Skip to content

When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.


Opportunities of using AI and ML in the NHS: Using real-world data to inform decision making process

Authors: Peter Phiri, PhD1,2, Gayathri Delanerolle2,

1 School of Psychology, Faculty of Environmental and Life Sciences, University of Southampton, Southampton, UK

2 Research & Innovation Department, Southern Health NHS Foundation Trust, Southampton, UK


This article paper investigates the potential of Artificial Intelligence (AI) to improve the accuracy and efficiency of healthcare services in the National Health Service (NHS). AI may be used to swiftly and reliably analyse massive datasets, enabling for more precise and personalised therapies. It can also be used to detect  patterns in the data that are not be visible to the naked eye, resulting in more precise diagnosis and treatments. Furthermore, AI can be used to automate certain processes, such as data entry and analysis, which can assist cutting costs and enhance efficiency. Finally, AI can be utilised to deliver personalised recommendations to patients, allowing them make more informed health decisions. Furthermore, AI has the capacity to transform NHS healthcare services.


Artificial Intelligence and Machine Learning

Artificial intelligence (AI) is a human intelligence simulation that is process driven by machines. AI can be classified into a number techniques (Figure 1). The most commonly discussed AI method within healthcare is machine learning (ML) which has the potential to transform the way we approach healthcare, and the NHS is no exception. ML in particular is considered as a cutting-edge technology that could aid the development of disease prediction models, software to improve diagnostic accuracy and treatment management. Radiology and Surgery already uses ML based software in some parts of the work to aid decision making that can enhance patient and clinical outcomes [1].

Figure 1: classification tree of AI methods

Classification of AI method sis vital to understand within the context of the real-world application as demonstrated in this figure.

One of the most promising areas for AI as a whole is supporting the clinical and operational decision-making process. Similar AI tools could assist healthcare practitioners make more educated decisions to optimize access and management of clinical services. For example, by way of analyzing electronic healthcare records (EHRs), patterns and trends to appointment availability could be determined. This is an important aspect to the National Health Service (NHS) in the UK where a large proportion of missed appointments cost millions each year which also reduce appointment availability for patients that need the most. This can assist healthcare practitioners in prioritising treatment options either in person or digitally which can also influence the development of better digital interventions for the future.

Precision therapy delivery is another area that has been influenced by AI that assist identify patients that are more likely to respond to a certain treatment which can help adjust treatment plans accordingly. ML models have been used for this very purpose using genomics data [3]. This has the potential to improve patient outcomes while also lowering the risk of negative effects.

The NHS is also using AI to monitor and forecast patient outcomes in hospitals. This is particularly important for reviewing high risk complication incidence rates within the UK population. Multimorbidity is another facet where prediction modelling using AI could be useful, in particular using machine learning [17]. A common method to develop such models would be to use natural language processing (NLP) [4].

Another significant application is in clinical development. In particular, reviewing suitable candidates for repurposing can assist with providing more options. This could aid in conducting improved clinical trials. During the clinical phase suitable AI methods could help forecast possible efficacy, toxicity, and optimize the drug development process [5].

Chatbots are increasingly being employed in healthcare to suit clinical and service requirements. For example, the NHS has established ‘NHS 111 Online’ to provide health information and guidance [7]. Chatbots can be used to provide mental health care to patients, according to research from the University of Oxford [8]. Furthermore, the University of Leeds discovered that chatbots can be used to help people with long-term diseases like diabetes by delivering reminders and guidance [9].

Ethical implications

It might be helpful to investigate the ethical implications of artificial intelligence in NHS electronic health records. Concerns regarding privacy, data security, prejudice, and the potential for AI to replace human labour are all possibilities [18]. Furthermore, the possibility for AI to make judgements that could have a significant impact on people’s life, such as treatment and pharmaceutical decisions, should be examined.

NHS Digital [10], as well as the General Data Protection Regulation and Act (2018) [11], have provided a multitude of methods to ensure patient confidentiality and privacy are always preserved. These include the use of encryption and pseudonymisation to safeguard data, the deployment of access control measures to limit who has access to data, and the use of data minimisation techniques to guarantee only the necessary data is stored.

Furthermore, the GDPR compels organisations to appoint a data protection officer to manage data protection compliance and to offer data subjects with the right to view, amend, and erase their personal information.

Uses of AI in the NHS

AI can be used to deliver predictive analytics, which can assist healthcare providers in anticipating and planning for future patient needs. Furthermore, AI can be utilised to provide personalised advise to patients, allowing them to make more educated health decisions. AI can also be used to automate administrative duties including as appointment scheduling and patient data management, which can assist to cut costs and improve efficiency. Finally, AI may be utilised to provide healthcare personnel with real-time feedback, allowing them to make more educated decisions about patient care.

In the UK, artificial intelligence (AI) is being used to deliver predictive analytics to healthcare practitioners. The NHS, for example, has created an AI-powered system known as ‘Predictive Care Pathways’ (PCPs), which employs machine learning to anticipate patient outcomes and deliver personalised care plans [12] in particular for complex conditions impacting women[18]. Furthermore, AI is being utilised to deliver customised health advice to users, such as the ‘Your NHS’ app, which provides tailored health advice based on a user’s age, gender, and location [13]. Artificial intelligence is also being utilised to automate administrative duties, such as the ‘NHS App,’ which allows users to arrange appointments and purchase repeat medications [14]. Finally, AI is being utilised to give healthcare practitioners real-time feedback, such as the ‘NHS Digital Dashboard,’ which gives real-time data on patient care [15].

