Tom Hardie: The promise of greater automation in healthcare

There is an urgency around the use of technology to transform healthcare at the momentwith Matt Hancock’s technology vision, the development of NHSX, and clear aspirations for technology articulated in the NHS Long Term Plan.

The greater use of automation is part of that vision.

What is the potential of automation in healthcare?

Automation is not a new concept and has its roots in the industrial revolution. Traditionally, routine repetitive manual tasks were performed by machines with minimal human input. Think of car assembly lines and ATMs. In health, computers and other technology have been used to automate billing, pill counting, and the continuous monitoring of patients’ vital signs in hospital.

Advancements in robotics and AI (aided by big data and increased computing power) are extending the boundaries of what is possible, both in the realm of automating physical tasks, and increasingly, tasks that require cognitive skills and abilities to perform. [1] At the Health Foundation, we are seeing the potential of automation in a range of areas of healthcare:

  • Reducing the burden of administrative work: Research by Oxford University, due to publish in the next few months, shows that automated systems could be particularly helpful in high volume and frequent tasks such as telephone answering, letter writing, and information filtering. Many believe this could release time for clinicians to care for and interact with patients and make workloads more manageable. [2]
  • Improving diagnosis and treatment: Barking, Havering and Redbridge University Hospitals NHS Trust is developing a system to assist the emergency department triage process by quickly identifying high-risk patients needing urgent care.
  • Improving the management of resources: In the context of patient flow, Northwest London CLAHRC and Imperial College London are building a model to analyse a range of data in real time and make predictions about when high levels of acute bed occupancy might be likely, automatically alerting staff in advance.

Getting automation right

Although it is tempting to be swept up in the promise of automation, there are some important questions to considertaking a holistic viewif we are to reap the benefits.

Some of these revolve around quality and safety. For patient-facing application of automation technologies, such as symptom checkers, evidence from real-world testing will need to demonstrate they are safe, effective, and provide a good user experience for all. It is crucial that new technology does not entrench or worsen health inequalities, for example, by excluding people with low levels of digital literacy. As the Good Things Foundation note, 11.3m people in the UK do not have basic online skills, and are more likely to be older, poorer, live with disabilities and to need healthcare services. [3]

Another important question relates to the impact of automation technologies on the workforce. While much of the broader debate on automation assumes technologies will replace human roles, in healthcare it is more often about automation augmenting what staff can do. In this context, how do we ensure that automation is useful and empowering for staff and does not have an adverse impact on their workloads or their ability to perform their jobs?

How do we successfully embed automation?

Finally, how do we successfully embed automation which has been evaluated, and shown to be successful, on the frontline? Experience shows that effective implementation of technologies in a highly complex sector like health care is challenging and will not be achieved quickly. This goes beyond technical issues that will need to be addressed such as software, hardware, interoperability and training. Critical also will be ensuring staff have the time to engage in the work of redesigning clinical pathways and workflows to accommodate the technology, a crucial factor that is often underestimated. [4]

More work is needed to address the many challenging questions that automation presents while at the same time capitalising on the benefits. Alongside the BMJ, we’ll be exploring these issues as part of Twitter chat on Tuesday 21 May, 20:00-21:00. (#NHSautomation @HealthFdn @bmj_latest). I hope you can join us.

The opportunities and challenges around AI, automation and other emerging technologies in the NHS will be discussed at the Health Foundation annual event 2019 on Thursday 23rd of May.

Keynote sessions from the event will be livestreamed and you can register now to watch them here.

Tom Hardie is the Health Foundation’s Improvement Fellow.

Twitter: @tlhardie1

Competing interests: None declared


1] A Future that Works: Automation, Employment, and productivity. McKinsey Global Institute, January 2017
2] Willis, M., P. Duckworth, A. Coulter, E.T. Meyer, and M. Osborne. (2019). The Future of Health Care: Protocol for Measuring the Potential of Task Automation Grounded in the National Health Service Primary Care System. Journal of Medical Internet Research Protocols, vol. 8, no. 4. doi:10.2196/11232. Topol E. The Topol Review: Preparing the healthcare workforce to deliver the digital future. Health Education England, 2019; Fogel, Alexander L.Kvedar, Joseph C, Artificial intelligence powers digital medicine npj Digital Medicine 2018/03/14
3] Davies, R. Topol v Marmot. Medium.
4] The Spread Challenge, Health Foundation, September 2018.