On 28-29 March 2023, the second ever Data Saves Lives Ambassador Training Bootcamp was held on the top of Artificial Intelligence (AI) in Lisbon, Portugal. The purpose was to equip regional and national level organisations with a better understanding of the opportunities and challenges presented by the use of AI in healthcare and the potential role of patient groups in influencing the design and use of AI to improve the lives of their members. This blog gives an insight into the hot topics that were discussed as well as a summary of the key insights.
Over the two days, the group explored the basic principles, terminology, governance, and legislation of AI in addition to discussing working examples through interactive workshops - all with guidance from AI and communications experts.
To set the scene on day one, expert trainers, Stefan Phillips from OneVision and Tanushree Tunstall from EATRIS, put the group through their paces with an overview of the fundamentals of AI – a whistle-stop tour of some of the key terminology that we would use across the two days. Afterwards, the group went on to learn about the benefits and limitations of AI in healthcare, touching on hot topics such as data bias, accountability, trust and data privacy. This theory-led session was then brought to life through real-life examples, which demonstrated how AI can and is already improving outcomes and quality of life for patients in conditions such as stroke and brain haemorrhages, chronic disease and cancer.
A virtual panel discussion with experts, Dr Elena Sügis, Assistant Professor of Health Informatics, University of Tortu for Computer Science; Dr Jovan Stevovic, CEO, Chino.io; and Dima Belich, CEO, FountainHead IT, brought the AI developer perspective to the discussion. The conversation focused on the different considerations for developing AI and gave patient group organisations a chance to ask their burning questions! We will be sharing the live recording of this discussion on our YouTube channel very soon, stay tuned!
Learnings from the morning were applied in the afternoon of day one, through interactive workshops centring around the question: how AI can be leveraged by patient communities? All attendees had the opportunity to work through real-life scenarios and learning from each other’s experiences in small group settings. Key topics emerged such as the need for more information on the AI already being used, so that patient groups can help identify true gaps to support co-creation; encouraging and supporting awareness of AI solutions that increase inclusivity (e.g. voice activated tools); how to ensure inclusivity of data to collected to make sure AI solutions truly help diverse patient populations; and ways of addressing misinformation that causes public distrust and resistance to AI.
Some of the common questions throughout the workshops included:
How do we motivate developers to see patient organisations as potential partners from the very beginning of the idea, not just dissemination?
How does protecting personal data rights limit the potential for community co-design
How can we manage consent?
As a group, it was agreed that capacity building within the patient group community, transparency, a strong governance process and accountability are key to helping to address such questions.
On the second day, Milana Trucl (Digital Health Policy Officer, EPF) presented the EPFs’ work on AI to date – including the patient organisations and patient advocates survey that is currently live, which will help to inform the EPFs’ policy paper on AI and subsequent recommendations for the community. The group then went on to discuss our main concerns and worries about AI through a series of interactive poll questions. The top three areas the group was most concerned about were:
Biased and nonrepresentative data used to train AI
Lack of involvement of patients in design of AI solutions
Patients not having equal access to AI-supported systems
The Data Saves Lives toolkit was presented – a fantastic resource to equip patient groups and health influencers with the information and materials they need to have a positive dialogue with their communities about health data, and to potentially launch their own initiatives. Ambassadors were asked to contribute ideas for resources and materials to support patient groups navigate working within the AI space. Armed with fresh ideas and insights, the Data Saves Lives team look forward to developing the Data Saves Lives Toolkit 2.0 later this year!
The Bootcamp training was rounded off with an ‘AI Hospital’ session – an interactive board game that explores the patient experience from diagnosis to management and how AI can be implemented along this journey. Everyone took value out of the scenario based game allowing people to think about the practicalities of using AI at each stage of the patient-healthcare professional (HCP) journey, considering a range of factors such as cost and patient/HCP preferences.
The second Data Saves Lives Ambassador Training Bootcamp was an inspiring two-day event, jam-packed with rich discussion, learning and ideation! All came away feeling inspired by the trainers and fellow 20 patient group ambassadors from across Europe, who are dedicated to promoting and fully harnessing the power of AI in their own communities, to ultimately make a difference to patients. The Data Saves Lives team plans to be back with more events similar to this, as well as tweet chats and webinars, so stay tuned by following us on Twitter and LinkedIn.
Here are some of the key takeaway points from across the two days:
AI in healthcare is not about replacing doctors, it is about helping doctors to practice better, more accurate and efficient medicine!
AI is a tool that we need to understand, respect and learn to use properly: because ultimately medical intervention is a joint responsibility between clinicians, the health care provider, and most importantly, the patients and society at large!
End users (i.e. patients) need to be at the heart of developing AI solutions! Capacity building and knowledge sharing (like this bootcamp) is vital to support patient groups to ‘take a seat at the table’ when it comes to discussions around co-designing AI solutions
Effective AI solutions can only be built with strong and diverse data sets! We need to build trust and awareness around health data sharing to help find better solutions
AI implementation is moving at lightning speed but there are dedicated regulations and frameworks in place for AI- geared toward building a more transparent, trustworthy, and explainable AI
Don’t fear that you are not an AI expert (most people aren’t!) – ask for clarity and credentials if you’re unsure!