How AI Can Improve Healthcare Outcomes

In this edition of Future Together, we sat down with Dr. Sree Chaguturu, EVP and Chief Medical Officer for CVS Health, and Dr. Karen DeSalvo, Chief Health Officer for Google. They discuss how the safe and responsible use of AI can drive better health outcomes and the significance of health equity, data privacy, and responsible AI adoption.

Future Together brings the Digital Futures Project to life through diverse conversations around the impact and opportunity of AI.

Sree Chaguturu: Healthcare artificial intelligence is at an inflection point in its evolution. We see it helping us across four domains: 1) improving clinical care, 2) administrative complexity, 3) financial forecasting, and 4) operational excellence.

If we think about administrative complexity and healthcare, artificial intelligence can help us with automating those time-consuming tasks. We see artificial intelligence helping to improve financial and accounting practices to better predict the expenses and costs of healthcare. Then lastly, applying AI to operations to make sure that we have smart, integrated operating models.

Karen DeSalvo: I like that language of “inflection point” a lot. For us at Google, AI has underpinned a lot of our work that we've done in health. And we think similarly about how we're working with the external ecosystem to support health broadly; our framing is consumers, caregivers, and communities. We have ways that we’re interacting directly with consumers around the world, hundreds of millions of people every day asking us health questions on our platforms like Search, and we use AI in those contexts to help deliver to people high-quality information. We also have been using AI in our hardware products, like Fitbit, to detect atrial fibrillation. And we’re working with the healthcare system, with caregivers like CVS Health, on ways that we can help use AI to read mammograms or retinal images or do analytics. We’re also working to develop a longitudinal patient record across siloed data that allows for generative AI solutions designed specifically for healthcare providers and their patients.

Generative AI has accelerated and enhanced the opportunity to do more for folks, to take data to do something more sophisticated and personalized. Health is complex, and the decisions that people make every day need to be more customized and personalized if we're really going to help people on their health journey.

Sree Chaguturu: Karen, as you talked about customization and personalization, I think about some of the examples of how we're using AI at CVS Health, and it's to do exactly that. And it's also as a way to augment clinical and human decision-making — not to replace that. We still believe that whether it's a client or patient relationship, that human touch is really important.

At Oak Street Health, our primary care practice, we're exploring using generative AI to help with case summarization and history taking, which allows our clinical teams to be present in conversations with patients and have a more human touch. It's been really helpful for our clinicians in doing what they want to do, which is provide clinical care.

Sree Chaguturu: With artificial intelligence, we have a real opportunity to help improve healthcare, so as we roll this out, we want to make sure that individuals can have trust in the tools. I know that a lot of what Google is working on, as well as CVS Health, is to ensure we’re using these innovations in a safe, responsible, transparent, and ethical way.

Karen DeSalvo: The expression that we use around here is, “Health moves at the speed of trust.” And as we've been developing AI and generative AI, our approach has been, “First, do no harm,” inspired by the oath medical professionals take when becoming doctors. Google has been building and thinking about how to assess the performance of the models in the laboratory, but also in the real-world environment.

What I'm really excited about is the way the broader ecosystem is coming together more, like the Coalition for Health AI or the work that the National Academy of Medicine is doing, because it's just so important to not do it alone. It really requires the best thinking from across the industry to make sure that we're not only thinking through the ways that we can improve health outcomes, but also, the watch-outs we should all be looking for and learning from each other as we're implementing the technology.

Sree Chaguturu: The rigorous process and thought that goes into ensuring trust and validity in Google’s AI development is very consistent with how CVS Health is thinking about the deployment of artificial intelligence. We continuously check to ensure — not only at the initial deployment of the technology, but through ongoing monitoring — that it continues to fundamentally improve healthcare outcomes better than the status quo. This is not a one and done. This is about making sure that we have a very robust development and deployment framework, as well as a robust ongoing oversight framework.

Karen DeSalvo: We just talked about trust, and one of the ways we build trust with our users, our customers, our partners, and communities is by ensuring their data is safe, secure and private. With a partner like CVS Health, we use Google Cloud’s AI services, which provide organizations with privacy protections and fine-grained security controls across their data. This means CVS Health can dictate who can access their data and how it can be used, which supports CVS’ HIPAA compliance obligations. Most importantly, Google never uses customer data to train our models, or models used by others, without permission.

