One of the signs of autumn for this editor is the first email from Flusurvey. This is a brilliantly simple system that sends you an email every week asking if you have flu-like symptoms, then produces a map of the UK that gives advance warnings of epidemics. It costs nothing to join and is a great contribution to public health so why not sign up?. (They also have some exciting developments that may surface soon such as a small device that you blow into the connects to a smartphone and can tell almost immediately if you have flu’.)
Increasingly of concern to this editor, due to his deep involvement in digital health regulation, is who is working out how to regulate self-learning algorithms. It is therefore good to see the issue breaking cover in the general press with this article. For what it’s worth this editor’s view is that as technology begins to behave more like humans, albeit in a much faster, and narrow, way by learning as it goes along, perhaps an appropriately adapted use of the way human clinicians are examined, supervised and regulated, might be most appropriate. Sitting next to an AHSN CIO interested in the topic at a Kings Fund event last week, I was pleased to hear him offer precisely the same suggestion, so perhaps there is a little mileage in the idea.
DHACA (disclosure: run by this editor) has just renewed its website after a long delay, and will be updating content over the next few weeks. First off is the events page advertising:
Our Digital health safety conference on 7th November at Cocoon Networks, London, is being run jointly with DigitalHealth.London – the MHRA has now confirmed they will present so we have almost all the relevant organisations and experts in the UK speaking at this event which should be essential attendance for all involved with the development and use of digital health & care. Attendance has increased substantially in the past few days so do book soon to be sure of securing a place. Much more, including an almost-finalised agenda, is here.
DHACA Day XV – we are back to our usual location at the Digital Catapult Centre on 10th January where are building an agenda of some extremely interesting speakers. To check out the agenda development and to book in advance, go here.
Catching up in the back file of articles is another in Bradley Merrill Thompson’s (Epstein Becker Green) series in Mobihealthnews on how the FDA is biased, by its very structure, against novel healthcare technology even if low risk. He further reflects on what’s truly novel, and what’s not. ‘Novel’ means Class III clearance and potentially millions of dollars. to gain it. He reckons that 80 percent of new digital health technology doesn’t qualify as ‘new’ in a regulatory sense–it may be ingenious in transferring the color reading of a test strip to, for instance, a smartphone and an analytic back end. All the new technology has to do is to demonstrate equivalence to the clunky traditional test–in other words, incremental improvement. What he’s worried about is the 20 percent that don’t fit any FDA classification, in particular software that automates what professionals do, repurposing non-healthcare technology for healthcare use (e.g. videogames for ADHD) or algorithms that automate what’s been done manually through a different method. The social cost is that the most in need, who would benefit from novel health tech that cuts cost and improves quality for individuals and populations (that old Triple Aim), will forever be blocked from having it by regulation. “We need a new paradigm where new technology is quickly evaluated for potential risk, and placed promptly into an appropriate regulatory category.”
There are certainly regulatory parallels ex-US. Much more here to ponder for your Weekend Reading.
Previously and related by Mr Thompson: Avoiding the FDA health IT-medical device regulatory trap for general IT companies, FDA, new technology approval and the Ossification Tango
When it comes to new technologies–and drugs for that matter–the worst thing that can happen to your invention is to receive a letter from FDA that you have been classified into Class III. Based on regulations passed by Congress in 1976, there are three FDA classes primarily based on device risk. Exception: Class III. Anything not ‘substantially equivalent’ to an existing device is automatically put into Class III, regardless of risk level. Author and health tech legal advocate Bradley Merrill Thompson of Epstein Becker Green takes a comprehensive review at this flawed and outdated system that puts groundbreaking health tech at an extreme disadvantage in his latest article on regulation in Mobihealthnews. (more…)
If you are an IT company in the US or internationally with services which could be useful to healthcare companies or practitioners, it’s easy to be overly specific and stray into FDA-regulated territory. The always-informative Bradley Merrill Thompson of the Epstein Becker Green law firm delineates the fine regulatory line that general purpose IT companies must observe when working with healthcare customers. First there is intended use, based on how the manufacturer intends its customer to use the product; if the customer uses it for the diagnosis or treatment of disease or other conditions, FDA will regulate it as a medical device. This is less clear than it seems, and Mr Thompson explores where a general IT company can, in the old PR adage, ‘say it safely’ and avoid falling into the unwanted medical device trap by avoiding medical feature and advice claims, and keeping the context away from medical use. The Journal of mHealth (August)–online version, optional PDF download. Hat tip to Mr Thompson via the Continua LinkedIn group. Other articles of interest in the JMH are: Scottish company HCi Viocare and its ‘smart insole’ pressure sensors for foot ulcer detection following, Northwestern University’s research around patterns of smartphone usage detecting depression (page 19) and a lengthy article on transforming patient data into actionable insights (page 34).
Medical apps may not be strangers to doctors’ offices anymore but they also realize that apps are difficult to recommend responsibly to patients or even to find, because there is no real guidance or validation. This current article in JAMA online confirms the perception and the need for care integration that both Editors Charles especially and Donna have pointed out lo these many years. However this Editor is quite disillusioned at the attempts to date to ‘curate’ apps with the Happtique failure and the relatively low profile to date of IMS Health’s AppScript and professional review site iMedical Apps and the stated intentions of SocialWellth which purchased Happtique. The reality is that the numbers are against it–IMS Health in their study estimated 40,000 medical apps–in 2013. For apps that want to take the high road, it’s economically difficult, but could be rewarding in the long term. The WellDoc BlueStar diabetes tracking and management support app did with FDA clearance and prescription-only use, but few so far can see a revenue model there. Also MedCityNews.
Intended use determines degree of health app regulation–and also how you communicate your attributes and performance claims. Bradley Merrill Thompson, who performs an invaluable service by advising our field on regulation, compliance and interacting with FDA, demonstrates how a developer can determine where the intended use of an app might fall (more…)
As anticipated, FDA issued final non-binding recommendations for guidance yesterday (Monday) that ease regulatory oversight of medical device data systems (MDDS), including image storage and communication devices, and mHealth devices.
In the MDDS guidance document, “(FDA) does not intend to enforce compliance with the regulatory controls that apply to MDDS, medical image storage devices, and medical image communications devices, due to the low risk they pose to patients and the importance they play in advancing digital health.” It defined MDDS as “a device that is intended to provide one or more of the following uses, without controlling or altering the functions or parameters of any connected medical devices: (i) The electronic transfer of medical device data; (ii) The electronic storage of medical device data; (iii) The electronic conversion of medical device data from one format to another format in accordance with a preset specification; or (iv) The electronic display of medical device data.” along with their hardware and software. It specifically excludes devices that are used in active patient monitoring.
Mobile health apps were covered in a separate and highly detailed guidance document, “Mobile Medical Applications”.
- FDA will regulate only “those mobile apps that are medical devices and whose functionality could pose a risk to a patient’s safety if the mobile app were to not function as intended.” (more…)