Skip to main content

AI can prove its value to society by slashing hospital readmissions

Image credit: Pixabay (Image credit: Image Credit: ElasticComputeFarm / Pixabay)
About the author

Orlando Agrippa has over 10 years of executive and CIO experience in the UK NHS system with relentless focus on the delivery of Clinical Systems, Business Intelligence and Analytics. Orlando spent time understanding the US healthcare system and worked in Australia as an analytics transformation director. With a track record of turnaround senior management success in both private and public sector organisations of varying scale and complexity. 

It is widely accepted that AI is on the cusp of revolutionising many industries, despite the fact there are still precious few practical examples of its success. So how can this technology, that holds so much promise, live up to the hype? It is clear to me that to deliver on this promise, it has to be used to make a real and positive difference, and do that in a way that transforms lives and fundamentally benefits society.

Some industries are more advanced in their application of this technology - retail and manufacturing for example have taken a lead, using AI to improve the customer experience, or taken the most menial tasks from humans to drive greater efficiencies. But there are other industries where there is the potential to solve some of the most enduring social problems with this new approach – none more so than in the pharmaceutical and healthcare space where the approach to drug testing and patient care is in need of change.

The perception is that industries, across the board and across multiple lines of business, are on the verge of a complete overhaul in the way they operate and will soon be structured completely differently. At the moment, 40% of companies believe their business models will simply not exist in the next five years. Transfer this sentiment to healthcare and the potential benefits are not just exciting, but truly transformative. Consider hospital readmissions - the perennial challenge for healthcare professionals and usually the result of getting the balance between efficiency and customer service very wrong.

Image credit: Pixabay

Image credit: Pixabay (Image credit: Image Credit: Rawpixel / Pixabay)

The readmission challenge

It might be considered controversial, but hospitals at their most basic form are very similar to hotels or factories. They need to work efficiently in moving people through the system, focus on delivering the most cost-effective processes and also linked very closely to that, deliver the best customer experience. Let’s face it, leaving hospital as quickly as possible is not just a benefit for the hospital, it’s what most patients really want.

But there is a fine balance. Releasing people too soon has an exponentially damaging effect on both the patient and the hospital. Patients released too soon will almost certainly be readmitted, some with more serious complications than before that put further strain on services. But there is huge pressure on healthcare professionals to maintain the flow of patients to reduce blocked beds or stranded patients - where people are kept in hospitals for longer than they should be. When that pressure leads to the early release of a patient, the impact on confidence and trust is massive, and the expertise of the hospital is called into question.

Consider the hotel and factory metaphor again. How costly is the loss of confidence when a batch of cars is recalled because of a problem with processes in a factory? Also, what is the impact on customer experience if someone is rushed to check out at the end of a dream holiday?

How AI can help

Using AI technology as an assistant in these circumstances has the potential to make a real difference. By using the vast amount of rich patient data available, hospitals have the potential to deploy advanced algorithms to predict the impact of the many different circumstances that healthcare practitioners become adept at handling. In this way, AI and machine learning prove to be an indispensable tool in maintaining patient flow. It has the capability to predict the impact of patient demographics on how quickly they will move through the system, allowing hospitals to have the right resources in place and even pre-empt peaks in demand or problems that might occur – predicting rather than reacting to circumstances.

Of course, the years of clinical experience and expertise healthcare professionals use to assess patient’s fitness - making strategic decisions on bed allocation and future resourcing - is difficult to beat. Having said that, using machine learning to process the same data and support clinicians of all levels provides even deeper insight into the patient journey. It allows professionals to focus on delivering the best treatment and making strategic decisions, based on data they would not have previously had visibility of.

There are a few forward-thinking hospitals that are already building strategic partnerships with businesses and suppliers that have an AI capability. However, speaking to my NHS colleagues it is clear there is a reticence that comes from a lack of understanding of what it actually entails and how to deploy it with ever depleting budgets. 

With around 1 million patients passing through the NHS every day and a further 4.3 million waiting for treatment, making that factory run efficiently is a task that has to be delivered on a vast scale. The pressure to move people through the system quickly is increasing and as a result, readmissions become more frequent. This not only makes the patient experience negative but also puts more strain on the system. Helping solve this problem will quite literally transform lives and benefit society.

Image credit: Shutterstock

Image credit: Shutterstock (Image credit: Image Credit: Everything Possible / Shutterstock )

The future of AI in hospitals

In healthcare, whether it is based on the insurance or publicly funded approach, there is universal concern about safeguarding the future of hospitals. The blind spot for AI and machine learning is hindering their progress in delivering models of business that focus both on efficiency and customer service – areas where this technology is already able to prove its value. This might be as a result of the lack of understanding of what AI actually is at the moment, but in time, we should expect it to be as inbuilt into the hospital system as our hospital’s hygiene process with hand sanitiser.

Orlando Agrippa, CEO of Drapper & Dash