Original article (IN FRENCH)
This is an article of information and opinion about the emergence of new technologies in the health world.
1. Introduction: New technologies knocking at health’s door
2. Smartphones, Smartwatches, Apps and Companies
3. Health online
4. Tech & Health trends
5. Google, Apple and IBM: big companies embracing health
6. Medical device connections
7. #Help: Communication drifts
For more information about these subjects, you can find a video (in French) of my appearance as a guest on an episode of Tech Café.
1. Introduction – New technologies knocking on health’s door
The subject of new technologies is incredibly expansive and this article discusses the favourite themes of eHealth: internet, smartphones, mobile apps, web giants, and emerging technologies such as 3D printing, VR etc.
Many of these technologies are developed; they are efficient, reliable, easy to use and low-cost, but some of them are still developing.
Users have adopted new technologies that everyone now uses to buy, learn, work and even our private lives have become inundated, too.
Industries and services became quickly involved in this popular movement by making their solutions “connected”.
The new entrepreneurial generation is growing up fully aware of these possibilities, and they are cognizant of the new wave of technology inspired by Silicon Valley. This impulse is encouraged more and more by governments through simplifications, financial help, accompaniment, federations, labels, etc.
However, health hasn’t yet undergone its digital revolution. This isn’t surprising when we understand that the health sector is highly regulated, people’s wellbeing is in the balance, evolution is long and expensive, and there is more resistance to change here than in other sectors.
That being said, here we are. New technologies are entering the health market, and this is the beginning of the end of the beginning.
2. Smartphones, Smartwatches, Apps & Companies
Even if some of the devices are equipped with activity sensors, heartbeat detectors, or even an SpO2 (to measure blood oxygen levels), their creators never positioned them as medical devices (MD), and as indicated in the European MD definition, the status of an object as a medical device is determined by the developer.
The Samsung Galaxy S5 epitomises this: all of the publicity surrounding the heartbeat sensor alerted South Korean authorities, and Samsung had to change its advertisements to avoid the device receiving a medical certification.
Apple has many contacts with the FDA, mostly concerning the Apple Watch, without the prior creation of an MD.
Watches and bracelets are the best candidates for medical use because the measured possibilities (due to a permanent contact with the user) may be invaluable to the patient.
Mostly medical applications.
Remember that an app can be considered as a MD: it has happened many times in Europe, specifically in 2007 on the Council Directive 93/42/EEC (concerning medical devices):
“'medical device' means any instrument, apparatus, appliance, material or other article, whether used alone or in combination, including the software necessary for its proper application (…)”.
In France, the ANSM (Agence Nationale de Sécurité du Médicament) specifies it is ici et là.
This notwithstanding, here are some of the apps you can find in appstores, from the most harmless to the most critical:
Quantified self (QS) apps for comfort purposes. Highlighted by the media a year ago, the buzz just seems to now be dying down. A QS app is used to monitor some lifestyle parameters such as the daily number of steps or metrics related to sleep quality or quantity, for example. General wellbeing is a part of health but it’s not specifically medical, so restrictions here aren’t more important than the ones on other popular products. Technical performances are variable.
Health journal apps: these simply record information without any application, so again there are no regulatory restrictions. They have basic functions such as medical history, weight charts, listing vaccinations etc.
Apps providing medical information: the target goes from the general public to professionals. These apps begin to approach something of a medical device concerning drugs, anatomy, pregnancy etc.
Apps using calculation or algorithm should really be considered as a medical devices. They are using data entered to provide a type of medical result (similar to a diagnosis or a therapeutic result). A common example is fertility (cycle) management apps, or apps that calculate medical scores.
Medical measurement apps (heart rate monitors, arterial pressure, glycaemia, SpO2 etc.): Most of the time, developers post an often unheeded warning for the consumers about the using the app as a tool for serious medical examination. Apps using smartphone sensors are incredibly unreliable, if not totally crazy.
Apps which accompany patients during therapy: observance and monitoring, aiding decision-making etc. Voluntis developed (for Sanofi) an app to help diabetes patients. It’s a medical device CE-market and reimbursement is studied.
The patient/consumer is helpless before all of these options, especially because it’s often a personal initiative, auxiliary to their current care. Their only basis of judgment is the opinions and comments left by other users. As a distributor, stores aren’t comprehensive enough, they should segment more, isolating medical apps and verifying their conformity (at least requesting a CE-mark declaration). They could also employ methods of “applicovigilance”: being able to alert the creator to issues, alert users and reject or remove questionable apps.
