Overview
With the rapid rise of artificial intelligence (AI), the incorporation of AI into different fields including healthcare, and its possible revolutionary effects has come under the spotlight.1 One of the recent developments includes remote patient monitoring (RPM), which involves using synchronous and asynchronous monitoring tools to enable healthcare providers to track their patients' conditions or biometric data remotely.
In the UK, the Regional Scale Programme (RSP) has been pioneering the implementation of new care pathways facilitated by remote monitoring technologies since 2020 in partnership with the NHS and local care system and involves over 487,000 patients.2,3
Overview of the potential of AI in the context of healthcare
Due to the advantages of AI over human performance in analysing a large volume of data and identifying patterns, it shows enormous potential in numerous areas of healthcare.
Artificial intelligence involves the use of multiple techniques such as machine learning (ML), deep learning (DL) and Large Language Models (LLMs). Machine Learning (ML) is helpful when it comes to taking into account the complexity of human diseases.1 This is because ML is trained based on data and algorithms to learn and make predictions without explicit human intervention.1 DL is another technique highly adapted to identifying data patterns that are useful for areas such as public health management, and predictive analytics for risks or treatment outcomes.1
The benefits of AI in healthcare include improved patient outcomes, identifying risks for developing diseases or hospital readmissions and cutting healthcare expenses.1
In what forms do RPM devices come in?
Monitoring devices include but are not limited to:
- pulse oximeters
- continuous glucose monitor (CGM)
- mobile devices
- blood pressure monitors
- home electrocardiogram (ECG) monitors
- weight scales
The data collected from the monitoring devices are then analysed by healthcare professionals to identify any concerning changes and provide the required early interventions.
Benefits of remote patient monitoring (RPM)
Enhanced patient engagement and compliance
The development of virtual assistants or AI-driven phone applications can aid in prioritisation of the patient’s needs for treatment based on the entered symptoms into the app.1 This can enhance patient access to healthcare services and improve the efficiency of healthcare delivery.1
Personalised treatment plans
AI enables a multi-faceted approach to patient care by considering granular information of individual patients, such as predictive analytics using ML algorithms that identify the risks of developing chronic diseases or hospital readmissions.1 This includes utilizing ML algorithms (process to be followed in calculations) to recommend specific medications or treatment strategies based on the specific genomic makeup of patients.1
Improved accessibility and convenience
Patient monitoring technologies can increase patient health awareness, which is linked to better healthcare outcomes and healthy behaviours.2
Enhanced efficiency of healthcare delivery
The development of virtual assistants or AI-driven phone applications can also help prioritise the patient’s need for treatment based on the entered symptoms into the app.1 This can enhance patient access to healthcare services and improve the efficiency of healthcare delivery.1 It contributes to higher efficiency of communication and collaboration between different healthcare providers by reducing the time required for administrative tasks.2
Case studies and real-world applications
Chronic disease management
The MyCare24 chronic obstructive pulmonary disease (COPD) service based at Airedale NHS Foundation Trust was established to support people with COPD.4 Patients on the scheme use oximeters and send vital sign data regularly to the service team through the mobile app called Luscii or by calling a service team member. The patients are contacted for intervention when the patient’s vital signs are abnormal Recently, the service has been extended to support patients with other conditions such as respiratory conditions and Parkinson's.4
By comparing the data from a sample of 232 patients six months before and after referral for the service, COPD-related hospital days decreased by 63.4% and emergency admissions reduced by 28.8%.4
Virtual healthcare assistants
In North London, the National Health Service (NHS) trialled an AI-driven phone application that helps determine the treatment order of patients based on their symptoms.1 The application encourages its current 1.2 million users to use the AI chatbot to seek healthcare-related information and assistance instead of dialling the NHS non-emergency number. This results in lighter demand on the 111 service. The trial has demonstrated the potential of virtual health assistants in enhancing the quality and efficiency of healthcare services.1
Optimisation and monitoring of drug dosing regimes
A pioneering study at WisDM at the National University of Singapore used an AI neural network called CURATE.AI to continuously adjust drug doses to optimise efficacy at a given time using AI mapping.5 The transformative effects of CURATE.AI on treatment outcomes have been validated by a clinical study for the treatment of metastatic castration-resistant prostate cancer and its prospective usage in the treatment of multiple myeloma and immunosuppression is being studied in clinical trials.5
Elderly care
As part of the Regional Scale Programme delivered by the NHS, 1,400 residents in over 30 Black Country care homes are adopting remote patient monitoring tools to enhance the care of frail and vulnerable residents.6 The use of remote patient monitoring tools at the care homes allows clinical staff to investigate any well-being concerns carers have about the residents and provide early interventions.6 Between March 2022 and December 2022, it was reported that there was a 62.8% reduction in accident and emergency attendances and a 13.7% reduction in emergency in-patient admissions.6
Mental health monitoring
AI-powered digital tools have shown effectiveness in monitoring patient progress and adherence to mental health treatment. With the ease of access, patients can also get support 24/7 thus reducing the demand for in-person appointments and waiting times.1
A case study of a mobile application called Woebot designed for patients suffering from substance use disorders has shown a significant positive impact.7 Woebot is based on cognitive behavioural therapy and AI to deliver customised digital content, such as mood tracking and psychotherapeutic tools.7 Users of Woebot have shown a significant reduction in substance use, cravings, depression and anxiety.7 This provides evidence that AI can be a useful supplement for monitoring mental health conditions.
