A Newly Candid FDA on the Future Of Medical Device Regulation
We all know the Food and Drug Administration is one of the most important entities for those who develop drugs and medical devices, but for the public, and even many of the people in industry, it remains a notoriously mysterious organization. So I was very excited about the opportunity to have an in-depth phone conversation with Jeffrey Shuren, who heads the agency’s Center for Devices and Radiological Health. Shuren has played various policy roles at the FDA since 1998 and has been at the helm of the center since September, first as acting director, and then as official director since January this year.
When we spoke, Shuren had just returned from a Town Hall meeting that the FDA had sponsored in Boston, and where medical device manufacturers were not shy about voicing their concerns about the agency. Chief among them was how difficult it is to predict how the agency will react to submissions for device approval. Shuren was very conscious of this being a major problem, especially for small companies.
I sense the agency is really trying to reach out to the public, to industry, and to members of the press. This is just my personal experience, but when I used to be a reporter for a major national newspaper, covering very high-profile medical devices, I was very rarely able to get on the phone with top FDA officials, in spite of numerous interview requests. At the time, the agency spoke to the press mostly through spokespeople, and often solely via e-mail. It is very refreshing to see Shuren, as well as Joshua Sharfstein, the agency’s deputy commissioner, giving interviews these days. This is the kind of transparency the agency has needed for a while.
Anyway, here’s an excerpt of my interview with Shuren. Among other things, we talked about how the agency is taking a more proactive role in not only ensuring device safety, but also balancing safety with efficacy. We talked and about the future of the controversial 510k program, through which many manufacturers get their devices approved following few or no clinical trials by showing that the new devices are “substantially equivalent” to ones already on the market. These are issues I bet many of you in the industry will be interested in.
Sylvia Westphal: How did things go at the town hall meeting?
Jeffrey Shuren: I found the session very valuable. Having a two-way dialogue for people to talk about whatever is on their mind and to have us engage in a back-and-forth is the best way to understand what’s most important to the people who we regulate and also the people we serve, because ultimately we are serving the public.
SW: I know manufacturers spoke up fervently.
JS: I think there were some very important comments that we heard and number one on the list is predictability. Providing predictable and consistent regulatory pathways to market and being predictable in other actions that the center takes is paramount for industry to plan for what they need to do to develop a device and get it to market, and also for venture capitalists to be willing to risk putting money to back the development of specific devices. In the medical device industry so many of these companies are small businesses that need to find sources of funding in order to drive their R&D, unlike the bigger companies which are able to generate that funding based upon sales of products they already have in the market.
SW: Do you think it’s going to be doable to offer industry more predictability? Does it boil down to clearer guidance, or more frequent dialogue between reviewers and each of the companies? How do you think you fix that problem?
JS: First of all yes, I think it is doable. Guidance for individual kinds of devices is one piece of it but there is a lot more that we need to be looking at in terms of how we actually run the pre-market review programs, particularly the 510k program, and how we provide input and expertise in the development of clinical trial protocols in support of the few 510ks that need clinical data.
Science constantly changes and our understanding of the risk/benefit profile of a device changes overtime. Sometimes, with greater experience and more data we realize that the device is safer than we originally thought. and you can imagine then we may need less information for a similar device coming in the door. In other cases we learn of a new safety problem and now our expectations for what a manufacturer would have to submit to us to demonstrate their device is safe and effective is going to change, because they’re going to need to address this new safety concern. What’s essential here is that when we recognize that the science has changed, and our regulatory expectations need to change accordingly, how can we communicate that to industry in a legally permissible manner that gives them as much advanced notice as possible? The way it’s done today is either a one-off ad-hoc approach, where a manufacturer learns about [the new issue] when they come in to see us—they may be coming in with their submission and we say “Well this was good before, but its not good anymore,”—or we could put out guidance. But the challenge with guidance is that it’s a very time-consuming process to develop, issue, get comment on, and finalize a guidance document. So we are exploring other options. All of these issues we’re going to be talking about in two reports that we’re hoping to release in the next few weeks. One is specifically on the 510k program and the second is on the use of science in regulatory decision making.
SW: And just to make sure, the one thing you’ve made clear is that the 510k process it going to stay, right?
JS: That’s right, the one thing I can say is that we are not recommending that the 510k program be abolished. As for other recommendations, it’s premature to say because we’re still wrapping up the report.
SW: Give me the broad perspective of what your hopes are for your office and what you wish 5 years down the road to have accomplished.
JS: I think, first and foremost, it’s having a smarter FDA and taking what we call a balanced public health approach. Our mission says that we are to protect the public health and to promote the public health. In the past, much of that focus has been on the protection side. There’s more we need to do there, but we also have to recognize there is the promotion side—that on the one hand we have to assure that the devices patients get are safe and effective and we use our enforcement tools wisely, but on the other hand we need to promote public health by facilitating innovation and fostering a culture of quality and prevention among industry. It’s striking that balance that I would like to achieve both on a culture level for the center and in practice.
