Will Comparative Effectiveness Research Finally Succeed?

No matter how CER is sliced, cost will be a factor. But, first, how to gather the right data?

Abstract

Autonomy, innovation, and drug access may be at stake along with personalized medicine. Can all stakeholders be satisfied and should cost be a factor?

The purpose of comparative effectiveness research (CER) is to provide information that helps clinicians and patients choose the best medical treatment option. The Department of Health and Human Services defines CER as “the conduct and synthesis of research comparing different interventions and strategies to prevent, diagnose, treat, and monitor health conditions in ‘real-world’ settings” (HHS 2009). Cost is not to be taken into account, nor should patient access to healthcare be limited. The goal is rational care — not rationing of care.

A standard methodology for compiling comparative data does not exist. Thus, ambiguity remains regarding how CER will be appropriately applied from both clinical and regulatory perspectives. Does CER decrease a physician’s autonomy and lead to patient access issues? How will CER affect the drug approval process, and how can comparative information be developed in a way that sustains drug innovation? Will CER affect personalized medicine? These are all questions waiting to be answered.

Erwin A. Blackstone, PhD

Joseph P. Fuhr Jr., PhD

Danielle Ziernicki, PharmD

The National Pharmaceutical Council’s second annual survey of 117 key healthcare stakeholders, conducted from October 2011 to February 2012, assessed the impact of CER on medical decision making (NPC 2012). Those surveyed included opinion leaders, government officials, health plan officials, and employers, as well as business associations and trade groups. The results indicated clear recognition among “key health care influentials about the importance of CER and its promise for the future.” The results also demonstrated “tempered optimism” regarding progress in efforts to provide or improve the tools necessary for CER and that “any significant effects of CER are yet to be realized, and much work remains to be done”(NPC 2012).

A brief history

CER is a concept that has been around since the 1970s under various names. First, it was known as health technology assessment. In the 1980s, it was called effectiveness research. In the 1990s, CER was referred to as outcomes research, and in the last decade as evidence-based medicine.

In the 1990s, Congress created the Agency for Health Care Policy and Research, later renamed Agency for Healthcare Research and Quality (AHRQ). In 2003, the Medicare Modernization Act allowed funding for CER through AHRQ. In 2006, the Centers for Medicare & Medicaid Services (CMS) issued a guidance document allowing the agency to integrate evidence-based decision making and research into its coverage determination policies. CMS is informed by the Medicare Evidence Development and Coverage Advisory Committee, a working group that supplements CMS’s expertise.

The American Recovery and Reinvestment Act of 2009 provided $1.1 billion for CER and sparked a series of activities, including the development of federal CER priorities and efforts to enhance the nation’s research infrastructure to conduct CER in real-world settings. The not-for-profit Patient-Centered Outcomes Research Institute (PCORI) was born out of the Patient Protection and Affordable Care Act of 2010. PCORI, an independent organization with a multistakeholder board of governors, has sustained public and private funding for CER that will reach nearly $650 million by 2014. The first iteration of PCORI’s proposed research agenda includes five areas: assessment of prevention, diagnosis, and treatment options; improving healthcare systems; communication and dissemination; addressing disparities; and accelerating patient-centered and methodological research (PCORI 2012a).

A few private-sector CER entities have been successful — for example, the Oregon Drug Effectiveness Review Project, the Cochrane Collaboration, and the Institute for Clinical and Economic Review — mainly because they produce data that are perceived as useful for clinical decision making, purchasing, coverage, formulary placement, and cost containment. They are also protected from political influence, thus improving their viability.

Payers and patients will confront contentious issues as a result of CER and other efforts to control costs.

Needed: a validated methodology

The Federal Food, Drug, and Cosmetic Act (FD&C) does not require the assessment of comparative effectiveness. The U.S. Food and Drug Administration is entrusted to make sure that a new drug is safe and effective but not to determine whether it is superior to existing drugs. Thus, CER comes into play only when inferior effectiveness has patient safety implications, such as mortality or irreversible morbidity, and where lack of treatment would be unethical.

In some cases, drug sponsors seek claims of superiority over another product. According to Robert Temple, deputy director for clinical science at the FDA’s Center for Drug Evaluation and Research, clinical situations that involve superiority could arise when multidrug use may not be optimal, safety advantages exist, or certain population subsets could benefit (Temple 2012).

The FDA has set a high bar for inclusion of comparative studies in prescribing information. From an industry perspective, randomized clinical comparative studies that identify or exclude small differences in treatments are expensive and challenging to perform, given the large number of patients required, length of time to completion, and the need to replicate results. For symptomatic conditions like depression, a placebo arm would be required to assure assay sensitivity. Consequently, pharmaceutical companies generally conduct comparative studies after FDA product approval, although the lack of such data could result in a delay between approval and optimal reimbursement.

