Draft analysis grid for the evaluation of medical devices embedding Artificial Intelligence
This analysis grid is intended to be used by the CNEDiMTS to contribute to its evaluation of medical devices embedding decision systems based on automatic learning or self-learning processes ("Artificial Intelligence")
The Medical Device and Health Technology Evaluation Committee (CNEDiMTS) is the Haute Autorité de santé (HAS) committee which evaluates medical devices (MD) in view of their reimbursement by the French health insurance scheme (article L. 165-1 of the CSS). It plays an advisory role to decision-makers recommending the reimbursement of the MDs or not (registration on the List of products and services qualifying for reimbursement – LPPR), helping to determine the conditions for their proper use and their role in the therapeutic, diagnostic, prevention or disability compensation strategy.
As part of the evaluations it conducts, the Committee receives applications for reimbursement of medical devices, some of which rely or will rely on automatic learning processes ("artificial intelligence").
Purpose of the public consultation
Previous work has led the CNEDiMTS to consider that the criteria for assessing the benefits of a technology remain clinical criteria and that it is not its place to evaluate the mathematical functioning of the model. The CNEDiMTS needs information, however, to evaluate the medical device-related decision-making process such as training, test and validation data characteristics, along with those relating to the model and algorithm(s). For this reason, the CNEDiMTS believes that an innovative and collective approach is necessary to validate its analysis grid, which will be used for the algorithmic part during its evaluations. This is the subject of this public consultation.
This phase of public consultation is also part of HAS intent on the one hand to continuously reinforce the transparency of its evaluation processes and, on the other hand, to allow the stakeholders in the development of these technologies to contribute to tool implementation.
Ultimately, the evaluation will strengthen the confidence needed to use these new MDs and to encourage their appropriation by professionals and users.
Public consultation is the key step in developing this grid. Following this consultation, the contributions received will feed into the reflections of one or more working groups which will improve the draft analysis grid before it is examined by the CNEDiMTS, which will validate the final grid.
Objectives of the consultation
Through this public consultation, HAS wishes to:
- Collect the opinions and considerations of stakeholders and experts involved in the development of solutions based on machine learning processes;
- Determine what is necessary and sufficient to enable HAS to evaluate medical devices based on these processes;
- Appreciate the degree of precision expected;
- Appreciate the contributors' understanding of the current project criteria.
Who is the public consultation for?
All stakeholders involved in the development or use of medical devices incorporating learning algorithms are invited to give their opinion: patient associations and healthcare system users’ associations, whether approved or not, national professional colleges and learned societies, public institutions and agencies, unions, companies, developers, researchers, etc.
The opinions and suggestions collected will be used for the final formulation of this grid.
Until when is participation allowed?
The public consultation runs from 20 November 2019 to 15 January 2020.
How to submit your input
- Download the support document
- Submit your contributions by completing the online response form (NB: all questions must be answered at one time; the form cannot be saved)
How are contributions taken into account?
All contributions will be analysed and discussed in order to achieve the final version of the analysis grid that will be used in the context of future applications for medical device reimbursement.
What happens to the personal data collected?
Personal data (surname, first name, email address, etc.) collected during this public consultation are used solely for the administrative processing of your contribution and are not transmitted to third parties.
The personal data collected are kept for a period that does not exceed the duration necessary for the purposes set out above.
In accordance with the French data protection act, you can exercise your right of access, opposition, rectification and deletion of your data by email at the following address: firstname.lastname@example.org
For further information, or to report any difficulties encountered: email@example.com