The Premarket Review Process: New Requirements for Manufacturers
Susan Alpert, MD
Director, Office of Device Evaluation
Center for Devices and Radiological Health
Food and Drug Administration
What we have today is a unique opportunity - a convergence of issues and attention from the clinical community, the manufacturing community, and government regulators, all recognizing that we have had a problem with both use error and with the products we use to take care of the population of the United States. In the 1970s, this issue resulted in new device regulations, where Congress attempted to introduce a level of safety and a level of performance assurance of products by implementing regulations for medical devices. This set the stage for guidances, standards, and approvals to assure that manufacturers were making devices well, and that users were getting good instructions, labeling, advice, and training on how to use devices. It was thought that this would add to the safe delivery of medical care.
Well, here we are 20 years later, and we've made some progress in device regulations. The manufacturing community has paid attention to what device regulations have said, and to what Congress was saying about device regulation. They have paid attention to safety issues, added fail-safes, made things easier to use, and wrote better and clearer language in labeling and directions for use. This has brought us far, but we are not there yet. We still have far too many patient injuries and deaths that we, the clinical community, and the patient community clearly believe are preventable if only we pay more attention to why those errors happen.
We have to recognize the environment in which we are working, and we work in many. We work in the home environment, our own work environments, and health care environments. Since medical devices are in all of these places, each of these settings faces human factors challenges, especially considering the distractions and the competing issues that arise. The devices that are used in these environments, therefore, bring with them different questions, issues, and solutions on errors in use, problems in use, and use itself. We have a unique opportunity here. We are all focused on the same thing at the same time, and already much groundwork has been laid to prepare us for preventing the errors we are currently seeing. We now know how to make health care delivery and the practice of using devices safer for all of us.
I would like to talk about where we stand on the issue of human factors at the Food and Drug Administration (FDA). For those of you in the manufacturing community, I think you will find that, although I speak specifically about the Center for Devices and Radiological Health (CDRH), what I'm saying really relates to what has been going on in your own companies in terms of human factors. I will also speak on how far we've progressed in our consideration of human factors----how they impact us and what we need to do as we move forward. I will try to tell you where we've been, where we are now, and where we hope with your help to go in the very near future.
In the past, how did we handle the issue of human factors? You have probably heard from manufacturers specific incidents or specific attention being paid to devices. The fact is that until now we have considered human factors in devices on an ad hoc basis. When we noticed there was a problem, we looked at it; if we thought human factors could help, we got involved. It wasn't an organized practice, and it was very device-specific. We also occasionally went global. For example, the Office of Communication, Education, and Radiation Programs (OCER) implemented global policies on how to write labeling for home use and over-the-counter use products. We focused on the fact that if you can't tell people how to use a product and to correctly follow directions, then you can't market the product. That's human factors. But we implemented this method only occasionally, and we did not globalize on many issues.
There was also limited participation at that time. We tended to do things inside FDA with only a bit of consultation outside; but we soon recognized we don't have all the expertise in the world inside CDRH. We need your help, the help of the manufacturing community, and particularly the help of the user community----that is an aspect of product assessment we haven't mined as well as we could have. Hundreds of thousands of people in the United States can teach us much about the use of medical products and devices, and about how to communicate with them so they can use devices more efficiently, effectively, and safely. It's our job to solicit input from them, have them examine devices before we finalize designs, and have them assess directions before we send the product out there into millions of hands.
The New Program
Today we have a convergence of attention from all of the affected community: regulators, manufacturers, and users. We have to make things easier to use, simpler to use, and safer to use so that we can all do our jobs better. Whether this effort is being driven by liability or lawsuits doesn't matter. What matters is that we all need to globalize our efforts to save each other time and to do a much better job together than any of us could do alone.
We are studying in the Office of Device Exemptions (ODE) the impact of human errors on use and injury. We don't simply look at premarket information; compliance and postmarket stages are also involved. For example, in compliance we are looking at design and how products get developed. We consider premarket, manufacturing, postmarket, our guidance development, and our staff's understanding of what factors contribute to both the good use and the risky use of products.
Global attention is now being paid on all levels in all areas. We are educating staff within ODE on human factors, and we offer courses through our staff college to teach reviewers. I realize certain engineering schools now teach human factors issues, but it has generally not been a part of the education for those who work at ODE. So we are training people by examples, with experience from industry, and with adverse events reporting from the community. We train them to think about how the devices they evaluate are actually used, and whether the design of the product, the directions for use of the product, and the clinical experience with the product support the assumption that attention has been paid to the user and the user interface.
We at ODE have always had excellent review on manufacturing, design, and engineering. We've paid less attention, however, to the final interface, to the clinical user, to the in-home user, and to the patient user. We are now focusing training and education on that aspect. After all, the use itself is the real end event. You can design and create the best device in the world, but if people can't use it, then it doesn't do what it's intended to do.
