Linking Insurance Coverage for Innovative Invasive Procedures With Participation in Clinical Research

A recent randomized controlled trial (RCT) evaluating a self-expanding intravascular stent (the Wingspan stent system), a device intended to open stenotic intracranial arteries via angioplasty and thereby prevent thrombotic stroke, reported surprising results.¹ Based on case series data, this type of stent system was expected to be better than usual care; however, patients who underwent the stent procedure were more than twice as likely to experience stroke or death within 30 days compared with patients treated with aggressive medical therapy (14.7% vs 5.8%).¹ These “negative” test results were disappointing. A promising new approach to a serious medical problem turned out to be worse than current best practice. However, within this failure was an important lesson worthy of notice and support from patients, clinicians, and policy makers. Through cooperation between the US Food and Drug Administration (FDA) and the Centers for Medicare & Medicaid Services (CMS), this trial represented the successful use of a policy mechanism that can balance the goal of rapid access to innovative medical procedures with the need to obtain rigorous evidence on the risks and benefits of these procedures before they become widespread in clinical practice. The key was to make initial coverage conditional on participation in clinical research.

Expensive invasive procedures often become widely adopted in clinical practice without high-quality evidence of effectiveness.² In part, this is because the FDA does not regulate invasive therapeutic procedures unless those procedures involve use of medical devices. Even novel interventions using FDA-approved devices often receive unconditional coverage from public and private insurers without rigorous evidence of effectiveness produced by RCTs, the standard required for new pharmaceutical products. These differences in regulatory scope and evidentiary standards contribute to the proliferation of expensive invasive procedures for which the clinical effectiveness, in comparison with other less expensive and often more extensively studied treatment options, is unknown. Early, unrestricted insurance coverage grants physicians and patients rapid access to “innovative” procedures, but with significant disadvantages—limited understanding of the actual risks and benefits of new procedures for different kinds of patients; barriers to recruiting clinicians and patients to perform subsequent RCTs; and creation of financial incentives that will often fuel overuse of new procedures, increase health care costs, and undermine attempts to gain better evidence of comparative clinical effectiveness to guide treatment decisions for future patients.

However, the policy tools to address this problem are available. When novel and promising invasive procedures lack adequate evidence of effectiveness, insurers can decide to offer coverage conditional for patients participating in RCTs designed to develop the needed evidence. Medicare has implemented this approach under the label of “coverage with evidence development in selected cases, including lung volume reduction surgery and home use of oxygen.”^{3 - 4} The example of the stent for cerebral arterial disease¹ demonstrates how this approach can achieve the difficult but critical balance between rapid access and better evidence on the safety and effectiveness of new procedures.

The investigational stent¹ was approved by the FDA under a humanitarian device exemption (HDE) in 2005.⁵ The HDE allows device manufacturers to market a device without going through the regular FDA approval process if several criteria are satisfied, including estimated use in fewer than 4000 patients per year in the United States and the existence of preliminary evidence suggesting a favorable risk-benefit ratio.⁶ The self-expanding stent system met these criteria, based on a small case series suggesting a low stroke rate compared with historical controls.⁷

However, even with the HDE, there was an important barrier to using this procedure in practice or to studying it in randomized trials—Medicare had a standing national coverage determination denying coverage for all intracranial stenting procedures. Once aware of the HDE approval, however, Medicare changed its national coverage determination to cover the device and procedure, but only in the context of an RCT. Unconditional coverage would have inflated the number of patients treated while simultaneously creating major barriers to any attempt to recruit physicians and patients for participation in an RCT. At more than $20 000 per procedure,⁸ costs would have increased while evidence on actual risks and benefits would not have been obtained, and thousands of patients might have been treated with a procedure that subsequently was shown to be less effective than medical management. This was avoided by the willingness of the FDA and CMS to work together to harness the interest in innovation with the needs of future patients and clinicians for better evidence. The clinical trial of the stent procedure was able to recruit patients quickly, and the results were published less than 3 years from commencement of trial enrollment.⁹ This RCT represents the fruit of a policy mechanism that strikes the right balance in fostering innovation while protecting patients and the financial sustainability of the health care system.

Ideally, making early coverage conditional on further evidence development would be a routine feature for invasive procedures that lack solid evidence of effectiveness produced by RCTs. However, neither the funding nor the infrastructure is available to perform an RCT on every new procedure. Additionally, Medicare and other insurers come under political pressure from clinicians and device manufacturers whenever they limit coverage to participation in RCTs. Accordingly, priorities must be determined when coverage with evidence development is most appropriate. Regulators and insurers may consider the following criteria as relevant to decisions to link coverage to requirements for comparative effectiveness research: (1) serious medical condition for which current outcomes are often poor; (2) large patient population potentially eligible for the procedure; (3) high risks potentially associated with the procedure; (4) early evidence from more than 1 well-conducted cohort or case series studies showing consistent results better than, but not substantially better than, historical controls; and (5) high cost of the intervention.

These criteria could help policy makers make reliable, transparent decisions regarding which innovative invasive procedures should be prioritized for early coverage with evidence development. The legitimacy of cost as a criterion might be disputed, because neither the FDA considers cost in approval decisions nor does the CMS in making coverage decisions. Nevertheless, having high costs as a criterion reflects the societal interest in having more certainty of clinical effectiveness when substantial health care resources are at stake. Targeting costly interventions with a wide eligible patient population also serves the interests of patients who may be burdened with copayments.

Coverage with evidence development is not a panacea. Not all patients potentially eligible for a given procedure will meet inclusion criteria for an RCT, and others will be unable to enroll for various practical reasons. There are residual concerns that this approach may feel coercive to some patients and physicians, although patients have no right to coverage for medical interventions that lack adequate evidence of effectiveness.⁴

The overall lesson of the stent procedure study should stand out amidst the major policy challenges facing the US health care system. The collective interests of patients and society can be served by rigorous evaluation of invasive procedures before they become widely available. Timely comparative effectiveness research, supported by linking initial insurance coverage with requirements for RCT participation, is a powerful tool for making medicine evidence-based and for improving the quality of health care.