Smart phones and their associated stand-alone software applications, or apps, have become ubiquitous within the field of Medicine and perhaps, nowhere more so than the forward-thinking, technology embracing, occasionally geeky world of Emergency Medicine. A survey in 2010 showed that 99% of medical staff had a mobile phone, and 81% of them used a ‘smart’ phone – figures that are unlikely to decrease any time soon. And crucially people don’t just have a phone, they have it within arms reach 24-7.
And, along with the smart phone technology, have come applications or ‘apps’. These little pieces of standalone software that run within the smart phone operating system and do everything from help you to find where you parked your car to take photographs like it’s 1977 (Ed – when I would have been watching the Fonz!).
Although individuals had written software for other hand-held devices such as the Palm Pilot, Apple was the first company provide the software and create a platform to facilitate writing and uploading programs for all software developers, from the bedroom amateur to multi-national companies. Crucially they also had a ready-made platform in the shape of iTunes, where hundreds of thousands of loyal users (with their associated bank details) found out how easy it was to download the apps onto your phone. Many of the apps are free, while most others are available for less then the price of a half-decent cappuccino.
A staggering 25 billion apps have now been downloaded from the Apple iTunes store from a selection of over 550,000 apps, generating revenue of over £2.5 billion for the developers. Apps are also available for other brand phones with software available from Google, RIM, Nokia, Palm and Microsoft all having their own app stores for different devices, though Apple are far and away the market leaders with an estimated 80% share of the market. And, along with all the apps for essential tasks such as fighting zombies or choosing your sofa, have come the medical apps.
The medical app section is enormous and encompasses a huge selection of apps than run the entire gamut from ABGs to Zimmer Arthritis. Apple recently split the Health section into a Health and Fitness section and a Medical section but even this section remains enormous. It’s also important to realise that the categorisation and descriptive purpose of the apps are still up to the software author. This is why you find the critical care apps jostling for space alongside the crystal care apps on the virtual bookshelves.
The available apps cover a wide variety of functions and purposes with variable degrees of utility and success. The best apps take advantage of the excellent memory, processing powers and graphics of these devices to perform functions so useful that you wonder how you used to live without them; other, more staid applications have cumbersome interfaces or simple replicate a text-book with four times as many pages as normal due to the tiny real-estate of the screen.
And so, as with any other category of apps, there are good apps and bad apps (crapps?) to which a natural selection process is applied. The good ones will be recommended by word-of-mouth or on-line reviews, the poor ones will be mercilessly criticised, or ignored. This process of peer recommendation works well for games or office utility apps but is not really good enough for apps in specialised fields such as medicine or aviation navigation.
While Apple have an approval process before an app will become available on their store, the purpose of this is to avoid code errors messing up anything else on your phone and also to make sure that there is nothing about your app that someone, somewhere may find offensive. What they do not do when you upload your app is check your qualifications, calculations or recommendations.
So, what happens if the writers of your app accidently advise you to use the amethyst instead of chalcopyrite, or put a decimal place in the wrong place in your drug calculation? Events at CERN and with the Hubble telescope have reassured us that even teams of the most brilliant international scientists can occasionally screw things up, so the odds of all 100,000 medical apps being completely error-free could be considered, at least, unlikely and some caution in applying… and that my friends is perhaps why the MHRA has decided to stick it’s oar in……!
Medicines and Healthcare products Regulatory Agency (MHRA)
In the UK, the MHRA has stepped into the breach to regulate the ever evolving and expanding universe of medical apps. This is a brave decision given the hundreds of thousands of apps that have already been downloaded millions of times (it could be argued that the bolting horses have already run off with the lid for Pandora’s box).
The first question to be asked is ‘what is a medical app?’ The MHRA have specifically addressed this issue firstly by distinguishing stand-alone software from software that is integral to a medical device. So, the software in a medical pump would not be considered a medical device as it would be considered part of the pump, which is itself considered a medical device; software running on a phone that you use to calculate the pump settings, however, might be considered a medical device in its own right.