Private labels for health apps.
Some start-ups have realized that a public interrogation is growing, and they have created an app labeling service.
This is the case in France with DMDSanté and Medappcare. These initiatives are useful for QS apps concerning medical information and more often these include low risks used to complement medicine.
The labeling standard relates to security, reliability and safety. These are reasonable criteria, but the requirement level is tiny compared to the standards of the medical industry. It would be wise to limit these to the non-medical as the addition of private labels could cause confusion amongst the public and developers, especially when the communication is based on institutional codes that could create the problem of finding fully certified apps which completely flounder the regulations.
To a lighted certification for apps?
App creators are mostly start-ups or professional individuals. Regulatory requirements are rarely respected because of a lack of information, funding or will on behalf of the creators. However these requirements are burdensome and following them can be incredibly complex, and consequently a lot of creators are now asking for regulatory simplification.
Here is the current situation:
- Public regulation is mostly self-certification, but the connected world has strict regulations concerning private health data.
- Devices trying to break into the medical domain have a regulatory burden disproportionate to their potential danger. A classification system of medical devices allows us to define the requirements for CE-mark. A major part of apps is a class I, in self-certification.
We can imagine two types of evolution:
- Limit the DM scope. DM regulation is leaning in this direction by (momentarily) relaxing contraception controls (contraception, pregnancy monitoring), and the autonomous apps could not be considered as active medical devices (but it’s all a question of point of view, cf. Cyrille Michaud post, MD101 co-founder).
- To tighten the quantified self and assimilated apps regulations. Some people have started to think that this will allow us to ensure a device’s quality on the market whilst being less onerous to MD. With a high level of borderlines cases…
3. Health online
The first health aspect improved by the internet is appointment booking. These tools aren’t medical, so opportunity for improvement has been quickly available and accepted. It is being used in private clinics and doctor’s offices to help patients, but the public offices haven’t quite caught up yet.
We’re still waiting for the Personal Medical File.
Patients as well as professionals are desperate to gain access to health information via a shared web platform, which would be an obvious benefit to time, efficacy, comfort and money. The personal medical file in France (dossier medical personnel –DPM) was created by law on August 13th 2004 (three years prior to the first iPhone). Its development has already run up costs amounting to several hundred thousand euros (or half-billion, depending on who is counting).
In all fairness, the DMP isn’t quite as simple as a Dropbox. The issues regarding security and quality are huge, notwithstanding the logistical deployment issues and how to go about creating the user base.
However, the DMP is finally open after many experiments, reports and re-definitions. The CNIL (French authority of private data control) gave its green light and the ASIP Santé will take care of the data security. Access is possible for the patient and for health professionals they will have chosen. The service will become generally available in 2017, and you can follow it with less than 600,000 files for the moment. To tide you over until 2017, there are currently private solutions such as Axigate, but we’ll have to see how they will evolve with the DMP generalization to come.
Evaluating health professionals
Just like your favourite restaurant or film, patients now have the ability to review health professionals and establishments online.
Evaluating doctors is very popular in the United States (this website has millions of reviews, for instance), but the practice remains very rare in France where the leader is notetondoc. The CNOM (National Conceal of Doctors Order) takes this position on the subject (translated from French):
“The CNOM reminds you that professional practices evaluation has been created for the benefit of the profession and with our implication (…). It’s up to the ordinal organization to make sure that establishment and doctor quality is adequate (…). Morality or behavioral problems coming from the doctor have to be judged professionally (…). The CNOM wonders if a notation without any serious data and rigorous methodology is really pertinent”.
The patients will have to make up their own minds.
The Internet is planning the road to recovery
Some still confidential services allow for the optimization of patient recovery and reduces costs. The start-up InnovSanté is full of these projects, for example :
- Improving ambulatory care management
- Optimizing side effects follow-up
- Providing tools dedicated to pharmacovigilance
- Evaluating patient satisfaction
This kind of approach brings a consensus: doctors, patients, companies, institutions, etc. Everyone is waiting to eat the fruits born from digital age.
4. Tech & Health trends
It’s one of the most promising technologies at the moment.
It allows us to simply produce some more or less complex devices with an important reduction of time and costs during development but also a large technical flexibility. Most of the projects concern developing countries.
The annoying aspect concerns the process of validation. Durability, precision and biocompatibility have to be controlled by the creator, and for the moment there is a lack of objectivity.