This situation illustrates a conundrum: Different stakeholders require different data to make decisions about patient care. What the FDA requires for drug approval often does not parallel what payers require for formulary decisions or reimbursement. Payers and patients, however, may be more likely to pay for a treatment with a safety or efficacy advantage. The quandary is how to obtain high-quality comparative data.

Randomized controlled clinical trials (RCTs) — the gold standard for determining the efficacy and safety of drug products — are homogenous by nature; however, the exclusion criteria of RCTs may omit patients with real-world comorbidities as well as elderly and pediatric patients. Often, RCTs include a placebo or comparator arms, not multiple treatments across various pharmacologic classes. Hence, CER data derived from RCTs may not provide insight into real-world comparative treatment.

It is even more important to examine the methodologies used to make evidence-based treatment decisions. Meta-analyses — observational epidemiologic approaches relying on insurance claim databases — can be used to obtain comparative effectiveness information, but as Temple (2012) warns, although this method is fast and may have “huge power,” those advantages “do not make up for potential bias and confounding.” Temple cites the unreliability of epidemiologic approaches to detect small effects, such as where epidemiology studies of hormone replacement therapy and calcium channel blocker toxicity have yielded incorrect results. On the other hand, large observational databases and pooled trial results could be used to learn more about the subgroups of patients who benefit from therapies (Garber 2009). Sox (2012) believes that CER must “identify clinical characteristics that predict which intervention would be most successful in an individual patient” as well as “subpopulations of patients that are more likely to benefit from one intervention than the other.”

Benefits and harms

An essential element of CER is to understand the benefits and harms of an intervention. Much of the practice of medicine occurs without measuring the outcome of a medical treatment.

The FD&C does not limit the manner in which a physician may use an approved drug — physicians are free to prescribe as they see fit, following what they believe to be best practice according to their training and clinical experience.

Ideally, CER may provide the data to eliminate care that provides no benefit. Chou (2007) claims that routine imaging for low back pain does not improve outcomes and may actually cause harm. Marko (2012) states that, overall, in excess of 50 percent of patient treatments have not been shown to have “clear evidence of clinical efficacy.” For example, vertoplasty has generated controversy about its worth; two studies have reported that the procedure is no better than sham (Genentech 2011). Similarly, treatment variations may result in less-than-optimal care. A classic article by Chassin (1986) shows considerable variation among regions of the United States in the use of medical and surgical procedures to treat patients with like conditions. In the United Kingdom, similar variations in medical practice helped to motivate CER efforts (Walley 2012).

Personalized medicine

What may be best practice for pediatric patients may not be best for the elderly. Comorbidities also can affect treatment outcomes.

Some believe that CER may lead to “cookbook medicine.” Garber (2009), for example, believes that CER could take decision making out of the hands of physicians and “may impede the development and adoption of improvements in medical care and stymie progress in personalized medicine.” But Garber also believes that CER offers a way to hasten the discovery of the best approaches to personalization and will provide better information with which to craft a management strategy for individual patients.

Conway (2009) says that the purpose of CER is to provide information that helps clinicians and patients choose the options that best fit the individual patient’s needs and preferences and believes that CER “can drive innovation and enable the practice of more personalized medicine.”

One area where CER evidence can be useful is coronary care. Mortality is similar overall for patients treated with either percutaneous coronary intervention (PCI) or coronary artery bypass, but results differ significantly by age: mortality is much lower for surgery among patients age 65 and over and lower with PCI for those age 55 or younger (Hlatky 2009). Regardless of age, people with diabetes seem to do better with coronary artery bypass (Hlatky 2009).

Impact on drug innovation

The impact of CER on drug innovation — and especially biologics — is a critical issue. CER aims to encourage desirable innovations and discourage those that add little or no value. In CER activities outside the United States, the impact has been, arguably, to encourage important or transformational innovations as opposed to small or “me too” innovations that add little value. Instead, sponsors are undertaking riskier, potentially transformational, innovations and are gaining health authority approval and reimbursement after CER (Carrier 2010). Here in the United States, although incremental innovations may be highly profitable in the short run, critics say that drug makers should focus on bold innovations.

With respect to government spending for research, Naik (2009) points out that the transition of investment into practice, enabling new laboratory discoveries to reach patients’ bedsides, is frustratingly slow. Naik uses the example of primary PCI during acute myocardial infarction, which was shown to be superior to fibrinolytic therapy in controlled clinical laboratory settings. Ten years after the first efficacy studies for PCI were published, however, less than a third of hospitals were performing PCI within 90 minutes of diagnosis (Naik 2009).

Vernon (2011) is particularly critical of CER and argues that many assume that the increase in healthcare spending arises largely from new drugs, medical devices, and diagnostics. Vernon postulates that CER could well require an increase in the number of patients for clinical trials, thereby increasing the cost and time for the development of new drugs and reducing incentives for such investment. New therapeutic biologics cost, on average, $1.2 billion and take about 12 years to develop (DiMasi 2003). Moreover, new FDA requirements for evaluating cardiovascular risk for such drugs are already adding to the size of clinical trials.

