Digital nursing 1: exploring the benefits and risks of artificial intelligence

Artificial intelligence is already utilized in healthcare; this primary article in a three-part sequence on digital healthcare seems at the benefits and risks

Artificial intelligence is already getting used to assist superior scientific choices, enhance the accuracy and security of care, and plan and handle NHS assets. It could make machines do issues that used to require human intelligence, and can draw on big quantities of information to make calculations which are past any human being. Artificial intelligence-enabled robots are being developed to tackle some nursing features. Nurses want to look at how their very own roles could also be modified and advocate for affected person involvement in mild of rising applied sciences. They will even want coaching and assist to really feel assured utilizing artificial intelligence instruments. This first article in a sequence on digital healthcare examines the benefits and risks of artificial intelligence.
Citation: Agnew T (2022) Digital nursing 1: exploring the benefits and risks of artificial intelligence. Nursing Times [online]; 118: 8.
Author: Thelma Agnew is a contract well being journalist.

The ministerial ahead to Joshi and Morley’s (2019) report, revealed by NHSX, supplied one key cause to be enthusiastic about artificial intelligence (AI) in healthcare: “put merely, this know-how could make the NHS even higher at what it does: treating and caring for individuals”. AI is thrilling, however what’s it, precisely? This first in a sequence of articles about digital healthcare will talk about the benefits and risks of AI.
Increasingly, nurses are inspired to guide and form rising digital applied sciences, to make sure the modifications made are match for objective and ward in opposition to unintended penalties, corresponding to elevated workloads, dehumanised care and the exclusion of already marginalised teams of individuals. The areas that might be improved, and even reworked by AI, based on Joshi and Morley’s (2019) report, embody:
Diagnostics – AI picture recognition instruments (pushed by information) can assist clinicians’ judgement by studying photos corresponding to mammograms, mind scans and eye scans. Diagnosis may grow to be quicker and remedies extra correct;
Planning and administration of NHS assets – AI could make advanced calculations, based mostly on big quantities of information, that make it potential to foretell, for instance, how a lot blood plasma a hospital will want or the probability of individuals lacking their outpatients’ appointments;
Medicines – AI comparisons of drug compounds, for instance, can obtain breakthroughs in weeks that may take human researchers years.
Despite this, it’s tough to method, not to mention lead on, technological developments if nurses don’t perceive them, and AI might really feel too huge to understand at occasions.
There are many definitions in the ever-growing literature launched about AI. Joshi and Morley’s (2019) report advised that one of the most helpful definitions in the discipline of healthcare can be the oldest; they defined that it dates from a analysis venture in 1955 and acknowledged that AI is “the science of making machines do issues that may require intelligence if executed by individuals”.
A publication by The King’s Fund – particularly, Mistry (2020) – gave a extra detailed, however nonetheless easy rationalization, stating that AI “is an umbrella time period encompassing a quantity of totally different approaches the place software program replicates features which have, till not too long ago, been synonymous with human intelligence. This features a huge spectrum of skills corresponding to visually figuring out and classifying objects [and] changing speech to textual content and textual content to speech”.
The origins of AI return a long time, so why are we listening to a lot about it now? One cause is that current developments in utilized arithmetic and pc science have made computer systems a lot better at studying patterns in massive quantities of advanced information, releasing AI’s potential (Mistry, 2020). The prospects of AI are being additional expanded by machine studying, which has been outlined by Mistry (2020) as “a sort of artificial intelligence that allows computer systems to be taught with out being explicitly programmed, that means they will train themselves to vary when uncovered to new information”.
AI in nursing
Most well being workers nonetheless lack direct expertise with AI applied sciences, as is highlighted in Nix et al’s (2022) report, developed by NHS AI Lab and Health Education England (Box 1). However, the growing use of AI applied sciences in nursing, corresponding to offering info for superior scientific choice assist, is considered inevitable (Booth et al, 2021; Robert, 2019).

Box 1. AI applied sciences: a wierd science is about to grow to be extra acquainted
A survey of >1,000 NHS workers in the UK by The Health Foundation – particularly, Hardie et al (2021) – discovered that three-quarters of respondents had heard, seen or learn “not very a lot” or “nothing in any respect” about AI. The survey additionally discovered:
Healthcare workers who have been extra aware of AI applied sciences have been extra constructive in direction of these applied sciences
Nurses and midwives have been much less constructive about AI than medical doctors and dentists, however extra constructive than healthcare assistants
Assistive purposes of AI, corresponding to picture evaluation and screening, have been perceived as a better alternative to enhance healthcare than robotic care assistants or different autonomous varieties of AI.
The Health Foundation survey recognized fears amongst well being staff that AI applied sciences current a menace to their jobs. This has been echoed in a number of different research, together with issues about information governance, cyber safety, affected person security and equity (Nix et al, 2022).
The reservations about AI are unlikely to place the brakes on their adoption in healthcare. Nix et al’s (2022) report factors to proof that use of AI is accelerating, with an growing quantity of AI applied sciences anticipated for use in healthcare in the subsequent three years. It highlights the AI roadmap report by Health Education England and Unity Insights (2021), which surveyed greater than 200 AI applied sciences: 20% have been estimated to be prepared for large-scale deployment in 2022, with an extra 40% prepared in the subsequent three years.
AI = artificial intelligence

AI-enabled choice assist programs probably present quite a few benefits; for instance, they’ve already dramatically improved the detection of sepsis (Horng et al, 2017). However, there are additionally risks, as AI is barely pretty much as good as its information. Nix et al’s (2022) report warns that confidence in artificial intelligence (AI) will not be all the time fascinating when utilizing it for scientific choice making, and nurses have to recognise when to steadiness it with different sources of scientific info (Box 2).

