As millions of individuals turn to their mobile devices to speak to chatbot-based health apps, researchers have shown that existing apps lack the functions to support a typical visit to a medical facility’s full diagnostic process.
Just five processes of an actual evaluation can be assisted by these chatbot-based symptom checker (CSC) applications: creating a medical history, assessing symptoms, offering an initial diagnosis, ordering more diagnostic tests, and providing referrals or other follow-up treatments, researchers from Pennsylvania State University study.
It is being said that these apps do not support conducting physical exams, providing a final diagnosis, and performing and analyzing test results, because these three processes are difficult to realize using mobile apps.
In the report, through a feature analysis, the researchers examined the functionalities of common CSC apps, then analyzed user interactions by reviewing user feedback and conducting user interviews. The team also noticed that users consider CSC apps to lack support for a detailed medical history, versatile feedback of symptoms, comprehensible questions, and diverse diseases and user groups through their user experience review.
Functional and conversational design updates for health care chatbots should be guided by the results, such as enhancing the features that enable users to input their symptoms or use understandable language to provide explanations during conversations.
Especially in health and medicine, [another question is] is there something else we should consider in the chatbot design, such as “how should we let users describe their symptoms when interacting with the chatbot?”
In addition, the results may help people understand the influence of AI technology, such as how conventional medical visits could be affected or modified by AI.
People usually trusted physicians in the past.
But now, people have more sources of knowledge with the advent of AI symptom checkers and the internet. “How would this information challenge doctors? Do people trust this information and why?”
This work is a starting point to think about the influence of AI symptom checkers.