An Open Letter To ARPA-H Designers: 5 Guiding Principles From Health Services Research

Proposals for an agency to drive innovative health research are not new. President Biden’s proposed Advanced Research Projects Agency for Health (ARPA-H) seeks to address longstanding challenges within the health research ecosystem, acknowledging that the traditional approaches to funding, conducting, and disseminating research are not achieving the results the nation might expect.

The reasons for our lack of progress are many, but among them are (1) the fact that innovations developed in laboratories or randomized trials don’t always work in the real world; (2) trust in government and health care has been damaged such that government-funded research may be disbelieved and ignored; and (3) the dominant paradigm by which we determine which research projects are funded and who conducts them is intrinsically flawed.

ARPA-H is being designed to break the mold of our current biomedical ecosystem. Modeled on the Pentagon’s Defense Advanced Research Projects Agency, or DARPA—which helped advance several breakthroughs, including the internet and Global Positioning System (GPS)—ARPA-H will seek to make the dramatic accomplishments we saw in the COVID-19 vaccine development more commonplace. However, the COVID-19 pandemic also highlighted our continued failure to provide equitable access, care, and health outcomes to all people. Clearly it isn’t just biomedical research that needs innovation. Research on how to ensure effective delivery of new therapeutics is critical to achieving better health.

That is where health services research (HSR) comes in. Our field provides the framework and process to ensure the “last mile” of the journey from discovery to delivery is safe and equitable. HSR examines factors such as how care is organized and delivered, the role of each clinician in the workforce, and who gets access to new therapeutics and why—and, importantly, who does not get access and why.

Systems operate as designed; therefore, we must design the ARPA-H system differently to avoid repeating today’s weaknesses. We offer the following five principles from HSR to guide ARPA-H’s design, as well as a to-do list for Congress as it considers ARPA-H’s creation.

1. 
   

   Prioritize Health Care Quality And Equity Domains That Have Stagnated

One of the key features of various ARPA agencies in defense (DARPA) and energy (ARPA-E) is that they do not use an investigator-initiated, hypothesis-driven strategy to select what to fund. Instead, projects are selected that address high-priority challenges chosen by the funding agency. ARPA-H must do the same.

We know where the greatest disease burdens are in this nation, and trends across National Healthcare Quality and Disparities (NHQD) reports reveal that the overall quality of care in the US remains suboptimal. In their 2010 report, the National Academies of Medicine (formerly the Institute of Medicine) noted: “Across all of the process of care measures tracked…persons received the recommended care less than 60 percent of the time. Furthermore, even when quality has improved…disparities in care often persist across socioeconomic groups, racial and ethnic groups, and geographic areas.” A review of the reports from 2019 show some progress with 84 measures of quality of care improving, while 76 measures showed no improvement and 10 measures actually worsened. Far less progress is being made on health disparities where, for example, data reveal that Black people are receiving worse quality of care in more than 80 measures.  

ARPA-H should prioritize health care quality and equity domains that have not improved, particularly in areas where there remains clinical uncertainty on the most effective approaches to health outcomes improvement. For example, while increases in diabetes rates among people below 20 years of age have been documented, the US Preventive Services Task Force (USPSTF) only lowered the age cutoff to 35 in its most recent screening guidance for this population; this delay occurred in part because more “research was needed” to confirm that younger people could meaningfully benefit from diabetes prevention programs. As a result of a lack of clear, actionable evidence derived from research, we lack a definitive USPSTF guideline for screening for diabetes in 18- to 34-year-olds at the same time that the incidence of diabetes in this age group amongst Black, Latinx, and Indigenous Americans is significantly higher than among whites of the same age.

2. 
   

   Each High-Priority Challenge Should Address Known Disparities, In Order To Achieve Innovative And Effective Health Care Treatments

Research innovation without a serious focus on addressing disparities would not only perpetuate health inequities in our nation but could also worsen them, as effective interventions would fail to reach those most in need. Continuing with the diabetes example, while youth onset diabetes rates are increasing across the board, type 2 diabetes rates have climbed fastest among Black and Hispanic youth.

We are pleased to see an acknowledgement of the importance of equity in early descriptions of ARPA-H. The best way to ensure equity is not only a core principle but also a tangible outcome of ARPA-H’s work is to ensure that each high-priority challenge addresses known disparities by centering equity at each step of the implementation process. For example, initial research project specification should include input from a range of stakeholders from communities and individuals who have been historically excluded, so that the real people who are impacted have a voice in priority-setting. Second, research teams should include members with a range of experiences, including the lived experience of being underrepresented investigators in the current biomedical enterprise. Third, pilot sites and partners should include a diverse spectrum of communities and organizations to ensure that the research designs are inclusive of a range of settings and understand if and how results vary by population subgroups. Fourth, once an intervention is proven to be effective, the dissemination and implementation plan must include specific investments in ensuring broad reach to communities that have suffered disproportionate burden from the disease in question.

3. 
   

