How a Chatbot Helped Adventist HealthCare Connect With Patients During COVID-19

August 14, 2020

// By Kyle Hardner //

Kyle HardnerWhen Adventist HealthCare, a three-hospital system based in Gaithersburg, Maryland deployed a chatbot on its website in late 2018, it did so to connect with two distinct audiences. First, it asked the chatbot to give moms-to-be answers to their pregnancy and childbirth-related questions.

Next, Adventist used the chatbot to provide customer support for the August 2019 move of its 100-year-old Washington Adventist Hospital into a brand-new home, White Oak Medical Center in Silver Spring, Maryland.

Richard Rinaudot, digital marketing director for Adventist HealthCare

Richard Rinaudot, digital marketing director, Adventist HealthCare

But those traditional chatbot applications soon took a back seat to the global pandemic. In January 2020, as emergency departments in Maryland began travel-screening patients for COVID-19, a multidisciplinary team at Adventist began to meet weekly to monitor the pandemic and plan community-related interventions. “With a chatbot already in place on the homepage of our website, we saw an immediate opportunity to leverage it as a knowledge base,” says Richard Rinaudot, digital marketing director for Adventist.

Rinaudot turned to his vendor, Loyal, an Atlanta-based tech firm and creator of the artificial intelligence-powered conversational chatbot called Guide. Working with Adventist and its other 21 hospital and health system clients, Loyal quickly created a solution to successfully assess and triage patients for COVID-19 risk, while also providing answers to healthcare consumers’ evolving series of pandemic-related questions.

Responding to Early Inquiries

As the pandemic started, the team at Loyal began monitoring conversation data from its chatbots. It quickly saw the nature of the discussion changing.

This content is only available to members.

Please log in.

Not a member yet?

Start a free 7-day trial membership to get instant access.

Log in below to access this content: