Better evidence for real healthcare reform

As healthcare costs continue to rise at twice the rate of inflation (6.9% in 2005), there is increasing focus on the value of what we spend on healthcare and the health benefits derived from these expenditures [1]. It is noted that although the USA outspends the rest of the world by a large margin [2] (US$6697 per person in 2005 [1]), its healthcare system is ranked 37th in the world by the WHO [101]. Given these statistics, it is no surprise that healthcare reform has been identified as the number one domestic issue for the 2008 Presidential campaign [102]. As the Presidential candidates begin to unveil their different healthcare proposals to control costs and increase access for more Americans, there is another policy discussion underway in Washington that may have an even larger impact on healthcare spending trends; proposals on ways to make clinical effectiveness research part of the national healthcare agenda. Part of that discussion is the topic of whether cost-effectiveness analysis should be included in this agenda. Whether or not one believes that it is important to include costs in healthcare discussions, all agree that it is crucial to examine comparative clinical effectiveness. Comparative effectiveness research, as noted in Section 1013(a)(6) of the Medicare Modernization Act (MMA), reflects the principle that physicians and patients should have the best available evidence upon which to make choices regarding healthcare items and services (which is the comparison of therapies). Effectiveness is the efficacy of medical intervention in actual practice. There is a growing consensus among policy makers and physicians that the USA must prepare to produce reliable and real-world effectiveness information about the types of care that work best for patients. In order to reduce spending on technologies of little or no clinical benefit, while increasing access to therapies and procedures shown to have patient benefit, we must have carefully carried out comparative effectiveness studies that can report not only whether the technology or procedure is beneficial for patients, but the magnitude of the benefit. Comparative effectiveness research should focus on evaluating clinical outcomes, in the interest of promoting what is best for patients. Costs of care become a consideration only after clinical effectiveness is established. While there is no consensus on the benefits of cost-effective analysis for any national healthcare policy, it is increasingly common and important to include cost analyses in healthcare studies. Thus, we can improve the evidence base for real-world clinical effectiveness studies.

Research continues to show that more healthcare spending does not mean better quality healthcare. A seminal study undermining the more-is-better hypothesis came from the work of Eliot Fisher and colleagues at Dartmouth University [3]. In an analysis of Medicare beneficiaries hospitalized with hip fracture, acute myocardial infarction and colorectal cancer, they found considerable regional variation in healthcare spending for these conditions, largely due to differences in numbers of physician visits, more frequent use of tests and minor procedures, and increased use of specialists. Higher healthcare spending was not associated with better quality of care or better access to care. A more recent study considering variations in payment concluded that patients had comparable length of stay and death rates, despite large differences in payments made to Pennsylvanian hospitals for heart bypass surgery [4,103]. These findings help to support the growing consensus that higher intensity of care and higher cost of care do not necessarily equate to better healthcare. Thus, market principles do not necessarily apply in the US healthcare market for many reasons, including a largely third-party payer system and lack of transparency in healthcare pricing.

The rapid increase in health costs is having widespread repercussions throughout the economy. Many recent labor strikes, such as at Safeway and General Electric, have been over the rising cost of health insurance. Healthcare costs continue to plague the US automobile industry. It is estimated that retiree health-benefit costs add approximately $1000 to the cost of each Silverado truck made by General Motors Corp. [104]. The high costs without evidence of corresponding improved health outcomes has led to greater focus on analyzing how we are spending our healthcare dollars.

