While key details are still missing, the picture that's emerging is surprisingly restrictive

How many companies stand to benefit from the FDA Pre-Cert program? Fewer than you might think

By Bradley Merrill Thompson
A generic startup

Photo by Caiaimage/Chris Ryan

Profile of companies that might find value in the program

While this is a bit simplistic, at a high level here are the characteristics that I think might steer companies toward or away from the program. I explain how I arrived at these in the next section.


Characteristics that make Precert look attractive

Characteristics that make Precert look unattractive


Company characteristics





Large companies

Small companies


Highly-experienced in medical devices

Thinly-experienced in medical devices


Domestic companies

Foreign companies


New entrants

Older drug, device and biologic companies with legacy quality systems


Product characteristics


Novel products

Derivative products


Medium risk devices

Either high or low risk devices


Multiple products within a given therapeutic area

Only a few products within a given therapeutic area


Products that are likely to need frequent, modest, modification

Products that are likely to be more stable than average


Software that makes use of machine learning

Software that largely relies upon programmed rules


Products that are unlikely to produce any noticeable, unhappy experiences

Products that are likely to produce a more than average number of unhappy experiences


Employee characteristics


Employees who enjoy a high level of structure and process

Mavericks and freethinkers


Those who like complexity

Those who like simplicity and predictability


Those who are comfortable with a pervasive level of regulatory oversight, with a confidence in government decision making

Those who are not comfortable with a pervasive level of government oversight, concerned that FDA may be overly cautious in its decision-making

Analysis of those profile factors

Company characteristics

1. Capitalization

This may be the biggest factor, because participation in the precert program is going to be expensive. Getting a company in shape to meet the excellence appraisal criteria is typically going to take money. It will require investment in lots of policies, procedures, training and auditing.

With startups, the typical approach is to implement a quality system when the product is nearly complete, just prior to the clinical study that is required before a 510(k).  They do this to conserve cash early in the development process. Indeed, from a business risk standpoint, investors may insist that companies focus on more fully developing the product before investing in a quality system. Trying to qualify for the precert program would require a much earlier focus on quality systems and the other elements of the excellence appraisal, burning cash earlier and creating un-welcomed business risk.  So investors will have to agree to invest more money earlier in the lifecycle of a startup if they want the company to take this route.

Throughout this spring, I’m visiting top engineering schools across the country to meet with startups to talk with them about, among other things, the precert program. While we are just getting started with the so-called Startup Roadshow: FDA Regulation of Artificial Intelligence Used in Healthcare, I’m learning quite a bit from the attendees. In the first program in Ann Arbor, in the panel of industry experts, there was a clear split in their view of the precert program on the basis of the level of capitalization they enjoyed. Those entrepreneurs who are trying to get a medical device company started by being very careful with their funds seem to have no interest in pursuing precertification. Those that had a high level of funding felt like they would potentially benefit.  My concern is that the program may be a barrier to entry deterring those who don’t have the resources.

2. Company Size

FDA says they want to make the program available to large and small companies alike. But when you read the FDA’s discussion of its excellence appraisal, and look at some of the questions that, for example, focus on a company having a proactive culture, if you are a couple of folks in a garage developing your medical device, you might have to laugh. The FDA’s approach to requiring systematic policies and procedures simply doesn’t fit the culture of a small startup. FDA’s concept is very focused on large organizations and the need for highly-structured policies and procedures designed to achieve organizational excellence.

3. Experience Level

In talking with entrepreneurs coming out of universities, many of them are engineers who are trying to learn the regulatory requirements as they go. For them, the precertification program makes no sense. The precertification program assumes that the company has wide and deep knowledge of regulatory requirements, and that the company institutionalizes that knowledge in policies and procedures, as well as human resources. So if, for example, a startup relies on consultants for regulatory expertise, as many of them do, it will be very difficult for that company to participate in precertification.

This issue was the focus of several questions directed at FDA during an agency webinar on February 7, 2019. Members of the audience were concerned that the program did not appear to be practical for startups.

The FDA officials on the call said that they, in fact, want the program to be useful to startups, and that they believe that while startups may not have a lot of traditional documentation required by the program, they may well have information “in other forms” that would be sufficient to prove a culture of quality and organizational excellence.  FDA did not elaborate on what that would be.

FDA pointed to the fact that a nonprofit called Tidepool is among the pilot participants. Tidepool has about 30 employees profiled on their website.  Among other sources of funding, in December 2018 the Helmsley Charitable Trust and JDRF granted Tidepool with the initial funding it needs – $6 million – to begin developing Tidepool Loop, a hybrid closed loop automated insulin delivery app.

FDA’s definition of a startup does not seem to include a handful of folks coming out of a university with little money and little experience to develop a new technology. The agency’s hope that the program will be useful to startups is mostly just that, hope. There is very little reason to believe that true startups will be able to navigate the program practically.

4. Nationality

One of the themes of the precertification program is that regulation is all about trust. If FDA can trust companies more, they can in theory give them more leeway. So the program requires trust, and uses things like transparency to create an assurance that FDA can trust that the patient is protected.