   Support Research That Includes Equitable Adoption, Adaptation, And Scale

Biomedical innovations typically focus on improving how a patient receives medication, services/protocols, or devices in a lab. We propose that research also include analysis of how that care is equitably distributed, accessed, and paid for in the real world. This means that we need to think about the impact of the innovation on individuals and on populations, with protocols themselves set up to ensure that the impact can be measured at both levels.

An oft-cited example of basic research transformed into health care services is the translation of a 1996 National Institutes of Health (NIH) study on prevention of diabetes into the Medicare Diabetes Prevention Program (MDPP), a structured intervention with the goal of preventing type 2 diabetes in individuals with an indication of prediabetes. Study findings weren’t incorporated into the MDPP model until 2018—a 22-year journey. And even though it translated into an expansion, thanks to further testing by the Centers for Medicare & Medicaid Services’ Center for Medicare and Medicaid Innovation, the program—now established within Medicare—has still not been successfully scaled or designed to address health inequity. For purpose of ARPA-H, the lesson learned from MDPP is that adoption, adaptability, scale, and equity should be part of the initial research to accelerate widespread adoption.    

4. 
   

   Conduct Regular And Independent Transparent Assessments Of Progress To Build Trust In ARPA-H

Over the last decades, trust in our national institutions has weakened while trust in science and scientists remains stable. Yet, a look below the surface of these data reveals significant differences in trust by race. In a 2020 study, the Pew Research Center found that Black adults are less likely than white adults to have a great deal of confidence in scientists to act in the best interests of the public. In the last two years, data, evidence, and interventions to address the pandemic have been met by mistrust, distrust, and misrepresentation, resulting in poor uptake of effective interventions such as the COVID-19 vaccine. Thus, a bold new initiative like ARPA-H will need to build trust among stakeholders and communities of color to ensure that its results are widely and appropriately used.

Trust in science grows when science is timely, addresses real problems, is transparent and accountable, and addresses the needs of all sub-populations. ARPA-H should provide regular progress reports to the public and Congress. Various models of accountability exist including agency reports as part of annual congressional justifications, occasional Government Accountability Office (GAO) reports, and advisory council reports. Far too often, these end up being simple exercises to make the agency or program “look good” with results cherry-picked and watered down to the point of near irrelevance. The key is for the reports to be publicly presented and discussed. These could take the form of a biennial report to the NAM delivered in a public setting so key milestones could be evaluated by key stakeholders including end users.

5. 
   

   Fund Differently To Drive Innovation

ARPA-H proposes to approach research funding very differently, with the use of active program managers who select projects based on agency priorities rather than the NIH’s traditional grant peer review system. This is not a surprising shift, since as much as half of current research published may not be reproducible and an estimated 85 percent of research funding in medical research is wasted on studies that are biased, unpublished, poorly designed, or inadequately reported. A number of these issues can be traced back to a flawed funding process. While grant peer review is the process by which the vast majority of funding decisions are made, a recent AcademyHealth report noted the following limitations, among others, to peer review as the only means of allocating research resources:

• Decisions may be subject to conservatism and a reduced chance of selecting risky, innovative proposals;
• Inconsistency, with assessments varying across reviewers;
• Possible bias on the basis of gender, age, or cognitive approach on the part of external reviewers, and a risk of cronyism.

The ARPA-H proposal does include language to allow the use of non-traditional research mechanisms, which should be fully explored and tested. Examples of alternate approaches to funding emerging across the field and outlined in the recent report include:

• Open peer review: Increase transparency in the peer review process—from sharing of reviewer identities to open publication of funding applications and accompanying review.
• Broadening participation/crowdsourcing: Involve different groups in the peer review process—from wider disciplines, through research users (e.g. patients, clinicians) and the general public, including “crowdsourcing” approaches.
• Innovation prizes: Rather than awarding funding for research, offer prizes for answering a particular challenge or problem.

ARPA-H should also study the impact of different processes for selecting which research to support and the ultimate success of funded projects. In this way ARPA-H will not only help advance solutions to priority health issues in our nation but will also advance our understanding of and ability to do science differently. New approaches could then be spread to other health research funding agencies, including other institutes at NIH.

So What Should Congress Do?

Based on the principles outlined above, we propose the following specific actions for Congress as they finalize the details of the authorizing legislation for this exciting new approach to health research:  

1. Require each project/program funded by ARPA-H to include an equity plan and stakeholder engagement.
2. Ensure that at least 5 percent of ARPA-H funding is dedicated to the “last mile” of research implementation in the real world.
3. Commit to having at least one in four projects designing innovations to eliminate a known health or health care disparity.
4. Establish a National Advisory Council to ARPA-H that is diverse and reports annually and publicly on progress in reach, impact, and implementation of ARPA-H developed innovations.

Progress in the health arena isn’t simply a matter of developing a new molecule or biomedical intervention. It is innovating a health care system that puts that innovation in the hands of patients, clinicians, and health systems to dramatically improve health outcomes and eliminate health disparities.   These principles and action items will increase the likelihood that ARPA-H’s efforts result in meaningful transformation of our health and health care system.