Expert Insight

Part I: BCBSA CEO Kim Keck on generative AI, high-cost drugs, and more

In part I of this edition of "Lessons from the C-suite," Advisory Board President Emeritus Eric Larsen sits down with Kim Keck, president and CEO of the Blue Cross and Blue Shield Association, to get her perspectives on generative AI, high-cost drugs, and the future of affordability, as the leader overseeing coverage for one-third of the U.S. population.

Eric: Kim, I’m excited to get your perspective on some of the happenings within our sector and beyond — generative AI, weight-loss drugs, rampant industry consolidation, etc. — because you have a unique and what I might even call ‘ambidextrous’ view on the industry. You spent 28 years on the for-profit payer side with Aetna before switching over to run BCBS-Rhode Island and, in January 2021, moving over to lead over the Blue Cross Blue Shield Association. The Blues are an American institution, covering 118 million lives and a provider network that reaches every zip code in the country. You’re also on the board of Oak Street and Evolent, so you get the risk-bearing primary care and the PCP enablement pieces. You have a uniquely panoramic view on our industry.

Kim: We’re off to a good start because you’ve already mentioned the two things that have been front-and-center of my mind for the last eight months: generative AI and high-cost drugs.

Generative AI: The good, the bad, and the uncomfortable

Eric: It’s hard to escape what’s happening with generative AI, with these advanced foundation models, transformers, and LLMs (large language models). What’s funny is that you and I did a panel together this past March, and we didn't mention a syllable on this. I didn't fully appreciate at the time how important November 30, 2022, was — the day that Open AI launched ChatGPT. I’ve now come to think of it as one of the most consequential dates in the history of technology.

But rather than commenting on civilizational change and other big philosophical questions, let’s narrow our focus “just” to the $450 billion in estimated administrative spend in U.S. healthcare, the potential of AI to augment (if not replace) some of what is viewed as the exclusive province of clinicians, and AI’s potential to compress the time and complexity it takes for new drug discovery and development.

How are you thinking about this watershed moment overall, and where do you see the biggest potential impact for the Blues?

Kim: First, I agree with you, Eric: It’s absolutely revolutionary. And as with any revolution, that means a lot of disruption, which can be good, bad, and uncomfortable.

To start off with the good, generative AI is only as good as the data that powers it. And the Blues have a powerful healthcare data asset to leverage as our starting point. We also, now, have an even greater incentive to share data across the system, build longitudinal records on our members, and determine everything from the next best action to the next best clinical program by testing the boundaries of this technology.

The question we get to answer is: What are the most impactful use cases we can drive to compliment all the work each plan is doing on its own to optimize efficiencies – whether in administrative, care navigation, or clinical tasks – with the right co-piloting? Democratizing this technology will ensure that any Blue plan, regardless of size, has access to the data and capabilities it needs to be competitive.

Eric: That makes a lot of sense — leveraging generative AI to strengthen connective tissue between the Blues. I’m curious where your mind goes with the “bad.”

Kim: If I think about this “revolution” at a macro level, I worry that we'll revolutionize vertically as opposed to horizontally. And I don't mean we, as in payers — I mean, we, the U.S. healthcare system. We're making clinical and administrative improvements, but mostly in our individual silos. I’d love to see us do it globally and cut out the risk of supercharging the incentive structures that drive so much inefficiency in our industry today.

Will providers focus on leveraging this technology to maximize their revenue, for example, because that’s the incentive model they have in fee-for-service? Take radiology: there is huge clinical opportunity with generative AI, but we need to make sure we’re driving improvement in outcomes, not just increasing the number of revenue-generating services provided. In a similar vein, will pharma focus on bringing more drugs to the world that carry a 25% profit margin, because that's their model?

Eric: How do you think about this specifically as a payer? Clearly there are use cases that could be impactful for the Blues. You can gain internal efficiencies by automating — or eliminating — things like administrative paperwork, regulation management, etc. And you can enable consumer-centricity by empowering members to use voice interaction or ChatGPT to receive hyper-personalized clinical counsel, demystify benefits and billing, allow direct scheduling, etc.

But the sheer velocity of the technological change and the exponential increase in data and compute make it hard to pause and say, “Here’s how we can best use this technology.” As a payer leader, how do you balance these things?

Kim: This is where we get to the “uncomfortable,” which I would break into two categories. First, we need to think about the defensive side and be mindful that these things are not without risk. We need to figure out how to put guardrails in place to mitigate patient harm and reputational risk.

