Scott Arnold, the hospital's Chief Digital and Innovation Officer, says Aimee is improving call center operations, and taking the heat out of spicy interactions.
For many health systems and hospitals the call center is the literal front line to healthcare access. Too often it acts as a barrier, with callers either hanging up, waiting too long to talk to someone or failing to be transferred to the right person.
At Tampa General Hospital, leadership is hoping Aimee can solve those problems.
Aimee is an AI tool developed by Hyro and launched last year at TGH to address call center operations. It’s Aimee’s job to answer phone calls that the call center’s human operators can’t get to, and to quickly and efficiently direct the caller to the right resource.
“These are people in our community [who are] trying to get access and haven’t been able to do that,” says Scott Arnold, the hospital’s EVP and Chief Digital and Innovation Officer, who estimates that thousands of phone calls to TGH went unanswered or were directed to the wrong location each month prior to this program. “We have a strategic imperative to improve access to care, and this is where we’re struggling to meet that goal.”
“We're not different than any other health system [in that] we struggle with finding people to answer the phones, work the call centers and do … complicated care coordination over the phone through a call center and do it consistently,” he adds.
The platform produced positive results almost immediately. According to TGH officials, Aimee not only reduced average call times by 58% and cut daily abandoned calls by 56% within weeks of deployment, but it also helped the health system increase scheduled appointments by 21%.
Learning the Lingo
Arnold says the key to making this particular AI implementation work is making sure every call to the call center is answered quickly. And that means preparing Aimee for a wide range of callers, from those asking to see a doctor to simple requests for directions.
Since the call center is often the first point of contact for angry or frustrated patients, Arnold said it quickly became apparent that Aimee should be trained to identify that anger. So the bot was brought up to date on a wide variety of expletives and curses.
“I think we underestimated how distrustful and angry [people can be],” he says. Sometimes “the expletives start flying” when Aimee announces herself as a digital host.
“Some people distrust technology – they don’t see any value in it at all and just want to get to a human,” Arnold adds. “That’s why we pivoted pretty quickly and taught Aimee that when she hears curse words or anything bad to quickly get it over to … a human on the line. That usually brings the temperature down.”
Arnold says Aimee not only acts as a buffer for angry patients, but also alerts staff to be better prepared for those emotional callers. This, in turn, reduces friction in care management and scheduling, and possibly even reduces any anger or frustration directed toward clinicians.
Giving Aimee an Identity
Many AI integrations struggle out of the gate over concerns that the technology will replace people. Arnold says that wasn’t the case in the call center, where hospitals typically have problems finding enough people. Aimee is a welcome addition, he says, handling the workload of three call center staffers.
“Those are probably three people that we can’t seem to hire because we can’t find qualified people,” he says.
In fact, call center staff embraced Aimee quite warmly. They gave her a cubicle, created a cardboard cutout of her and dressed her up for holidays.
“We just gave her a name; they took it a step further,” Arnold says.
Aimee might have an identity now, but she’s also teaching TGH about what healthcare really means. Arnold says it’s ironic that an AI bot is helping the hospital eliminate the barriers to care that technology often creates.
“We went into this to resolve an issue we had with one of our strategic imperatives, which was care coordination and access,” he notes. “And where we were failing the most was just some of the simplest things like answering the phone.”
Eric Wicklund is the Associate Content Manager and Senior Editor for Innovation and Technology at HealthLeaders.
KEY TAKEAWAYS
Many hospital call centers struggle to keep up with traffic, losing hundreds to thousands of calls each month when callers hang up or are directed to the wrong department.
AI bots can supplement call center staff to manage excess traffic, quickly redirecting callers to the right resources and handling basic queries like directions.
At Tampa General Hospital, an AI agent named Aimee was also trained to recognize vulgar language and help guide angry or frustrated callers to someone who could calm them down.