What is the best way to reduce knee pain? Which cholesterol medicine will fend off heart disease better than the others? What is the most reliable test to diagnose cancer?

Physicians, patients and insurers confront questions like these every day, but often lack solid information about how one treatment approach might work compared to another. While many different treatments are effective, which is best is difficult to determine.

As part of the sweeping U.S. health system reform, a new emphasis is being placed on comparative effectiveness research (CER), which pits treatments against one another to determine the method that leads to the best results for patients. While not specifically designed to address health care costs, CER could ultimately lead to changes in the ways people seek medical treatment, the development of innovative remedies and other trends within the industry that could impact costs.

But some worry that CER may also result in restrictions on what remedies insurers or government-funded programs will cover, or limit the treatment options available to patients. There is also concern that if CER is concentrated in the public sector, it could fall victim to political influence.

In response to new federal CER initiatives, a policy analysis of CER by Wharton health care management professor Scott E. Harrington argues that the government should avoid developing a monopoly on CER. Instead, Harrington’s paper, “Incentivizing Comparative Effectiveness Research,” says the United States should encourage privately funded research into comparative effectiveness. Refundable tax credits to stimulate such studies, and government investment in better data systems to fuel this research would also help, according to the paper. In addition, changes in the way health care is funded that shift more decision-making power to consumers and medical providers would stimulate an increase in private sector investment in CER, Harrington suggests.

While Harrington is not opposed to state and federal investment in CER, he argues that it should not be designed in such a way that discourages private sector research. Government-directed decisions about treatment that could eventually result from publicly funded CER “frightens many people,” he says. “There’s too little attention to how we might stimulate creation of more evidence in the private sector in ways that might improve medical care and lead to better decisions over time — without having the government play an ever-increasing role in deciding what types of care it might pay for.”

Indirectly, CER could play a role in the current budget debate in Washington, according to Harrington. “We obviously have serious budget problems, and a major part of that is federal spending on health care,” says Harrington. Increasing pressure on health care budgets in the years ahead might lead to a model that relies more heavily on government mechanisms that determine what works best, and then incorporates those findings into decisions about what treatments will be paid for by government health programs such as Medicare and Medicaid and, eventually, by private insurers. “We’re not anywhere near that now, and we don’t need to go that route until we check out some other avenues, including motivating consumers and providers to pay greater attention to utilization and costs,” Harrington states. “If we have additional information about what works and what it costs, that information will be reflected in care decisions and adopted over time.”

Two new government initiatives focus on CER. The American Recovery and Reinvestment Act (ARRA) of 2009, also known as the federal stimulus package, authorized $1.1 billion to fund CER. Last year, federal health reform legislation — the Patient Protection and Affordable Care Act of 2010 — called for creation of an independent agency, the Patient-Centered Outcomes Research Institute (PCORI), to fund and support CER. The PCORI will be funded by direct appropriations of $10 million, $50 million and $150 million during fiscal years 2010 to 2012, respectively. From 2013 to 2019, funding will come from a trust that will receive annual direct appropriations of $150 million and charges (estimated at $2 per Medicare enrollee) from participants in Medicare and privately and self-insured health plans. The projected value of the charges is estimated to exceed $500 million. Harrington notes that total CER spending was estimated at less than $1.5 billion in 2005, compared with total health care spending in 2009 of roughly $2.4 trillion, or 17% of GDP.

The Institute of Medicine (IOM), an independent, nonprofit organization that provides advice to health policymakers and the public, estimates that less than half of medical care provided in the U.S. is based on evidence of what works best. Meanwhile, research shows substantial variation in regional Medicare spending due primarily to different levels of care provided to patients suffering from the same conditions. Other data suggests that higher spending is not associated with better quality care, and may even lead to worse health outcomes. As a result, some analysts say that Medicare spending could be reduced by up to 30% annually without reducing quality of patient care.

Harrington is skeptical of such claims and emphasizes that linking CER to spending is controversial. “The fear is that significant government involvement in CER will give rise to some form of restrictions on access to care [or to] particular treatments based on research studies and bureaucratic decision-making,” he says. “It might not be tied to what people want to have and what they are willing to pay for.”

For example, the paper notes, decisions about medical treatment could become politicized with elected officials and bureaucrats responding to outside lobbying. Another problem, Harrington notes, is the traditionally slow reaction time of government bureaucracies, which might prevent patients from receiving valuable treatments earlier. He also adds that it is important that CER account for differences in subpopulations, such as variations in responses in relation to the presence of other health conditions, and factors such as age, gender and race. That would be another challenge for a public program, which would tend to focus on the effect on the average patient population that would benefit the most from public spending.

