Medibot
MediBot is a web-based training application designed for medical students and experts. Its objective is to enhance their proficiency in clinical reasoning. The research project is a collaboration with Unisanté Lausanne, University of Basel and Swiss TPH with the objective of developing a chatbot that can be utilised in medical education and that addresses the issue of gender bias in healthcare.
Discipline
Figma
Webflow
Hotjar
Usertesting.com
Role
Visual Design
UX Researcher
Design System
01
Problem Statement
Prejudice can arise even in healthcare workers, leading to discriminatory practices, racism, or an underdiagnosis of patients. The lack of human resources to educate medical students and a safe space to practise with patients has been a persistent issue for some time.
02
Research
A chatbot that takes on the role of the patient was conceptualised to address the identified gap. To validate the concept, we designed a prototype and conducted a moderated Wizard of Oz test with medical students, simulating real-world interactions to gather feedback. We found out that users preferred a conversational tone to engage, so we adapted the tone to include empathetic conversational responses.
03
Solution
Students can practise their clinical reasoning skills by interacting with a trained chatbot that takes on the role of the patient in various case studies and based on the results of the conversation, an evaluation is generated that provides suggestions for improvement and best practices.
04
Result
The study has demonstrated an increase in awareness of the consequences of biased diagnosis for patients from diverse racial backgrounds. The participants have been encouraged to consider the potential for prejudice in diagnostic questions and have expressed satisfaction with the opportunity to develop their soft skills through this training approach. The study is still on going and haven't been published yet.