How does digital transformation change the healthcare industry? What are successful digital business models, what’s the outlook? The following blogpost gives an overview, with a focus on the German market. We start looking at digital services used by the “end customer”: That is, either the patient or a person interested in preventive measures or monitoring of vital parameters. We then take a look at B2B digital business models and digital disruptions: It will be about hospitals, insurance companies and pharmaceutical manufacturers.
Procurement of information and transparency about health offers
Just to give you an overview, we usually refer to a first wave of digitisation (1985 to 1999: development of the Internet) and a second wave of digitisation (2000 to 2015: Emergence of big players such as Google, Facebook). Even in the early days of digitization the Internet established itself as the most important health guide: You can find anything from health tips to home remedies. From descriptions of clinical pictures to special forums for specific clinical pictures. You will find explanatory videos on YouTube as well as nutrition tips.
Of course, these information offers are by no means verified. The data sources range from the venerable Robert Koch Institute to hobby therapists. In order to get it right (and not get lost) you need a good pinch of media competence: you need to evaluate any piece of information carefully. But it is true that the availability of information has increased massively – and it is still increasing. It’s a paradox, actually: One the one hand there’s a flood of information. On the other hand we lack the information on the reliability of that information. This lack of information also applies to the quality of service providers in the health care market. Actually, this was (and is) the core idea of a digital business model. Take an evaluation platform such as Jameda or the White List (an offer of the Bertelsmann Foundation): It’s about collecting reviews on doctors, hospitals, etc. – which provides a “heat map” of excellence in health services.
If you assume that sick people try to find answer on Google this leads to yet another “big data business case”. These search search queries (symptoms, remedies) result in “heat maps” for diseases (or rather suspected cases of diseases). Some time ago, Google had shared the vision of being able to detect pandemics at an early stage and thus initiate emergency measures; indeed, Google is ahead of the reporting system to public health authorities (the obligation to report diseases such as measles, mumps, diphtheria, etc.), not least because of the official inertia of an institution such as the public health department.
Monitoring, initial diagnosis, therapy support
Did you also know, by the way, that Apple is now the largest watch manufacturer in the world? But this has to be put into perspective, because the buyers of the Apple Watch are not so much interested in reading the time. These users regularly check the pedometer, the heart rate or the information how long you have been standing. Another example is the StartUp KENKOU (Freemium version available): the APP measures the stress level, and it does so in only one minute (I tried it, it works). It’s all about self-monitoring, not least with the aim of self-optimization or health prevention and so on. Mostly, this kind of monitoring is non-invasive (Apple Watch, fitness tracker, etc.), but there are also invasive variants (e.g. implantable microsensors, for example for diabetics).
In many of the digital offers, monitoring is only one component – or rather the first step. The next step is a diagnosis, and – third step – therapeutic offers (but only if possible; and although not as a complete substitute for medical treatment). The app Kenkou therefore not only measures the stress level, but also offers techniques for stress reduction. Moodpath is an APP that assists affected persons during depression, burnout and stress. The application developed by psychologists provides a (first) assessment of the state of health, helps to keep a mood diary and provides more than 150 psychological exercises (based on behavioural therapy) to improve emotional well-being. Apps like Woebot (founded by Stanford University scientists) have a comparable target group, but additionally use the chatbot technology: the app becomes a conversation partner for people affected by depression and anxiety. Affected people report that it works. Since people with depression in particular have difficulty in talking to others, the hurdle to exchange with chatbots is lower and even gives chatbots an advantage.
Telemedicine, the e-prescription and online pharmacies
It took a while before eCommerce extended to selling of drugs (Note: Amazon started its offer in Germany in 1998): The online pharmacy DocMorris was founded in 2000, it initially operated from abroad to deliver drugs to customers in Germany. Since January 2004, the dispatch of prescription drugs is also permitted in Germany, but the following restriction applies: it is still required to submit the original drug prescriptions (on paper). So, the digitization got stuck halfway. This will only change with the introduction of the e-prescription. Whereas more than a dozen European countries have already introduced the e-prescription (e.g. UK, Switzerland, Spain, Romania), this will probably arrive in practice in Germany in 2021.
Unfortunately, Germany lags also behind in telemedicine. Consultation by video only as well as drug prescriptions in the wake of teleconsulting has only recently become possible (thanks to a decision by the German Medical Association in 2018). A pioneer of telemedicine in Germany was the company Teleclinic, and since the end of 2019 a powerful competitor from Sweden has gained a foothold on the German market: Kry, which is strategically well positioned, especially through a cooperation with DocMorris.
Let’s go one step further: from telemedicine to tele-surgery. Already in 2001 (namely on September 7th) the first tele-operation was carried out. Unsurprisingly, it attracted a lot of media interest. The first ever tele-surgery was about the removal of a gall bladder. The patient was lying on an operating table in Strasbourg, the surgeon (Jacques Marescaux) virtually handled the scalpel from New York. Astronauts and rural areas could benefit from this.
