The NHSx, digital unit of UK National Health Service (NHS) announced the launch of a new multi-agency regulatory advice service for artificial intelligence (AI) at the AIMed NHS AI Lab virtual conference took place last Thursday (24 September). The NHSx will oversight and be funding the initiative making up of four core agencies: the Care Quality Commission (CQC); National Institute for Health and Care Excellence (NICE); Health Research Authority (HRA), and Medicines and Healthcare products Regulatory Agency (MHRA) as equal partners.

The new service is still at its infant stage

The service has yet to have an official name but primarily, over the next three years, it aims to offer support, information and advice on the regulation and assessment of AI and related digital health technologies targeting at the UK’s health and care system. The service hopes to provide an easy to access and comprehensive guidelines to all innovators so that they can meet the robust safety and quality requirement as they developed their respective technologies. At the same time, the service will also ensure healthcare providers have the knowledge and tools to help them adopt and deploy the best technologies.

First stage of the new service involves mapping of stakeholders and NICE will be the coordinator to host the core team and others who are coming on board. It is expected each player will bring their own AI experiences and advices on the table. According to Jeanette Kusel, Director of Scientific Advice at NICE, a key challenge that the agency faced right now is AI developers do not necessarily realize the additional evidence requirement on top of those needed for CE mark.

Kusel said in order for technologies to penetrate into the NHS, additional prerequisites are needed to that they do actually work in practice and are considered effective in the present standards of care. This is when NICE usually steps in to help innovators plan the evidence generation at early stage when they are preparing their regulatory evidence and lead them not to misinterpret certain regulations. The multi-agency regulatory advice service will enhance that and provide a more coordinated guidance on evaluations and regulations.

Nevertheless, the actual scope of the new service is not fully determined yet. So, some audience research will have to be done to understand what are the needs, the barriers and knowledge gaps faced by the users of such as service. After which, solutions will be designed and built to address the challenges raised by potential users. Ultimately, as the service matures, it will also identify gaps in the present regulatory and evaluation processes for AI and hope to influence and inform policymakers to fix those issues.

The importance of a joint-advice service

Naho Yamazaki, Head of Policy and Engagement at HRA said the joint-advice service is particularly important in two aspect: to reassure a more appropriate use of patient data and to facilitate research approval including ethical reviews for developed products. Yamazaki believes these are crucial in protecting the interests of the public and patients as well as to maintain trust in the AI technologies that will eventually be deployed in healthcare setting.

On top of advice, the service will also like to streamline the evaluation and regulatory processes for fellow developers. To achieve that, HRA is now undertaking two projects. One targets at offering a more seamless and coordinated application process to shorten the time needed for approval. The other is to take a more efficient approach to review studies that made used of patient data.

On the other hand, Zoe Fyffe, Interim Head of Strategy at CQC pointed out the tricky balance between helping to get technologies that are meaningful and have positive impacts on people’s care to the frontline and protecting it from the risks associated with changes brought about by new innovations and development. Over the last 18 months, CQC had devoted its energy to improve its regulations in the area. With that, Fyffe said it’s time to think about what’s on the horizon and what’s coming way so that the industry is more ready for AI and other new technologies. Thus, this new service acts just in time, to provide applicants the modernized application portal they need.

Graeme Tunbridge, Director of Devices at MHRA thought there are huge benefits of bringing different parties and regulators together. All along, different agencies perceived regulations fit together like a jigsaw but in reality, this don’t tend to be the case. To prevent developers from filling similar forms again and again to meet the requirement put forward by different agencies, this new service becomes a pivot to do something different.

Tunbridge urged developers who will like to engage with the service to start thinking about how they will like to deploy their products so they will not miss the many regulatory barriers along the way. At the end of the day, the team hopes to set up some form of digital resources or website for developers to get in touch with really quickly and helping them to work out the best routes to approve and deploy their products.


Author Bio

Hazel Tang A science writer with data background and an interest in the current affair, culture, and arts; a no-med from an (almost) all-med family. Follow on Twitter.