The interest in improving value of healthcare spending has sparked enthusiasm for clinical effectiveness research in the USA and worldwide. Although there have been calls for effectiveness research and even for an effectiveness board within the Centers for Medicare and Medicaid Services (formerly Health Care Financing Administration) dating back to 1988 [5], it has gained momentum in the last few years as healthcare costs continue to soar to unsustainable levels. In the 2004 fiscal year, Congress appropriated $15 million for the US Agency for Health Research and Quality (AHRQ) to conduct research and set priorities as authorized in the 2003 MMA relating to the clinical effectiveness and appropriateness of health services. This year, policymakers have introduced several new proposals to expand the USA’s capacity to develop unbiased, credible information on comparative clinical effectiveness for drugs, medical devices, procedures and other healthcare interventions. In Congress, Representatives Allen (D-ME) and Emerson (R-MO) introduced legislation that called for $3 billion over 5 years to fund comparative effectiveness studies. In its June 2007 report, The Medicare Payment Advisory Commission (MedPAC) recommended that Congress should charge an independent entity to sponsor credible research on the comparative effectiveness of healthcare services. In the same month, the director of the Congressional Budget Office, Peter Orszag, testified in front of the Ways and Means Committee that improved comparative effectiveness research could positively impact future spending trends. Well-respected health economist and former Health Care Financing Administration administrator, Gail Wilensky, outlined a comprehensive and feasible plan for a Center for Comparative Effectiveness [6]. In addition, large payer organizations such as America’s Health Insurance Plans (AHIP) and the Blue Cross Blue Shield Association published independent proposals for a new public/private comparative effectiveness research institute. The growing awareness in Washington of a need to improve the evidence base with more comparative studies of health technologies is exciting news for those that hope to improve our mediocre standing in terms of our nation’s health status to our leading position in healthcare expenditures.

Cost–effectiveness is not currently included in the public and political healthcare discussions or in current proposals for a center for clinical effectiveness research for several reasons. First, any economic considerations in medicine, such as a true cost–benefit analysis, are difficult because they normally require value-laden and controversial tasks (e.g., valuing human life according to factors such as income potential). Therefore, instead of a cost–benefit analysis, health economists are more likely to use a cost-effectiveness analysis, which compares one therapy against another in terms of incremental costs and benefits of Therapy A compared with Therapy B. However, well-designed studies that can inform us about the comparative benefits or harms necessary to evaluate incremental benefits are rarely available in the published literature. Although progress has been made in developing models that can analyze the cost–effectiveness of a treatment option, these economic tools are still complex and are thus less transparent and trusted by the public than traditional comparative trials. In addition, applying a cost-effectiveness analysis in medicine has developed a bad reputation because any healthcare strategy that implies considering cost as a criterion for coverage is associated with rationing. Americans have a deep mistrust for rationing healthcare based on costs, mainly due to the disbelief that an organization, rather than a physician or patient, can make a proper medical decision [7]. This was exemplified in the late 1990s by the backlash towards managed care plans and the thousands of bills introduced in Congress and state legislatures to protect patient rights. Attempts to introduce cost as a consideration for Medicare coverage in 1989 and 1997 were defeated owing to the political volatility of the issue and industry and patient group lobbying. Similarly, cost–effectiveness was not included in the 2003 MMA [8,9].

However, it is hard to ignore costs, as healthcare budgets, although rapidly increasing, are not infinite. Thus, implicit rationing occurs, as 45 million Americans do not have health insurance because they or their employers cannot afford it. Even insured people may have limited access to healthcare owing to high co-payments and deductibles. There is clearly a connection between the increased demand for more care and higher technology care without regard to health benefit or costs of care and rising health costs, rising health premiums and out-of-pocket costs (cost-shifting). Currently, there is the implicit assumption that more care is better care. This association seems particularly true if more care involves a high-tech procedure or piece for equipment. Americans generally make decisions on healthcare technologies very differently from other spending decisions. As David Eddy stated in a recent interview on evidence-based medicine, “Imagine making a decision about a stereo, a car or a house, without considering the cost. What the heck, let’s all go out and get airplanes. We would never do this in our personal or business lives because it would be dumb. And yet we try to do it in healthcare. The result is trivially obvious: we end with a healthcare system that everyone agrees is too expensive. The only people who like it are the people for whom costs are incomes” [10]. In addition, our third-party system insulates the purchaser (patient or doctor) from the costs that are covered by the payer. Thus, traditional market principles do not operate well in healthcare.

The first step in clinical effectiveness research is to be sure we are receiving information on patient outcomes and how a new therapy or test can lead to patient benefit. If the answer to this question is no (or we don’t know), then there is no reason to consider costs, as this test/therapy clearly needs further study to show benefit before widespread use. This principle has not been applied consistently to healthcare. One example of a new technology where use exceeds evidence of benefit is cardiac CT scans.

