As one of the most centralised health providers in the world, the NHS possesses arguably one of the largest sources of healthcare data. Artificial intelligence (AI) should be adding extraordinary value for both patients and the organisation. However, in practice, the NHS is comprised of many local and regional organisations with multiple legacy systems, some of them still paper based.

Developing a consistent approach to data sharing could unlock this potential and allow analysis and improvement of health service operations and outcomes. Operationally, it could enable optimised supply chains for materials and scheduling, better use of supplies, facilities, and people, and help guide patients through the care pathway. Clinically, it could support and improve diagnostic capabilities. So, how do we liberate all that insight and realise the transformational potential?

There are two parts to the NHS’ data challenge. One is technology, which must be improved and updated to allow data sharing in the practical sense. The other is governance—ensuring appropriate data privacy is maintained so that trust and permission are explicitly given by the individuals whose data is being shared and creating the processes to facilitate that secure sharing.

Overcoming technology challenges

Currently, data is stored on multiple, incompatible and aging systems. It is also stored in different formats that make it difficult to connect data between systems.  Whilst some human intervention will always be needed to rationalise the most complex data between these systems, more than ever this could be automated. Additionally, finding the perfect combination of standards, compatibility and data security doesn’t necessarily mean creating a massive central data lake. It’s entirely feasible to create a national ecosystem of federated systems that talk to one another as needed.

This merging and standardisation is essential for future transformation, but it would also have immediate benefits. For a start, it would enable broader research that supports more effective population health management and could expand insight into social determinants of health. It would also help patients directly by giving care providers a full patient view, so information doesn’t need to be re-supplied every time a patient goes to a new provider.

Bringing it all together through trust

Trust will present an equally big challenge to data sharing and one that, arguably, needs to be tackled first. For an example of how to do this we need only look to the airline industry. Whether they consider it consciously or not, millions of people put their trust in AI to fly them from A to B every year. This trust has been established over time by adhering to rigorous industry wide safety guidelines and taking radical, public steps to identify and fix failings whenever they occur. We need a similar benchmark in healthcare.

Patient safety, and patient data privacy are paramount and require transparent dialogue with citizens. Currently, 42% of patients we surveyed said that privacy concerns were the main reason they would not use chatbots, computers or digital devices to answer health questions or access care. This ranked higher than issues like service effectiveness. We need to make the benefits tangible and provide patients with reassurance that data is secure, anonymised and being used to their benefit, as my colleagues have discussed recently

Advocacy from physicians will be essential in this process. Our research, done before the COVID-19 pandemic, indicated that English patients, regardless of age, (87 percent of people aged 55 and over and 71 percent of those aged 18-55) trust healthcare professionals significantly more than any other source of care advice. However, this relationship between physicians and AI needs to be carefully nurtured. Our research also found that, currently, trust plummets from 79 percent for physicians alone to 46 percent for physicians supported by AI.

To overcome this, advocates for insight driven healthcare will need to convince physicians of the value and give them tools to build trust with patients, such as simple interfaces to present potential  diagnoses to patients based on analysis of their symptoms and national healthcare data. This would improve individual diagnoses and feed data back into the system. It would also allow physicians to demonstrate the benefits of AI to patients and show how data is anonymised. Tools like this could enable physicians to become both advocates and beneficiaries of AI powered insight, whilst simultaneously improving transparency and trust amongst patients.

<<< Start >>>

Doctor providing virtual care.

<<< End >>>

The art of the possible

These applications would just be the start. Once patient trust is established and the data is securely available across the system, teams will have the opportunity to challenge traditional ways of working and find other creative use cases that deliver benefits. For example, giving a hospital administrator access to an optimised bed schedule based on patient forecasts, staff availability and physical capacity. With AI, planning processes that previously took months could take minutes.

Capabilities like this are no longer a luxury, they are essential to the future of the NHS. The service has the opportunity, now is the time to embrace it.

Health is changing and providers will need the full value from their existing data to realise the opportunities this change brings. Accenture’s expertise in data and AI could help you improve operations and enhance patient experience. Connect with Ashish to find out more.

Ashish Goel

Accenture UK Health Lead

Subscribe to Accenture's Humanizing Healthcare Blog Subscribe to Accenture's Humanizing Healthcare Blog