Box 2. Confidence in utilizing AI for scientific choice making
A current report from NHS AI Lab and Health Education England recommends:
During scientific choice making, contemplate what’s an acceptable stage of confidence in info derived from AI know-how and steadiness this with different sources of scientific info
Appropriate confidence will differ relying on the know-how and scientific context
It could also be cheap to belief an AI know-how, whereas having low confidence in a selected prediction from that know-how as a result of it contradicts sturdy scientific proof or is being utilized in an uncommon state of affairs. As acknowledged by Nix et al (2022): “The problem is to allow customers to make context-dependent worth judgements and repeatedly confirm the acceptable stage of confidence in AI-derived info”.
The fundamental suggestion from the report is to develop instructional pathways and supplies for all well being professionals to equip them to confidently consider, undertake and use AI (Nix et al, 2022).
AI = artificial intelligence

AI programs that evolve themselves might mirror or reinforce societal biases (for instance, racial biases) and different inequities current in the information (Obermeyer et al, 2019; Gianfrancesco et al, 2018). It is necessary for nurses to be concerned in improvements corresponding to AI to ensure they’re creating programs in keeping with moral frameworks and to advocate for affected person involvement (Booth et al, 2021). There can be a threat that AI programs that carry out extraordinarily properly in managed circumstances will probably be much less spectacular in the actual world, and there are unanswered questions on their security and value effectiveness in healthcare settings (Maguire et al, 2021).
The NHS is already utilizing AI and machine studying, at a inhabitants stage, to assist establish older individuals in native areas who’re in danger of frailty and hostile well being outcomes; one instance of that is the Electronic Frailty Index, which pulls on information that’s routinely recorded by GP practices (NHS England, 2017).
Predictive analytics in digital affected person information must also, more and more, assist medical doctors and nurses to diagnose and deal with the particular person affected person in entrance of them (HEE, 2019). AI has additionally performed a key function in informing the authorities’s response to the coronavirus pandemic: the launch of the NHS Covid-19 Data Store by NHSX has aided the evaluation of huge quantities of information to:
Reveal how the virus is spreading;
Ascertain how the NHS is coping;
Suggest the handiest present and future interventions (Maguire et al, 2021; Gould et al, 2020).
AI can be central to the authorities’s new digital well being and social care plan (Department of Health and Social Care and NHS England, 2022). This consists of utilizing AI to develop “new diagnostics capability to allow image-sharing and scientific choice assist… These applied sciences assist testing near house, streamlining of pathways, triaging of ready lists, quicker analysis and levelling up under-served areas”
Robotics in nursing
With the growth of smaller and more-sophisticated digital elements, robots embedded with AI will seemingly grow to be extra extensively utilized in healthcare. (Mistry, 2020; HEE, 2019). The extremely influential Topol evaluate predicted that robots would grow to be the “{hardware}” for AI, performing handbook and cognitive duties, and releasing up healthcare workers to spend extra time doing issues which are “uniquely human”, corresponding to interacting with sufferers (HEE, 2019).
This image is sophisticated by the incontrovertible fact that robots have additionally been developed to offer social and emotional assist to individuals, arguably blurring the line between machines and people. Examples presently in use embody:
Sophia, a companion robotic for older individuals;
Miko 2, a robotic for youngsters that may reply to feelings;
Paro, an animal remedy robotic (Robert, 2019).
A standard theme in the literature on AI and robotics in healthcare is the expectation that sufferers, in addition to workers, will obtain a number of benefits from the introduction of clever machines, with enhancements in early analysis, and the accuracy and security of care (Mistry, 2020; HEE, 2019). Unlike human healthcare staff, robots by no means get bored or drained, are unaffected by hazards in scientific settings, corresponding to X-ray radiation, and can endlessly repeat duties that require precision with no drop in efficiency (Mistry, 2020). The potential is there for robotics to assist with all the pieces from transferring sufferers to surgical procedures which are past the capabilities of surgeons (Mistry, 2020).
With the enter of nurses, robots are additionally being developed to tackle some nursing features, together with:
Ambulation assist;
Vital-signs measurement;
Medication administration;
Infectious illness protocols (Robert, 2019).
What AI means for nursing
Developments don’t imply that nurses are about to get replaced by clever machines, however they do counsel that nursing, as it’s presently understood, will change. A 2019 examine by former American Nurses Association govt vice chairman Nancy Robert advised that the arrival of telehealth and good robots in individuals’s properties will see nursing evolve into extra of a training function, guiding sufferers to enhance their well being and offering continuity of care, however nonetheless being bodily current at the bedside when it actually issues (Robert, 2019). A nursing dean quoted in the examine mentioned they might not think about ever selecting a robotic over a human to take care of them in the event that they have been dying, and acknowledged that: “Nuances in human behaviour will hold nurses on the entrance line of care” (Robert, 2019).
It is hoped that AI and robotics will work collectively as assistants to nurses by, on the one hand, supporting superior scientific choices and, on the different, automating primary duties which are time consuming however might be carried out by somebody – or one thing – else. In this imaginative and prescient of an AI-enabled future, machines unlock nurses professionally to make use of their training, expertise and expertise (Robert, 2019). A barrier to nurses utilizing the know-how to fulfil their potential might be different healthcare disciplines’ resistance to nurses practising at the high of their licence (Robert, 2019).
There can be the threat that, as AI instruments grow to be extra extensively utilized in healthcare, they’ll affect how nurses’ practise, with out nurses having the alternative to affect them. A 2021 examine on how the nursing career ought to adapt for a digital future referred to as for an “speedy inquiry” into the affect of AI on nursing apply for the subsequent 10 years and past (Booth et al, 2021). The authors identified that the elevated use of AI is bringing with it new coverage, regulatory, authorized and moral points; they referred to as on the nursing career to:
Investigate the risks and alternatives;
Examine its personal function;
Develop frameworks and tips to steer nursing apply.
Robert (2019) means that nurses have a accountability to ask about the information used to coach AI programs they use, and guarantee they’ve been checked for bias.
As outlined in Box 3, nurses will want coaching and assist to really feel assured in, and overcome the obstacles to, utilizing AI instruments – which is able to solely work correctly if the ageing know-how infrastructure of the NHS improves (Joshi and Morley, 2019). The pleasure about AI is justified however it’s important to not get dazzled by the “hype” (Joshi and Morley, 2019).

