AI can help clinicians to better address health risks from climate change

From Deloitte, this is Government’s Future Frontiers, the podcast that asks questions today to help create tomorrow. I’m Bill Eggers, the executive director of Deloitte’s Center for Government Insights.

This episode comes from HLTH Europe in Amsterdam, the Netherlands: It’s Europe’s largest health innovation event, where global leaders in health, business, innovation, science, public service, and philanthropy all come together, with the aim of transforming health for all.

In the conversation you’re about to hear, my Deloitte colleague David Rabinowitz sits down with Dr. Chethan Sarabu, director of clinical innovation for the Health Tech Hub at Cornell Tech’s Jacobs Institute.

David Rabinowitz: Imagine it’s seven o’clock in the morning, and a patient with asthma is getting ready for work. Her wearable is about to nudge her to exercise today, but her weather app says it might be a bad–air quality day. Her doctor’s office knows that she’s had two recent asthma exacerbations, and her health plan knows that she’s pretty stressed about the cost of care and her own affordability issues.

Her employer knows she’s got time off [that] she hasn’t been using. At the same time, her daughter’s school is debating whether or not to cancel outdoor activities. The emergency department is also preparing quietly in the background for a surge in heat-related health issues.

Now imagine all of those signals actually talk to each other. That’s the premise we’re going to explore on this episode of the podcast: Artificial intelligence being focused on one of the most defining health challenges of our time—the intersection of climate and health.

I’m thrilled to be joined here with Dr. Chethan Sarabu. Chethan is a landscape architect, a pediatrician, and a leader in responsible use of AI, and one of a team working to bring clinicians, researchers, innovators, technologists, patients, and others—together—into the same room to talk about and work on the nexus of climate and health.

Their work is focused on catalyzing innovation at this important intersection, and the work also reveals a tension that we’re going to come back to throughout the discussion: AI can be a powerful tool in this topic, but trust, clinical oversight, community centricity, and environmental stewardship have to be part of the core conversation.

So, Chethan, welcome.

Dr. Chethan Sarabu: Thank you, thank you, David. It’s really great to be in conversation with you today on this topic, for me, that’s been, you know, a really personal journey. I got started in this space, living in California, working as a primary care pediatrician, taking care of vulnerable patients at a federally qualified health center.

During some of the worst wildfires in 2020, we’re in the midst of COVID-19, and what I realized was the children I was seeing in clinic. I knew that wildfire smoke played an important role in their asthma management, but in the electronic health record, there was no information about air quality, no information about heat or other environmental variables that I needed to take into account. And that's really what started this journey for me.

Rabinowitz: So, that brings up a really interesting way to get started. You know, we’re here at Health Europe. We’re surrounded by people who are talking about and investing heavily in innovation and technology. You were just talking about the potential at the start of your journey in this space. It still feels though like this climate and health intersection isn’t part of the main conversation. What do you think would happen if that really changed?

Sarabu: Yeah, I think we’re getting there. We’re seeing more momentum around people realizing that, as we have more and more severe weather events that are happening with greater frequency, and those are really affecting the health of people in ways that are so visceral. And [it’s] not just individuals, but the systems that take care of them—hospitals are often on the front lines of responding to a natural disaster that happens.

And being able to use data effectively to predict when an extreme weather event may lead to a surge of patients so that you could more proactively outreach those most vulnerable populations and make sure they’re good, make sure that they’re not gonna end up in the emergency room at a time where capacity may be throttled.

I think [that] is really the promise that we’re hoping to see by connecting the dots between environmental information, weather information, with health information—all of the signals that you talked about. From personal wearables that are all now augmented by AI and the promise of AI to stitch these together and not only create better predictions, but turn that into better personalized action, I think [that] is the hope […] and the sister conferences have started to invite us into this conversation over the past few years. But we hope that we can actually accelerate more action in this space.

Rabinowitz: So, thinking about more action in this space, help us understand what’s different about this moment. What? Why now? It’s been very clear in the literature about the impact that weather, climate, and environmental factors have on health outcomes for individuals and for communities. What’s different about the innovation agenda, at the moment, that gives you a lot of optimism and hope that now is really the chance to drive more action?

Sarabu: Yeah, I think there are a couple of pieces that have started to come together. So one, I think there’s greater awareness of the health impacts of climate change, which is not only being recognized for its health impacts, but its financial impacts. Hospitals that are on the front lines of climate disasters have more and more costs that they’re paying, and that’s driving an increasing shift in investments upstream. Still not as much as we’d like to see.

And we have much better data interoperability with being able to connect health data from personal devices to your medical records at your doctor’s office. And the environmental data is getting better and better too, with AI prediction. And so, we’re able to better stitch all this information together.