There is no politically correct way to say this, but FDA has inherently less trust of foreign firms because the agency has inherently less authority over them. This is true both because of the legal difficulties of bringing enforcement actions against foreign firms, but also the budgetary limitations that make it difficult for FDA to inspect and otherwise work with foreign firms.

The pilot program does not include any foreign establishments. It’s difficult to understand right now exactly how practical this will be for firms overseas. It’s quite possible that the program will not treat foreign firms equally with domestic ones.

5. Company Age

There are drug, device and biologics companies that have been around for quite some time and are heavily invested in their current quality management systems. Many of these organizations are large and these legacy quality management systems are extremely complex, and interwoven throughout the company’s business processes. The new excellence appraisal process upends that traditional quality management system, and requires radical revision to it. Older companies that are heavily invested in their existing quality management systems may not view the cost and disruption that would be required to dramatically those systems as worth the uncertain benefits.

Frankly this might seem like a bit of a contradiction from the discussion above regarding large companies favoring the precert program, and to some degree it is. Large companies stand to benefit more in some ways from the program, but the precert program also in some ways will cost large companies more. So in a sense, both are true. 

But this also means that new large entrants into the industry will stand to benefit the most. So, for example, Apple may well have future plans that include a scale of medical devices that would benefit significantly from participating in precert, and at the same time, Apple doesn’t have a lot of legacy good manufacturing practice policies and procedures from years of medical device operations that will need to change. So in a sense, a company like Apple stands to benefit most from the new program.  In contrast, large medical device companies that have been in the industry for a long time will have to weigh the benefit against the large cost of revamping their extensive legacy systems.

Alternatively, large companies might want to create new units to specifically focus on software as a medical device in order to minimize the disruption on their legacy systems. So, for example, companies in the medical imaging space that have large hardware units may want to create special, dedicated business units to develop image analysis software programs. By creating dedicated business units, they can avoid disrupting some of their extensive legacy quality systems.

Product characteristics

6. Product Novelty

This factor might change if FDA goes to Congress to get statutory authority when it comes time to implement the program beyond the pilot. But as of right now, FDA is only making this program available for those products that are technically, currently in class III on account of them being unprecedented, but which should be reclassified into class II as explained in the next section. Again, the two Apple Watch apps FDA reviewed in August of last year are an example.

As explained above, FDA is using its de novo authority as the legal basis for the program. But the de novo process only applies to those products which cannot be compared to other products already on the market. Thus, at least until we see legislative authority, this program will be restricted to only that one percent of new products that is unprecedented.

7. Product Risk Categorization

The primary benefit of the precertification program is quicker reviews for class II medical devices. If your firm makes class I medical devices, there is absolutely no benefit to participating in the program.

Further, it’s unclear how the program would treat class III medical devices. By their very nature, if a product actually belongs in class III due to its safety and effectiveness profile (as opposed to simply due to its novelty), it is not a candidate for de novo reclassification. Therefore it would be excluded from the program.

Within class II, some products are 510(k) exempt, and they likewise would obtain no benefit from being in the precertification program. Thus we’re really talking about the higher end of risk within the class II category. It’s those products that would ordinarily trigger a 510(k) that might one day find themselves either newly exempt or subject to a quicker review under the precertification program.

8. Breadth of Product Line

In some ways, the precertification program is all about finding a more efficient path for FDA oversight. The idea is that by creating an early gateway process called excellence appraisal, the FDA can have the opportunity to review important policies and procedures that apply to more than one device. Thus, to really get benefit out of the program, a company will need to have a reasonable number of different SaMD products that can all be reviewed quicker or even not at all on the basis of the company’s investment in the excellence appraisal. The more products, the more benefit.

9. Frequency of Modifications

One of the major selling points of the precertification program is that FDA plans to adopt new criteria for determining when a new clearance will be required for product modifications. So if you have the kinds of products that need to be modified frequently, you will be a greater beneficiary than one whose products only require occasional updating.  Companies spend much time and energy debating if a “note to file” is adequate or if a special or full 510(k) is required: the precert approach might help with this. I say might because the FDA hasn’t released any details on how it will implement this aspect.

10. Development of Artificial Intelligence-Based Products

FDA came up with the idea precertification in part to address the need to regulate products that evolve over the course of their lifecycle. FDA found it challenging to clear software based on AI (including machine learning, rules engines, expert systems neural networks, expert systems) that could evolve after the clearance. It simply made the agency nervous because performance can degrade over time if the company isn’t careful. Further, FDA felt as though its postmarket authorities were simply too weak to give the agency adequate assurance that it could remedy products that did not perform well in the marketplace. So the idea is that FDA can more readily clear devices upfront, if the agency has greater assurance that it can both monitor and remedy problems that arise postmarket.

This is, of course, theory. Those of us who are old have seen the agency try to do this before with postmarket surveillance. Most of the folks I know who went through an expedited clearance process only to agree to postmarket surveillance regretted it later, because the surveillance turned out to be enormously burdensome. But the theory is, FDA can require less premarket assurance if they have greater postmarket authority. I guess we shall see.

