Maintaining a Sustainable Environment for Innovation: Ending the Evidence Arms Race for Regulatory and Reimbursement Decision-Making

Increasing attention on containing healthcare costs has spurred new business opportunities in data aggregation, real-time analytics, and rapid translation of data into evidence on the value of new innovation. This has become a race for more and faster evidence. But is the investment made by manufacturers, payers, and the government resulting in better evidence to aid patients and clinicians in making better decisions?

The Affordable Care Act crystallized the need for manufacturers to quickly move beyond traditional price negotiations and keep pace with a system evolving to paying for value and creating a more transparent market of information to consumers around what constitutes high-quality care. Though ACA, the government made foundational investments with an increase in funding of comparative effectiveness research through the Patient-Centered Outcomes Research Institute (PCORI) and development of new performance standards for incorporation in future federal programs.

Over the same timeframe, the Food and Drug Administration is engaging beyond its traditional regulatory function to include additional and deeper collaborations with other stakeholders, including sponsors of research and consumers of technology. Indeed, integral to this year’s user fee debates are the ways in which, over the next five years, they will define post-approval evidence standards. This will solidify the FDA’s role in post-approval evidence generation.

Gillian Woollett, Allison Colbert, and Tanisha Carino focus on FDA regulatory strategy and policy issues within Avalere Health LLC’s Center on Evidence-Based Medicine.

As the FDA grapples with how data adds value in determining the safety and efficacy of technologies, payers attempt to determine appropriate reimbursement paradigms in parallel. As a consequence, multiple stakeholders are asking new questions on the relative value of new market entrants. However, disparate demands from multiple masters may yield an “internet without Google” - an excess of data of widely variable quality, lacking standards for what data is most useful to integrate into decision-making.

Life sciences manufacturers then face enormous and often unquantifiable risks when making large research investments that they will encounter changing expectations from any or all stakeholders. At best, demands will be incessant; at worst, they will conflict.

It is therefore time to assess the direction and trajectory of these “evidence standards” and their likelihood to deter life science company investments in those therapeutic areas deemed, in a post-blockbuster era, to be “higher risk” of regulatory or reimbursement uncertainty. These evidentiary standards will also affect industries’ inclinations to innovate with new products or better methods for manufacturing.

Key considerations for ensuring market success have become whether companies are able to:

• Anticipate many years ahead of FDA submission the appropriate evidentiary base at launch, and post-approval, to enable timely access for patients;
• Understand and be able to navigate the respective roles of FDA and payers in determining pre- and post-approval data requirements, and resolve conflicting evidence requests;
• Engage with new entrants, and influence the evidence thresholds throughout a product’s lifecycle, ideally in a manner that represent good health policy and increases access over time;
• Create new mechanisms for communication to ensure that all stakeholders both understand, and then help reduce, uncertainly in the evidence standards necessary for regulatory and commercial success (or at least stop believing more data is always better).

A review of the changing landscape of evidence will help us to consider the potential consequences of a failure to define expectations that facilitate the collection of actionable data. To collect data ad hoc will likely compound the current confusion and lead to lower stakeholder confidence in health care while continuing to escalate costs. Therefore, concrete goals reflecting input from all stakeholders should determine the value of different types and extents of data investments to improve health.

Evidentiary Base for Regulatory and Commercial Approval is a Moving Target

Historically, product sponsors were able to treat market entry and commercial success using distinct functions. The regulatory team got the FDA approval and the commercial team launched the product. To do this, R&D and regulatory teams focused on ensuring that their products met evidence requirements of safety, and efficacy, while commercial teams could focus on engaging with payers through contracting based on price and volume.

In the historical model, the evidentiary base of controlled clinical trials demanded by FDA and available at launch was frequently sufficient for early commercial adoption. Payers required limited, if any, observational studies to support favorable coverage and positioning of certain technologies at launch (See Exhibit 1). The FDA required robust randomized clinical trials for approval while the Centers for Medicare & Medicaid Services (CMS) and other payers had limited need to conduct their own observational studies or to create policies that provided incentives for manufacturers to create data relevant to questions of payer interest.