The MHRA guidelines from 2010 further define such as medical device by stating that ‘the software must facilitate some transformation of entered data, software that simply facilitates the storage and retrieval of patient data does not count as a medical device.’ So software that stored a patients blood results and then allowed retrieval of these results, would not be a medical device; but software that calculated an estimated GFR from the blood results might be. Or, software that stores an image and allows retrieval of the same image would not be a medical device, but software that manipulated the image or performed calculations based on it might be.
Or, if I read the calculation for acetyl cysteine dose for patients with paracetamol overdose on the College of Emergency Medicine website then this is not considered a medical device. If I then perform this calculation on the back of an envelope with a stubby pencil then these do not constitute a medical device. If I do my sums with a hand-calculator or even use the calculator app on my smart-phone then this does not constitute a medical device. However, if I write a piece of software that asks someone to enter the data then performs the calculation for them then this may be considered a medical device.
The guidelines then go on to state that the status of the software may depend on the qualifications of the person using it: ‘Decision Support software will generally not be considered a medical device if it exists to provide information to enable a healthcare professional to make a clinical decision. However, if it performs a calculation or the software interprets or interpolates data and the healthcare professional does not review the raw data, then this software may be considered a medical device.’
My interpretation of this is statement is that if an app has a disclaimer stating that it is only for use be a suitably qualified person who knows what they are doing, and then provides clear information about each calculation it is performing, then this may not need to be registered as a medical device.
However, European guidelines (which would also apply to the UK) from MDDEV in 2012, state that, ‘with respect to Decision Support Software, where the software is intended to provide healthcare professionals and/or users with recommendations for diagnosis, prognosis, monitoring and treatment of individual patients then it does qualify as a medical device.‘
The guidelines distinguish the ‘individual patient’ status to mean that if an app states ‘nobody should smoke tobacco’ then this would not constitute a medical device whereas an app that compiled data about Patient X and then recommended that ‘Patient X should not smoke tobacco’ would be a medical device.
So what is involved in registering an app with the MHRA?
Registering an app with the MHRA in order to get a CE mark of approval can be done for the fairly nominal sum of £70, (although this may be enough to dissuade the casual coder who already devoted a few days of his life hunched over a computer screen in an altruistic effort to save mankind) but what is actually involved in the process is less clear.
Most medical apps that perform some sort of data transformation related to the care of an individual patient would appear to qualify as Class 1 Medical Devices. This means that they need a quality assessment to be performed, but how and by who are not specified. There is also a mention that the assessment of quality may sometimes need to be conducted by a Notified Body, an independent company registered with the MHRA (or European equivalent) whose involvement adds very stringent quality control and several thousand pounds or more to the whole process. The requirement of a Notified Body assessment would therefore limit the application process to larger companies who would be willing /able to spend such sums.
I can only find one app that has actually received MHRA approval to date, the very helpful Mersey Burns app. This app, which aids in estimating the size of burns and then calculating any fluid requirements, is available for free so I would assume that it did not cost thousands of pounds to process. The associated website states that the authors have carried out their own quality assessment study to assess the performance, and have built in quality control checks for the calculator aspects that performs numerous self-checks calculations before proceeding with any user data.
I suspect that this app will be the first of many to be approved and as experience in this field grows, the rules and guidelines around this approval process will become easier to traverse.
It can be argued that this new layer of bureaucracy is necessary to prevent patients being exposed to the dangers from clinicians blindly following the advice of error-strewn apps, although this level of scrutiny has never been applied to textbooks. People may also provide poor or wrong information or advice on websites without coming under the auspices of the MHRA but it appears to be the inclusion of calculations that triggers the need for regulation.
As is to be expected for such a fast-moving field, some questions remain:
- Where is the line drawn between medical devices and non-medical device apps?
- What happens to apps that are not MHRA approved?
- Are apps from anywhere in the world subject to the same scrutiny before being released in the UK?
- Will MHRA approval be required for the hundreds of thousands of apps that already exist and have been downloaded?
- What form of quality assessment needs to be performed and who by?
- Who is responsible if a medical error occurs as a result of using an app?
I have emailed the MHRA seeking clarification on these points and will hopefully provide an update when a clearer picture emerges.
bw
Craig Ferguson