Among the current projects: spirometer, braces, prosthesis, cranial implants, and even vertebrae are already possible.
3D printing will also be used for pedagogic purposes. The company BioModex offers 3D printed imitation organs to allow training before a surgical operation with some or all of an organ’s biomechanical operation.
There are also big expectations for the “biological” 3D printer, where it won’t just be printed plastic.
If real organ printing is not going to happen soon, some more modest projects are emerging such as a pen to print cartilage, or the possibility to print meniscus, heart valves, ears, skin, bones etc.
Video games and porn are not the only things being revolutionized by virtual reality, and there are plenty of useful outlets for this technology in the medical world.
A first aspect is once again pedagogic. You can imagine a filmed surgical operation is 3D/360° and the student is able to follow it with their virtual reality helmet. It would be much more convenient than looking over the surgeon’s shoulder. In April, the company Medical Realities allowed people to watch a colon operation this way.
Some other projects have therapeutic goals: hypnosis for little operations (like in dentistry), overcoming phobias (fear of heights for example), curing addiction (by simulating trigger factors) and even managing pain.
Robotic surgical assistants are already here. They are controlled by the surgeon and allow them to ensure safety and precision (for example they can filter trembling). Right here we can mention the company Medtech, one of the world leaders in this field, who are French, and, even better, come from Montpellier.
Not really robots but cool anyway: exoskeletons are developed to offset a handicap, this is the Wandercraft leitmotiv.
Pet robots: the idea here is to offer a robot for elderly people, Alzheimers patients, autism sufferers, etc., to stimulate and reassure them. Plus it’s really cute.
The assistance robots. We’re not really that close to getting them yet, but we can imagine a robot which helps grandma to walk up and down the stairs, reminds her to take her pills; in fact, everything that grandma’s family doesn’t have the time to do. Currently, utilitarian robots are already used in some hospitals in Asia, but they are limited to really simple tasks. We’re talking more about carrying a meal tray than patient clinical analysis.
Let me try to define the uberisation:
“Trying to turn an economic model upside down by relating supply and demand through a web platform, without really caring about the current regulations”.
Health uberisation is very marginal still, and for many reasons:
The activity is strictly controlled
It requires a high level of qualification
Margins are not that big (the supply is limited and the costs are controlled too).
It’s hard to create a buzz (it’s difficult to get patients to like metastasis pages on Facebook).
However, health uberisation is showing signs of slow growth. MedecinDirect goes in that direction: services are offered by partnership (insurance, bank, companies are many potential entries) and they virtually link patients with doctors.
This is out of the classic care pathway, so the CNOM showed their reaction in this release. Translated extract:
“(…) to risk drifting to a non regulated electronic commerce which would reduce medicine practice to a simple electronic delivery service against money, through market sector platforms.”
They don’t call into question the professionals’ skills who are employed for these services, it’s really just a financial question with an obvious risk for health to become dominated more and more by private companies.
More disturbing but even less avoidable is medical tourism in the style of Airbnb with Medical Departures. There’s nothing illegal here and the phenomenon has existed for some time, it’s just making its way onto a new form of media.
5. Google, Apple & IBM : Big companies embracing health
My apologies to Mark and Jeff, but there’s not a lot to say on Facebook projects around an AI in health and on medication delivery from Amazon.
Verily, the Alphabet’s V
Alphabet is the entity regrouping a different branch of Google, where Verily is dedicated to life sciences.
Alphabet is full of health projects, with the modest objective to triumph over death.
Here are some of the main projects (available on l’observatoire des projet de google en santé):
- Lenses to continuously measure glycaemia (the glucose rate in tears), and maybe added LEDs to communicate with the user by colors.
- Motorized spoon for Parkinson's sufferers.
- Disease detection through nanoparticles (here, I’ve got no idea).
- A project around big data to define “Good genetic health”.
- Bracelets to track health: heart rate monitors, body temperature, SpO2 and environmental condition barometers(sound, light, pollution, etc.). The gossip says that it’ll be a medical device certificated and designated for clinical studies.
- Assistant chirurgical robot in partnership with Johnson & Johnson (whose anesthesia robot is actually borrowed from a US operatoring room, but that’s an other story).
- Clothes that measure heart rate.
- A partnership with Sanofi, confirmed by their European counterparts, to develop tools around diabetes care and follow-up, or sampling doctor’s interfaces to get simpler, miniaturized devices to the patients, …
- Partnership with Dexcom around the fabrication of connected devices to continuously follow glycaemia with a reduced cost and size.