Exhibit 1

Representation of the Historic Environment for Pre- and Post-Market Evidentiary Standards

Avalere Health

Over the last decade, the evidence thresholds for regulatory approval and payer coverage have become less and less clear. FDA now requires on-going data collected post-approval and payers demand additional data that in many cases could only be incorporated by companies as part of their registration trials. These evidence demands blur the development schedule and decisions on what are necessary study investments (See Exhibit 2).

High-profile market withdrawals based on new evidence have further undermined the confidence of some stakeholders in FDA and caused the Agency to redouble its efforts in post-market surveillance– irrespective of the quality of the new evidence. FDA received novel statutory authority in the last round of Prescription Drug User Fee Act renewal in 2007 – to impose civil monetary penalties for companies not complying with their post-marketing commitments.

While the majority of post-market data collection initiatives focus on safety (such as the Sentinel Initiative and Risk Evaluation and Mitigation Strategies—or REMS), FDA recently showed its commitment to evaluate evolving efficacy evidence by withdrawing the metastatic breast cancer label for Roche/Genentech’s Avastin (bevacizumab), an indication originally receiving accelerated approval based on limited data. While the product remains on the market for other indications, the case is likely a bellwether for evolving drug regulatory programs that need to embrace explicit tradeoffs between evidence generation and access.

However, if we lose the value of regulatory certainty, we risk requiring the initiation of excess studies simply to mitigate perceived regulatory risk. FDA has recently re-articulated this commitment to the iterative nature of evidence that, while unquestionably valid as a scientific matter, risks causing policy confusion and uncertainty for investors and sponsors. This may translate into additional spending that may not serve consumers well in the end. In the meantime, however, the FDA is requiring on-going post-marketing surveillance where there are known safety issues with particular products, but few post-approval formal clinical trials.

Exhibit 2

Representation of the Current Environment for Pre- and Post-Market Evidentiary Standards

Avalere Health

Over this same past decade, CMS has endeavored to become a more prudent purchaser of novel medical technologies though its use of its national coverage determination process and in particular, their Coverage with Evidence Development policy to shape research programs to prove direct benefit for the Medicare population. CMS intends for its CED policy to allow for Medicare coverage of promising, but not yet proven, technologies.

However, some argue that CED also creates a “perpetual” data collection environment for those new products subject to requirements involving data collection via registries. While these registries may include the questions most important to the Agency, they can also provide the basis for collecting data on a host of new research questions that may crop up as the product becomes more widely used. In a notable recent example, the proposed national coverage determination for transcatheter aortic valve replacement would involve CED requirements, including a post-market registry negotiated with the FDA. (https://www.cms.gov/medicare-coverage-database/details/nca-proposed-decision-memo.aspx?NCAId=257)

However, for now, such cases using an existing collaboration remain the exception. Most cases of CED seen to-date have required development of novel trials or registries to meet CMS requirements while allowing limited access to promising medical technologies. In the meantime, private payers are slow to adopt analogous limitations on their coverage decisions, but there is an increasing payer desire for additional “real-world” evidence relevant to their particular population prior to granting coverage for a new technology. Nonetheless, for product sponsors this represents a potential delay in reimbursement and for patients a potential delay in access.

For all stakeholders, the ideal evidence is difficult and increasingly expensive to obtain. PCORI includes a Methodology Committee whose charge is to guide the development and dissemination of comparative effectiveness research standards. The group is required to engage stakeholders to help shape their thinking and recommendations; however, unless clear goals can be aligned among all groups, it will be difficult to truly shift the clinical research enterprise in this country.

“Mind the Gap”

Recognizing the gap emerging between traditional randomized clinical trial standards accepted for approval and appropriate market access, FDA and CMS have initiated a pilot “parallel review” process in the medical device arena in an attempt to provide input on evidence needs for the Medicare population prior to product approval. FDA has also begun investment in a CER infrastructure programs such as the Partnership for Applied Comparative Effectiveness Science and the Janus Data Repository, although previous investments in their Observational Medical Outcomes Partnership and mini-Sentinel have to-date yielded limited return.

One can begin to envision a “future state” (See Exhibit 3) with the line for pre- and post-market evidentiary thresholds rotated even further to illustrate the influence of both entities throughout the process, with sponsors needing to consider both FDA and CMS throughout development and life cycle management for any given product.

