How Harnessing the Power of Data Can Improve Americans’ Health

David Hyun, M.D., and Rachel Zetts, MPH, lead The Pew Charitable Trusts’ recently launched state health solutions project, which focuses on using data to improve health outcomes, narrow health disparities, and reduce health care costs.

This interview has been edited for clarity and length.

Q. What is Pew’s new state health solutions project hoping to achieve?

David Hyun: The goal of our state health solutions project is to harness the potential of currently siloed health data to improve the health outcomes of specific populations and, ultimately, public health overall.

Rachel Zetts: We’ll focus on getting different stakeholders and state agencies that have specific sets of health data to forge sustained data-driven partnerships that will allow them to not only share what they have but also analyze and act on that data together. State public health and state Medicaid agencies in particular have complementary data sets and are well-suited to working together in this way.

Q. Why did Pew decide to work on this issue?

RZ: Our state health solutions work is actually part of a broader emphasis on public health data at Pew. We also have a public health data improvement project, which focuses on improving the reporting and use of data from health care providers to help state public health agencies better detect, prevent, and  treat diseases. Together, these efforts aim to leverage the untapped potential of the data that’s out there to improve public health in the United States.

Q. You mentioned wanting to help forge data-driven state public health partnerships. Could you describe what those are?

RZ: The data-driven partnerships that we’ll be working to set up will help states establish infrastructure and processes for combining and supplementing their existing health datasets. By bringing together distinct but complementary groups of data from different state agencies, policymakers and state leaders can do more to maximize the utility of data they already have. Being able to draw on and analyze these larger pools of information can provide a more complete picture of health needs and priorities within a state, and inform and enable coordinated health interventions for specific populations that are more targeted and effective.

Q. Can you give an example?

DH: Say a state’s Medicaid office has data that shows maternal mortality is much higher among specific populations or in certain zip codes. If that data was accessible to the state’s public health department, it could help the department better target efforts to connect pregnant women with prenatal care via mobile clinics, for example. It could help public health staff focus their community-based education campaigns or hone the effectiveness of other interventions. Then, Medicaid could work to supplement these public health efforts, based on the needs of the at-risk populations, by incentivizing specific improvements to clinical practices through reimbursement, for example, or updating continuing education requirements for health care providers. The idea is to create a cyclical, reciprocal collaboration that is continually leveraging shared data to improve health care quality and value.

RZ: Yes, and that’s just one example. States may focus their data-driven partnerships on a specific disease or condition to start, but ultimately we want these partnerships to be sustainable, disease-agnostic mechanisms that can be used to tackle the full spectrum of population health issues.

Q. Why focus on state-level partnerships instead of a nationwide effort?

RZ: There are a couple of key reasons. First, a lot of the public health and health care data in the U.S. is collected at the state level. We saw that very clearly during the COVID-19 pandemic, when organizations like CDC (the Centers for Disease Control and Prevention) and Johns Hopkins had to aggregate and harmonize data from different states and localities to knit together a national landscape assessment. For better or for worse, that’s the way things are currently structured, and that's the framework we have to work within.

Second, there is great variation across states in terms of top health issues, resources available, etc., so a state-level focus allows for more targeted and customized interventions that align with the needs of different patient populations.

DH: That said, even though our focus is on the state level, we hope—and expect—that over time these data partnerships will have a collective positive impact at a national level.

Q. Why isn’t this kind of data-sharing already happening?

DH: Well, it actually is happening in certain instances. Rhode Island’s Department of Health and the state’s Medicaid agency, for example, established an ongoing data-sharing agreement, and the two organizations worked together to gather and assess data on the Home Asthma Response Program (HARP), an evidence-based intervention designed to reduce preventable asthma emergency department visits and hospitalizations among high-risk children. The data showed that after participating in HARP for a year, patients had about a 75% reduction in asthma-related hospital and emergency department costs. Based on this and other positive data, Rhode Island is now working to expand access to HARP by making it a covered Medicaid benefit. We’ve also seen good outcomes from these kinds of data partnerships focused on vaccination and other health issues.

But, since the data—as well as the agencies themselves—is often so siloed, and given the typically inconsistent nature of public health funding, these partnerships have often been one-off efforts or have fizzled out.

Q. So how will you overcome those challenges?

RZ: Our goal is to create partnerships that have staying power and that are easily adaptable, so that states have established collaboration mechanisms in place and can jump right in and tackle priority population health issues when needed.

To do this, Pew plans to create a blueprint to help guide states in setting up these partnerships. The blueprint will identify best practices—including approaches to structuring these kinds of partnerships, data governance, and leveraging shared data to support health care quality improvement—with the goal of helping to streamline the process and facilitating the creation of sustainable, long-term data partnerships that help states avoid reinventing the wheel.

Q. Are there privacy concerns with using Medicaid claims data for public health?

DH: Any time you’re working with health care data, privacy is essential, and it’s a top priority as we work to build these data-driven partnerships. We’ll be working with legal experts and data privacy experts to establish and share best practices as part of our blueprint, which we envision will include concrete guidance and customizable resources, such as templates for master data-sharing agreements, protocols for data governance, and policies for HIPAA compliance.

Q. What are you most excited about as this new project begins?

DH: It’s hard to pick just one thing. I guess I’m excited about how much potential there is here. We’ve seen the promise of data partnerships, and we know it’s possible to make a big and meaningful difference. That’s exciting.

RZ: I agree. If we can reach our goals and broaden the scale and sustainability of these state-based public health data partnerships, the result is going to be improved quality of life for many people—and lives saved, too. It’s hard to think of a better reason to get up and go to work every day.