Why Hospitals Must Use AI Responsibly

00;00;00;04 - 00;00;22;28
Tom Haederle
Welcome to Advancing Health. How can we make sure that everyone benefits from the efficiencies and improvements to health care that are expected from artificial intelligence? In this podcast, we hear from a theological ethicist who says the guiding principle should be this: all efficiency should serve compassion.

00;00;23;00 - 00;00;46;29
Joy Rhoden
Joy Rhoden here with the American Hospital Association, where I serve as the senior vice president for the Division of Health Outcomes and Care Transformation. And I'm joined today by Daniel Daly, who's the executive director for the center for Theology and Ethics in Catholic Health at the Catholic Health Association. And it's a pleasure to have you here with me, Daniel.

00;00;47;02 - 00;01;11;26
Joy Rhoden
We're going to be diving into a topic that sits at the crossroads of evolving technology and health care, and that is the rise of artificial intelligence in health care. But specifically, given Daniel's role as a theological ethicist - and I'm going to ask you to explain to our listenership what exactly is that and what do you do in your day job -

00;01;11;27 - 00;01;41;02
Joy Rhoden
but we're going to be talking about essentially, how do we think about AI from through an ethical lens? What are those considerations as health care leaders are really launching these tools? And because I sit and lead the division of Health Outcomes and Care Transformation, I also want to talk to you about some concerns that have surfaced about maybe unequal access to AI driven benefits.

00;01;41;02 - 00;01;56;04
Joy Rhoden
And so with that, I'm going to give you the opportunity to do a better introduction of your work at the Catholic Health Association. So what is a theological ethicist? Who is that? What's your role?

00;01;56;06 - 00;02;15;19
Daniel Daly
Well, first, Joy, thank you for having me. I'm excited to be with you today to talk about this really important topic. So as you said, I'm a theological ethicist. I also teach at Boston College. And a theological ethicist looks at goodness and rightness in light of theology. When we think about this in terms of health care, you know, I'm a Catholic theological ethicist.

00;02;15;19 - 00;02;38;25
Daniel Daly
As our listening audience knows, Catholic health care is throughout the world, but you know, a lot of it here in the United States. It basically emerges from this understanding of who we're called to be, which is to care for the suffering, the sick, the poor, the forgotten. So an ethicist, the theological ethicist is looking at that tradition. You live out your faith through healing the sick.

00;02;38;26 - 00;02;49;20
Daniel Daly
And so we need to, you know, how do we do that? How do we do that? Well, we need people to be thinking about that. And that's essentially the role of an assist in this kind of Catholic health care space.

00;02;49;27 - 00;03;17;20 
Joy Rhoden
Very unique role. Thanks for sharing that. So many of our organizations, our hospitals and health systems are really excited about the efficiency gains that AI can bring them. In a field where folks are called to care for others, how should leaders think about balancing the efficiencies that they can gain from AI with compassion and responsibility?

00;03;17;22 - 00;03;43;02
Daniel Daly
And Joy, it's a great question, and I think the way that I look at this is that it's not so much that we're balancing efficiency against something like compassion or responsibility, but rather all efficiency should serve compassion. It should serve responsibility because, as you rightly noted, the goal of health care is to care for, to provide healing and health and even enable human flourishing for patients.

00;03;43;08 - 00;04;08;28
Daniel Daly
Really, the threat of AI, one of the one of the threats is that we turn more to kind of optimized care that the focus is on efficiency for efficiency sake, and that would get us away from person centered care. So efficiency has its place, but its place it's not the final goal of health care. And so what we have to be careful as we bring in these tools that help us be more efficient.

00;04;08;28 - 00;04;22;16
Daniel Daly
And that's great. We don't let the means drive the end. The end is always going to be person centered care for human health and well-being. And I think this is a critical, critical point as we as we take up AI.

00;04;22;19 - 00;04;51;19
Joy Rhoden
You've surfaced a few key concepts and constructs, and I want to go a little bit deeper around person centered care. So many of the use cases thus far in health care for AI are around admin processes, right? But as AI becomes more integrated into the care delivery space, how might providers leverage AI in support of person centered care?

00;04;51;26 - 00;05;19;12
Daniel Daly
I think the first thing is, is that we need a culture that emphasizes the importance of person centered care, the importance of that professional patient relationship. These should be sacrosanct in health care. They should be untouchable, and the team needs to know when that leaders value person centered care, that the leaders value the relationship of the professional and the patients.

00;05;19;17 - 00;05;41;24
Daniel Daly
We're certainly going to be integrating this in clinical settings. But, you know, I was at a conference last year, and in the room there was basically this this adage that we need to fall in love with the use case, not the technology. That's right. And I had to push back on that. The use case is valuable only insofar as it helps us to heal and care for patients.

00;05;41;24 - 00;06;01;03
Daniel Daly
So we shouldn't be falling in love with the technology, nor that specific use case. But all of our evaluation of AI and health care has to be in light of human well-being, human flourishing, whether it's the flourishing of patients or as we think about, you know, we're also concerned with the flourishing of professionals.

00;06;01;04 - 00;06;02;21
Joy Rhoden
That's right. The workforce. Yep.

