The UK’s Outlook On AIaMD And Need For Regulatory Improvements
One thing is clear, the future of Artificial Intelligence as a Medical Device (AIaMD) is very promising. The global market for AI health technologies is expected to expand at a compound annual growth rate of 38.4% from 2022 to 2030 to reach USD 208.2 billion by 2030.1 The questions that remains are: how will regulatory agencies respond? And who will lead in innovation given the challenges of the current climate?
The Regulatory Horizons Council (RHC) report regarding the AIaMD industry can be summed up into a few areas. These areas they’ve highlighted as critical to the success of AIaMD development, patient safety and the regulatory pathway (pre- and post- market).
- The first area is building the foundation for a long-term plan to build out the regulatory capacity and capabilities. The governing bodies need to be aligned on creating the appropriate investment into AIaMD with adequate urgency. The objective should not be to just increase the number of agents but to have well-qualified industry subject matter experts. Having the right personnel will ensure that the regulations put in place will be specific to AIaMD and help close the gaps regarding pre- and post- market phases.
- The second area, and probably the most important one, is providing a revamp to the product lifecycle. AIaMD creates new and different challenges than other medical devices and therefore should have specific regulations regarding safety, effectiveness and equitability. The biggest concern the RHC raises is around generalizability and AI bias. Meaning the scope (or intended use) of an AIaMD needs to take into account the population groups that were tested and in which groups their device may produce erroneous readings. Use of the device should be limited to the applicable areas that were properly tested and dissemination of this information should be readily available to the public upon market entry. Initiating threshold for use of the device should be established prior to launch and clearly stated in the regulatory filing documentation. Furthermore, the risk associated with the tested population groups and more importantly those that were not tested should be clearly identified and all mitigation opportunities should be identified by the manufacturer. This helps keep patients safe and allows for proper performance of the device. The post- market surveillance process should also address these concerns proactively with more standards around generalizability and AI bias.
- The third area is transparency between all parties involved, from regulators to manufacturers to patients. There can be a synergistic effect if all parties started working more collaboratively and cross globally. The first step would be for regulators across multiple countries to work together on creating their standards, to not make it overly complicated by each creating their own extensive standards and verbiage. The Medical Device Single Audit Program (MDSAP) is a fantastic example of how everyone “wins” when agencies come together. Leveraging expertise globally instead of locally will create better more effective standards and reduce the agency burden too. These standards, along with additional guidance documents, will allow for manufacturers to be more successful and allow for innovation to reach the market sooner and wider. Once the product has reached the market, the public should be able to easily access the intended use of the device along with the associated risks and mitigation plans when using the device. By helping to educate the public only furthers the success of the device and efficacy.
To answer the second question as to whom the leader will be in AIaMD industry will largely depend on how much of the above concerns are addressed. Currently, the UK along with several other countries are no longer a part of the EU. This provides them with additional freedom to set their own standards but does present additional challenges for manufacturers to have to engage in yet another regulatory approval process to gain access to these markets. The RHC believes that the UK has great potential to step up as leaders in the AIaMD industry as they’re already well respected in this field and well-respected regulators and gate keepers (NHS and MHRA). The upcoming year(s) will be extremely telling as the UK and the EU undergo major regulatory transitions.
Definitions:
AIaMD: the AI component of AIaMD may be variable in complexity and significance. Broadly AI may be defined as ‘the science of developing computer systems which can perform tasks normally requiring human intelligence’. Most medical devices deploying AI can be categorized into one of three broad use categories: diagnosis and screening, prognosis and supporting treatment.
Regulatory Horizons Council (RHC): is an independent expert committee that provides the government with advice on regulatory reform. They create reports providing recommendations on where the market is heading and how the regulatory agencies can best respond, support and collaborate with the industry leaders.
