Most chatbots are custodians of information, designed to get the user to the answer they seek. The in-development app PatientX flips that process – the user is the one with the knowledge, and they ask questions of the chatbot in order to arrive at a conclusion. The aim is to provide medical students with a simulation of the diagnostic process and help them develop patient-friendly ways of articulating medical concepts.
PatientX is part of a global incubation program run by Elsevier that builds on the Elsevier Hacks 2017 event in Helsinki. The original event paired medical students with technologists and tasked them to solve some of the biggest challenges in medical education.
Emerging from that event, the team behind PatientX is one of three working together across the globe. Over a 12-week period, they will be coached by experts from Elsevier and other businesses in design, development, communication and business, taking the ideas and prototypes created at the hackathon and building them into fully matured business plans and MVPs (minimum viable products), ready to release in the market.
This week, the teams are gathering in the UK to present their prototypes to a global panel of Elsevier senior managers on July 19. One team will be awarded an additional $10,000 from Elsevier to continue to develop their prototype and grow their startup.
What is PatientX?
PatientX is a realistic and safe simulation of a patient case that allows students to practice clinical reasoning and receive feedback. Utilising chatbot technology, this app will enable medical students to improve their diagnostic process at an early stage.
What the mentors say
Tim Morris, Product and Partnerships Director at Elsevier, commented: “What the team is focussing on here is giving the experience of transitioning from understanding the science and technology associated with being a doctor, to meeting a real patient and understanding what’s happening in front of you. That’s an important issue because, for the most part, patients speak in layperson terms, whereas medical students learn the medical language. It’s an ambitious project because patients don’t present in the same way every time.
“What we’ve encouraged them to focus on is picking a few clinical presentations that allows the student to explore the problem and ask questions in a way that’s comprehensible. The chatbot needs to be able to respond like a human and give answers that guide the student to a diagnosis. I’ve been working with chatbots in patient-facing decision support, where patients do testing on themselves, and one of the key elements is understanding how patients question the chatbot systems they use. PatientX kind of reverses that process, and the team is thinking about how to create a bot that they need to question and that will simulate gaps in the patient’s knowledge and understanding.”
What the hackers say
Eliza van Wulfften Palthe, who was a medical student at the Utrecht University in the Netherlands and Chief Marketing Officer of PatientX, commented in the team’s weekly updates: “The problem we’re tackling is that a large majority of students witness the essential decision-making process needed to help patients, but few are given the autonomy to make their own decisions in terms of ordering investigations and deciding on a diagnosis. When they graduate and are working as a doctor, they’re expected to gain those skills overnight. So, we meet their needs for a safe environment to practice and gain confidence in the decision-making process.
Alice Leung, a medical student at Western Sydney University in Australia and PatientX Chief Operating Officer, added: “Working on this project has been out of this world, and I don’t think many medical students get the opportunity to venture out into the world of technology and startups. It’s been such a privilege to be able to share our work an ideas with experts in this field of medical education. Elsevier’s mentors have been so approachable and helpful and have really helped shape our team’s vision for our project. We’ve learnt what creating an MVP really means and how making sure it does one single function well is the most important part.”