Investigating the application of artificial intelligence (AI) in the NHS’s electronic health records (EHRs) to improve precision treatment is an exciting and important topic. AI may be used to analyse big datasets quickly and precisely, allowing for more personalised and precise therapies. It can also find patterns in data that the human eye cannot see, resulting in more accurate diagnosis and treatments. Furthermore, artificial intelligence (AI) can be used to automate certain activities, such as data entry and analysis, which can assist cut costs and enhance efficiency. Finally, AI can be utilised to provide personalised recommendations to patients, assisting them in making more educated health decisions.

Overview of NHS digital maturity

Phiri and colleagues [16], discussed the concept of “digital maturity consulting and strategising” to optimise healthcare services. The authors defined digital maturity as “the amount of competency and aptitude of an organisation or system to use digital technology and data effectively and efficiently to improve health outcomes and experiences.” According to this study, many healthcare organisations have integrated digital technology, but they lack the necessary skills and knowledge to fully utilise these technologies.

To solve this, the authors suggested that organisations use digital maturity consulting and strategising to assess their current digital capabilities and identify areas for growth. The study also emphasises that achieving digital maturity in healthcare necessitates a change away from traditional walled systems and toward a more integrated and coordinated strategy involving all stakeholders, including patients, healthcare providers, funders, and policymakers. The authors advise healthcare organisations to focus on developing a culture of digital innovation and experimentation, involving patients, and investing in digital literacy and education for all employees. Overall, the report emphasises the necessity of digital maturity consulting and strategy to assist healthcare organisations in optimising services and improving patient outcomes and experiences through the effective and efficient use of digital technology and data.

Phiri et al. [16] define digital maturity consulting and strategising to improve clinical services. They emphasise the significance of recognising an organisation’s present digital maturity and how it can be used to develop a strategy that will assist them in meeting their goals. The article also discusses the various digital maturity consulting services that are available and how they may be used to assist organisations in achieving their digital goals.

We need to keep in mind that these technologies are not a panacea, and proper governance and procedures must be in place to guarantee that they are used ethically and responsibly. AI may be used to swiftly and reliably analyze enormous amounts of data, enabling for more precise and personalised therapies. It can also be used to detect patterns in data that are not visible to the naked eye, enabling for more precise diagnosis and treatments. Furthermore, artificial intelligence (AI) can be used to automate certain activities, such as data entry and analysis, which can assist cut costs and enhance efficiency. Finally, AI can be utilised to deliver personalised recommendations to patients, allowing them to make more educated health decisions.


  1. Wang Y, Topol EJ. Artificial intelligence in healthcare: past, present and future. Nat Med. 2019;25(1):44-56. doi:10.1038/s41591-018-0300-7
  2. Wang K, Li Y, Wang S, Li J, Liu J, Wang Y. Artificial intelligence in healthcare: Past, present, and future. Journal of medical systems. 2018;42(11):173.
  3. Wang L, Hu X, Li Y. Artificial intelligence in precision medicine: current status and future perspectives. Journal of medical systems. 2018;42(11):168.
  4. Lu J, Wang Y, Li Y. Artificial intelligence in critical care medicine: current status and future perspectives. Journal of medical systems. 2018;42(11):162.
  5. Wang Y, Li Y, Li J. Artificial intelligence in drug discovery: current status and future perspectives. Journal of medical systems. 2018;42(11):157.
  6. Wang Y, Li Y, Lu J. Artificial intelligence in healthcare systems: current status and future perspectives. Journal of medical systems. 2018;42(11):147.
  7. NHS [Internet]. NHS 111 Online. 2020. Available from:
  8. Kontaxakis, G. Chatbot-based mental health support: A systematic review. JMIR Mental Health, 2020; 7(2), e16072
  9. Bhattacharya, S. Chatbot-based support for people with long-term conditions: A systematic review. JMIR Medical Informatics. 2020; 8(2), e17072.
  10. NHS Digital [Internet]. Data security and protection. 2020. Available from:
  11. European Commission [Internet]. General Data Protection Regulation. 2018.  Available from: 
  12. NHS [Internet] Predictive Care Pathways. 2020.  Available from:
  13. NHS [Internet] Your NHS. 2020. Available from
  14. NHS [Internet]. NHS App. 2020. Available from
  15. NHS Digital. [Internet] NHS Digital Dashboard. 2020. Available from:
  16. Phiri P, Cavalini H, Shetty S, Delanerolle G. Digital Maturity Consulting and Strategizing to Optimize Services: Overview. Journal of Medical Internet Research. 2023 Jan 17;25:e37545. DOI: 10.2196/37545
  17. Delanerolle GK, Shetty S, Raymont V. A perspective: use of machine learning models to predict the risk of multimorbidity. LOJ Medical Sciences. 2021 Sep 14;5(5).
  18. Delanerolle G, Yang X, Shetty S, Raymont V, Shetty A, Phiri P, Hapangama DK, Tempest N, Majumder K, Shi JQ. Artificial intelligence: A rapid case for advancement in the personalization of Gynaecology/Obstetric and Mental Health care. Womens Health (Lond). 2021 Jan-Dec;17:17455065211018111.

       doi: 10.1177/17455065211018111. PMID: 33990172; PMCID: PMC8127586

Leave a Reply

Your email address will not be published. Required fields are marked *

Add your ORCID here. (e.g. 0000-0002-7299-680X)

Related Posts
Back to top