Bill Klaproth (Host): This is Today in Nursing Leadership, a podcast from the American Organization for Nursing Leadership. I'm Bill Klaproth. And with me is Connie Clemmons-Brown, System Senior Vice President, Patient Care Services, and Lauren Bulin System Vice President, Clinical Excellence and Nursing Operations, both with CommonSpirit Health as we discuss Unveiling The Unseen: AI for Human Trafficking Detection in Healthcare, where we dive into this critical issue and explore innovative solutions. Lauren and Connie, welcome.

Lauren A. Bulin, DNP, MSN, MBA, RN, CNML: Thank you, Bill. It's great to be here.

Connie A. Clemmons-Brown, DNP, MBA, RN, CENP, NEA-BC: Thanks, Bill. Thanks for having us.

Host: Absolutely. Lauren and Connie, thank you. Lauren, I'm going to start with you. So, what strategies were effectively employed to cultivate staff engagement and secure participation in this AI pilot program?

Lauren A. Bulin, DNP, MSN, MBA, RN, CNML: User-centric design is critically important in the AI detection model, and that's because we really want to engage the clinicians and seeing it through their eyes and helping them to have a tool to enhance detection of human trafficking red flags.

I think the other piece that is critical is really having the support of an intercollaborative team, a team composed of IT, data science, social services, and nursing as pivotal to help us see this work from multiple perspectives, and also to understand that not one of us alone could do this work, but we required the whole team to come together so that we can create something really unique and that addresses the needs of human trafficking.

Host: Well, this is a big topic. And it's really interesting how you're looking to solve the problem. So, Connie, how were ethical principles integrated and upheld throughout the design and development phases of the AI human trafficking detection model?

Connie A. Clemmons-Brown, DNP, MBA, RN, CENP, NEA-BC: Well, I think that's an excellent question and it's a topic that we don't often think about. When we look at vulnerable populations like human trafficking victims, the ability to protect their autonomy, their dignity, their privacy, especially because this information can be very sensitive, we built the model based on some non-negotiables. And those non-negotiables included taking a very deliberate approach to ethics integration across the lifecycle of the AI. What that means then is that we focused on privacy, data security. We placed emphasis on governance to make sure that we had the right stakeholders at the table to bring that lens, so that we didn't lose sight of really what was important.

We also felt that minimizing bias was important, that mitigation, because we didn't want to overrecognize victims and we didn't want to underrecognize victims. And so, we were very deliberate in that respect.

And then, lastly, a human in the loop. We can get caught up in automation. But really, this is a very personal experience for these victims. And we wanted to make sure that our clinicians were actively involved, as Lauren said, in the design and then in the implementation and ongoing evaluation.

Host: Again, this is very interesting. I like how you say a human in the loop. That's so important. Can you just really briefly explain to us how the eye helps detect human trafficking?

Connie A. Clemmons-Brown, DNP, MBA, RN, CENP, NEA-BC: Sure. So, when you build a detection model, the algorithms have to be built on reliable and trustworthy data. So, we were very deliberate in the data that we used to train the model. And then, . we also used documented cases, right, known cases of human trafficking. So, we didn't introduce any additional bias, especially systemic bias that may lead us down a path that would unfortunately retraumatize a victim.

Host: Okay. This is really interesting. So Connie, what were the most significant challenges and barriers encountered during the integration of the AI model into the existing healthcare environment? That had to be an interesting integration. And how were those addressed?

Connie A. Clemmons-Brown, DNP, MBA, RN, CENP, NEA-BC: Sure. Well, there were many challenges, if I was being honest. The first one I touched on earlier is around the data integrity challenge. Like, how do you curate the appropriate training material, the training data and evidence so that we didn't introduce unnecessary bias?

The second was that we had to try to integrate this very advanced technology into some legacy electronic health records. And how did we do that without creating additional noise for our clinicians and without adding more cognitive burden or disruption to their workflows.

And so, as Lauren said earlier, we included key stakeholders in the design of the model in the first place. So, they gave us real-world perspective. And they helped us problem solve through many of those challenges. So, we didn't have to face them later on.

And then, lastly, it's about addressing clinician skepticism and trust, right? People are afraid because they don't understand AI. Nurses are afraid they're going to make a mistake by overreliance on AI. And so, by including them on the front end, they understood why a notification would come forward indicating that this victim may be involved in human trafficking. So, they understood the why and they then developed trust as a result.

Host: So, not only did you have to integrate this new technology into older legacy systems as far as people, you had to explain to them the value of AI and overcome maybe their negative perceptions of AI or what AI can do.

Connie A. Clemmons-Brown, DNP, MBA, RN, CENP, NEA-BC: Absolutely. Healthcare is a human business. We need technology. We need to learn how to use technology. We need to transform and lead the development of technology. But we can't lose sight of really what our core business is. And our core business is about taking care of people.

Host: Yeah. And Lauren, I would imagine that you learn some things along the way. So based on the AI innovation, what key insights and lessons did you discover that will inform future iterations of this?

Lauren A. Bulin, DNP, MSN, MBA, RN, CNML: That's a great question. And what I would say about that is that AI is a tool, and our clinicians have this incredible ability to assess and have these senses that help them to know that something might not be quite right, right? So, they're assessing this patient. They're seeing some things, they're questioning whether they're a physical or a psychological impact. And then, they have a tool in AI to be able to say, "You know what? What you're seeing is this combination of many of these risks or indicators that could signal human trafficking for this particular patient."

So, I think it helps them to build confidence. And it helps to close that gap where they might have a sense, but now they have a tool as well to help confirm some of their judgments.

