The Speed of Innovation
Mobile radiology apps are evolving at the speed of imagination — but can regulators keep up?
Apple's iPhone kicked off the mobile device revolution just five years ago, and now there are well over a million apps in existence. Tens of thousands of these apps are designed to either improve or maintain user health, doing everything from measuring blood pressure to monitoring a patient's adherence to certain medical treatments.
A fraction of these medical apps are designed for radiologists, transforming a smartphone or tablet into, for instance, a portable workstation on which to perform occasional primary diagnoses.
The number of these mobile medical apps is growing rapidly. According to Research2Guidance, a research firm specializing in mobile technology, 500 million smartphone users will be running health-related apps by the year 2015. Before a subset of these apps can be marketed to the public, however, software developers must make their existence known to the federal government. But with so many new tools being developed so fast, some have begun questioning whether the government can regulate them quickly and efficiently enough to avoid stifling innovation.
Regulations for the 21st Century
Since health-related apps are being created at a breakneck pace, some software vendors worry that government oversight could slow the developmental process. The U.S. Federal Drug Administration (FDA) regulates medical devices, which include a subset of health-related apps known as "mobile medical apps." Important benchmarks of the agency's efficiency in regulating these devices and apps had steadily worsened over the past decade so, in 2011, the a plan of action that has improved the review process. Later that year, the agency put forth a draft guidance, to be finalized in 2013, that designated a small subset of mobile medical apps that the agency will oversee. In addition, the guidance outlined into which pre-established medical device categories mobile medical apps should be placed, further clarifying the process for developers. These categories are named Class I, Class II, and Class III.
When evaluating the market readiness for mobile medical apps, the FDA applies a risk-based approach to classification: Class I represents apps that pose the least risk to a patient, while Class III designates the apps that carry the greatest risk. Apps that fall into Class I or Class II, which are deemed substantially equivalent in safety and effectiveness to an already legally marketed app, use what is called a 510(k) pathway, or a premarket notification, to gain FDA marketing clearance. In other words, these mobile medical apps do not represent a total departure from other apps already in existence that have proven safe and effective for users. Software developers who submit information about their apps for an FDA premarket clearance through the 510(k) pathway are usually notified within 90 days that their product meets the requirements, after which time they may proceed to market their product with the FDA's clearance.
If, on the other hand, the vendor is developing a Class III app, they must apply for premarket approval (PMA). Class III apps are comprised of "those that support or sustain human life, are of substantial importance in preventing impairment of human health, or which present a potential, unreasonable risk of illness or injury." PMA represents a more rigorous review process, and developers must demonstrate safety and effectiveness of their product to the FDA in order to receive approval to market it. This designation applies even if a smaller app is already being legally marketed.
Innovation Under Threat?
According to the FDA, all of the radiology mobile apps it has approved have fallen into the Class II category, which, explains David S. Hirschorn, MD, director of radiology informatics at Staten Island University Hospital, has caused little disruption in innovation. This largely because, to his mind, the 510(k) process is so short. "Of the software vendors I've talked to, no one has said that the FDA is dragging its heels during the review process," he notes, saying that the organization strives not to over-regulate medical apps. Indeed, the agency treads lightly when it comes to medical software. "With the exception of mammography," he asserts, "states regulate the practice of medicine, not the FDA."
But while the FDA generally cannot regulate the practice of medicine, medical app manufacturers do come under its purview. The agency distinguishes whether or not it has oversight authority over a product based on a product's intended use, which largely depends on how a vendor markets it. Hirschorn illustrates the fine line the FDA walks with an example: "If a magnifying glass is advertised as something that can help you see smaller things, it's not a medical device. If, however, it's advertised as helping you see micro-calcifications in a mammogram, then it's a medical device," and the agency has right approval over it.
Bradley Merrill Thompson, general counsel of the mHealth regulatory coalition in Washington, D.C., agrees. While he would like to see the premarket notification process for mobile medical apps cut down, he doesn't believe that the FDA is overly surpressing innovation. "The pace at which the agency moves is dependent upon its available resources," says Thompson. "It's a balancing act between speed and resources — it can only hire so many staff members without getting more resources either from Congress or from user fees. The FDA has to go through a deliberate product-approval process but also try to move as fast as possible."
Time Is of the Essence
Others, however, are not convinced that the FDA's approval process paves the way for innovation. Joel C. White, executive director of Health IT Now, in Washington, D.C., states that the life cycle of many mobile apps is shorter than the time it takes to go through the FDA approval process. "Some apps are updated weekly, and in some cases daily," he says. "Turnaround time will be crucial to ensuring the latest, best information is available for providers to treat patients." He recommends a more hands-off approach to regulation: "At a minimum, software that poses the least risk should fall outside the FDA review process. Only when a product interacts with a variable like human physiology should the FDA be involved."
But there are steps developers can take to reduce turnaround times, argues FDA policy advisor Baul Patel. "One point I can emphasize to help reduce frustration is the need to talk to us early in the development of mobile medical apps," he notes. "The FDA can discuss data needed to show safety and effectiveness with developers so they can plan those requirements into the development of the app. This can help developers to find the quickest path to market, help them use resources more efficiently, and shorten review times."
Although finding the quickest route to market may be helpful for developers, however, it is the end product — and the end users — that matter most. In Hirschorn's experience, radiologists tend not to suspect the safety of the software they use, often assuming it is safe unless told otherwise. "They focus more on the clinical task and less on the software itself," he explains. "As a consequence, though, they are very unnerved if they learn that what they were shown by a computing device was not what they thought it was. This implicit trust can lead to a strong distrust if the radiologist feels she has been betrayed by the technology," he says, emphasizing the need for government oversight to ensure that products work as they should.
The rapidly changing landscape of mobile medical app development is a cause of both optimism and unease. By working with developers and vendors, the FDA is making an effort to alleviate this concern. While innovation will only increase the speed at which new technologies are born, spurring unimaginable new treatment options, the public's safety must also be taken into account. Radiologists would do well to ensure that the apps they use are tested thoroughly so that their trust in the technology remains strong.
By Chris Hobson