Kenes Group – Reality Check

About Kenes Group

Kenes Group is one of the world’s leading Professional Conference Organisers (PCO) and the only global PCO to medical and scientific events. It’s hosted over 3,800 conferences in more than 100 cities worldwide. Alongside its extensive conference programme, Kenes offers Continuing Medical Education content, focussed on improving the delivery of medical education.

We worked with Kenes Group on an activity that gives healthcare practitioners (HCPs) a way to practise empathetic conversations with patients who have diabetes. Learners simulate a conversation with a fictional patient, and emotional recognition and generative AI provide feedback on how the learner is communicating with the patient. The focus is on tackling biases and stigma around weight in patients with diabetes.

Helping healthcare practitioners have better conversations

We first met Kenes Group at Learning Technologies in 2023, where we were demonstrating our emotional recognition proof-of-concept. The demo invites users to practise delivering some bad news to a manager. It uses emotional recognition AI to feedback on how you’re coming across, giving you an opportunity to reflect on how you might adjust your delivery for future conversations. It’s an evolution of our work with Lloyds Banking Group on ‘Hear to Listen’, an activity that gives learners a safe space to practise talking to a colleague going through a tough time.

Kenes Group saw the potential for improving patient outcomes by using this kind of learning to help HCPs better handle sensitive conversations with their patients, particularly those living with diabetes and obesity.

Stereotypes about people with obesity lead to stigma and discrimination. When that stigma comes from healthcare practitioners, it can mean that people end up avoiding healthcare altogether. The challenge lay in overcoming biases and stigmas surrounding weight, so that practitioners would approach conversations with sensitivity and without judgment.

A space for practising and reflecting

So, we set out to create a piece of learning that would give HCPs a chance to practise these conversations as many times as they wanted, in a safe space for self-reflection.

We worked closely with Kenes Group to establish a fictional patient and the challenges they’re facing, who would be representative of a typical patient a practitioner might meet. From there, we moved to develop the script that would give practitioners several opportunities to simulate a video call with the patient.

We took time to carefully select the ideal actor to portray the patient for the most impact and alongside Kenes Group we came up with Lucy: a patient looking for advice to manage her recently-diagnosed diabetes, while juggling parenthood and her mental health.

Throughout the conversation, the experience gives learners advice on what to consider in their responses, based on guidance we researched from leading organisations.

The activity analyses both the delivery and the content of what the learners says, as well as how attentive they seem when speaking with Lucy. A key point is that the activity doesn’t offer medical advice, or seek to tell you exactly how you’re feeling – it acts more as a mirror, to help you reflect on how you may come across to a patient.

Our takeaways – intelligence augmented, not artificial intelligence

This is the first project in which we’ve deployed AI, whether generative or emotional recognition – and we were so excited to have the opportunity!

It also proved a valuable experience in finetuning the use of AI for the best results.

• We took time to refine the prompts for the generative AI – when it analyses and feeds back on the content of the learners’ responses, the AI isn’t working from scratch. It’s working with a plenty of references for good practice that we provided, based on our research into what leading organisations and charities recommend.
• To make the feedback even more powerful, the AI picks up on specific quotes from the learner, so that the analysis is personalised and isn’t providing generic advice.
• We had to ensure it was clear how these AI technologies were being used in the experience, as this kind of learning is likely to be new to the learners. For example, the emotional recognition AI tended to refer to fairly scientific or academic terms for emotions. We tweaked the wording so that it would use more day-to-day language for easier understanding.

The nature of the activity means it can be expanded upon, adding new scenarios with new patients, and adding variety to the conversations that HCPs can practise.

For more on how the project came to be, check out our study with Lloyds Banking Group on turning learners into activists, and how that evolved into our emotional recognition proof-of-concept.