PROVIDENCE – There is a lot of talk about data driven policy these days, but far too few examples of the concept in action.

As Lauren Greenawalt of New America [a good government think tank] pointed out last year: “We, as a society, often invest significant resources into ambitious public policies. But despite the time and money we spend doing this, we struggle to determine whether these policies have successfully met their goals. In no small part, that’s because we typically lack monitoring and evaluation mechanisms that can help us decide whether policies are really effective.” [See link to article below.]

Nowhere is this paucity of evaluation tools more apparent, or more tragic, than in health policy, where we funnel so many of our public and private resources, yet have so little data-driven feedback on results.

This is one reason I am excited about an effort underway here in Rhode Island utilizing data related to health care services to improve quality and reduce costs. The data source is called the All-Payer Claims Dataset [APCD], but is also known as HealthFacts RI. Researchers and policymakers have only recently begun to realize its potential.

An APCD is the data set that results from a state gathering all available information about claims paid for services by all private health insurers [Blue Cross and Blue Shield of RI and UnitedHealthcare, for example] and almost all public health insurance programs [Medicaid and Medicare]. Twenty states now have laws on the books establishing authority for an APCD and several are actively using their APCDs to support decision-making. In Rhode Island’s case, the data are scrubbed so we can’t identify individual patients, but it is still a rich source of information on patterns of health care cost and use.

Brown University, the Office of the Health Insurance Commissioner, the Executive Office of Health and Human Services, and the Governor’s Office have teamed up, with financial support from the Peterson Family Foundation, to use Rhode Island’s APCD to examine health care costs, and drivers of health care cost growth, in the state.

Why? Because health care is eating our lunch. If we look at trends over the past several decades, health care cost growth has well exceeded growth in our economy, outstripping wage growth and crowding out other priorities for investment, including infrastructure and education.

Addressing the health cost crisis
But quantifying that cost growth in anything close to real-time and understanding what drives it is much harder than one might think.

Once a year we get an assessment from the federal government of how much health care costs grew, overall, nationwide, for a period about two years prior, and a projection of that rate of growth going forward.

Closer to the ground, we can see annually how much Medicaid program costs rise at the state level [not much lately, due to program innovations and budget cuts], how much Medicare costs rise at the federal level [usually a little more than Medicaid] and how much commercial insurance premiums rise [usually at least twice the rate of growth in Medicare].

Unfortunately, these overall trends tell us very little about how to address the cost crisis. Ideally, we would have data that helps us, across all payers and programs, answer such questions as:

• How does our health care use, and the prices we pay for health care services, compare to other states?

• Do Rhode Islanders use so-called “low value care” [services such as unnecessary diagnostic testing] and to what degree?

• Are some providers more efficient or effective than others? If so, what could we learn from their practices that might improve overall system efficiency and effectiveness?

Our group is undertaking exactly this type of analysis, guided by a broad group of stakeholders, called the Cost Trends Steering Committee. The Steering Committee began meeting last year to establish a target for health care cost growth reduction. That work culminated in representatives of hospitals, physicians, insurers, the business community and others signing a voluntary compact to hold cost growth to 3.2 percent through 2022. [See link below to ConvergenceRI story, “Annual cap of 3.2 percent put on health care costs in RI.”]

Gov. Gina Raimondo issued an Executive Order on Feb. 6 pledging support for this effort. [See link below to Executive Order.]

Calculating adherence
How will we adhere to this target? Our group will use analysis of the APCD to pinpoint waste in Rhode Island’s health care system – and to find opportunities to improve efficiency, improve care and avoid unnecessary cost.

We expect this work will prompt changes in how some of Rhode Island’s health care organizations, payers and perhaps even consumers do business. The exciting part is that we have:

• First, a dataset that supports this effort; and

• Second, an engaged, broad set of stakeholders who are committed to learning from and acting upon the analyses. This holds great promise.

APCDs, and analysis of APCD data, are not new. I began my adventure with all-payer claims datasets more than a decade ago, when I was working in Massachusetts and appointed to the Massachusetts Healthcare Quality and Cost Council.

Massachusetts was one of the first states to pursue collection of all-payer data to feed public policy goals. At the time, the focus in Massachusetts was on how to use the APCD to support consumer decision-making. Specifically, the Council wanted to create a website that provided comparative price and quality information, so patients seeking care could “shop” based on price and quality.

Several years and many millions of dollars into this endeavor, the Council concluded that the price and quality transparency website was not achieving its intended goal. Few consumers were using it as a shopping tool. Most of the visitors to the website were providers, scouting for the prices of the competition. So what was intended as a tool of consumerism was really just undermining, or at least having no effect on, competition in the marketplace.

My next experience with an APCD was in Vermont. Vermont began to establish an APCD around 2008, but it took until 2014 to include Medicare data. The primary use of their data was [and still is] to support a state health-improvement program called the Blueprint for Health, partly with reporting to providers on cost and quality measures, and partly for program evaluation. This is a worthy cause, but very limited in scope and with a narrow audience.

Rhode Island’s effort, to me, seems to have much broader appeal and potential utility. It is the first example I have seen of using an APCD to support a system-wide effort at cost control.

We have a long way to go, but so far we have shown that the data are usable, have presented some preliminary findings to the Steering Committee, and have received guidance from the committee about how to prioritize analyses in the future.

In the next phase of the project, we will work with the Steering Committee to further validate our analyses and begin to pinpoint areas for improvement and cost reduction. This is a great example of using data to guide public and private actions and public policy – and to measure the results.

Anya Rader Wallack, Ph.D., is the Associate Director and Professor of the Practice at the Center for Evidence Synthesis in Health at the Brown University School of Public Health.