In the study, 23 patients with a median age of 15, all of whom were being treated for weight management and prediabetes symptoms at Nemours Children’s Health System in Orlando, communicated with an artificial intelligence-powered chatbot named Tess via methods such as text message and Facebook Messenger. Tess offered individualized guidance about the treatment program, behavioral changes and overall wellness.
As a result, the adolescent patients reported that Tess helped them make positive progress toward their goals 81 percent of the time and was useful 96 percent of the time. Those reactions, combined with the sheer number of messages the group exchanged with Tess over the course of 10-12 weeks, demonstrate the bot’s ability to engage with and help patients.
Additionally, the chatbot offers solutions to other frequent barriers to care. Since Tess was available 24/7, patients were able to consult with the bot outside of regular office hours. The virtual solution also cut down on the cost of traveling to and from treatment, and saved the hospital an estimated $8,933 in staff hours. Finally, researchers suggest that the anonymous nature of the conversations helped neutralize the stigma often associated with seeking behavioral support.
More articles about AI:
Microsoft announces 7 new AI for Accessibility grant recipients
Machine learning detects high-risk lung nodules in radiology reports
Optum funnels $16M into AI startup Health[at]Scale