The second category is financing. When we take the low-hanging fruit out of clinical inefficiencies or clinical diagnoses, it will be both beneficial and cost-saving in the long run. But how do we financially manage it in the short run? We’re not set up for it.

Eric: Let’s talk about your first point, which touches on this idea of ‘data sovereignty.’ The Blues have this huge repository of structured and unstructured data (claims, clinician notes, labs, images, behavioral write-ups, etc.) that is proprietary, and I would think is a competitive advantage. But you don’t necessarily have the in-house tech capabilities to take full advantage of it in the ways you’ve described. Do you feel comfortable partnering with the multi-trillion-dollar big tech company who will get access to that data? Are you confident your data will stay “yours” and not accrue advantage to competitors down the road?

Kim: I don’t know that we’re ready to declare the answer to that just yet. There are great analytic capabilities that exist within Blues either through plans themselves or collectively owned entities, and we need to ask ourselves what are the incremental capabilities we need on top of those? We are working to figure that out.

Eric: Fair enough. I agree with your skepticism.

Now to the idea of financing, the second category of “uncomfortable.” How are you grappling with this in practice?

Kim: A perfect example is personalized medicine. If large language models can help us decipher low-value versus high-value care and how we pay for it, terrific. Part of the reason we don’t operate that way right now is because it’s so hard to distinguish what is “low-value” for one person versus another. Does Kim need to get that pre-op test and go to the physician’s office 14 times before her elective surgery? Or is that only “valuable” if she has these 75 different comorbidities? We don’t deliver personalized medicine in the U.S. because we’re not set up to handle the nuances and complexity. Being able to offer and pay for care like that would be great in the long run, but is it prohibitively expensive in the short run?

And in the future, if I’m an employer, what do I do? Do I offer 32,000 flavors of benefits for my 32,000 employees? I don’t know what that looks like yet, or how we pay for it.

GLP-1s are clinically significant, politically inevitable, and financially untenable

Eric: This is probably the right cue to transition to our next acronym: GLPs. This has emerged as an enormously consequential therapeutic class of drugs, with the ability to target not just obesity — which is an epidemic in the United States that will affect 50% of the population by 2030 — but, as we continue to see, chronic disease more broadly. And the market is on fire.

As of late 2022, analysts were projecting a $30 billion to $50 billion market for GLP-1s by 2030. And since the first readout of Novo Nordisk’s SELECT trial, I’ve seen the TAM (total addressable market) estimates for this therapeutic class climb as high as $130 billion by 2030. And, of course, the capital markets have fallen in love. Eli Lilly added $100 billion in market cap in a single day and became the most valuable healthcare company in the world in July of this year. Novo is the single most valuable company in Europe, regardless of industry.

But back to the Blues — your association provides coverage for nearly one-third of the U.S. population, so it doesn’t surprise me this has been front-and-center on your mind. How are you thinking about this?

Kim: First, the commonality among GLP-1s and large language models is the macro challenge they present. As you pointed out, the fields of science and technology are moving at a pace that our financing for healthcare in the United States simply has not kept up with.

To your point, let’s say Medicare approves GLP-1s for the treatment of obesity. Treating just 10% of Medicare beneficiaries with obesity would raise spending by $27 billion per year. Employers say they want to cover it. But if you think about small employers, they can’t afford it. Twenty percent of small employers cover it today and less than half of large employers cover it.

Eric: And yet Ro came out with a consumer survey that found that 44% of employees are willing to shift employers to get access to GLP-1s. And when I look at what’s coming down the pike, I don’t foresee demand slowing at all – just the opposite. There are 85 phase II and III trials underway for medications that address obesity. And the clinical impacts are compelling – the SELECT trial found that Wegovy cuts the risk of heart attack, stroke, or cardiovascular death by 20%; Eli Lilly’s retatrutide, which is awaiting phase III study, has demonstrated an average of 24% weight reduction over 48 weeks.

Kim: We want people to have access to medicines that treat chronic conditions, and the early data looks phenomenal. But they’re not that old, so we don’t really know the long-term impacts, and we’re actively learning about the side effects.

What I come back to when I think about this from a payer lens is the cost. ICER has said that even with the cardiovascular benefit, the price tag for these drugs should be 70% lower than they are. That recommendation hasn’t gotten as much traction as I expected. If we’re concerned about access, the fundamental question is about making this affordable.