Exhibit 3

Representation of the Emerging Environment for Pre- and Post-Market Evidentiary Standards

Avalere Health

Indeed, informal FDA/CMS collaborations already point to the beginnings of parallel review. In the transcatheter aortic valve replacement coverage decision, CMS accepted a request from the American College of Cardiology and The Society of Thoracic Surgeons to initiate a coverage analysis even though the procedure had yet to receive FDA approval.  The professional societies proposed a CED policy requiring evidence collection through a registry they co-owned.  When the FDA published its approval letter for the first TAVR therapy a month later, it called for post-market surveillance through that same registry. 

This may be a sustainable steady-state, but only to the extent these data sources (and those generated by third-party entities such as PCORI) can agree to sufficient interoperability and cooperation. However, this will take active collaboration and prioritization by all stakeholders focused on shared goals. Otherwise, life sciences companies might end up needing to make parallel and potentially wasteful evidentiary investments (See Exhibit 4), which would not incent sustainable innovation, and potentially create an excess of meaningless “evidence” generation. If the costs of this evidence compound, there will be an associated effect on access and affordability of life-saving medical technologies, with patients as the ones ultimately affected.

Exhibit 4

Representation of the “Evidence Arms Race” Scenario for Pre- and Post-Market Evidentiary Standards

Avalere Health

Agency Roles in Determining the Nature of Evidence in the Value Equation

Both FDA and payers such as CMS (whose public determinations often lead private payer decision-making) have a stake in defining both pre- and post-market evidentiary requirements. While FDA is traditionally oriented toward placebo or standard-of-care controlled clinical studies in defined populations, there has been a recent push toward so-called real-world registry designs to support the agency’s public health mission, at least on the safety side of the risk/benefit equation.

Payers, on the other hand, expect to see observational data to provide them with “real-world” information such as representative contraindications and patient compliance. Such data necessarily comes from less controlled situations and is subject to less scientific precision and, therefore, open to some criticism as well as different interpretations. This creates a vulnerability to conclusions perceived as compromised by different stakeholder priorities.

Nonetheless, it is increasingly important for each sponsor to understand and harmonize their evidence generation goals and activities among stakeholders, or decide early in development that they need to chart separate courses for each constituency. CMS has indicated that its strategy may be to harmonize with its fellow stakeholders, but this is far from predictable and the ramifications create a level of regulatory uncertainty with significant consequences for all stakeholders – sponsors, payers, providers and patients.

FDA and CMS have different authorities by statute that require them to pursue different research priorities, but without harmonization or a bright line in the transfer of authorities at the point of initial market entry, their legitimate respective efforts have the potential to conflict in a manner that has not occurred previously and may have little public health benefit.  FDA approves technologies based upon “safety and efficacy,” while CMS covers and pays for technologies only if they are “reasonable and necessary.” 

The biggest danger may be that many stakeholders, particularly consumers, assume that data is available to answer any questions, and that it is just its accessibility and interpretability that is in doubt.

That there is no data available, let alone that its collection is impossible (ethically or economically – or that statistically meaningful conclusions cannot be made) is simply not considered.

Indeed, this is where FDA may be most adamant about data quality. Just as stakeholders have called on both agencies to maintain their independent standards and judgments within the context of parallel review, there may be merit for FDA and CMS to continue identifying and answering research questions for current as well as new technologies based on their separate and distinct authorities. However, no approach will provide the “perfect knowledge” sought by many as the basis of fair healthcare decisions as a national policy objective, nor can any decisions be as timely as would be ideal for very real patients facing life and death decisions.

It is important to note that FDA’s interpretation of their mission, despite their public health focus, excludes data related to cost or cost-effectiveness in pre- and post-approval studies, which leads to a risk that some study demands will lead to a product never reaching the market at all. That payers are increasingly demanding these endpoints as part of their value judgment for a product, or for comparative judgments between products that the FDA likewise will not and cannot make, does not change the premises under which FDA operates. Only Congress can do that through a change in statutory authority. Therefore, even though payers want certain information as the basis for coverage, formulary placement and utilization management, this will remain a reasonable but unachievable request if directed at FDA.