And large language models—which I know are a hot topic—are actually a really critical ingredient here, because one of the core challenges with connecting the dots here is [that] we have different types of data. And we’re talking about, you know, heat data combined with knowing someone might be taking a specific medication that puts them at greater risk, because of the extreme heat, but whether or not they have air conditioning—all of those data points exist in many different modalities.

And the power of large language models is [that] they can actually work across different modalities of data. So to be able to stitch all of that information together, [in that] LLMs actually play a powerful role.

And, moreover, they could then translate that to individual clinicians or patients. So taking this example of medications, we now have a list of about 60 to 70 medications, many commonly used ones. So diuretics—medications that can help lower your blood pressure by allowing your body to pee more, basically, to lose more water—are very commonly prescribed, but during a heat wave, [they] can put you at much greater risk for heat stroke.

And, so, a lot of clinicians aren’t really educated on that. And the literature starts to indicate that, if there is a heat wave, you should maybe adjust the medication, adjust the dose of the medication. And clinicians aren’t necessarily educated on it.

But AI tools that can look at the latest literature [and] bring it to the point of care are starting to surface this knowledge [and] shorten that research—that time from research to clinical action. And moreover, make it much more personalized where it’s not only the education for the clinician, but you can now push this to patients with personalized chatbots that understand their unique nuances and circumstances. So, in short, AI is actually a really big gamechanger for this moment.

Rabinowitz: Speaking of AI as a gamechanger, I want to dig in a little bit more to the model and the work that you and your team are really leading that’s bringing a broad range of perspectives and expertise, and leaders, together, to think about how to solve this challenge. I know that last September, you hosted a hackathon in New York. Go back to that room. What was it like? Who was there? How did it work?

Sarabu: Yeah, so, the way this hackathon started really was one year before—September 2024, New York Climate Week—[when] my colleague and friend Seema Wadwa, who had previously led sustainability for Kaiser Permanente and a Time 100 climate leader, was reaching out to me after Climate Week 2024, saying a lot of great conversations about centering health and climate, but not a lot of action.

She reached out, said, “I think we should do a hackathon. But I’ve never been to one.” And so, that’s where this partnership started. I’m someone who has been to many hackathons, [and] helped to organize them. And so, that was really this nexus. And what a hackathon is for someone who is not aware is [that] it’s a time-bound event, usually a weekend long, where you bring together a diverse group of people around a specific challenge—a diverse group of people from the perspectives and trainings that you have.

 And so, that's what we really set out to do and we did. We brought together about 150 participants from over 70 institutions, brought together computer science students with clinicians and really across a pretty wide age range from high school students to people who are near retirement.

And what was really powerful about that moment was that AI tools actually enable a whole new level of participation and engagement, even for people who don’t have a deep technical background. Hackathons, by their nature, tend to be a bit more on the tech-heavy side, but they don’t have to be. There’s different flavors of them. But AI vibe-coding tools actually allow someone who is not a computer scientist to open up a data set.

And what we did for the hackathon [was that] we focused on four different challenge areas where extreme heat, poor air quality, [and] severe weather events impact health. And [thought about] how do we make healthcare more sustainable? We had a lot of partners, including Deloitte, provide a lot of human support, technical support, [and] data support.

We were able to get healthcare organizations, environmental data organizations to provide proprietary data sets, which really created a vibrant ecosystem and substrate for these teams to be able to come up with creative solutions.

And over the span of one weekend, we saw, you know, teams of computer science students working with sustainability experts, rapidly exchanging knowledge and coming up with not only great ideas, but actually very functional, early prototypes and solutions that could be deployed with not that much more refinement.

Rabinowitz: So what kind of solutions, what kind of prototypes came out of it? Give us an example, one or two, that you thought was particularly exciting, cutting-edge, or impactful.

Sarabu: Yeah, so one that we saw there was focused on how extreme heat impacts outdoor workers. And we know that employers in particular are a group that’s very concerned with the health of their workers. And so, one of the award-winning teams developed a dashboard to better help employers manage the health of their workers who are exposed to extreme heat.

We had another sponsor of ours that was able to take one of the ideas. This idea was looking at “pharmacy deserts,” where this organization helps to support building pharmacies in parts of the country that are vulnerable, that don’t have enough pharmacies.

But they also wanted to look at climate vulnerability or climate/health vulnerability and this team developed a tool to do that, and students from this group were then brought into this organization to help build that out beyond the hackathon. So this is our first time doing this, but we already started to see the signal that the ideas that came out of this had legs beyond that initial weekend.

Rabinowitz: So, given the first time was such a success, what’s next?

Sarabu: Yeah, so this has led us to realize that this model has a lot of potential and we’ve turned this one hackathon into an AI innovation cooperative, the Health and Climate AI Innovation Cooperative, where we’re going to be hosting multiple hackathons, including one in London and New York this year, along with London Climate Action Week [and] New York Climate Action Week. And we are bringing together our sponsors [and] partners to help shape the challenges through a series of roundtables.