Eric: Right, we talked about Medicare and employers, but it’s also worth noting that these drugs today run $900 to $1,400 per month, on average, for patients. Even with cost-sharing, health plans are left holding the bag for a lot of this. How do you approach cost mitigation for something this massive?

Kim: I worry about that exact question, because the tools in our toolkits to manage costs for things like this are pretty cumbersome – think prior authorization.

Our early data suggests that less than 20% of our members who switch to Ozempic are still taking it 12 months later. By 18 months that number drops to 15%. To me, this signals the need to dig deeper. Is there some degree of inappropriate use, unsustainable financing, or a combination?

A recent BCBSA study found that we can save an enormous amount of money if we reduce even a fraction of inappropriate prescribing by providers. Our clinical data suggests that we’re currently seeing approximately 20% inappropriate prescribing, on average, for GLP-1s, based on the traditional clinical protocols as recommended by the American Diabetes Association. For example, we see huge percentages of prescriptions coming from dermatology and ENT practices. So we have some additional work to do here.

What I don’t love about these tools is that they’re all incremental approaches.

Eric: I'm curious. I know you're not an ‘incrementalist’ — you want to get things done, and now — but you're also a pragmatist. This is sincere question: what’s the solution? We've got the Inflation Reduction Act attempting to deputize Medicare to negotiate (if it manages to ultimately scrape through the courts, which I’m skeptical about). So, the $900 to $1,400 a month is inevitably going to come way down. But as you pointed out, even if only 10% of Medicare takes these drugs, it's $27 billion, which is 20% of the Part B spending. It’s going to break the bank.

The question basically revolves around how we de-risk Medicare. Let’s say the Blues and the seven publicly traded MCOs cover this: it’s going to lower comorbidities and complications, mitigate cardiovascular disease and stroke, and lower cost for Medicare that way. So should the government pay for this? I don't know.

Kim: I think we need a completely new financing structure in this country. And I do think the emergence of new weight-loss drugs is going to be a catalyst, because this is not something any one stakeholder can do on their own. This is not a problem that scale alone can solve.

If we think about the long-term impacts, we have to ask ourselves — from employers to payers and even providers — how do we do things fundamentally differently?

Eric: You're right; this is not a Blues issue. It is a societal issue. The budget deficit this year is $1.6 trillion. Our debt is $33 trillion. We're spending $600 billion on our debt interest alone this year. Ultimately, this is a national policy question that you and I are not going to solve today. But it does get at the massive impact drugs have on our healthcare spending, which I know is something you’re personally passionate about and has been a huge area of focus for the Blues both individually and as an association.

Cutting down the Blues’ $80 billion cost burden of high-cost drugs

Eric: We should talk about Paul Markovich and his recent announcement that Blue Shield California is diversifying its PBM business away from total CVS control. Give us your take on this move and how translatable it is to the other 33 Blues plans?

Kim: The amount of money we spend supporting the big three PBMs is not trivial. We want best-in-breed capabilities; we also want a more transparent pharmacy model that makes drugs more affordable for our members. It’s directly relevant to the work we’re advancing with programs like the Synergie Medication Collaborative we launched this year. The announcement from Blue Shield of California is furthering the same objective – we are not waiting for the pharmacy model to change for us.

Eric: This strikes me as a perfect example of that “superconductor” energy the BCBS Association has. I imagine this is something you’re closely assessing and will potentially export to the other Blues. Should we expect other Blues to emulate Blue Shield California?

Kim: I welcome what Paul has done as a catalyst for the rest of the Blues. We’ve started this work of looking at the bigger picture of the $80 billion we spend on drugs as Blues and saying: Let’s disrupt that. Let’s get scale here. That was the impetus for launching Synergie. Now Blue Shield of California is pushing further, which will inherently drive our collective investments, like Synergie, to keep up.

Eric: You’ve mentioned Synergie — I want to double click here. GLP-1s aside, there are over 2,000 cell and gene therapies currently in the pipeline, with the entire market projected to hit $50 billion by 2026. These are super high-ticket, seven-figure therapies. How does Synergie factor in here?

Kim: The whole goal of Synergie is to leverage this notion of “building capabilities that give the Blues national scale” to reduce the cost burden of high-cost drugs. And it does that in two main ways that I categorize as volume-based and value-based. On the volume side, it’s using scale to negotiate better deals for medical drug benefits.