Box 3. AI and robotics in healthcare: obstacles and studying
Health Education England’s (2019) Topol evaluate recognized important obstacles to the deployment of AI and robotics in the NHS. These included:
NHS information high quality
Information governance
Lack of experience in AI and robotics.
Along with a code of conduct and steering on the effectiveness of the applied sciences, the evaluate referred to as for workforce studying in three key areas:
Knowledge and expertise in information provenance, curation and governance
Knowledge and understanding of moral concerns
Critical appraisal of digital healthcare applied sciences, and understanding how the know-how works.
AI = artificial intelligence.

Key factors
Technologies powered by artificial intelligence may rework healthcare
Artificial intelligence is already informing superior scientific choice making, corresponding to in the detection of sepsis
In future, artificial intelligence-enabled robotics might act as assistants to nurses, releasing workers to practise at the high of their licence
Nurses ought to advocate for sufferers and be certain artificial intelligence programs don’t reinforce information bias
The nursing career wants to research the risks and alternatives of artificial intelligence and develop frameworks to information apply

ReferencesBooth RG et al (2021) How the nursing career ought to adapt for a digital future. BMJ; 373: n1190.Department of Health and Social Care, NHS England (2022) A plan for digital well being and social care. 29 June (accessed 29 June 2022).Gianfrancesco MA et al (2018) Potential biases in machine studying algorithms utilizing digital well being report information. JAMA Internal Medicine; 178: 11, 1544-1547.Gould M et al (2020) The energy of information in a pandemic., 15 April (accessed 21 June 2022).Hardie T et al (2021) Switched On: How Do We Get the Best out of Automation and AI in Health Care? The Health Foundation.Health Education England and Unity Insights (2021) AI Roadmap: Methodology and Findings Report. HEEHealth Education England (2019) The Topol Review: Preparing the Healthcare Workforce to Deliver the Digital Future. HEE.Horng S et al (2017) Creating an automatic set off for sepsis scientific choice assist at emergency division triage utilizing machine studying. PLoS ONE; 12: 4, e0174708.Joshi I, Morley J (2019) Artificial Intelligence: How to Get it Right – Putting Policy into Practice for Safe Data-driven Innovation in Health and Care. NHSX.Maguire D et al (2021) Shaping the Future of Digital Technology in Health and Social Care. The King’s Fund.Mistry P (2020) The digital revolution: eight applied sciences that can change well being and care., 13 November (accessed 21 June).NHS England (2017) Supporting Routine Frailty Identification and Frailty Through the GP Contract 2017/2018. NHS England.Nix M et al (2022) Understanding Healthcare Workers’ Confidence in AI. NHS.Obermeyer Z et al (2019) Dissecting racial bias in an algorithm used to handle the well being of populations. Science; 366: 6464, 447-453.Robert N (2019) How artificial intelligence is altering nursing. Nursing Management; 50:
9, 30-39.

Recommended For You