Medical spending has resumed its long-term rise. After several years of deceptive stability in the last, deep recession’s wake, health spending rose by 3.7 percentage points more than general inflation in 2014, then by 5.8 percentage points more in 2015, to a 17.8 percent share of the US economy.

Not only does this spending rise threaten the United States’ fiscal stability and capacity to address other needs; it is undermining the promise of health care for all. To manage rising costs, insurers are hiking premiums, narrowing their networks, and raising deductibles and copayments, making purchase of coverage less appealing. Millions of Americans are thus eschewing subsidized coverage via the Affordable Care Act’s (ACA’s) insurance exchanges, choosing instead to pay the ACA’s tax penalty. And the skimpier coverage and scaled-back subsidies envisioned by congressional Republicans as an ACA replacement would add millions of people to the ranks of the uninsured.

The central, intractable obstacle to long-term cost containment, we contend here, is the near impossibility of saying “no” to ever-more-expensive care that yields small marginal benefits. Public and private payers have made myriad unsuccessful attempts to surmount this barrier. We propose, instead, to circumvent it—through redirection of cost-control policy away from efforts to limit use of existing, low-benefit technologies and toward strategies for influencing the emergence of new technology. To this end, we urge: redesigning value-based payment to emphasize future rewards for tests and treatments that haven’t yet emerged, and varying the duration of intellectual-property protection so as to tie its rewards to therapeutic effectiveness.

Why Don’t We Just Say ‘No’

Many factors influence health spending, including coverage expansion, the balance of bargaining power between payers and providers, and the aging of the United States’ population. But the persistent driving force behind rising medical spending is technological advance, fueled by health insurance’s promise of rich reward and unchecked by insistence that advances deliver clinical benefit worth their cost.

Drug and device developers, clinical researchers, and their financial backers anticipate coverage for new tests and treatments with little concern for whether they add substantial therapeutic value, and they make research and development decisions accordingly. Once a new technology emerges, its developers are vested in pursuing its broad adoption. Hospitals, physicians, and others, in turn, commit to equipment and training that lock them into the new technology, even when the technology is merely a “me too” or marginal advance over existing clinical methods.

Patients and their loved ones then come to expect the technology’s availability. The psychology of rescue makes it anguishing to say “no”—or to accept “no” as an answer—when care might save lives, however high its marginal cost. Medical ethics tightens this lock-in: Hundreds of years of Hippocratic tradition stand in the way of withholding care believed to do more good than harm. And both public and private payers risk charges of “rationing” if and when they push back.

History underscores the difficulty of constraining access to existing, expensive technologies. Under both Republican and Democratic presidents since George H. W. Bush, the Centers for Medicare and Medicaid Services proposed, then backed away from, regulations that would make national coverage decisions under Medicare explicitly cost sensitive. Republican efforts to limit Medicare spending by reconceiving the program as a voucher scheme for purchase of private coverage have likewise been repeatedly stymied. Private insurers’ aggressive efforts in the 1990s to limit spending through tight administrative constraints on access to pricey treatments foundered in the face of the late 1990s consumer backlash against “managed care.” And the ACA’s misnamed “luxury tax” on high-cost health plans (really, a way to phase in taxation of employment-based health benefits as medical costs rise) is the object of hostility from Democrats as well as Republicans.

Getting Beyond ‘No’: Reimagining Therapeutic Progress

By contrast, clinical methods not yet available, or even envisioned, aren’t the object of vested expectations. Developers, investors, and health care providers haven’t yet made large financial and human-capital commitments. And while most Americans might bristle at being denied today’s care on account of cost, they’re not made livid by lack of access to future technologies. People yearn for cures to illnesses that terrify, but they don’t rail against health plans, providers, or public officials for failing to deliver, say, the snap cures “Dr. McCoy” offers to his Star Trek shipmates.

This expectations gap opens a cost-control pathway that circumnavigates the central obstacle to success: the practical impossibility of saying “no,” via regulatory or market means, to marginally beneficial care. By reining in the development of ever more expensive, minimal-benefit technologies before expectations vest, our health care system can restrain future spending growth without withholding beneficial tests and treatments from patients today. Selectivity here is, of course, crucial: Health policy should, at the same time, encourage advances that yield high value relative to cost.

To this end, we propose a dramatic redirection of cost-containment policy, toward long-term efforts to influence the trajectory of technological change. By reshaping medical innovators’ incentives at the research and development stage, before stakeholders make large financial commitments and stoke public expectations, health policy can promote high-value advances, while discouraging me-too technologies and marginal innovations.