What else do we need to do? We need to educate not only on the inside, but on the outside as well. This seminar itself is one of the "education issues," but that doesn't mean that we think we have the answers. We have many questions, and we're here to ask as much as to answer. We need to develop educational information, not just for the users or manufacturers who understand, but for the thousands who are making their first product or getting into medical device use, manufacturing, or oversight for the first time. Together, we have to develop the right educational tools by interacting with you, with human factors experts, and with user experts.
We need to develop guidances. In the past, we didn't always go out and ask for input when we developed guidances and the device evaluation program. That has clanged over the last several years. We are spending more time getting advice and sounding out our guidances in the manufacturing community, in the clinical community, and in the user community. The best guidances play a very important role in identifying the easiest path for a product to reach the marketplace. That’s what they are intended to do. They are not intended to be restrictive or oppressive, but to pave a nice, clean path. If you follow the guidances and pay attention to the issues involved, your product will wind up in the marketplace as soon as you are done developing and testing it. You'll have already paid attention to the issues we will examine when we look at your dossier, 510(k), or pre-market approval (PMA) submission.
We are looking for evidence that you have consistently made your product, been concerned with how the product works, and tested the product either in animals or in actual use. We have not included this method of evaluation in 95% of our guidances, but it's an important issue for all medical devices. We have to be able to put a human factors requirement into our guidances in all areas where there is a human interface with the device. You need to help us do that, because we can't devise those guidances by ourselves. The guidances are intended for you to use to help each other and the user community.
Our goal, with your help, is to evaluate human factors in investigational device exemptions (IDEs), 510(k)s, and PMAs. Please don't panic! It is not our intention to add "new" things. We are simply going to be looking at this area in a way that we haven't done before, and a little more consistently. We are going to expect to see some documentation a little more consistently. But we also think we need to pay attention to whether or not a problem or potential problem exists, and not do knee-jerk types of checks. It is important, when we make an assessment, that a device has been established to be safe and effective for use. Safe does not mean it won't shock you when you put your hand on it; it means users can actually use the device safely, both for themselves and for the patients on the other end.
IDEs allow manufacturers to tell us how they've designed and tested their products and why it is now appropriate to start clinical trials. We are going to be checking to see whether a device fits a small, large, or left-handed person; whether you can actually do all you need to do with the device while you are paying attention to the patient; whether you need multiple people to operate the device; whether bells, whistles, or lights indicate when things are not working; and whether these signals are located in places that make it easy to find errors or device failures.
We are also thinking about where it would be appropriate to pay attention to design controls under IDE. As any of you who deal with IDEs know, when manufacturing a product for IDE or clinical trial use, the good manufacturing practices (GMPs) don't have to be in place yet. At that stage, you are exempt from GMPs. You don't have to pass inspections, but you do have to show us in your IDE that you can actually make the device consistently and well. This tells us that you have actually designed and evaluated your device and that it can be used in the way you are telling people to use it. In some cases, you may want an evaluation of human factors issues in your studies.
For 510(k) submissions, verification of paying attention to design controls and GMPs are important for Class II products and for some Class I products. Again, we recognize there are no preapproval device-specific inspections. In this area, you need to establish that your human factors issues have been paid attention to. We do not think it is necessary to have elaborate documentation and oversight for all products. Still, you need to think about human factors and see if it is necessary in your product area. In PMAs, human factors consideration is clearly always an issue because the device is a brand-new product. We are going to be interested in seeing how you dealt with human factors. We are not going to be telling you how; we are not going to impose strict criteria. But we will expect to see that you have paid attention to the issue. We are also going to try to develop guidances and labeling standards.
In developing guidances, we will incorporate the work of AAMI and examine the guidances being developed by AAMI and other standards organizations to see how to assess human factors. We are not going to try to create for ourselves a new and different independent standard. We are going to, as we do in many areas, incorporate and involve standards organizations, voluntary groups, and information developed by industry consensus and regulators to determine what tools to use in evaluating devices. We are currently trying to prioritize within ODE to decide where to place our efforts. We are putting in priority order devices that exhibit user interface problems or potential problems. We will also decide if more attention paid to human factors issues can ensure safer, better, and easier use of the product. We need help in prioritizing. We are going to be reaching out for advice and comment from the manufacturing and clinical community to tell us if we have the answers and if we have focused on the right devices.
We are all in the same situation. We are in a time of reduced resources. We cannot do everything on every product in every instance for every manufacturer or every user. None of us has the capacity to do that. We need to prioritize and decide when our oversight on an area or issue is value-added. Because we in FDA are seeing a broad view of the whole industry, we need to determine when we can provide oversight that you as an individual manufacturer cannot. When do we need to get involved and when should we be leaving it up to you? When is it inappropriate for us to be involved?
In the end we all share the same goal: No preventable injuries and deaths. We should not have deaths and serious injuries that could have been prevented. It is unacceptable, and we have to do the best we can in each of our roles in the production of safe medical devices to prevent it. This is what ODE sees, what we expect to see, and what we expect from you in terms of helping us get there.