How can AI help the workforce?

With an overburdened and depleted workforce on top of a growing elective backlog following the pandemic, it is essential that the NHS looks to support the workforce and its patients using digital solutions. Christina Sothinathan, NHS Navigator at DigitalHealth.London discusses how Artificial Intelligence (AI) could be used to support NHS staff and patients. Drawing on her own experiences as a Clinician, she reflects on how Limbic, a company on the Accelerator programme, are specifically tackling this issue across mental health pathways.

Limbic is a digital AI solution addressing key touch points across the entire mental health pathway. Limbic Access (a class 1 medical device), allows patients to self-refer to the NHS, Improving Access to Psychological Therapies (IAPT) services. It employs machine learning techniques, such as Natural Language Processing (NLP), to help patients record their data with an experience similar to speaking to a non-judgemental human. The chatbot uses conversational AI to engage users more as it has a more caring and humanistic approach compared to a static online form. This improves useability and streamlines the referral pathway, as reflected by Limbic’s high completion rates of 91%.

The tool also collects clinical screening information such as the PHQ-9 (a depression screening questionnaire) to stratify referrals so patients can be put on the most appropriate treatment pathways faster. Reflecting on my time as a Clinician, we introduced digital pre-appointment screening using PHQ-9 and musculoskeletal (MSK) outcome measures. This made a huge difference as it allowed me to spend the time saved in screening by adding more value to the patient’s appointment time.  

Saving time and removing the burden of paperwork

Limbic also identifies patients who won’t benefit from IAPT due to presenting with more complex mental and physical health issues. Referrals highlighting suicidal risk and other complexities are filtered by Psychological Wellbeing Practitioners (PWPs) in the Clinical Referral Team (CRT) who specifically manage referrals which come from Limbic Access. If a risk is flagged, then the CRT makes contact with the patient and can help them get more appropriate support straight away. Limbic Access helps to ensure that referrals booked in for IAPT support are suitable for this service – which reduces wasted time for clinicians and patients utilising appointments in the incorrect service. It also improves access across IAPT and helps each patient to find the right support options quickly, while reducing burden on service staff.

It can be incredibly challenging when clinicians are faced with highly complex patients with potential suicidal risk during routine IAPT appointments. Front loading the mandatory data sets and screening questions to the referral, can free up clinical time during the assessment to allow risk to be addressed appropriately. Furthermore, for IAPT-appropriate patients, removing the burden of completing mandatory screening and questions within a 45-minute assessment, frees up clinical time for PWPs and supports them in their clinical decision making. From having this information available pre-appointment, PWPs can focus on the quality of assessment and care. PWPs are also able to prepare their assessment based on questions asked by Limbic Access and clarify unclear areas in order to optimise their assessment, add more value-based care and enhance patient experience.

In addition to the support of clinical assessments, Limbic Access seeks to make admin and booking processes smoother and quicker. Through integrating with appointment systems, patients can be automatically booked into open slots – thus improving efficiency and reducing administrative burden. From personal experience, having underutilised appointment slots can be extremely frustrating as unfilled slots ultimately leads to longer waiting times for those in need of treatment. When patients wait longer for care, their treatment outcomes are often poorer.

Improving patient engagement and experience

Limbic Care uses the same AI engine but for a treatment support tool across the full patient pathway. Patients can use the tool from when they are on the waiting list, all the way through to after discharge, and can be used 24/7 by the patient. For example, the user can log on stating “I’m feeling anxious” and the chatbot can access the treatments and recommendations from their Clinician and reply with “your therapist has recommended breathing exercises, let’s try that now.” More generally, the app is used as an interface between patients and clinicians for example to administer CBT exercises and homework. Following discharge, users can access Limbic care to help reduce relapse risk and therefore reduce burden to already overstretched mental health services. Strategies such as these, accessed 24 hours daily can help to create distance from unhelpful thoughts which can reduce the overall impact they have on emotions [1].

As a Clinician, it is extremely rewarding when patients have access to high quality resources in between treatment sessions to help encourage session success carry over. Easily accessible solutions for patients 24 hours daily offer invaluable support as we know problem solving abilities can be inhibited by emotional distress [2].

AI does not replace clinicians, it supports clinicians

AI is not here to replace humans as we know it lacks human values such as empathy and gut instinct [3]. Additionally, NLP is also limited in terms of its ability to pick up on sarcasm and such intrinsic human characteristics [4].

AI has huge benefits in terms of user experience, as it can be accessed 24 hours daily and 7 days per week. Humans on the other hand, need to sleep and have rest days from work (and rightfully so).

Limbic’s user feedback has shown people feel less judged speaking to a chatbot and it’s less burdensome than speaking with a human which makes it easier to access, particularly for those with social anxiety.

Furthermore, by supporting clinicians through streamlining processes, this reduces admin burden, which we know is a huge challenge in the NHS. From personal experience, I know how useful it is for Clinicians to have relevant information prior to an appointment as it can allow for appropriate preparation, whilst saving time and reducing stress.

Ultimately, understanding what AI can do well and where its limitations sit, is key to finding where it can fit best across NHS services to improve patient care and support the workforce.

References

1. Barrett, L. F. (2018). How emotions are made. Pan Books.
2. Smith, J. (2022). Why has nobody told me this before. Penguin Books Ltd.
3. Christian, B. (2020). The Alignment Problem: How Can Artificial Intelligence Learn Human Values. Atlantic Books London.
4. Mitchell, M. (2019). Artificial Intelligence: A Guide for Thinking Humans. Pelican Books.

DigitalHealth.London is delighted to publish blogs by the NHS staff and digital health companies we support through our programmes, as well as sector thought-leaders, experts and academics. Any opinions expressed within blogs published on our website are those of the author and not necessarily held by DigitalHealth.London. For more information, or if you would like to write a blog for our website, please email info@digitalHealth.london.

Limbic is part of the sixth cohort of the DigitalHealth.London Accelerator programme. The DigitalHealth.London Accelerator is a collaborative programme funded by London’s three Academic Health Science Networks – UCL Partners, Imperial College Health Partners, and the Health Innovation Network, MedCity, CW+ and receives match funding from the European Regional Development Fund.