New Entrants Influencing the Evidence Thresholds Throughout a Product’s Lifecycle

For the past few years, the activities of the federal government in funding or conducting comparative effectiveness research have been limited to a few key organizations.  These agencies have both funded CER studies and established CER priorities for funding new data sources for research as well as programs to translate and disseminate the findings of CER to payers, clinicians, and patients. However, new entrants, such as PCORI, are poised to influence the evidence thresholds throughout a product’s lifecycle. Indeed, PCORI has the potential to advance the CER enterprise even further with its large and sustained budget through 2019.

Since the current CER enterprise focuses on the patient, and the FDA and payers base decisions at the population level, the methodology standards used in their respective decision-making risk distinctions beyond reconciliation. Patients have differing levels of cost-awareness for their treatments, but they usually focus on preserving timely access and maximizing therapeutic options.

The idea of no effective treatment is rarely an acceptable outcome for anyone, but particularly not for a patient facing a life threatening condition. Unfortunately we have many conditions for which no therapies are available, let alone cures. Indeed, that is why we need on-going investment in innovative products.

FDA has made some investments in CER infrastructure, but they lack the statutory authority to demand data in head-to-head studies of different products approved in the same therapeutic area, and each approval remains a standalone decision of safety/efficacy (drugs), safety/effectiveness (devices) or safety/purity/potency (biologics, including vaccines).

The relative differences between products may be such that even given if the authority, as Center for Drug Evaluation & Research Deputy Director for Clinical Science Robert Temple noted in an article published in Clinical Trials in October, feasible studies may not be able to show a meaningful difference. However, this still leaves payers to draw interpretations from product labels, based on independent studies, and to determine whether and when to reimburse whom for use of the products in what situations.

The need for flexibility and competition make it unlikely that payers will be able to agree to evidence standards for benefit design – and indeed, the life sciences industry risks locking into specific requirements. However, as third parties such as like the cardiology specialty societies emerge to offer solutions to FDA and CMS, manufacturers must anticipate and engage them meaningfully on the development, financing, and governance of their data infrastructures.

On the one hand, there is value in coordinating evidence generation activities among stakeholders to avoid duplicative research.  On the other hand, manufacturers often create infrastructure for data collection in the hopes of answering unique research questions, running the risk of failing to fully meet any of its demands if they harmonize too much with other research activities.

Ending the Evidence Arms Race Will Require Clarified Expectations for Evidence Necessary to Demonstrate Value

Through better collection of evidence, there can be a more informed dialogue regarding the use of all types of evidence to support rational and fair public health policy decisions. This requires care and discipline in demanding answers to the right questions that can then support sound policy outcomes. It may also require acceptance that some questions are unanswerable because of the very nature of the studies that we are able and willing as a society to conduct in patient populations, as well as the intrinsic limits of the defined population sizes in which we will be able to place our confidence.

Further, any questions about medicinal products (drugs, biologics and devices) will be in the context of the overall health care system, where other variables such as the lifestyle of any given patient may have a far greater effect on the outcome of the therapy than the drug itself. There is no perfect knowledge in health care, and the complexities will compound as we deal with aging populations and concurrent conditions all increasingly subject to various medicinal interventions, each of which may affect the other.

In the meantime, through careful study of potentially precedent-setting decisions, life sciences companies can contribute to the evolving expectations for evidence and to some extent anticipate data needs pre- and post-approval to ensure the continued success and commercial viability of innovative medical technologies. They can better articulate what is and isn’t feasible from the perspectives of companies that will only remain viable if they can be cost-effective. Consumers have to become aware of the limitations of data, and their demands may move beyond just more and better products to even cause earlier project cancellations. A balance of expectations will be important as much as a balance in the data itself.

A key remaining element of uncertainty for life sciences companies is the population-based nature of decision-making at FDA and payers. Physicians and patients strongly advocate for individual clinical decisions on appropriate medical technology, which may pose inherent conflicts with regulatory and reimbursement decision-making that is based on populations. Communication between stakeholders is of vital importance and will serve to benefit all stakeholders.

Perhaps most importantly, stakeholders must keep the cost of evidence development appropriately in check, as this has a direct impact on the access and affordability of life-saving healthcare technologies ultimately available.

Comments? E-mail the author at [email protected]