In the past 10 years, cardiac imaging of all types has grown by 26% per year, which contributes to the total cost of imaging services in our healthcare system growing at an alarming rate [11]. The growth of multi-slice CT imaging, especially 16- and 64-slice machines, has been attributed to multiple factors, such as improvements in technology with the promise to noninvasively visualize coronary arteries, marketing and direct-to-consumer advertising, and positive media attention. Recently, the cover of Time magazine showed a picture of a cardiac CT scan and its headline proclaimed that the test can prevent a heart attack [12]. This headline is misleading because even proponents of the technology do not claim the test can prevent a heart attack. Instead, proponents tout several advantages of the technology, including the high-resolution color images of the coronary arteries and the fact that it is less invasive and costly than a traditional coronary angiography. Despite limited evidence in published trials, many private insurance plans and all regional Medicare carriers offer some coverage for multi-slice cardiac CT imaging.

Although CT scans can often provide beautiful images of the coronary arteries, sharper and more precise pictures do not necessarily lead to patient benefit or better healthcare. The most important question providers should ask themselves is whether CT can lead to better patient outcomes than the alternatives. The real value of any diagnostic procedure, such as a cardiac CT scan, is not its ability to provide more information, but whether the information it provides suggests a better course of treatment than the physician would have recommended without it. Many studies help support this suggestion that new multi-slice cardiac CT scans are accurate and precise, but none have demonstrated that CT leads to improved patient care over better studied alternative diagnostic techniques [13]. The evidence on cardiac CT scans was summarized in 2006 by the Duke Evidence-Based Practice Center for the Medicare program. In their report, investigators noted that there were limited data available consisting of small, single-center studies that suffered from selection bias and excluding poor images to inflate accuracy estimates. They concluded that more studies are needed to examine CT technology in larger populations and in clinical settings in which they are expected to be used [105]. Shortly after, the first multicenter study of multi-slice CT was published and found that a high number of scans (~25%) could not be evaluated and there was a high false-positive rate. The authors of the report stated that routine use of CT scans in clinical practice is not justified [14].

Despite rapid adoption in the market today, evidence for the clear benefits of cardiac CT imaging is not yet apparent. In the absence of outcomes data for the CT scan, proponents rely on the images themselves as evidence of the benefit, highlighting their superior resolution. The problem is that there is no evidence that improved images lead to better care. Outcomes data from a well-designed clinical effectiveness trial are needed. This story is not unique to multi-slice CT technology. ‘Pay now, benefits might follow’ is the pattern that applies to many healthcare technologies in America. For example, once groups have invested in capital equipment for new and expensive technologies, such as a CT scanner, there is pressure to use it. It is a cause of great concern that our healthcare system adopts technologies before there is evidence that their use leads to patient benefits. The American College of Cardiology has begun developing appropriateness criteria to encourage use of new technologies where patients will benefit most, and to label use of new technologies as inappropriate where patient benefit is not established or harm may result [15]. Data from comparative clinical effectiveness trials are essential to the development of appropriateness criteria.

The policy proposals concerning clinical effectiveness research for healthcare technologies, which are currently being discussed in Washington, will help keep patient benefit as the ultimate goal. In our current system, new and more expensive pharmaceuticals, medical devices, biologics, diagnostic technologies and other procedures continue to be introduced without evidence of a comparative benefit to current treatments. To ensure the optimal allocation of healthcare resources and appropriate use of new technologies prior to and after widespread adoption, we must focus on sound systematic studies that evaluate comparative effectiveness. Comparative clinical effectiveness research is an important part of good healthcare. Understanding the most effective applications of new medical procedures and therapies will enable us to use them to avoid unnecessary, potentially harmful and therefore wasteful expenditure. Thus, we can use our healthcare resources to the greatest good.

Financial & competing interests disclosure

Until July 2007, Ryan Padrez worked for the nonprofit Center for Medical Technology Policy. Rita Redberg is a Consulting Cardiologist for that group. The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.