Real-World Evidence Might Take a Decade To Be Fully Established

A conversation with Jennifer Graff, National Pharmaceutical Council VP

Managed Care recently caught up with Jennifer Graff, vice president of comparative effectiveness research at the National Pharmaceutical Council. In an email exchangeThe exchange has been edited for clarity and length., Managing Editor Frank Diamond got some of her thoughts about the future of real-world evidence.

Do you think that RWE has been overhyped?

The potential for RWE and real-world data [RWD] sources to better understand how treatments work across care settings and more diverse patient populations is real. However, there is potential overwhip on when RWE will have an impact on patient care decisions, health system efficiency, care delivery, and patient outcomes. Will RWE be of sufficient quality and will it be trusted in the next 10 months, the next 10 years, or will this be a 40-year evolution?

What do you think, Ms. Graff? 10 months? 10 years? 40 years?

I believe it will take the next 10 years for RWE to be used consistently. Whether this is ultimately a 10-year or 40-year evolution will be determined by how quickly cultural changes occur. Will data be interoperable to routinely allow RWE analyses? Will there be enough data scientists who can use credible methods? How long will it take for decision makers to consider “best available evidence” vs. relying on small slices of information based on decades-old hierarchies of evidence which place the design of the study (e.g., systematic reviews then RCTs then observational studies) above whether the study can credibly answer their question of interest?

The FDA seems to be wary about RWE; CMS less so. Is that a fair assessment? Can the government do anything to move RWE along?

The FDA under Scott Gottlieb has made a good, but wary, first step with the recent FDA framework for a RWE program. The FDA requires a high level of certainty in their decision making, and we see that in their measured approach. However, FDA is not just putting forth a single guidance; they have a robust program including many forthcoming guidance documents. Their 2019 budget request includes $100 million to support a modernized RWE data system, and in 2017 the FDA finalized a guidance related to RWE and medical devices.

At CMS, RWE is critical to measure the effectiveness of treatments in Medicare and Medicaid beneficiaries, to design value-based agreements, and to test whether alternative payment models succeed or not. For drugs, it will be interesting to see how Medicare Part D plans and managed Medicaid groups consider RWE when determining formulary coverage, patient care programs, and care delivery.

But despite all this, 10 years?

The private sector will play an important role in the 10-year time frame. Patients are downloading their data from patient portals and sharing their information between their specialists. Provider groups and clinical societies will become more adept at using data to inform care pathways. Vertical integration across providers and health systems creates incentives, if not the infrastructure, to remove the silos created at doctor offices and hospitals. Plans will compete on their ability to manage care more efficiently. And we can’t underestimate the impact that big technology companies could play as they expand their health care footprint.

Where do you think RWE will break through first? Oncology? Where do health insurance plans fit in all this?

Oncology has the potential to lead the way in using RWE due to the implications for patient health and health care costs. In oncology, we are learning that tumor expression, not the site of the tumor, is associated with treatment benefits. RWE can help inform what treatment pathways work best and for whom. In addition, oncologists are familiar with using registries and documenting information to inform care, but some elements (e.g., radiology reports and some lab data) are still hard to access and integrate.

In contrast, for conditions where patient symptoms are critical to the assessment of care, the use of RWE may be further out. For example, in mental health or musculoskeletal disorders more routine capture of patient symptoms and integration into care systems is needed.

Clinical and IT execs at health insurance plans are routinely using RWD. How RWE is used varies across plans. For example, RWE is routinely used to conduct utilization management, assess treatment adherence, or inform network design.

RWE is used by more sophisticated plans to develop clinical pathways, to develop predictive analytics about which patients may develop a condition or benefit from a treatment, and to develop the foundation for value-based contracts. Clinical executives within a health plan will not only have a vested interest in RWE, but RWE will be a critical component for differentiation and competition.