How Watson Assistant helped UAMS answer the public’s health questions

By | 2 minute read | December 1, 2020

The University of Arkansas for Medical Sciences (UAMS) is the state’s only health sciences university. Located in Little Rock, Arkansas, its mission is to improve the health, health care and well-being of Arkansans and of others in the region, nation and the world. When the outbreak of COVID-19 spiked in March, concerned citizens overburdened UAMS staff with the surge of inquiries. “We had lots of people calling any number they could find at UAMS trying to get help. We knew what we had in our internal physicians call center was not going to be able to scale,” said Joseph Sanford, MD, Chief Clinical Informatics Officer and Interim Director of the Institute for Digital Health & Innovation, University of Arkansas for Medical Sciences.

Due to information around the COVID-19 virus changing in real time, the medical professionals at UAMS wanted to choose an adaptable AI assistant. Watson Assistant’s ability to train quickly was a critical deciding factor for the team, allowing UAMS to turn an idea into a pilot within a week — and successfully communicate the latest COVID-19 information to users over the web and phone.

As UAMS saw success in their communication efforts with Watson, the team found expanding their use case critical during these times.
According to Kevin W. Sexton, MD, Associate Director, Institute for Digital Health & Innovation, University of Arkansas for Medical Sciences, “One of the benefits is looking at the responses and data that Watson collects. You’re able to see where the population has questions that you may not be answering directly on your website.” As users engaged with Watson and UAMS could see the conversation topics come in, medical professionals uncovered a considerable concern in the community among pregnant women and COVID-19. After analyzing chat log data, doctors trained Watson Assistant specifically on pregnancy questions and screenings for pregnant women.

Also, after one month of going live, the team saw an increase in COVID-19 tests and an influx of questions regarding patients’ results. With Watson Assistant, UAMS quickly created a workflow to provide COVID-19 lab results to individuals. Now, a small team analyzes the data weekly to ensure no patient question gets missed.

UAMS will continue providing reliable information to users across multiple channels, in both English and Spanish.
UAMS has found that Arkansans depend on Watson Assistant, and about 20% of interactions are from repeat visitors! With success like this, the team recognized a need to support Spanish-speaking communities, so UAMS quickly trained Watson Assistant in Spanish.

And it doesn’t stop with the patients; since UAMS is a university, the school also has to train Watson for student questions. UAMS has nursing, medical, pharmacy, and health professions students, all of whom want to help patients — so Watson now can provide information on how students can interact with patients safely.