NHS Shared Business Services, which provides back-office and IT services to NHS Trusts, recently launched a framework tender to help them purchase ‘intelligent automation’ technologies and services. But can robotic process automation and AI really help cut back the notorious burden of bureaucracy that weighs on the UK’s healthcare system?
The framework tender, which was announced in February and opened to suppliers last month, aims to support the “adoption, implementation and ongoing development” of intelligent automation by NHS Trusts and other public sector organisations.
“By ‘intelligent automation’ we mean a combination of some or all of artificial intelligence, robotic process automation (RPA) and machine learning to streamline and improve an organisation’s decision making, in particular in the healthcare context to improve the effectiveness and quality of the patient experience,” a spokesperson for NHS SBS told Tech Monitor.
The framework includes a lot for “intelligent automation as a service”. This lot is designed for buying organisations “who want to completely outsource the day-to-day running of an intelligent automation service and solution due to barriers to entry into this area such as funding and lack of internal expertise”, the tender says. “Our research in designing this framework showed that NHS/healthcare organisations, for varying reasons, want the flexibility to be able to appoint partners on a managed service, effectively outsourced basis.”
This might include setting up an external ‘centre of excellence’ for intelligent automation, “a new independent organisation established to support healthcare organisations – typically NHS Trusts – through facilitating, co-ordinating and implementing a programme approach to rollout and promoting best practice”.
Automated admin in the NHS
There are hopes that AI and RPA might help to relieve the burden of bureaucracy that blights the NHS. “Frontline staff report that too much of their time is still spent on administrative tasks, meaning less time for the direct care of patients and service users,” a consultation by the Department of Health and Social Care carried out last year concluded. “For example, around a third of a community-based clinician’s time (or 88 days per working year) is estimated to be spent on administration and patient co-ordination.”
The causes for this administrational burden include duplicative requests for data from multiple bodies and “time-consuming staff processes and information management”. The consultation found that patient records management is the greatest source of bureaucracy for frontline NHS workers, with “excessive forms and multiple systems” frequently cited.
Meanwhile, in a 2018 survey by the British Medical Association, 53% of UK doctors said that 'more effective IT systems that are interoperable' would improve their day-to-day working lives, second only to 'guaranteed safe levels of medical staffing' (57%).
While the Department of Health did not identify intelligent automation as a potential solution to this bureaucracy, NHS Trusts are already using AI and RPA to cut down on admin. Earlier this year, Tech Monitor spoke to Kettering General Hospital NHS Foundation Trust's digital director about its use of RPA to automate Covid-19 reporting and to power a chatbot that helps with HR admin.
Some are sceptical of the maturity of AI and RPA for use in a healthcare setting, however. The NHS Business Services Authority, another service provider for NHS Trusts, referred to both as "hype tech" in its digital strategy (although it added that "we may adopt some of these technologies").
Others are more circumspect. "RPA is good for some things, but dreadful for others," says Dr Peter Bloomfield, head of policy and research at Future Care Capital, a healthcare innovation think tank and investor. "Where there are very set forms and structured [documents] that never change, it works really well. But for clinical notes written across three different boxes, and things like that, it starts to fall down a bit."
Meanwhile, RPA is often used as a short-term solution for integrating legacy systems, says Reetika Fleming, VP for research at analyst company HfS Research. This can lead to new IT management issues in future. "That's why we declared the death of RPA in 2019," she explains. "We saw this type of tooling being used to band-aid legacy [systems], in such a siloed way. And now we've created new legacy, because there's all these 'bot estates' that need to be managed over time." Bloomfield agrees: "There's definitely that thing of putting plasters on top of plasters," he says.
Fleming advises that intelligent automation should instead be considered in the context of an overall transformation of an organisation's processes, alongside data analytics and staffing. "We need to be more creative with how we design our business processes," she says. Organisations should ask: "How much of this work actually needs to exist and of that, how much of that is automatable? And then of that remaining work, where should this work be done and by who?"