Host: Which can really be an effective tool as well, because it can uncover and detect things that we just can't, or it can do it a lot faster than we can do it. So, given its pervasive role in patient care, Lauren, how does nursing leadership strategically catalyze and sustain truly effective interpersonal collaboration between clinicians, IT, data scientists, and ethicists to ensure AI adoption is not just technically sound, but deeply integrated, clinically relevant, and patient-centered?

Lauren A. Bulin, DNP, MSN, MBA, RN, CNML: That is no small task. What I would say is that, again, I cannot overstate the importance of really having a collaborative approach. And so, we're looking at it from a nursing perspective. And we're also having to understand different perspectives of the multiple experts that we have in this space that we're learning from as well, they're learning from us.

And I think together we can create this detection model and this summary that really do come together. And as Connie referenced, it's really important to get that continuous feedback. And so, we're asking the clinicians and a pilot, how did that go? What could we learn from this? What could we do differently? And the clinicians really brought forward some of the concerns. And I think through that, we're able to make iterations and make it an even stronger model to help us, you know, with our decision-making and our interventions for those patients.

Host: So, kind of learning as you go in all the departments, IT, the clinicians, data scientists, ethicists. So, you're all learning together.

Lauren A. Bulin, DNP, MSN, MBA, RN, CNML: Yes.

Host: Yeah.

Connie A. Clemmons-Brown, DNP, MBA, RN, CENP, NEA-BC: Bill, could I just add, I think an important—I want to kind of double click on what Lauren was saying. Traditionally, IT is a function in silos. Nursing is a function in silos. Data science has been kind of on their own. And what this initiative has taught us is that we have expertise and imagine the possibility when we bring it together to solve a real-world problem.

And so, we invite our technicians and our technical colleagues into the clinical space, much like they're sharing their knowledge and expertise, teaching us terminology, teaching us concepts that then help us reframe how we think about a clinical problem and then therefore the solution.

Host: Absolutely. So then, Connie, question for you. How does AI fundamentally elevate and redefine the professional role of the nurse? How does it do that?

Connie A. Clemmons-Brown, DNP, MBA, RN, CENP, NEA-BC: I think historically, and this is my opinion, historically, nurses have been the recipient of innovation. And so, we've had to learn to adapt, adjust, take on things that were created by somebody else.

And what AI has created or afforded us is the opportunity to transform the profession. So, we go from—I don't want to say simple, but we go from data consumers where we pick out pieces and parts from the chart and try to put it all together ourselves to being very integrated and strategic innovators in informing clinical decisions and in transforming healthcare, compared to how we know it today. It's an opportunity for us that I think we should open the door and walk through. Certainly, we should exercise caution and skepticism, but we shouldn't be afraid.

Host: Yeah, that makes a lot of sense. I love how you said the nursing profession has generally been the recipient of innovation. It sounds like it's time to turn the tables where AI can help turn the tables where nurses can become strategic innovators.

Connie A. Clemmons-Brown, DNP, MBA, RN, CENP, NEA-BC: We have over 4.7 million nurses in this country. They're all well educated. They have a lot of experience. They experience the troubles, the inefficiencies in our processes, the delays in care. We very much are in tune with the human piece of healthcare. Why not ask them? Why not give them a forum and an opportunity to express the ideas that they've had all along and never really had an opportunity or never took the opportunity to say, "Hey, we could do this better."

Host: Yeah. So, leadership is always asking ROI, how is this benefiting us? So, Connie, what strategies can be used to demonstrate its compelling return on investment for future philanthropic and organizational support?

Connie A. Clemmons-Brown, DNP, MBA, RN, CENP, NEA-BC: Great question. I think nurses need to step up. And when I say nurse leader, I don't just mean the titled individual, somebody who has a manager, director, chief nurse, executive after their name. This is really about each and every individual professional nurse stepping forward and understanding the value they bring to the table. And how their ideas and innovations can change healthcare. And I'll give you a couple of examples. We, as nurses, are very much champions of patient safety and patient quality. And those translate very well to administration, to the organization when it comes to value-based care. But it also translates to donors in the philanthropy space where they're looking for lives saved impact to community. And so, I think those are messages that we need to craft and develop and then share.

The second thing I'll say is it's about professional development and clinical confidence. When your clinicians are confident in the work they do and the care they deliver, outcomes are better. And that means healthier communities, that means a better bottom line. That quality goes straight to the bottom line. And so, we need to leverage that data, leverage those outcomes to garner the additional resources we need to provide good care to our communities.

Host: Yeah, the frontline can certainly influence the bottom line.

Connie A. Clemmons-Brown, DNP, MBA, RN, CENP, NEA-BC: Absolutely.

Host: Yeah. Well, this has been a great discussion. Lauren and Connie, thank you so much for your time today.

Connie A. Clemmons-Brown, DNP, MBA, RN, CENP, NEA-BC: Thank you for having us.

Lauren A. Bulin, DNP, MSN, MBA, RN, CNML: Thanks for having us.

Host: Yeah, absolutely. Good. And keep up the good work. This is really exciting what you're doing.

Connie A. Clemmons-Brown, DNP, MBA, RN, CENP, NEA-BC: Thank you for the audience. We appreciate it.

Host: For sure. Once again, that is Connie Clemmons-Brown and Lauren Bulin. And if you found this podcast helpful, please share it on your social channels and check out the full podcast library for topics of interest to you at aonl.org/nursing-leadership-podcast. This is Today in Nursing leadership. Thanks for listening.