This one in London is in partnership with the United Nations Global Innovation Hub. And, by specifying the challenge tracks from sponsors, the ideas that then come out of the hackathons could help solve real-world problems, and that’s where we’re also by the end of the year hoping to launch an incubator that can support the ideas coming out of the hackathon and connect them to the organizations that are ready to deploy these solutions.

Rabinowitz: I know that thinking deeply about not just the problems to solve, but the principles by which you and the team help the participants solve those problems is really kind of central to the work that you’re doing. What are you doing to bring that responsibility layer to the forefront?

Sarabu: This is such a critical question. I think you’ve highlighted so many of the dimensions of responsible AI, but one that I want to dig into further is the environmental footprint of AI itself.

We know AI has huge potential to help, but it also has huge potential for harm on so many dimensions. So, AI is powered by data centers; these data centers need a huge amount of electricity to run the GPUs—the graphical processing units that are doing all the number crunching that makes the AI happen; and when you’re doing all this number crunching, they generate a lot of heat. So, you need water to cool it down. And these data centers are consuming a massive amount of resources around the world and it’s leading to a lot of alarm from not only that environmental strain, but the impacts that they’re having on the communities where they’re based, from the economic impacts of the change in prices of utilities to some of the health impacts of the noise living around a data center.

And so, for us, thinking about climate and health and where AI can do a lot of good […] we are particularly focused also on that other side of how we do this in a way where we’re maximizing the efficiency of what AI can do for a particular use case, but while making sure we’re not overusing it.

And so, we have a research team that published an article called the Sahai Framework or Sustainably Advancing Health AI led by Anu Ramachandran who is at UCSF and Udit Gupta at Cornell, Shomith Ghosh and Vivian Lee at Harvard […] working with the Health and Climate team. And this framework paper really set out to show how we can use AI responsibly, that there are different levers from the how it’s applied to the model that you’re using, to the efficiency of the data center that you’re using—all those have a huge difference in the overall emissions, even getting the same outputs that you might want.

And so this framework is being taken up by organizations like Healthcare Without Harm [and] the American Medical Informatics Association to bring this up to national policymakers. And so, we incorporate those principles into our hackathon, but are also working with the national and international leaders to bring that to healthcare systems around the world.

Rabinowitz: So, I want to end with how to make this a little practical. And if you’re a clinical executive, a leader in a variety of ways across the multitude of organizations that you’re talking about; and maybe, you haven’t been exposed to the work that you are doing or just the topic, in general. What would you want more people to be doing?

Sarabu: Yeah, so I think, get engaged, learn more about the topic. Climate change obviously affects so many different aspects of our world, life, but really think about it as a healthcare issue first: It’s about your child’s asthma. It’s about whether or not you can show up to work and [have] economic productivity in times where there's increasing heat waves everywhere. And the power of the digital health community, the promise of digital health care and augmented by AI is that we can better connect the signals, we have wearables and apps that can help us understand, in more real-time, [the] dynamic nature [of] how the environment and health have this complex interplay, and so, really encourage everyone at the health community, and in the broader digital health community, to get engaged because this is the greatest health challenge of our lifetime, and we need all of you there to be part of it.

Rabinowitz: You know, I love you bringing it back to the patient and the family and the community at the center of this and thinking about how much promise there is to use the tools that you and others are developing to really make such an impact around not just the experience of care, but the cost, the quality, and the outcomes associated with it. So, thank you for what you’re doing, and thanks for joining us today, and thank you for such an insightful and thoughtful conversation.

Sarabu: Thank you so much. Appreciate it

Thanks for listening to this episode of Government’s Future Frontiers from Deloitte. You’ll find all our previous episodes wherever you get your podcasts. And to make sure you don’t miss new ones, be sure to follow the show on your favorite podcast platform.

This podcast is produced by Deloitte. The views and opinions expressed by podcast speakers and guests are solely their own and do not reflect the opinions of Deloitte. This podcast provides general information only and is not intended to constitute advice or services of any kind. For additional information about Deloitte, go to Deloitte.com/about.

About Deloitte

Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee (“DTTL”), its network of member firms, and their related entities. DTTL and each of its member firms are legally separate and independent entities. DTTL (also referred to as “Deloitte Global”) does not provide services to clients. In the United States, Deloitte refers to one or more of the US member firms of DTTL, their related entities that operate using the “Deloitte” name in the United States and their respective affiliates. Certain services may not be available to attest clients under the rules and regulations of public accounting. Please see www.deloitte.com/about to learn more about our global network of member firms.

Copyright © 2026 Deloitte Development LLC. All rights reserved.