The value side is where we’re moving beyond the models that exist today. Early in my tenure, I brought up the idea of creating a pooled risk model to share the cost burden of high-cost drugs across the BCBS Association. Let’s take Zolgensma, the $2.2 million drug for spinal muscular atrophy (SMA) in babies and toddlers. If BCBS-RI has to pay that $2.2 million claim, that’s a big deal. If a larger Blue Plan has to pay it, it’s not trivial, but it’s not a financial threat to the organization. The idea is to do a collective, Blues-wide, outcomes-based arrangement in which we’re willing to collaborate on clinical protocols — we’re willing to compare data so we can feed it back into the clinical evidence base and understand what’s actually working.

It’s worth underscoring that the benefit goes both ways here (to payers and pharma). On the payer side, we can pool risk and think comprehensively about how to drive outcomes. On the pharma side, this could be a good way to test drive the real-world evidence for these drugs, which is not something these companies necessarily have. The idea is they might actually get some data that helps them understand the effectiveness of their treatment.

Eric: So, in part it’s essentially functioning as a reinsurance or a risk distribution network for the high-cost cell and gene therapies like Zolgensma for SMA, Zynteglo for beta thalassemia, etc. And by functioning this way, empowers the Blues to collaborate around best practices to drive better outcomes.

This makes total sense to me. It’s a way to get a lift through the coalition of the Blues that isn’t dependent on something like consolidation.

Kim: It’s a place to start, Eric. Right now, it’s focused on many physician-administered drugs and value-based approaches for cell and gene therapies, and we’re actively thinking about where to go next.

Getting tactical on affordability

Eric: Let’s turn to the question of affordability. BCBS Association has shown a lot of energy here. Earlier this year you published what is essentially an affordability “manifesto” that identifies ways to save a collective $767 billion over the next decade. It covers things like fixing the patent thicket and the 12-year biologic exclusivity, but also non-drug savings like advancing site-neutrality, etc.

What stood out about the piece to me was the candor and the unapologetic declaratives – ‘this is what needs to happen’ – and also the fact that the piece is so tactical. You linked the savings to specific legislation that's either been mothballed or waiting or otherwise deprioritized. How would you decipher the tone for me?

Kim: I think your framing is spot on. We all have “affordability fatigue.” We’ve been talking about how unaffordable health care is forever. It’s time — past time, actually — to get serious. We said to ourselves, fundamentally: Do we care about affordability or not? We do. So, we’re taking positions that drive solutions to the crisis.

We took some great work that had been sitting as draft legislation or on the cutting room floor and hired an economist who used to work at the Congressional Budget Office to crunch the numbers on what impact these ideas could have.

For example, site neutrality. The dialogue has meaningfully changed on this issue, and I think we’re likely to see some form of legislation on it this year. Site neutrality is the single largest source of potential cost savings — up to $471 billion over 10 years. I don’t know that what passes will be our full recommended solution, but getting something passed is still a positive.

To be continued.

This interview was edited by Abby Burns.

Check out the full series

Browse our library of the "Lessons from the C-suite" series, featuring conversations with the most influential leaders in healthcare.

Related resources


  • All healthcare organizations

  • You’ll understand how the head of the Blues views the opportunities – and risks – of generative AI.
  • You’ll learn about the Blues’ approach to tackling high-cost drugs.
  • You’ll gain perspective on how the Blues are leaning into their ‘superconductor’ energy to promote affordability.



Don't miss out on the latest Advisory Board insights

Create your free account to access 1 resource, including the latest research and webinars.

Want access without creating an account?


You have 1 free members-only resource remaining this month.

1 free members-only resources remaining

1 free members-only resources remaining

You've reached your limit of free insights

Become a member to access all of Advisory Board's resources, events, and experts

Never miss out on the latest innovative health care content tailored to you.

Benefits include:

Unlimited access to research and resources
Member-only access to events and trainings
Expert-led consultation and facilitation
The latest content delivered to your inbox

You've reached your limit of free insights

Become a member to access all of Advisory Board's resources, events, and experts

Never miss out on the latest innovative health care content tailored to you.

Benefits include:

Unlimited access to research and resources
Member-only access to events and trainings
Expert-led consultation and facilitation
The latest content delivered to your inbox
Thank you! Your updates have been made successfully.
Oh no! There was a problem with your request.
Error in form submission. Please try again.