Functional classification, according to their intended use, of digital solutions used in the context of medical and paramedical care

Which classification has been proposed?

HAS drew up a draft version, and subsequently submitted it for public consultation over a 3-month period by all stakeholders in the sectors involved in the development or use of digital solutions. Following the analysis of the 76 contributions, indicating strong interest in the draft classification, the French National Authority for Health has drawn up an easy-to-use classification. It includes a total of 11 types of digital solutions classified in 4 levels (A, B, C, D), according to their intended use supplemented by:

• the capacity of the digital solution to take into account user/patient parameters (capable of resulting in personalisation of the response),
• the autonomous or non-autonomous nature of the digital solution, making a distinction between technologies requiring human intervention to implement a therapeutic, screening or diagnostic action, and those generating this type of action themselves, i.e. without prior human intervention.

	Description	Number of categories	Personalisation	Autonomy
Level A	System services for patients, caregivers or healthcare professionals within the scope of care or care pathway optimisation or medical/socio-administrative management with no direct action on patient health.	1 category
Level B	General non-personalised user information on living conditions, healthy living, diseases/disabilities or any health state (in the broad sense of the term), health, care or life pathways, etc. Also provides training materials or tools.	1 category
Level C	Aid for living, for prevention, screening, diagnosis, compliance, monitoring or treatment of a disease, health state, or in the context of a disability. No autonomy of the digital solution in care pathway management.	8 categories
Level D	Autonomous decision management after data analysis and diagnosis with a view to automatically adjusting the treatment to be administered, without human intervention.	1 category

What does the future hold for such a classification?

This classification was devised as a reference foundation tool intended for the various potential stakeholders and for multiple purposes. This function-based approach will only be beneficial if, over time and according to its use, it is accompanied by developments of matrix approaches including other aspects. For example:

• applicable regulations: technology falling under MD, IVDMD status or not (if yes, which risk class), subject (or not) to confidentiality/personal data protection regulations (GDPR) or health data hosting regulations (L.1111-8 CSP), considered as high-risk or not (according to the upcoming European Regulation), etc.,
• technical guidelines: need for interoperability or not, etc.,
• possible funding models: assessment with a view to funding or not? by HAS or by other stakeholders?, etc.

On a national and European level, the digital framework is taking shape, particularly in terms of autonomy and artificial intelligence. The grid proposed by HAS could help structure interactions and, ultimately, help achieve effective integration of digital solutions in the healthcare system, in terms of healthcare and medico-social aspects.

This will require in-depth research by the stakeholders concerned in a given field to potentially link this grid with one or more maps of the characteristics of their field. For HAS, this is a sine qua non condition for facilitating the understanding of each party (competent authorities and companies seeking a better understanding of their expectations, users, etc.). All this work will be fully coordinated with the work of the Digital Health Council.

Under these conditions, this grid will reach its full potential.