Using AI Chatbot to Support Real-time Assistance for Patients in Hospitals

What is it about?

Patient experience (PX) has emerged as a vital component of healthcare quality, linked to clinical outcomes and hospital reputation. Traditional methods via surveys have limitations including low response rates, biased samples, recall errors, and ineffective timing. Chatbots could be used to engage patients in natural dialogues and has shown the potential as a tool for collecting PX data. We co-designed an AI chatbot to improve patient experience in hospitals for patients' real-time assistance.

Featured Image

Photo by National Cancer Institute on Unsplash

Photo by National Cancer Institute on Unsplash

Why is it important?

They used the human centered design cycle as a basis for the research and design. In the 1st iterative design cycle, they first interviewed a hospital patient experience team and gathered design feedback. This study came up with the initial concept that is to design a web-or app-based chatbot to encourage users to share their PX in the hospital or emergency department as the things were occurring. In the 2nd iterative design cycle, they mined patient reviews, performed a qualitative analysis, and developed 3 AI techniques to code natural language from patients. Finally, based on the insights from these two design cycles, they developed a patient-facing AI chatbot through a pre-trained LLM, which recognizes complaints and can respond in natural language.

Perspectives