Gematics, electronic patient file
A further decisive change in the health sector is achieved by the introduction of the Electronic Patient Record or Electronic Health Record. This will be offered by health insurance companies from January 2021 onwards. Data sovereignty lies with the patients. All data and information on the treatment and diagnosis history of a patient can be stored on this health file, including medication, allergies and data relevant for medical emergencies.
If the patient releases (selected data entries or all) health data for a treating physician, this physician can access this data via the so-called Telematics Infrastructure (TI). This is a closed network that connects doctors, alternative practitioners, hospitals and dentists with each other. The TI is subject to particularly high security requirements specified by the BSI.
In an interview Jochen Werner, Medical Director and Chairman of the Management Board at the University Hospital Essen, gives insights on the Digital Transformation at his hospital. The interview takes approximately one hour, it’s in German. Mr. Werner provides intriguing answers to a lot of questions: What are the challenges? To what extent is Artificial Intelligence used for diagnosis? How does Mr. Werner assess the chances of success of IBM Watson in the healthcare sector? PODCAST: Smart Hospital
Diagnostic support with artificial intelligence and Big Data
The telematics infrastructure (TI), however, is just a single step towards a more systematic use of data in healthcare. The electronic patient file will initially consolidate only the data relevant to a patient, nothing more. But, of course, there are already more far-reaching concepts for the intelligent use of data. One example is the project KIKS (artificial intelligence for clinical studies). This is a digital ecosystem that serves at least two purposes: First, it allows to fulfil documentation obligations with less effort (qua EU regulation: evaluation of medical devices). Second, it enables the use of clinical data across several institutions. A consortium (including B.Braun, University Hospital Leipzig, University Hospital Jena) is driving the development, the managing director Frank Trautwein of one of the consortium members explains: “It should offer functions to exchange anonymized data, to understand relevant data in unstructured text or to analyze radiological images fully automatically.”
Speaking of analyzing radiological images: pattern recognition by AI is very well suited for this purpose. The subsidiary Healthineers of the Siemens Group has, for example, launched the “AI-Rad Companion Chest CT” program, which can mark round foci in the lungs or measure the aorta on CT images.
Another vision for the healthcare market is that artificial intelligence (in conjunction with Big Data) will support doctors in making diagnoses: A “search engine” compares a patient’s symptom picture with disease diagnoses in a comprehensive database and makes suggestions to the doctor in charge. This idea is easy to formulate, but by no means easy to implement (main reason: lack of sufficient amount of curated data): The application of AI Watson from IBM in the hospital sector failed at least at the first attempt .
Nevertheless, the market for AI in the health care sector is expected to grow massively. The research company MarketsandMarkets assumes that this market will soar from approx. 2bn US-Dolar in 2018 to 36bn US-Dollar in 2025
Allow me to share a few thoughts on the practical challenges of using AI for diagnosis:
Researchers at the New York Ichan School of Medicine (located at Mount Sinai Hospital) developed an AI algorithm for cancer prediction, which they were able to train with extensive data from about 700,000 patients. Each data set contained hundreds of different variables. The AI algorithm (called “Deep Patient”) identified many new patterns in the data, which the researchers could not always comprehend, but which formed viable indicators to identify early stages of many diseases (e.g. liver cancer) in patients.
Mysteriously, “Deep Patient” could also provide indicators of psychiatric disorders such as schizophrenia. But even the researchers who built the system did not know how the AI algorithm makes decisions (“black box problem”). So, “Deep Patient” makes clever predictions, but without any explanations. What does this mean in practice? How can such a system be used by medical teams that have to make decisions or recommendations with far-reaching consequences? Decisions about discontinuing or changing medication? Decisions on the administration of radiotherapy or chemotherapy or a surgery?
Optimization of administration, processes, scheduling
If you look at the daily routines in hospitals and doctors’ practices, you will probably notice: There is not only potential for optimization of tasks such as anamnesis, diagnosis, therapy or surgeries. Everyday life in hospitals and medical practices is also determined to a large extent by documentation, scheduling, referral management and so on. There’s huge potential to make these workflows more efficient, or even automate these by using digitization.
Companies like Doctolib offer solutions for scheduling and appointment management with reduced administrative effort. And companies such as Nuance offer solutions for simple documentation via speech recognition. Relevant information such as symptoms, vital signs, medical history or diagnosis and therapy can be easily dictated and transferred to the electronic patient file. The efficiency gains are quantified on their website: “A comparative study by the University Hospital Düsseldorf provides meaningful results: Speech recognition accelerates documentation by 26%, the satisfaction of doctors increases by 23% and 82% more data was recorded.”. This coincides with other studies, which put the time savings for doctors at around 20% and for nursing staff at around 40%. This is a sensible investment, especially in view of the shortage of skilled nursing staff.