European health care systems are mainly based on one simple principle – solidarity. As part of the six principles of the Charter of Fundamental Rights of the European Union, solidarity is known to be a core value in the way social health systems in Europe work. Whether it is health insurance systems like in Germany, where I come from, or tax-funded systems like in Sweden, all of them are built upon the idea that all members of a society carry the risk and burdens of health care costs due to their economic situation rather than taking into account their medical status. Thus, it is possible for everyone to receive the same kind of treatment without discriminating the ones who are disadvantaged by their genes or anything else.
As I am more familiar with the German system, I will draw my thoughts by means of the German health insurance system but the main idea is of course transferable to other systems too. In Germany everyone has to be medically insured. Normally as an employee a certain percentage of your income is paid to the statutory health insurance. And this is where the solidarity principle comes into effect. Regardless of their state of health, the contributors bear the costs for all of the people in the country, thus also for the unemployed, children or pensioners. The statutory health insurance consists of different insurance companies, which have to provide publicly regulated care services to their members but can additionally offer an extended service range for optional tariffs. On the other hand for certain occupational groups or if your income is above a certain level, there exist the private health insurances. In contrast to the solidarity principle, private health insurances apply the principle of equivalence, which implies that premiums are calculated independently of the income of a member. So as main difference to the social health insurances, private health insurance companies set their payments of contribution according to for example the age, health status, individual risk situation and scope of services a member wants to obtain.
So where do AI and Big Data comes into play now?
The power of AI and Big Data is slightly changing the way medical diagnoses are made and diseases are determined. Whereas in the past doctors and medically trained specialists were the ultimately basis of trust for making diagnosis, nowadays technologies make it possible to obtain even more precise medical findings in some areas. With the ability to consider a far wider range of characteristics, algorithms can automatically analyze really big and complex data sets and discover hidden patterns that a human brain might not be able to capture. Of course, these developments are not going to replace the work of medical professionals; what I want to point out is the fact that in some medical areas AI already is and in the future will be a big part in medical research and supporting medical decisions. It offers the possibility to bring in new perspectives and information about patients, that have not even been considered before as they were not stored digitally or too complex to analyze without computer technology.
Therefore, AI has high potential to help to improve medical diagnosis and predictions. One example that demonstrates this potential quite well is the analysis of genetic material for different diseases. Highly complex structures in the genome and the possibility to compare it with millions of other samples of healthy and ill humans can for example be used to predict whether someone will face a specific disease or not.
On the one hand this development is of course a really big chance to help patients with early detection and treatment but on the other hand some ethical questions arise that you cannot ignore when you talk about the use of AI in the health care. And to come back to my initial question, the role of solidarity in this development is one of them. In an article about the use of artificial intelligence for making diagnosis a German professor, Prof. Dr. Karsten Weber, who is dealing with technology assessment, raised the question about the risk of desolidarization of our health system when using AI for medical predictions. He questioned what if AI would be able to help medical professionals to precisely assign and predict diseases? If we know we have “good” genes and a small risk to get ill are we still willing to pay for everyone, including those having “bad” genes and more prone to acquire certain diseases?
These questions simultaneously stimulate the discussion about the risk of introducing more personal information into the health care systems. If your medical future becomes more and more transparent to insurance companies where do fairly personalized premiums end and discrimination start? Does this lever the whole concept of risk and uncertainty that supports a solidary health care system? Current discussions in Germany often criticize that private insurances are supporting a two-tier health care that provides care with better quality and faster access to the ones who can afford it. Looking at it from an AI perspective, introducing more and more information about each individual into the system would give the opportunity to make it even easier to assign someone to a certain risk group. And from this point the danger of unintentionally discriminating certain groups in the health care system is not far away.
For now, predicting a transparent picture of our medical future is of course not possible but living in a modern society, you should be aware of the fact that personal data is collected everywhere. This enables personalization, which no longer exclusively plays an important role in customized advertisements on the internet. Also insurance companies make use of this trend to create a more precise picture of their members. If you are, for example, willing to share whether you are doing medical checkups, go to the gym or track your health and fitness with the use of apps, the insurance companies offer bonus payments. By sharing those habits, personal data about different lifestyles can be gathered and used for categorizing different behaviors. In this context it should also be mentioned that one might argue that these programs unintentionally embed some kind of discrimination due to errors in the data capture. One example is the way how a "sportive" lifestyle is measured in these bonus programs. Often the possession of a gym card is a criterion for receiving money from the insurance companies as a reward for doing sports whereas at the same time other sport activities like swimming or jogging are not captured or rewarded.
In this context the general problem of sparse and scattered data capture when gathering personal data should be kept in mind. As it is not possible to measure everything and equally for everyone, it might lead to distorted views of reality which hold the risk of drawing wrong conclusions about the habits of an individual. Maybe you have experienced that an ad on social media has been displayed to you where you had the feeling that this does not fit into your preferences. In this case, the underlying algorithm probably drew a false conclusion from your personal data. Health care data and the way insurance companies work is of course not comparable with ads on Facebook and co.; what I want to point out is that you should be cautious with what you conclude from personal data. Drawing a false conclusion in the medical field has for sure a lot more impact than just a wrong ad on social media as it involves a lot more invasion of privacy.
Additionally, the discussions about the rights to healthcare and the right to decide what kind of lifestyle you want to live come up. If AI and Big Data is supporting the opportunity to create a more precise picture about the way of life and state of health of every individual, who will then decide what is considered to be a healthy or conscious lifestyle in order to get a fair insurance premium?
To sum up my thoughts, I want to point out that the use of AI and Big Data in the medical area is a big opportunity and a powerful tool for actually helping people. I personally think the development is more than fascinating and we should not stop here. The introduction of AI and Big Data involves so many possibilities to revolutionize the health care sector and the way we think about it. The use of virtual health assistants, for example, can change the way we access medical help enormously. By offering the opportunity to provide personalized information, based on age, medical history and other individualized settings, at every time, these systems address the need for instant guidelines that people are looking for. At the same time those systems can help patients to keep track of their actions. Thus, they can also play an important role in treating chronic diseases as they empower and motivate the patient to manage their disease and symptoms. This could also transform the whole idea of being medically insured. For now, a big part of being medical insured is the opportunity of getting medical advice from professionals in a timely manner and as often as you want. But if AI technology will be able to deliver this advice even faster, a major part of how health care is organized could be replaced in the future. Simultaneously patients would get more power over their health actions and do not exclusively have to rely on the services of the health insurances. As the need for a wide range of general services would then decrease, the concept of insurances might shift to the supply of advanced and unusual procedures in medical treatment.
The described developments show that AI and the access to Big Data in the medical field will play a big role in how health care will be viewed and accessed in the future. But at the same time we should not forget the ethical consequences that the transformation involves. When discussing the opportunities, there is also a need for discussing access rights to the analysis of personal data and how we want to use the information that AI generates. This has to be an essential part of future developments in order not to exclude anybody from equal treatment in the medical care and to maintain the solidarity in our health care systems.
Elger, Bernice. Ethical issues of human genetic databases: a challenge to classical health research ethics?. Routledge, 2016.
Kirch, Wilhelm, ed Encyclopedia of Public Health: Volume 1: A-H Volume 2: I-Z. Springer Science & Business Media, 2008