandemic raised a host of challenges to timely implementation.   Addressing legal issues involved in formalizing the partnership between the state and university also stoked delay.   Initiating the project involved review by state cyber security authorities and necessitated state attorney general’s office review of concerns related to data privacy and security as well as the legal liability of the university, governing board, and the Department of Health.

Raising funds. Second, the legislative requirement that the iPHD be self-supporting financially presents an enduring challenge. The initial iPHD budget is $550,000 per year, covering the cost of staff time, secure data processing infrastructure, and other needs. To generate these revenues, the Rutgers Center will seek grant support. It will also have to thread the needle in setting the fees it charges to researchers seeking to use the state’s administrative data. Fees need to be high enough to generate meaningful revenue without erecting undue barriers to data access. The iPHD law requires that the fee schedule be set on a breakeven basis and thus Rutgers has established fees complying with NIH shared facilities guidelines.

Fueling Responsive, Effective Policy

The iPHD governance model is, to the best of our knowledge, unique among the 50 states. We believe that it has great promise for yielding important outcomes. As the iPHD implementation progresses, it will be important to assess whether the university partnership model can yield better outcomes than the more common model of data integration and analysis by state administrative units. Answers to this and a spectrum of related questions could kindle innovation in governance models to fuel responsive  and effective policy solutions that enhance population health and health equity.