A New Role For Value-Based Payment

The current trend toward value-based payment for services has long-term promise in this regard. Proponents of value-based reimbursement have focused on near-term efficiencies, including better coordination of care, improving the quality of existing services, and more rational use of these services. Such improvements, though, are “one-offs”—desirable changes, to be sure, but with limited potential to slow future spending increases.

By contrast, tying payment rates for future tests and treatments to emerging evidence of their therapeutic value could channel research and development investment away from me-too and marginal advances and toward decisive innovation. Introducing wide variation in future economic reward, for technologies not yet developed, would meet much less resistance than would similar payment variation for existing technologies backed by vested expectations. Current clinical services could be “grandfathered in”—subjected to more modest pay-for-performance variation—while much wider reimbursement variation is set in place for the future. This could achieve what well-functioning markets do: closely link innovators’ incentives and the value they deliver.

Such linkage wasn’t practical in the recent past. Long lag times between initiation of prospective clinical trials and publication of results, plus narrow patient inclusion criteria, made their “gold-standard” evidence mostly unavailable to reimbursement decision makers. But medicine’s ongoing big-data revolution, made possible by electronic records and interoperability, is transforming this picture. Now, comparative efficacy can be tracked in near-real time for large numbers of patients. Suspected confounders—influences on outcomes that randomization removes but that bedevil retrospective studies—can be effectively factored out as data become richer. Possible side effects can be tracked, as can multiple measures of clinical success.

Tying Intellectual Property Protection To Therapeutic Value

These big-data tools open another, transformative policy possibility. Patent law has long rewarded innovators with monopoly power for a fixed period, regardless of the value an innovation delivers. Patents thus allow markets to determine an innovator’s reward, with purchasers’ ability and willingness to pay serving as the measure of an innovation’s value. This works well enough for most goods and services but not for medical care, where the dynamics discussed above cut the connection between sales and value.

We therefore propose to vary the length of intellectual property protection for new health care technologies based on therapeutic value. Uncertainty about the clinical impact of new tests and treatments precludes advance bestowing of differing patent terms based on clinical value. But the emerging ability to assess therapeutic efficacy in near-real time by tracking electronic records opens the way for setting and adjusting patent terms as understandings of clinical value develop.

Variable intellectual-property protection, tied to therapeutic value, would transform research and development incentives by more generously rewarding therapeutic breakthroughs while shrinking payoffs for me-too and marginal advances. This incentive structure could be designed to roughly maintain patent protection’s aggregate economic value across the medical care, drug, and device industries; this should reduce industry resistance arising from fear of diminished overall intellectual property protection. There would, of course, be winners and losers, but the transition could be phased in over a long enough time to avoid effects on research and development commitments already made.

Much needs to be done to develop this concept into a detailed, practical reform proposal. Judgments must be made about how to assess therapeutic value in ongoing fashion, adjust for changes in a technology’s use, and translate metrics of value into variable patent terms. We’re already undertaking some of this work, beginning with an attempt at proof of concept. In particular, we’re working with a multicenter aggregator of electronic medical records to track the effectiveness of several new, high-cost cancer treatments that target molecular markers. Our aim is to develop a multiaxial model of value (incorporating prolongation and quality of life, side effects, and other suffering) that could translate into variable reimbursement levels and intellectual-property protection terms.

Our proposed redirection of cost-control efforts will need to take account of real-world constraints. To this end, we’d be grateful for any feedback from readers of this post, especially those involved in decisions about development and adoption of new clinical technologies. We suspect, for example, that most who make these decisions are both loss and risk averse, which might color their responses to the changed incentives we envision. Would multinational drug and device makers with ongoing research and development budgets tend to “play it safe” by pursuing less ambitious projects that are more likely to pan out? Or would these firms’ large research and development portfolios make them more willing to take chances, since pursuit of many possibilities offers some “insurance” against the failure of one? Conversely, would the expanded range of risk and reward that we envision inspire academic researchers and start-up investors (without large, ongoing research and development commitments) to try for home runs or to steer away from biomedical innovation?

We also expect the relative influence of value-based reimbursement incentives and value-based intellectual property protection to vary among differently situated decision makers—for example, drug and device makers versus surgeons who invent new procedures. For pharmaceutical firms, patent terms matter greatly, but for new surgical techniques, reimbursement levels matter more.

Our approach doesn’t assure that medical care’s share of the gross domestic product won’t continue to grow. Some major breakthroughs are hugely costly: Drugs that cure hepatitis C are a current, controversial example. Breakthroughs, moreover, beget aging populations, with greater clinical needs. But the strategy we’re pursuing holds out the promise of finessing the vested financial and human capital commitments and the popular alarm over “rationing” that have so far stymied long-term efforts to rein in rising costs. At the least, it could change the trajectory of clinical innovation to yield greater value from future medical spending.