Challenges and limitations
Data privacy and security concerns
Robust data governance and investigation into healthcare cybersecurity are essential to ensure that patient data is not compromised.1
Possible bias and inaccuracy of AI
AI is trained using a set of data and algorithms, which may have limitations in representing diverse populations or accounting for disease complexity.1 Therefore, it may produce inaccurate and biased results, especially during diagnosis.1
A need for human expertise
Patients require empathy and personalisation that needs to be provided by healthcare professionals, which proves to be especially important in the context of chronic illnesses and end-of-life care. Public perception and trust in the role of AI in healthcare systems can significantly affect the extent to which AI can be integrated into healthcare.1
Integration with existing healthcare systems
A large effort is required to train and secure engagement from healthcare staff. A close collaboration between computer scientists and healthcare providers will be required to implement and maintain the technology.1
Lack of digital literacy
Low levels of digital literacy are a major barrier to participation. However, challenges encountered during the COVID-19 pandemic have encouraged engagement with digital platforms for both healthcare professionals and patients.2
FAQs
What is the difference between telehealth/telemedicine and remote patient monitoring?
Telehealth/telemedicine is a broad term that describes the remote delivery of healthcare services to patients. On the other hand, remote patient monitoring is one of the applications of telehealth. It is the extended care provided outside of healthcare settings by healthcare professionals who monitor and evaluate patient health data using monitoring technologies.8
How often should patients participate in remote patient monitoring?
The amount of monitoring required will depend on the individual patient’s health condition and guidance from their healthcare provider.
Will the patient's health data be kept safe?
In the UK, the storage and usage of patient data are regulated by legal frameworks including the Data Protection Act (DPA) 2018, EU General Data Protection Regulation (GDPR), and the Common Law Duty of Confidentiality (CLDC).9 Healthcare providers and monitoring technology providers are legally obliged to adhere to the legal frameworks.9
Summary
AI has enabled the development of remote patient monitoring, which allows healthcare providers to track and respond to their patients’ biometric data remotely. Remote patient monitoring offers numerous benefits, including enhanced patient engagement and compliance, personalised treatment plans, improved accessibility and convenience, and enhanced efficiency of healthcare delivery. The transformative effects of AI and remote patient monitoring have been validated by case studies and clinical trials. Although the implementation of these technologies into healthcare has taken great strides, it is still a highly debated topic with many limitations, challenges and questions that are yet to be tackled.
References
- Alowais SA, Alghamdi SS, Alsuhebany N, Alqahtani T, Alshaya AI, Almohareb SN, et al. Revolutionizing healthcare: the role of artificial intelligence in clinical practice. BMC Medical Education [Internet]. 2023 Sep 22 [cited 2024 Jun 28];23(1):689. Available from: https://doi.org/10.1186/s12909-023-04698-z
- NHS Transformation Directorate [Internet]. [cited 2024 Jun 28]. Evaluation of the regional scale programme and the national innovation collaborative - executive summary. Available from: https://transform.england.nhs.uk/key-tools-and-info/evaluation-of-the-regional-scale-programme-and-the-national-innovation-collaborative-executive-summary/
- NHS Transformation Directorate [Internet]. [cited 2024 Jun 28]. Supporting care with remote monitoring. Available from: https://transform.england.nhs.uk/covid-19-response/technology-nhs/supporting-the-innovation-collaboratives-to-expand-their-remote-monitoring-plans/
- NHS Transformation Directorate [Internet]. [cited 2024 Jun 28]. Confidence around the clock – how technology-enabled remote monitoring is empowering patients and transforming lives in Yorkshire and beyond. Available from: https://transform.england.nhs.uk/covid-19-response/technology-nhs/remote-monitoring-is-empowering-patients-and-transforming-lives-in-yorkshire-and-beyond/
- Blasiak A, Khong J, Kee T. Curate. Ai: optimizing personalized medicine with artificial intelligence. SLAS Technology [Internet]. 2020 Apr 1 [cited 2024 Jun 28];25(2):95–105. Available from: https://www.sciencedirect.com/science/article/pii/S2472630322010317
- NHS Transformation Directorate [Internet]. [cited 2024 Jun 28]. Protecting care home residents from the impacts of frailty: digitally-enabled innovation in the Black Country. Available from: https://transform.england.nhs.uk/covid-19-response/technology-nhs/protecting-care-home-residents-from-the-impacts-of-frailty/
- Prochaska JJ, Vogel EA, Chieng A, Kendra M, Baiocchi M, Pajarito S, et al. A therapeutic relational agent for reducing problematic substance use (Woebot): development and usability study. Journal of Medical Internet Research [Internet]. 2021 Mar 23 [cited 2024 Jun 28];23(3):e24850. Available from: https://www.jmir.org/2021/3/e24850
- Wang T, Ho MH, Tong MCF, Chow JCH, Voss JG, Lin CC. Effects of patient-reported outcome tracking and health information provision via remote patient monitoring software on patient outcomes in oncology care: a systematic review and meta-analysis. Seminars in Oncology Nursing [Internet]. 2023 Oct 1 [cited 2024 Jun 28];39(5):151473. Available from: https://www.sciencedirect.com/science/article/pii/S0749208123001171
- NHS England Digital [Internet]. [cited 2024 Jun 28]. Protecting patient data. Available from: https://digital.nhs.uk/services/national-data-opt-out/understanding-the-national-data-opt-out/protecting-patient-data