Impacting Treatment

The paper outlines how four other countries — Canada, Australia, the U.K. and Germany — approach CER. In most other nations, the government plays a much larger role in health care finance than in the United States, and the links between CER and government health spending are tighter. For example, in the U.K., the National Institute for Health and Clinical Excellence (NICE) sets standards for the use of health technology including pharmaceuticals, clinical practice and prevention. NICE guidance on medical technology and pharmaceuticals is mandatory for caregivers, although clinical care guidelines are only advisory. The organization bases its determination on quality-adjusted life years (QALYs), a calculation of the estimated additional life expectancy from a medical intervention. NICE assesses the cost-effectiveness of new treatments compared to the next best treatment currently in use. The National Health Service, which pays for health care in the U.K., then generally bases payment decisions on NICE’s recommendations. If patients want to deviate from the standards, they must pay out-of-pocket.

Harrington points out that the American health care system, with a much greater emphasis on private markets, has insulated patients and providers from much of the cost of medical care. As a result, there has been less demand for information about which treatments are most effective. In the paper, Harrington notes that tax exemptions for health insurance have encouraged high levels of coverage with little incentive for consumers, insurers and employers to be as focused on costs as they might be in other sectors of the economy. The National Commission on Fiscal Responsibility and Reform, co-chaired by Alan Simpson, a former Republican senator from Wyoming, and Erskine Bowles, a former White House chief of staff to former president Bill Clinton, put forth a proposal to eliminate this deduction as part of comprehensive tax reform.

Another proposal, made by Rep. Paul Ryan and Alice Rivlin, a former director of the Office of Management and Budget, might also encourage private sector investment in CER. Ryan and Rivlin have suggested that Medicare become more like a defined contribution program, rather than a government insurance initiative, with consumers directing how their health care funds would be spent. “The extent to which consumers have more skin in the game will create enormous demand for additional information,” says Harrington. “There’s every reason to believe that much of that demand could be met by private investment.”

In the paper, Harrington writes: “Changes in the tax/insurance system to reduce moral hazard would reduce excessive utilization of low-value care and increase the demand for information to help identify such care.”

Harrington suggests that the public sector could help encourage the development of independent, private sector CER. One way would be to expand research and development-related tax credits for qualified investment in CER, including the collection and analysis of health care data. While government would have an indirect role in subsidizing the research, he notes, it would not be in a position to determine what projects are most important, or picking “winners and losers” among different medical treatments.

In addition, government investment in data infrastructure and availability would also help generate new private sector CER. Harrington says better medical information systems would help broaden research into effective health treatments beyond the current medical research community. Access to medical claims information, for example, would give researchers new data sets to explore — at no additional cost to taxpayers — that would yield findings about which medical interventions work best. If data were more widely available, he says, skilled researchers in the social sciences, including economists and academics specializing in business and finance, would be drawn to CER and would produce research without expending additional government funds.

“The availability of, and ease of using, data has a major impact on … research agendas,” the paper states. “Initiatives to make medical claims and related data more widely accessible to researchers, without pre-approval by a government agency of the research topic, would likely cause many additional researchers to allocate effort to CER and related research, without requiring direct government funding of their time and research support.”

Under the new health reform legislation, PCORI-funded researchers will have access to data from Medicare, Medicaid and the Children’s Health Insurance Program. Their findings must be submitted to the peer-review process and released within 90 days to the public. Traditionally, a significant proportion of CER research has been done by medical school faculty using grants from the government, nonprofit centers such as the Rand Corp., hospital systems and pharmaceutical manufacturers. Harrington adds that advances in technology will make CER faster and easier. For example, he says, full-scale randomized trials are the current gold standard in research. Going forward, analysts may be able to focus more on observational studies that look at samples of patients and different treatments to make inferences about what is most effective without the time and expense of trials. The paper notes that “concerns about privacy of medical data and associated restrictions on data availability pose barriers to CER that might be reduced by revising regulations and legal rules.”

The paper also argues that better CER would encourage developers of medical treatments and technologies to invest in breakthrough ideas that could ultimately become clear market leaders. These innovations could be protected by intellectual property statutes, including patents that would prevent knock-off competitors from eroding the investment made into discovering a promising cure. Additional information about the effectiveness of treatments would also create more opportunities to brand, license and certify efficient treatment protocols that are not patentable.

“With the right environment,” Harrington says, “we could create a broader market for CER in the private sector.”