International regulators need and want to rely on each other's regulatory decisions for efficiency but don't want to lose sovereign autonomy. Here's an approach to unlocking "reliance"
To achieve reliance: separate scientific review from value judgment.
This week, the International Medical Devices Regulators Forum (IMDRF) convened in Washington, DC, to discuss issues and opportunities for greater efficiency in medical device regulatory review. Ideally, once a new device is available from a manufacturer, a patient should have access to that device as soon as possible.
In practice, devices often become available in one country well before becoming available in another. For example, people who use insulin living in Europe had access to Dexcom's much-awaited G7 well before people in the U.S. If reliance was in effect between Europe and the U.S. FDA, the G7 may have become available to Americans more than a year earlier.
There are many barriers to achieving trustworthy reliance. There is also reason to make sure we don't reach over-reliance (I believe over-reliance would cause fragility; as reliance increases, one missed risk has a broader impact). At this IMDRF session, trust was raised as a major theme. How can a country trust another country's review? How would we measure the performance of one country's regulatory review to build and maintain that trust? Speed and cost of regulatory reviews are obvious KPIs to measure. But what about quality?
How do you measure the quality of a regulatory review? It is hard. Very hard. Regulatory reviews require a lot of expertise. (I had the honor of leading one of the assessments of the FDA's review performance under MDUFA IV. Trust me, it's not trivial.) To start, break it down.
Ideally, regulatory reviews assess a product's benefit-risk balance. If the benefits outweigh the risks, the product should be approved. I'm going to ignore a lot of complexity for the moment (e.g., 510(k)s versus PMAs, instructions for use, etc.). Consider the review in two separate steps. (1) Scientific review and (2) value judgment.
Scientific review: Does the evidence support the claims? What is the remaining uncertainty?
Value judgment: Assume the evidence supports the claims. Do the benefits outweigh the risks for the indicated context of use?
Separating these pieces of a review can achieve clarity even within a single review team. At CDER and CDRH, in cases of tough decisions, my teams helped FDA reviewers break deadlocks and structure discussions by disentangling uncertainty from value judgment. One team was locked in disagreement about a drug that had a few middling benefits and one rare but serious risk. My FDA decision science colleagues and I designed and facilitated a working session to separate the review team's perceptions of scientific review versus their individual value judgments. The results were immediate and stunning: the team finally could see that they were not disagreeing on the scientific merit but on the individual value of each benefit and the major risk. At CDRH, in a retrospective review of a tough de novo decision, our team studied a case of a device with high uncertainty in the benefit but negligible risk. The review team initially decided not to approve the device for marketing. They were ultimately overturned on appeal because the novel technology was so beneficial to the world in its potential and the risks so low that it was deemed worthy to prove its benefit in the real world.
This concept of separation is already well-established in the medical regulatory world. In fact, it's a best practice in decision science. The Center for Drugs (CDER) has a benefit-risk framework that supports this deceptively simple approach. The European Medicines Agency has been field testing the PROACT-URL approach, which is also deceptively simple. Both methods are rooted in and driven by decision science research.
Separating scientific review from value judgment may also have significant positive effects on providers and patients. Individual patients with the same disease and demographics may have very different value judgments on the same scientific risk and benefit profile. For example, imagine a treatment for a skin disease that may cause a risk of muscle soreness. One patient with the exact same measurable demographic profile (e.g., age, gender, socioeconomic status) may have a drastically different value judgment of the skin disease treatment versus the muscle soreness. The patient may benefit from a reading of the regulator's judgment of the scientific evidence of the risk (how certain was the risk signal?). Providers could better guide patients in their treatment choices with a deeper understanding and separation.
I want to be cautious: The issue of reliance is MUCH more complex than what I offer here. I do not suggest that reliance is easy or that this is the only way to gain faster access to safe and effective medical technologies. I also do not pass judgment on the current methods of communicating review results. I deeply revere our regulators and appreciate the work they do on behalf of human health and our economies. Rather, I offer this concept as an approach to continuous regulatory innovation.
~Shannon, the Optimistic Optimizer
Ps. Scientific review is also more measurable as a KPI than a value judgment. KPIs were a concern and topic at the session.