CER has the potential to increase required studies post-FDA approval in order for drug sponsors to obtain market access and reimbursement from private insurers and CMS. Vernon estimates that over a 10-year period, CER may contribute to a $31.6 billion reduction in research and development costs. Such a result would be clearly undesirable in view of the great benefits provided by innovative drugs.

Changing behavior

Changing the behavior of patients and physicians is a major obstacle to the success of CER. People are creatures of habit, and physicians also are often reluctant to change their prescribing preferences. Patients are slow to act on critical health information. For example, the importance of lifestyle changes such as a healthy diet, exercise, and smoking cessation and how these changes can significantly reduce morbidity and mortality is well known. Yet, poor habits continue — rises in obesity and diabetes prevalence can attest to this.

Patient compliance with a prescribed therapy and the impact on health outcomes is another major issue. One drug may be more effective than another, but patients may decrease compliance or stop taking the prescribed drug because of side effects or for other reasons. Thus, the more-effective treatment becomes a less-effective treatment. A real-world example of this is illustrated by WellPoint’s retrospective outcomes study in patients with mild and severe asthma (Tan 2009). The results indicated that members who were compliant with inhaled corticosteroids achieved the best outcomes. However, the study also found that those patients were less compliant than patients taking oral medications. As a result, the oral medication may have better outcomes.

Also at issue is the patient’s option to choose. If two drugs are equally effective, will insurers offer the same out-of-pocket cost for both and allow the consumer to choose the drug?

Benefits vs. costs

Americans are afraid that if CER takes cost into consideration, it will lead to the rationing of care. Providers are concerned about CER because they are usually paid on the basis of services delivered — CER could reduce the demand for their services and thus their earnings (Walley 2012).

Taking cost into consideration does not necessarily mean that high-cost drugs would be disfavored. For example, even though warfarin is more expensive than aspirin, use of warfarin has been shown to have lower long-term costs for high-risk patients with nonvalvular atrial fibrillation (Eckman 2009). On the other hand, CER could increase overall healthcare costs if those treatments or drugs deemed more effective are also more expensive (DeMaria 2009, Brixner 2011, Walley 2012). When CER was first introduced in the United Kingdom, cost was not a consideration. Eventually, a rule for quality-adjusted life-year values that took cost into consideration was developed, and the National Institute for Health and Clinical Excellence (NICE) has developed a cost-saving guide (NICE 2012). In the United States, the Center for the Evaluation of Value and Risk in Health’s Cost-Effectiveness Analysis Registry contains details of nearly 3,000 published cost-utility analyses (CEVR 2012).

If two drugs are equally effective, will insurers offer the same out-of-pocket cost for both and allow the consumer to choose the drug?

Standardizing data

Despite the progress achieved so far, consensus is needed on standardizing CER. With various organizations working on CER, how will stakeholders evaluate all of the data to make the best treatment decision for patients and who would decide on the standards for CER? PCORI’s methodology report was accepted by its board of governors in May and a public comment period began in July (PCORI 2012b).

In an effort to address the standardization issue, Bryan Luce, of United BioSource Corp., along with other researchers, has developed 13 principles that could improve CER. They include setting explicit and meaningful objectives, actively engaging stakeholders, incorporating methods for assessing clinical outcomes, transferring study findings across patients, and creating a plan for disseminating information (Luce 2012).

Moving forward

CER offers the potential to reduce inappropriate or ineffective medical treatments. Both private and public payers could save billions of dollars through the reduction of ineffective treatments (Marko 2012). But CER could also add to reimbursement delays and stymie drug innovation.

The types of studies and endpoints required for regulatory product approval differ from those that a clinician may consider in a clinical practice setting and also from the endpoints that payers consider in a cost-benefit analysis. Although it would be ideal to align all clinical trial endpoints for each specific audience, building alignment into the equation during premarket evaluation of a product often is not feasible. Thus, comparative effectiveness enters the equation after FDA approval and payers are left to make coverage decisions based on a plethora of data from RCTs, observational and outcomes studies, and systematic reviews. Moreover, each payer evaluates the data in various formats provided by various sources.

Payers and patients will confront many contentious issues as a result of CER and efforts to control costs. For example, patients will want payers to cover treatments that may be shown to be ineffective and payers may be more aggressive than is warranted by CER findings in denying coverage (Blackstone 2010).

Ultimately, both the medical community and patients will benefit when CER is in a standardized format, properly reflected in health benefit design, and incorporated into clinical decision making. Physicians should retain the ability to adjust care based on a patient’s individual clinical characteristics.

Proper communication is vital. If the methodology and results of CER studies are not communicated clearly to the medical community, patients could be adversely affected — for example, by the lack of treatment or incorrect treatment because of the healthcare provider’s misinterpretation or confusion about the data.

Properly applied, however, CER has the potential to help achieve better health outcomes and lower overall healthcare costs.