Sree Chaguturu: What we've done within CVS Health is we’ve built an internal oversight process that governs all of our artificial intelligence work at the company, ensuring that when we are deploying these technologies that we do this in secure environments, considering patient privacy.

We've also aligned with the White House’s ongoing work to ensure the responsible, safe, secure, and trustworthy use and purchase of AI in healthcare. In signing this voluntary commitment, we are working with peers and partners across the industry to ensure that everything that we do is Fair, Appropriate, Valid, Effective, and Safe — the FAVES AI Principles.

Sree Chaguturu: When we talk about health equity, it is fundamentally about improving the quality of care and outcomes in this country, full stop. When we think about how we ensure health equity at CVS Health, we have a three-part approach. First is empowering our colleagues through education and training. Second, it's important for us to measure what matters — or measure what you treasure — and make sure that we have a common framework around improving health equity and healthcare outcomes. Lastly, it's about taking bold actions. We've made commitments across our company to help reduce and eliminate disparities in blood pressure control and depression screening — really focusing on hearts and minds.

When it comes to artificial intelligence, we want to understand how similar the training data is to the population. Understanding that artificial intelligence has been developed with the community in mind is really important, because that means that when it's deployed, it is likely to perform in the way that it has been tested and developed.

Karen DeSalvo: I know you have a really strong commitment to equity. We were on a Leveraging Data to Drive Outcomes in Health Equity podcast episode together, and we talked deeply about this. At Google, we do equity by design. One of the early things that I did when I joined the company was to stand up a Health Equity team so that we could build out knowledge and understanding of the fact that in healthcare, there are historic inequities. So what can we do as a company to see that technology doesn't exacerbate those, but eliminates those? It requires that you're upstream and doing that by design.

The opportunity here with AI is pretty extraordinary. On a global scale, when we think about the people on the planet that don't have access to any healthcare and the inequities that brings, AI offers an opportunity to augment the workforce and fill gaps. We must ensure that everyone, everywhere — and not just some people, some places — get to benefit from the promise of AI.

Sree Chaguturu: If we want to make sure that this incredible opportunity doesn't slip through our hands, we need to make sure that as we roll this out that we explain, clearly and meticulously and repeatedly, what AI is, why it's important, and how we're ensuring that it's safe, effective, and is going to be designed with equity in mind to improve healthcare outcomes. And when we see it not doing that, we will address it as quickly as possible. So it's not about just the innovation; it's how we implement the innovation that's going to be really important in this upcoming year so that we can continue to build on that early success.

As we are at this inflection point, we want to make sure that we have a regulatory framework that helps spur on innovation, but simultaneously helps us ensure the safety of the work. This is going to be a really important time as the private sector works closely with regulatory agencies and legislators to make sure that we help support this important time in healthcare technology.

Karen DeSalvo: I think that the government has been really open about the fact that because generative AI is a relatively new technology, they need partnership with the private sector and civil society to understand the potential and the possible risks as they're thinking about developing their framework. Governments and the private sector will need to get aligned to share knowledge and experience and expertise so that governments can make decisions about the right framing for regulation. I have always been a believer that when we partner together we can go further, and we are seeing that governments, regulators, and companies like Google and CVS Health want to be part of the process to create frameworks and guidelines that result in better health outcomes for everyone.

This interview has been edited for clarity and brevity.

Dr. Sree Chaguturu

Dr. Sree Chaguturu, MD is Executive Vice President and Chief Medical Officer of CVS Health. He leads the CVS Health Medical Affairs organization spanning Aetna, CVS Caremark, CVS Pharmacy, MinuteClinic, Oak Street Health, Signify Health, women’s health and genomics, data and analytics, patient safety, behavioral health, and health equity. Sree is focused on advancing the highest possible clinical quality standards, increasing access to care, improving patient outcomes, and reducing overall healthcare costs across the CVS Health enterprise. Prior to joining CVS Health, Sree was chief population health officer of Mass General Brigham. He is also a practicing internal medicine physician at Massachusetts General Hospital.

Dr. Karen DeSalvo

Dr. Karen DeSalvo is Chief Health Officer at Google, where she leads the team of health professionals who provide guidance for health research, products, and services across Google, including those for Google’s own employees and their family members. She is a physician executive and public health expert who served in the Obama Administration as National Coordinator of Health Information Technology and Acting Assistant Secretary for Health. She was formerly New Orleans Health Commissioner and Vice Dean at Tulane School of Medicine.