One of the realizations we have come to as a center is that we have a unique experience and expertise that’s not available to industry, the venture capital community, the healthcare community, and even many of our foreign regulators. We see every moderate- and high-risk device before it goes to market. We get proprietary data on devices that isn’t available to anyone else but the individual manufacturer, so the manufacturer knows what it knows about its product, but we know about their product and their competitors’ products and we know about similar products. Even many of the foreign regulators don’t review medical devices before they’re marketed; they have third parties who do it. The third parties see some devices, other third parties see other devices, and much of that information is kept with the third parties and not reviewed by the regulator. We don’t do that here in the US. We review everything ourselves and we look at all of it, and having that depth and breath of experience and expertise puts us in a place where we can give insights back to the manufacturers that can be helpful in addressing problems with existing devices and that can be used in developing the next generation of medical devices.
SW: Can you give me concrete examples of this new approach where you give insight to industry in a way that wasn’t done before?
JS: A great example of how we are doing business differently already is what we have done with external infusion pumps. We were seeing that the number of adverse event reports pertaining to these devices was going up—we were seeing many recalls, about 87 recalls over the past five years—so external infusion pumps as a group were in kind of the top ten devices for adverse events reports and the top ten devices for recalls. Of the adverse event reports that came to us, many of them, according to the manufacturer, were mistakes on the part of the doctors and the nurses. But when we drilled down deeper we found it really wasn’t their mistake, it was problems with the design of that device.
SW: What kinds of problems?
JS: One was in the interface between the pump and the person who uses the pump. For example, we found that the on/off button was right next to the button you push to start the infusion, so sometimes people spent all this time to program the infusion pump and then they hit the wrong button, so instead of hitting it to start they turned it off and then walked away. Another example we saw was something called ‘key bounce,’ where you push a button and you would think if I hit the button for a number the number gets pressed once. Say I’m going to give 20ml of some drug every hour and I hit 2 and 0, it should be 20, but in a number of instances you’d hit 2 and 0 and it wasn’t 20, it was 200. If you’re dealing with a drug that in higher concentrations has significant safety concerns, like the blood thinner heparin, you’ve put that patient at significant risk.
Normally our standard operating procedure would have been that we’d identify problems and we’d go back to the manufacturer and say “Your problem, you figure out what to do and then come back and talk to us.” Here we said “We’re going to come out and make our expertise available to work with you to solve this problem.” And that’s very different.
SW: How so?
JS: We have worked with the University of Pennsylvania to develop a generic infusion pump software code so a manufacturer can take that software and build on it to make its own device or they can use it as a benchmark for their own software to try to avoid common problems.
Also, we are offering to perform what is called static analysis, where we use software to look for some of the most common problems and go back to the manufacturer and say “We identified this problem, you should fix that now.” We’ve offered to perform this at any time in the life cycle of the product, so a manufacturer can come to us as they’re developing the device. Normally they would come to us with a submission of data when they’re done with the device and they’ve analyzed and assessed it and they’re asking us to review this information to determine whether or not to approve it. Now we’re willing to do a diagnostic on their technology while they’re developing it, so this potentially could help some manufacturers save a lot of time and effort. We have not done this before.
SW: Is this unprecedented for the agency to be coming in and giving manufacturers a specific path to product design? It seems very unique to me that you’re coming in and saying “This is how everybody should design their pumps,” and that some of the changes are geared towards working around the possibility that an operator might make a mistake.
JS: You are exactly right, it is unique. This idea that you have to design around the way people behave is exactly on target. It’s one of the reasons why we came back and said that for many pumps, before they can come on the market, they have to test them in the environment in which they will be used, or simulate an environment in which they will be used, with the people who would be using them. It’s what we call usability testing, so you can assess whether it was designed in a way that’s going to lead to problems with how people use it.
SW: If you had your pick of something you don’t have and you’d like to get right now for the FDA and for your office what would that be?
JS: One of the challenges that I face is that if have an expert in an area, particularly in the medical disciplines, I may at best have one expert and in some cases I may have none. That creates a challenge for our ability to stay on top of the science, particularly for cutting-edge science. If you want to understand the technology you need to have people who understand the technology itself and then you need to understand how it interfaces with the body.
One of the things that’s a priority for us is setting up what we call a network of experts so that we can more readily tap into people outside of the agency who are knowledgeable in critical areas to help us answer scientific questions. This is different from how we use our advisory panels. There are times when we need to pose a question to an expert to try to understand a new technology. The challenge, of course, is setting this up in a way where we’ve identified the experts and we understand their relationship to the technology and the makers of the technology, so we’ve looked at potential conflicts. Then we need to assure we have the right experts in the agency who can, if you will, talk the talk. So for example I’m a neurologist; you aren’t going to send me in to have discussions on orthopedics. I’ll understand it but I’m not going to understand it in the same way that another orthopedist would, and vice versa.
SW: So you would hope to maybe recruit more people within various areas of expertise?
JS: We’re looking at several options. It is one of the things that we will be assessing.