Challenge

Artificial intelligence (AI) has the potential to make a significant difference in health and care. We’re already seeing great examples of impactful AI, but more needs to be done to support the use of these technologies safely and ethically across health and care sectors.

The NHS AI Lab was set up to accelerate the safe and effective adoption of AI in health and care. Its mission is to create a sustainable health and care system which achieves better outcomes, equality and fairness for all. By supporting AI technologies that have the potential to improve the quality of health and care services, while building a robust ethical and regulatory framework to ensure patient and citizen safety. FutureGov was brought on as strategic partners in February 2020 to provide advice and human-centred design support to NHSX on the work for the lab, as well as the AI in Health and Care Award.

COVID-19 changed the operating environment and focus of the NHS AI Lab, as government priorities changed and NHSX became responsible for various elements of the UK response. We supported NHSX and the Accelerated Access Collaborative (AAC) adjust to changing demands and focus the work of the lab, so it could be most useful for a health system under pressure.

Approach

For the NHS AI Lab and Award, FutureGov provided strategic advice and support around user-centred design and research, as well as service, organisation, content and product design. Human+ provided expert advice on machine learning and implementation of RPA at the trust level.

AI Award

The AI in Health and Care Award benefits patients by testing and scaling the most promising AI technologies that can improve health and care outcomes.

Selection

We were involved in executing the round 1 process for accepting, reviewing and judging applications for phase 4 of the award, through to final selection. We worked alongside AAC staff and LGC to map phases of the award, as we would a service, identifying interaction points, needs and actions required at each stage.

Our involvement continued into producing guidance for assessors and designing a user-centred longlisting process around the COVID-19 pandemic. Alongside these activities, we co-coordinated multiple due diligence processes and a user-centred approach for the final selection panel.

We identified and recommended improvements to the award process that could be enacted for future rounds. Our designs aimed to make its running smoother for staff and applicants and make sure the Award is thoughtfully joined with the NHS AI Lab and other activities.

Evaluation

Working with NHSX and the AAC, we designed a process for evaluating winners of the phase 4 AI Award as the technologies are implemented in the real world. The evaluations would then support informed decision making about scaling winning technologies across the NHS.

We designed the overall governance structure for evaluation, along with the in-depth processes to follow and the capabilities required. An Expert Task Force advised on this work, supporting the development and review of processes and deliverables. These included the AI Evaluation Playbook (a guide to good practice for new teams conducting the evaluation of AI products). As part of this, we supported parallel site selection and regulatory process mapping, as well as the COVID-19 Fast Track.

COVID-19 fast-track

Early in the COVID-19 pandemic, the NHSX AI team wanted to ensure its work had a positive impact for the unfolding response. We worked with other strategic partners to set up a streamlined process for technologies that could help during the pandemic. An end-to-end process was designed, including detailed assessment criteria at each stage. This helped to quickly bring stakeholders together to consider ways to streamline the regulatory process, and further informed our work on the evaluation process for the AI Award.

Change management

This piece of work was originally envisioned as developing tools and guidance to support the implementation and adoption of AI within hospitals and local care providers. We carried out a programme of user research to understand the needs of the organisations, and capture existing best practice. This resulted in a picture of what guidance and tools people would find useful, and fed directly into the development of the resources section of the website. This was used in structuring information around use cases and jobs to be done, and will inform future content for the Lab’s interactions with local providers.

NHS AI Lab website

We created a useful website for the NHS AI Lab, based on existing content, our user research, stakeholder engagement work and the existing structure of the NHSX website. This site would then be used as a template for future expansions of the NHSX site. We used a blended team of designers and developers from FutureGov and DXW, who maintain the main NHSX website, to make sure continuity and integration of the site continues.

RPA/Skunkworks

Human+ was also commissioned by the NHS AI Lab, to undertake a discovery to determine the suitability of Intelligent Automation (IA) within NHS hospitals and Trusts across England. Automating the daily COVID-19 situational report (a requirement that covers multiple departments) aimed to save time and reduce human error.

Impact

We worked with the NHSX AI team and the AAC to understand their needs before and throughout the COVID-19 response, and into recovery. Helping make sense of the changing landscape, adapting to a different focus and authorising environment for the health sector. We provided them with organisation design options and guidance on capability gaps, while building the team’s capability in user-centred research, design and agile methods of delivery.

AI Award

Our people-centred approach equipped the AAC team with the tools and processes needed to run a second round of the AI in Health and Care Award that is more smooth, efficient and strategic.

We also provided the AAC team with governance, processes and job descriptions, as well as a procurement approach so that they could quickly and effectively set up the teams and structures needed to effectively evaluate award recipients. We also worked with the expert panel to create an Evaluation Playbook designed to provide a baseline for leading practice in evaluating AI technologies in the health system. This playbook will be used in evaluations for Round 1 of the award, adapted based on lessons from its application, and subsequently published. View the AI in Health and Care Award winners.

NHS AI Lab website

Our user research helped focus the resources and content being produced by the Lab on what its stakeholders need, rather than taking a program-focused approach. We built the NHS AI Lab website within six weeks that aligns with the AI Lab’s strategy for the coming years and will be easy for the NHSX team to build-on and maintain.

RPA/Skunkworks

The estimated staff time saving for the hospital once the full automation project is complete will be 4400 hours saved across all processes per annum and a staffing cost equivalent savings of £153,846.00 over one year.