11. Anticipated Product Experience

Every company I talk to will say that their product is outstanding, very safe, and their customers love it. What else would a company say? But by definition, not every company can have above average customer experiences. More to the point, there are simply some clinical conditions and some technologies that are more likely to produce some sort of frustration and increased risk on the part of customers.

Take, for example, the business of blood glucose meters. These meters are very common, and they play an incredibly important clinical role in helping people with diabetes manage their blood glucose levels. But if you simply look at FDA databases regarding experiences of customers, and if you look at the complaint files of companies that produce these products, frankly you might think the world is about to end.

A big part of the explanation is simply how big the market is, meaning how big the denominator is. Lots of products means lots of opportunities for problems. Another part of the explanation is that people with diabetes often have challenges associated with the heavy time burden of their disease, so they don’t always follow basic steps like washing their hand or changing their lance before checking their glucose. They also can have complications like finger numbness and poor   vision. Those real world situations can make operating medical devices difficult. Many people with diabetes are also older, and may have difficulty with new technology with which they are unfamiliar. There are lots of reasons associated with a given clinical need and a given technology that might mean we should inherently expect a higher level of complaints and bad experiences for that product category compared to other product categories.

Every company that enters the precert program is going to have to love the sunshine. They’re going to have to love sharing with FDA, and the public, everything including complaints, and even experiences that do not amount to a complaint but simply in some way suggests product performance is imperfect.  They are going to have to enjoy being the subject of FDA and public scrutiny and questioning. So if you’re going into an area where you can expect some hiccups along the way, you are going to need to think really long and hard about whether you want all of those hiccups scrutinized and debated publicly.

Employee characteristics

12. Company Culture

FDA is trying to change company culture. The agency is pretty upfront about that. One of the main advantages to FDA is that they see this program as encouraging companies to adopt a culture of quality and organizational excellence. So, make no mistake, your company will have to change its culture to meet the standards. And the kind of culture FDA wants is a highly structured one based on lots of policies and procedures and metrics and audit training. So, if you have a liberal arts degree, you may not like it. As already explained, the agency’s focus is on large organizations where policies and procedures tend to define the company’s culture more than the people who work there.

13. Complexity

There’s an old joke in Washington about the tax code: If the tax code is “simplified” one more time, it’s likely no one will understand it. 

Regulations tend to get more complex, not less, and the precertification program is certainly no exception. FDA is taking what might be only a modestly complicated system of getting new products cleared through the 510(k) process, and is transforming it into a multilayered, multistep, multidimensional, total product lifecycle program that frankly makes my head hurt. 

The consequence is that the program will favor those who have an appetite for complexity, which amounts to another barrier to entry. Firms that invest in figuring out how to navigate the system may well have, in the end, a competitive advantage. Regulatory lawyers are the only ones who are sure to benefit – more complexity means more hours billed interpreting and explaining that complexity. Wait a minute. I guess I like the precert program.

14. Political Philosophy

I really didn’t set out to engage in political commentary. But there’s just no getting around the fact that a person’s worldview may very well color how he/she perceives this program. FDA wants to get much more deeply involved in your business, moving away from a pre-market encounter to a total product lifecycle partnership. FDA wants to be involved actively in the entire lifecycle of your products. This whole program is about how FDA can be more involved in your business. FDA will have broader visibility into what you are doing through the excellence appraisal, and will have greater insight through the transparency requirements both as they relate to the continuing excellence assessment as well as the product performance through shared real-world data. And FDA wants to be right there with you as you decide how to respond to the real-world data you collect.

Some people will find that acceptable, and some won’t. Some will be concerned that while people at the agency do their best, they are still human. Not only do humans make mistakes, but we are all subject to bias. Government regulators have built-in, job-related reasons to be cautious in their decision-making, and FDA is no different. So as FDA’s role expands in the decision-making that occurs throughout the product lifecycle for software, we need to anticipate that FDA’s involvement will be more cautious than business people would be on their own.

So, all in all, you have to be prepared for FDA to be your partner, and make no mistake, they will be the senior partner. They will be the ones with the legal authority to tell you what to do. If you’re okay with that, and if you like the agency’s approach to decision-making, you will love the precertification program. If you’re skeptical of government, and believe that they don’t always make the right decision, this program is not for you. 


I already gave away the punchline. I think a lot fewer companies will find the precertification program attractive as they study it. And in some ways, that’s very deliberate on FDA’s part. The program seems at least partly intended to define a certain club in the industry, those that are willing to trust the FDA and disclose everything to the public; those the FDA believes operate with this agency-defined culture of quality and organizational excellence. 

The program will have competitive implications, helping some and hurting others. That doesn’t make it wrong, though. In the end, we all are trying to ensure that patients get what they need and are safe while getting it. But at the same time, I think it’s clear that the program will competitively hurt many small companies that have been the source of great innovations, even if they aren’t in some sort of management nirvana. That’s why in the end, this proposal will need to be debated in Congress, and by the stakeholders that influence them, in order to make judgments about whether the benefits are worth the cost, and about whether the program is optimally designed.

Nike Air Presto Fly