00;06;02;23 - 00;06;27;12
Daniel Daly
Yeah, exactly. We need to be concerned not just because they're the ones who provide the care, but because they have intrinsic value themselves. They are human beings who deserve to be treated with respect and dignity, and that the AI shouldn't be changing their work to the point that they are not respected as healers, as carers, as professionals in what they do.

00;06;27;13 - 00;06;33;28
Joy Rhoden
To think about how to leverage the technology to strengthen those relationships.

00;06;33;28 - 00;06;57;01
Daniel Daly
That's exactly it. It can be leveraged. I think it has been. You can think about something like ambient listing technology, which the early returns are pretty strong, if not little mixed, but that ambient listening technology, you know, that famous AMA study from a couple of years ago that it really reduced pajama time for professionals. It leads them to be less burned out, have greater job satisfaction.

00;06;57;04 - 00;07;18;04
Daniel Daly
But patients early reports are that patients like it as well because they get that eye contact. They feel listened to, they feel seen where the iPad is not the fixation of the professional, but rather the person. Now that's a huge win that promotes the professional patient relationship and makes care more person centered.

00;07;18;11 - 00;07;59;14
Joy Rhoden
That's really all exciting and sounds directionally correct, but we know that technology has the potential to exacerbate disparities if it's not deployed in a responsible manner. And so some patients may benefit more from AI driven tools than others, particularly those with access to digital devices and Wi-Fi, broadband and other technologies. What do you see as the responsibility or the role that hospitals and health systems play to ensure that while they're moving forward with these advancements in AI, that they are not leaving some patients behind?

00;07;59;16 - 00;08;18;29
Daniel Daly
You know, Joy, it's an incredibly important question, and I think it's underappreciated, under discussed, so I really appreciate you bringing it in. I think systems have a mission. I think their mission is if you if you ask them and they is to provide care for all in a community, they're not making distinctions between who should be cared for and who shouldn't be or who doesn't deserve care.

00;08;19;04 - 00;08;52;22
Daniel Daly
They're looking to reduce those inequalities. Or they at least they should be, and I believe they are. However, as you noted, the well-resourced are positioned to benefit the most from AI. So what does that mean? I think it means we need to really focus on the kind of tools that we adopt and use, and to test them and to monitor them for the expansion of access that they benefit, not just the well-resourced, but they benefit everyone who is coming into the medical setting.

00;08;53;00 - 00;09;09;00
Joy Rhoden
You said health system leaders should actually maybe think about how they can influence the development of new AI tools, right? What are some practical steps that you would give to health system leaders on the development front?

00;09;09;02 - 00;09;30;15
Daniel Daly
So yeah, on the development front, I think what we need is we need strong governance structures here to evaluate and monitor AI across its life cycle in the in the system. So I think we need people in policy. We need committees. We need groups, interdisciplinary groups that are that come together, that are evaluating regularly. And they need to be diverse.

00;09;30;16 - 00;09;35;03
Daniel Daly
Right. You need to bring in multiple kinds of people with different backgrounds to do that.

00;09;35;04 - 00;09;42;04
Joy Rhoden
It sounds like we should actually be moving upstream and working with the vendors in the development phase. Correct?

00;09;42;05 - 00;10;01;09
Daniel Daly
Well, I mean, ideally, yes, that does happen and it can happen. I've talked to many people that it's often the vendor coming after the product has been developed. There can then be modifications that happen in light of the values that maybe I have as a system, but certainly, yeah, I mean, the earlier you get in  - ethics by design is a best practice.

00;10;01;10 - 00;10;20;00
Daniel Daly
Exactly. And I think if we can make that normative in the in the field, we'll go a long way to avoiding more surprises that happen as these tools get deployed. And often there are surprises. You can you can cut off those problems by invoking ethics by design.

00;10;20;03 - 00;10;32;03
Joy Rhoden
As AI is introduced across care settings, you know, to really think about that alignment that you spoke about with mission and ethical commitments.

00;10;32;05 - 00;10;51;18
Daniel Daly
I think when you think about making ethics more than lip service, because it often is, and it needs to be, it needs to be more than that. It's about getting it into the structure and the culture. What are the ideas that we endorse? What are the ideas we support, and then what structures have we put in place to ensure that we are aligned with those values that we have?

00;10;51;20 - 00;11;15;09
Daniel Daly
So things like performance reviews or the way the organizational chart is set up. Position descriptions, all of that has to have those things embedded, because we reward what we value and we value what we reward. And if we value the ethical use of AI, we need to reward that in our associates, in our workers, in our professionals, in our administrators.

00;11;15;09 - 00;11;23;25
Daniel Daly
And that's the surest way to get those values to be lived out on the ground in the care for the patients and communities.

00;11;24;01 - 00;11;44;21
Joy Rhoden
Well said. Thank you so much, Daniel, for being with us today and for really helping our audience think about the both and how might we pursue potential efficiency gains leveraging AI without compromising person centered care and improving health outcomes for all. Thank you.

00;11;44;25 - 00;11;46;07
Daniel Daly
Thank you. Joy.

00;11;46;09 - 00;11;55;02
Tom Haederle
Thanks for listening to Advancing Health. Please subscribe and rate us five stars on Apple Podcasts, Spotify, or wherever you get your podcasts.