Driving Innovations in Healthcare Using AI

Dr James Risley is a Consultant Emergency Medicine Physician with two decades of experience. He has a diverse background which includes surgical training, qualifications in law and an Executive MBA from Cambridge University. Currently, as the Deputy Executive Medical Director for Betsi Cadwaladr University Health Board, Dr Risley leads initiatives to meet the healthcare needs of North Wales. He provides senior leadership across the entire healthcare system, notably in relation to the subject of this article in Research & Development and Digital, Data & Technology, and he also sits as an appraiser for Health Technology Wales. His expertise and leadership continue to profoundly impact the medical field.

In an exclusive interview with Healthcare Tech Outlook Europe, Dr Risley shares his valuable insights into challenges and opportunities in the industry for start-ups, and how AI can be effective in the healthcare sector. 

Does your background in law and entrepreneurship align with your current role?

My legal background has proven particularly valuable in clinical governance, where being able to systematically problem-solve and use deductive reasoning helps to identify the root causes of clinical incidents and patient harm. This then allows the right remedial actions to be implemented to ensure that we can minimise any further risks to patients.

Completing the Executive MBA program involves developing, amongst many other things, a comprehensive understanding of how an organisation should function. My current position in Betsi Cadwaladr focuses on helping to support and run clinical services across North Wales, and drawing on the combination of knowledge of key business practices and my clinical experience really helps when delivering at an executive level.

More broadly, Cambridge University’s specific emphasis on entrepreneurship and venture creation resonates strongly with my interest in healthcare technology. This has facilitated my involvement in assisting and guiding healthcare sector start-ups, where my role is to support ventures to identify their true value-creation potential, and how they can effectively convert their concept or product directly to improving patient care.

Since you focus on innovation, what potential difficulties do you see for start-ups and entrepreneurs in this space?

A significant challenge faced by start-ups in the healthcare sector is navigating the intricacies of interacting with the National Health Service (NHS), and being able to effectively demonstrate the value of their product. This issue has arisen because the NHS actually comprises many disconnected individual organisations in the form of, for example, Hospitals, Trusts and Health Boards. Consequently, entrepreneurs often have to engage in multiple discussions regarding the value of their product with a number of different procurement departments, each of which have legacy systems and processes that differ across organisations. These discussions can be difficult. They may result in the need to adapt a product to suit the distinct needs of each organisation, a process and compromise that some founders may be reluctant to undertake despite its importance to maximising the likelihood of their product being adopted.

A further issue arises when a start-up’s product cannot not be immediately translated into cost savings. Many innovations are geared towards enhancing patient outcomes. However, in a time of critical financial strain on the NHS, a big challenge lies in quantifying the financial value of such outcomes against a backdrop of constrained resources and existing services competing for limited financial resource. 

How can start-ups mitigate these challenges?

Start-ups in the healthcare sector face different challenges to established companies when trying to contract with the NHS, and may find it advantageous to focus initially on the private sector rather than pursuing a direct approach with the NHS. Although counter-intuitive, this strategy involves directly engaging with patients, allowing for exploration and product development in the private sphere, which can serve as a compelling use-case when later approaching the NHS. Attempting to introduce new technologies directly into the NHS without prior implementation at scale can lead to uncertainty for those making decisions on procurement because it is difficult quantify the products value without robust evidence demonstrating improved patient care outcomes.

The increasing demand for private healthcare, driven partly by increasing waiting lists, presents a burgeoning market for start-ups. By addressing the needs and preferences of patients in this expanding landscape, entrepreneurs can position their innovations for success and foster positive customer relationships directly with patients. This approach allows start-ups to validate their solutions and provides valuable insights through direct customer feedback. Further, this approach may also help to relieve pressure on NHS services in the event that their product succeeds.

With this  robust evidence in hand, start-ups can then approach NHS organisations with a compelling value proposition bolstered by a proven track record of successful implementation. By acknowledging the interplay between private healthcare and patient care, start-ups can strategically position themselves for long-term growth and future collaboration with the NHS.

What general advice would you give these start-ups?

It is crucial to reframe setbacks as opportunities for growth and improvement, rather than dwelling on perceived failure. I know it’s not easy, but I’d say always focus on a ‘win or learn’ mindset. Embarking on the journey of developing a start-up is often far more challenging and time-consuming than anticipated, with periods of no movement followed by moving at a frenetic pace. Many individuals who venture into this sphere possess a history of being highly motivated, driven, and dynamic in their personal and professional lives, accustomed to achieving their personal goals through sheer determination. However, the landscape of entrepreneurship introduces a unique dynamic where you may at points need to rely heavily on the collaboration and support of others. Consequently, achieving milestones or timelines solely through accustomed personal willpower may be difficult, which can easily lead to frustration and demotivation.

