Like chatbots, new digital healthcare tools are changing patient care. Their promoters tout the assistance they provide to doctors, who can focus more on the consultation process. To ensure ethical use, humans must remain the central agent in the practice of digital medicine.
Digital health tools have gradually become a part of our daily lives. Laws passed in 2018 and 2019 incorporated telemedicine into the legal framework to foster innovation and improve access to care. The Covid-19 health crisis, during which the French became more familiar with digital health tools, marked a new milestone. Today, the “My Health 2022” plan reaffirms the key role they intend to play in reducing inequalities in access to care. And according to the 2024 edition of Harris Interactive’s Digital Health Observatory, 78% of French people have a positive view of the development of digital technologies in the healthcare sector.
Some French people are digital experts, others have little mastery
The Harris Interactive survey, however, distinguishes between different profiles according to the ease or, conversely, the difficulties encountered by users in familiarizing themselves with different digital tools. Thus, 16% of the population would be classified among the “digitally vulnerable” because they lack mastery of digital tools. This profile would be more feminine, older, and more rural than the average. On the other hand, 28% of French people are said to be “digital experts.” Very comfortable with the technological tools they use on a daily basis, this group is said to be more masculine, younger, wealthier, and more urban than average.Digital health is very diverse today
But when we talk about these new digital tools that are now becoming essential in the field of medicine, what exactly are we referring to? In 2022, in a special issue of its Bulletin devoted to this topic , the National Council of the Order of Physicians drew up an exhaustive list of the main digital uses in medicine. Some so-called telemedicine tools help caregivers in their professional practice, particularly by soliciting their peers. The patient benefits indirectly since it impacts their care (in terms of diagnosis, treatment, follow-up in the community or in the hospital, etc.). This is particularly the case for:- tele -expertise , which allows a general practitioner to remotely call on a specialist with specific expertise (for the interpretation of an image, monitoring cancer, etc.)
- medical teleassistance , which corresponds to cooperation between health professionals, and during which a medical professional remotely assists another health professional (to treat a complex wound, provide emergency care in a nursing home, etc.)
These digital tools focus on the caregiver-patient relationship
Some digital tools that have become part of the daily lives of the French directly affect the caregiver-patient (or doctor-patient) relationship. Examples include:- teleconsultation , which allows a healthcare professional to provide a remote consultation to a patient, if the latter’s health situation allows it (if necessary, the patient can be assisted on site by a healthcare professional)
- medical telemonitoring , which allows the patient to remotely transmit their health data to the medical team monitoring them so that the latter can interpret them
- e -prescription or electronic prescription, which corresponds to a dematerialized prescription circuit
- conversational agents called chatbots
The rise of health chatbots
The first health chatbots were primarily developed to refine medical pre-diagnoses, such as the Watson program created by IBM in 2007 , which is capable of understanding human language, analyzing data and formulating online responses. It notably distinguished itself in Japan by identifying a rare form of leukemia. Since then, their presence on the market has experienced significant growth, driven by innovations in AI , and now in generative AI, with for example Sophie the avatar robot , a humanoid robot designed to interact with patients at home or Ada Health , a mobile application for medical self-diagnosis based on artificial intelligence, or even assisted robotic systems for surgical interventions. We are witnessing a veritable proliferation of new applications and uses that are transforming medical practice and giving rise to AI-assisted medicine or precision medicine, such as Stryker’s Mako . Essentially focused on pre-diagnosis or therapeutic monitoring, the benefits expected by those promoting chatbots are numerous. For healthcare professionals, the exchanges that take place before a consultation – in situations where the patient asks the chatbot about their state of health and the latter, in return, can recommend an appropriate approach after analyzing their responses – shift the “Useful Medical Time” towards the consultation . Additionally, virtual pre-diagnosis performed by a Health Chatbot offers a user-friendly interface that can facilitate faster care. Many patient categories and pathological conditions are now targeted. For example, the market is expanding toward the elderly. Dr. Antoine Poignant, a doctor, judges, for his part , that:“Connected devices allow us to move data rather than patients, and for the elderly, this is not insignificant. Their home becomes their place of care.”Part of the explanation for the effectiveness of conversational agents lies in the very structure of language. Generative AI language models such as LLM ( Large Language Model ), trained on very large corpora of text, can simulate these relational networks. They can understand and generate language based on complex contexts.
A change in the GP-patient relationship
Patient-professional relationships are potentially transformed by these digital medical assistance tools, and this in three major directions, according to Anthéa Serafin. First, these tools lead to a redefinition of the roles of patient and physician. Digital health tools strengthen patient autonomy and accountability in the care relationship, and they could facilitate better therapeutic adherence, which is a significant challenge in the face of the rise of chronic diseases. In addition, they help professionals to coordinate care pathways more easily, as has been shown, for example, in the case of cancer patients . Finally, they promote earlier detection of diseases, for example in the case of autism or addictions .Let us keep an ethical view of these uses.
It is essential that digital health tools based on artificial intelligence remain assistance tools and not autonomous decision-makers. Humans must remain the central agent in the practice of digital medicine, retaining control and initiative over medical decisions . The results produced by AI systems must be subject to human supervision and final validation by a healthcare professional, who remains responsible for them. Several questions are crucial : how far should we allow algorithms and those who design them to control medical decisions? How can we guarantee the confidentiality of our private lives? These are all issues that healthcare professionals and patient associations are addressing, particularly in the Regional Ethical Reflection Spaces (ERER). We suggest some warning points to ensure the ethical and safe use of digital therapy:- Establish a framework adapted to the uses of AI in health, defining rules and best practices
- Ensure transparency of AI to its users, to enable healthcare professionals and patients to understand the processes and data that influence the results generated by AI
- Clarify the responsibility of the different actors involved in the design and use of AI
- Ensure the strongest possible privacy protection and data confidentiality
- Combating digital illiteracy (or the digital divide) , which represents a particularly worrying risk of social exclusion.
