This article was originally published on Forbes.
When it comes to healthcare, regulation is vital. Without it, doctors and patients cannot know what drugs, products and services are safe, exposing us all to greater uncertainty and unacceptable levels of risk.
However, regulation is far from easy - especially in healthtech, where the rise of digital and mobile technologies has seen the pace of innovation accelerate markedly. Regulatory processes are rarely designed with speed and agility in mind and, as a result, existing structures and institutions have struggled to cope with this rapid influx of new players and products.
This is understandable, but with the newest wave of healthtech innovations set to pose a series of regulatory questions never faced before, we have to overcome these challenges. Failure to do so will mean not only stifled innovation, but also worse outcomes for patients, doctors and health services.
A game of cat and mouse
The current regulatory landscape is complex and not set up to respond rapidly to the proliferation of new healthtech products and services that rarely fit neatly into existing categories.
In the UK, where I trained and worked as a doctor before entering the world of tech, we have the Medicines and Healthcare Products Regulation Authority (MHRA), which is responsible for ensuring that medicines and medical devices work as advertised and are acceptably safe; and the Care Quality Commission (CQC), which ensures that practices and services provide people with the right care. In an ever-evolving field, these roles are becoming blurred and regulators are left trying to play catch-up with innovators, whose focus is on solving complex health challenges rather than designing products to fit the established regulatory categories.
For example, symptom checkers were not regulated as medical devices by the MHRA until June 2018 when a small addendum was attached to the CE mark guidelines - the main certification for health products. This stated that symptom checkers were now, with immediate effect, considered class I devices, requiring appropriate registration with the MHRA, although no changes were made to the definition of class I itself.
Class I puts symptom checkers in the same category as plasters, bandages, and compression hosiery. To the man or woman in the street, this classification would clearly be wrong-headed: how can you devise regulations that apply to both AI-powered symptom assessment software and a band-aid? However, the reality is that there is currently no good fit within existing structures for these kinds of mobile apps.
The difficulty of measuring corporate liability with AI was recently demonstrated by the PHG Foundation, a think tank affiliated with the University of Cambridge. It showed that, with physical equipment, there is a clear link between manufacturing quality and accountability. But with an algorithm that responds dynamically to an infinite array of inputs (such as symptom combinations), it becomes much harder to assess. That’s before we even get on to the so-called ‘black-box’ problem, which makes it impossible to know how some AI applications reach the conclusions that they do.
Despite this link between manufacturing quality and accountability, there is emerging evidence of inadequate and ineffective regulation of medical implants. And, if there are clear shortcomings of rigor and transparency in the testing of surgical devices, is it unrealistic to expect effective measurement of corporate liability in AI in the near future?
Regulatory grey areas are thorny problems to solve and they are being compounded by a lack of communication between healthtech companies and regulators. The pace of innovation in healthtech is such that regulators often have no choice but to be reactive – but unfortunately their approach can at times tip over into the reactionary.
An example occurred in the UK earlier this year when the CQC was criticized for its response to internet pharmacies. Subsequently, and without prior warning, the CQC made sudden changes which required significant amendments to the way digital providers operate. The majority of providers were then deemed to be unsafe, because the CQC conducted an immediate round of inspections under the newly introduced criteria.
This reactive approach undermines public confidence in digital providers striving to provide safe effective care and the wider ecosystem, and fuels tension between regulators and tech companies. Better and more proactive communication could prevent many issues before they happen. The CQC has since made efforts to engage in a more proactive and positive way with digital providers, to create ways of working together to define how regulations and requirements should evolve. This is a step in the right direction following what was a frustrating episode for everyone involved.
Planning, talking, sharing…
When rethinking how we regulate healthtech, it’s useful to start from first principles. We obviously need regulations to prioritize the safety of the patient and to ensure that the products and services available on the market can be trusted by individuals and clinicians. However, we also need a regulatory environment that encourages and guides innovation; after all we do want companies to devote time and resources to developing new products and services in this space.
This means we need more transparency on both sides, and a more open dialogue.
Regulators should publish a roadmap of how guidelines and structures will change over time, rather than simply firing out updates without warning. This means adopting a proactive, rather than reactive mindset. This might not come naturally to regulators – but healthtech companies spend a lot of time trying to predict the future, so they can and should work with regulators to educate them on the shifts that they see shaping it.
This in turn will require greater collaboration and information sharing within the sector, but having a clear framework for how the next generation of health technologies might be applied and regulated will be to the benefit of everyone.
In the US, the FDA is setting a good example which Europe may be able to learn from. It is taking a proactive, innovation-friendly approach, which involves such things as: incorporating a pilot period to give companies the necessary space to develop their products before facing regulation; and working alongside established tech companies to keep the dialogue open between themselves and the industry.
Healthcare is going to get personal
Healthtech regulation affects us all and, increasingly, it will need to be designed and published with public education in mind.
In the digital age, healthcare will become more centred on the individual. Patients will have more choice when it comes to their health: over which app or service to use, what information to believe and which technologies they can trust. This means that regulators will have an even more vital role to play to provide clear guidance to consumers over which choices to make.
But giving the right guidance requires understanding of how these new technologies work. That’s why it’s so important for industry, regulators and academics to work together to ensure that we’re creating an ecosystem that is focused on creating the best outcomes for everyone: innovators, regulators, healthcare systems, doctors and most importantly patients.