The UK government has launched a consultation to gather views on the imminent updates of the current regulatory regime for medical devices, which include the application of technologies such as artificial intelligence (AI).
Ahead of the introduction of the amendments to create the new regime, scheduled to be in force at the beginning of July 2023, the consultation’s scope is informed by the government’s early engagement with stakeholders. The purpose is to enable early access to innovation relating to medical devices through data-driven assessments of the risks and benefits to patients.
“Patient safety and innovation are not mutually exclusive, and we want to ensure that our regulations maximise both,” said the Medicines & Healthcare products Regulatory Agency (MHRA) in the executive summary of the consultation.
As well as creating access pathways to support innovations, the new regulatory regime aims to introduce a framework for regulating software and AI as medical devices. According to the MHRA, this will ensure the UK attracts life sciences businesses and innovators “without compromising on safety”.
The MHRA acknowledges that the field of software as a medical device (SaMD), both in relation to being standalone software and software included in wider hardware – including AI as a medical device (AIaMD) – has grown in market share and complexity.
According to the agency, there are few provisions in the existing regulations for the application of SaMD and AIaMD in health and social care, and the proposal is to amend the current setup to support “responsible innovation” in that field while protecting patients. The MHRA noted that the majority of the changes required to do this will likely be around guidance rather than legislation.
Changes that would be specific to or have implications for SaMD include the introduction of a new definition for UK medical device regulations. Under the amendments, software would be defined as “a set of instructions that processes input data and creates output data”. This definition is part of the debates to be had under the consultation, as well as the potential expanding of the definition of in vitro diagnostic (IVDs) to explicitly include software.
The MHRA noted that SaMD can be deployed to the UK by websites, app stores and via other electronic means including deployment from websites hosted in other jurisdictions, so the consultation is also aimed at discussion whether regulatory change is required for placing SaMD on the market in these circumstances.
Moreover, the agency is considering to introduce an “airlock classification rule”, a temporary classification likely to involve monitoring and restricting the SaMD as if it were a high-risk device. This change, aimed at applying regulation when the risk profile is unclear, is understood as a measure to allow early access to the technologies while ensuring patient and user safety while risk assessments are carried out.
In addition, the government wants to ensure SaMD receives pre-market scrutiny to assess aspects such as safety, quality and performance, and to ensure the requirements in place meet this set of needs. The requirement for SaMD to have a hyperlink to MHRA-endorsed websites where users can report an adverse incident with a medical device is also part of the proposals.
Moreover, the consultation will discuss proposals that relate to SaMD cyber security and information security. This covers the direct safety of the device – and potential risks such as tampering – and the security of personal data held on or in relation to the device.
To manage such risks, the government is planning to require that manufacturers of SaMD meet certain minimum security requirements, especially in relation to protection against unauthorised access.
In relation to AIaMD, the MHRA noted that, as a subset of SaMD, the changes relating to software should also benefit AIaMD. Additional regulations specific to the application of AI in medical devices include the requirement of performance evaluation methods for diagnostic AI.
According to the MHRA, such AI-specific methods “would take a comparable approach to performance evaluation methods used for IVD medical devices in terms of requiring demonstration similar to that of scientific validity along with analytical and clinical performance”.
The introduction of new regulation around software and AI in the context of medical devices relates to programmes that go back as far as 2019, when the healthcare service’s digital delivery unit NHSX was set up, which aimed to lay the foundations for what the government described as a “new era of evidence-based self-care”, with technologies including innovative devices and AI.