Verily is very credible and by creating partnership with MD and pharma industries, who are already aware of the context and by starting with widespread chronic diseases. They would even be profitable for Alphabet.
Apple, users of regulatory software gap to do medical.
Apple focuses on software solutions without any medical demand (in fact, they delay their obligatory reports to developers).
There are two important services:
Reasearchkit: A tool to make clinical research easier (collecting, analysing and sharing data) by using the data collected in the store.
Carekit: A very recent development (March 2016) which is meant to be a place to develop health/medical applications dedicated to follow-up, diagnosis and telemedicine.
But there’s a big limitation on these tools: They are only compatible with Apple devices, which means less than 15% of the French population. But not such a big deal for the Reaserchkit tool, because the medical industry loves Apples (60% of doctors questioned for the 2016 Vidal study owned an iPhone).
Second problem: Everything pushes developers to work in medical, sometimes even by using the Apple Watch sensors, a general public product, to detect epilepsy crises with Epiwatch, for example.
IBM, my dear Watson
In development for 10 years, Watson is the super calculation system created by IBM. It involves a lot of very complicated stuff with a neuronal network and some AI. This is a very high level technical definition.
Not just created to beat everyone at Jeopardy, Watson can also use its calculation capacities in many different areas: weather prediction, logistical transport organisation, financial simulations, and, obviously, health.
If the technical part is really complicated, then luckily the working principle is simple:
For the calculator entry data: A huge data mass: information about patients (personal and family), medical images, clinical data, genetic data, …
As an exit point: Advice or even indications for future diagnostics and therapy.
Realistically, we’re talking about offering an AI as support for health professionals.
Some concrete examples:
Sorting between medical publications on a case by case basis
Medical image analysis
Evaluating the nature of cancer (oncology is the principle area of exploration by Watson)
Complementary examination suggestions
This project is full of promise, the efficacy has been proved, solid partnership with US hospitals and medical devices developers (such as Medtronic) have been created, and IBM has spent a lot of money buying health companies (ex: Truven Health Analytics at the beginning of this year, data providers and medical analysis, for $2.3bn). Plus, the tools needed to analyze the huge data quantity which will be collected soon let us safely say that Watson has a future.
The only issue is the reluctance of doctors, who aren’t ready to be
replaced assisted by an AI, so things are going to change slowly.
6. Medical device connections
New technologies are not only offering new products, they can also be used in devices already present on the market.
All the MDs using electronics can easily be connected, it’s even more justified when the device uses real measures.
The results can be transmitted through a local or distant host on an interface to display and analyse the data for the patient, the doctors, the service providers, the distributors, etc.
Some common MDs already exist in a connected version: thermometers, glucometers, tensiometers, scales, PPC devices, pillboxes, otoscopes, stethoscopes, …
Chronic diseases are the most attractive, a lot of money is used to solve the world’s growing diabetes problem. It’s also here that the follow-up and pedagogic tools are the most useful for the patient.
The elderly are also targeted, but not all of them feel comfortable with new technologies.
In an ideal world, connected MD could:
Improve care efficacy thanks to the data
Improve patient engagement, with “companion” apps
Improve medical knowledge
Reduce health costs
Improve device efficacy with automatic monitoring and maintenance
Develop prevention to complete therapy
7. #Help – Communication 2.0 applied to health
By adopting technology, health also inherited excess communication and technology (which is part of the appeal, to be fair).
The most visible is in the form of short and quick information, the point is to draw attention and create a buzz. High-tech mark is used as an example.
The most painful is in the content: It’s easy to forget the context, where honesty and respecting patients are a necessity.
That’s why we need:
The Qualitiso Medical (almost) Worldwide Institute 10 Commandments:
1. You shall have no other gods before the legislative health code (here in French).
2. You shall not make idols of your products.
3. You shall not take the name of the HEALTH in vain.
4. Remember before talking about revolution, to keep it holy.
5. Honor your patients and your regulations, so that you may live long on the market you are given.
6. You shall not confuse public health with business.
7. You shall not conduct misleading and exploitative surveys.
8. You shall not use hype expressions such as 2.0, 3.0 and 4.0-connected-turbo-diesel-whatever.
9. You shall not bear false witness against the company you work for.
10. You shall not covet your client’s data.
Written by Guillaume PROME,
MD101 Expert & Qualitiso.
Translation by Juliette OLIVIER.