Understanding that their success often depends on numerous external factors and cooperative relationships with other stakeholders is crucial for maintaining realistic growth expectations and being able to adapt to unforeseen challenges. This acceptance fosters resilience, ensuring that setbacks are treated as learning experiences which can ultimately increase their chances of success in the long-term.

How can AI be utilised in the context of the challenges you mentioned earlier?

The benefit and use of AI lies in accurately defining its purpose, and currently it often seems to be used as a vague catch-all term as the solution for all problems. In healthcare, selling AI as a complete replacement for individuals is premature and unrealistic. Instead, for now and the near future, I believe AI should be viewed as a tool to aid and expedite and support, rather than entirely replace, human decision-making.

A significant challenge concerning AI relates to regulation, establishing governance around its use, and ensuring the accuracy and fairness of input data. AI's iterative nature means that its outcomes and performance may change significantly if regulatory approval is achieved prematurely before sufficient algorithmic learning has occurred. The continuous learning process with the ongoing influx of information must be carefully considered when it comes to responsible AI regulation.

Is staff shortage still a pertinent issue in the healthcare sector?

The healthcare sector faces a substantial supply-demand imbalance concerning essential staff such as doctors, nurses and other allied healthcare professionals. This shortage will not be resolved quickly, prompting consideration of AI as a potential solution. However, as I mentioned earlier, fundamental questions around the role of AI in healthcare remain, and I am sceptical about its ability to replace individuals at this time, in this context. 

For me, acknowledging the irreplaceable value of person-to-person interactions in healthcare is important. Patients often disclose intimate information to their healthcare professionals, using verbal and non-verbal cues, fostering a bond and rapport that is essential for accurate diagnosis and treatment. How this trust and communication with patients is established is a complex and unique process, and one which significantly influences the effectiveness of healthcare provision. The challenge emerges when contemplating how algorithms can replicate such intricate relationships to elicit the necessary information from patients and achieve optimal outcomes. Consequently, the focus should be on viewing AI as a decision-making tool to enhance the capabilities of healthcare professionals, facilitating informed decision-making and augmenting their productivity. 

What is your advice for companies developing AI-based solutions to cater to the healthcare sector?

The key with AI lies in the expertise of those developing and implementing it. Many companies, driven by overconfidence and lacking sufficient expertise, often promise more than they can deliver; the recent challenges faced by Babylon can be seen as evidence of this. The mismatch between promises and reality poses significant challenges for investors when considering the potential value of a product, especially those in the prototype phase of development. So, make sure you have the right people who have a true understanding  not just of medicine, but also of healthcare delivery, and beware over-promising.

Another important concern is determining liability when AI-driven systems fail. For instance, if an AI platform misses a critical diagnosis, who bears responsibility? Is it the software creators? The healthcare organisation that purchased it? The doctor who was using it as a decision-making aid? Addressing liability will become even more complex when questions concerning algorithmic development and the training data used come more into the spotlight, which I believe they will. Companies must therefore ensure that they have extremely clear governance around the development of their products in the event that this is challenged in the future.

In certain domains, like robotic assisted surgery, I have seen companies attempt to sell the idea of future autonomous systems that can eventually replace human operators. Given the current technological and regulatory limitations, such ambitious claims require a realistic assessment and investors must be cautious when evaluating these companies for their true value. Ensuring responsible and transparent development and assessing the true feasibility of AI's capabilities is crucial to avoid undue risks for investors for what is essentially often unproven technology. 

Accountability and authenticity are also critical factors. Investors seek transparency and honesty, not exaggerated claims, and being forthright with investors is essential for sustainable and meaningful partnerships. A balanced and informed approach to AI adoption is essential when considering its potential benefits and the practical challenges accompanying its implementation in healthcare. Scrutiny, ethical considerations and realistic expectations are paramount to realising AI's true value in diverse settings, particularly in healthcare. For me, it’s really important that we don’t lose sight of these principles if AI is going to most effectively serve its ultimate function: improving patient experience and outcomes.

I am still unclear as to whether it should be seen, as many would have us believe, as a unicorn solution to our healthcare problems, or indeed whether it is actually a chimera in disguise.