Laura Roy, Tampa General Hospital

"This past fiscal year, our neuro ICU had zero catheter-related infections.

If you work in healthcare, you know that's almost unheard of—especially in a unit that uses catheters as frequently as ours.

Twenty years ago, ICUs were seeing those infection numbers in the double digits. That was normal. Expected, even.

So how did we get to zero?

It wasn't one big change. It was a thousand small ones—made possible because we finally had the right information at the right time.

Every morning, I pull up a screen that shows me every patient in our 32-bed unit. I can see instantly who has a Foley catheter, who has a central line, who's on a ventilator. More importantly, I can see why they have it, and how long it's been there.

That second part is critical. Because a catheter that's medically necessary on Day One might be an unnecessary infection risk by Day Three. But if you don't have that information easily accessible, if you're buried in manual chart reviews and scattered documentation, you can't have those conversations in real time.

Before we had this tool, getting that information took hours. I'd be trying to organize it, sift through notes, pull details from different systems. By the time I finished, I was already behind.

Now it takes minutes. And that means multiple times a day, we're asking: Does this patient still need this device? Is there still a valid indication? Are we creating risk where we don't need to?

Those conversations happen because we have the data. And we have the time.

That's the thing people miss when they talk about AI in healthcare. They focus on what it might replace. They miss what it enables.

It's not just infections. Our length of stay is down. Our quality metrics are up. We're catching things earlier—before they become problems instead of after.

I can see whether a patient is stable enough to leave the ICU before the orders are even written. That matters because hospital capacity is one of the most critical challenges we face right now. If someone's mom comes through the emergency room with a stroke, I want that ICU bed available. I want her to get the care she needs immediately, not wait because we didn't move someone out who was ready to step down.

The AI isn't making those clinical decisions—my team is. But the tool is giving us the information we need to make them well, and quickly.

And it gives me the space to actually be the leader I'm supposed to be instead of drowning in administrative tasks. I know my nurses now. Not just their names—I know their lives. Healthcare is hard. The stress is real—at home and at work. If I can reduce even a piece of that stress, if I can give someone a reason to feel supported and valued, that changes whether they stay in this field.

And we need people to stay. We need good nurses to become exceptional nurses. We need people who went into healthcare because they wanted to help to actually have the time and space to do that.

I've been at Tampa General for 18 years. I've seen a lot of changes. But I've never seen anything shift the culture of a unit the way this tool has.

My team used to joke—darkly—about walking around with pens and paper, scribbling notes on their legs during rounds. Now some of them say they can't even remember what it was like before. That's not because the technology is flashy. It's because it works.

We're piloting other innovations now too—a pupilometer app that's the first of its kind in the U.S. My nurses are giving feedback, and the creator is flying in to make adjustments based on what they need. That's the kind of collaboration that happens when you're not drowning.

I keep thinking about the next generation of nurses. What if they started their careers with these tools? What if they never had to learn the old way—the way that burned people out before they even hit their stride?

My hope is that they come into a system where they can focus on care from day one. Where they have time to sit at the bedside and talk to a patient. Where they can look someone in the eye instead of staring at a screen.

Because that's what this is really about. AI didn't fix healthcare—healthcare is still a challenging, often broken system. But it gave us one critical piece back: the ability to deliver on our promises. To work at the top of our license instead of drowning at the bottom of someone else's paperwork.

My hope is that every nurse, in every hospital, gets access to tools like this. Not to replace us—but to let us do what we signed up to do in the first place."