Increasing productivity in healthcare: An interview with Allen Karp of Horizon Blue Cross Blue Shield

While this podcast was recorded before the full effects of COVID-19 were known, analytic capability, value-based care and labor are among the pressing topics for healthcare systems and services. Below is an edited transcript between McKinsey’s David Knott and Horizon Blue Cross Blue Shield’s Allen Karp.

David Knott: Good afternoon, this is David Knott. I’m a senior partner in McKinsey’s healthcare practice, and today it is my pleasure to have Allen Karp, executive vice president of healthcare management and transformation at Horizon Blue Cross Blue Shield join me for a conversation around productivity in healthcare. Allen, welcome, and thanks for joining us.

Allen Karp: Thank you, David.

David: Allen, perhaps just to start, I think at the highest level, we have seen the application of technology and other measures that improve the productivity of industries over time. And I think one of the vexing challenges in healthcare is that we continue to throw a lot of technology at various problems, and yet we see precious, if any, improvement in productivity over time.

And we all know that affordable quality is one of America’s greatest challenges. Any top-of-mind observations on why this is such a vexing problem so entrenched for healthcare?

Allen: Yes, and that’s a great question, David. I will focus on two that I think drive the fact that our productivity in healthcare is not as great as some other industries. One is the economic model historically has been a fee-for-service–based model, where incentives are not aligned between payers and providers. And there are significant administrative processes that are in play today as a payer interacts with a provider and that causes inefficiencies.

The other area would be within a health system. For example, the administrative and operational processes are cumbersome and inefficient. If you look at the healthcare industry, it is a very labor-intensive industry which of course drives the cost up and creates some of this inefficiency.

There was a study done by Johns Hopkins, where they were looking at two areas of focus. One was the way they assigned beds to patients who came into the hospital. And the second was how quickly they could assign beds to those who came through the emergency room. Utilizing AI [artificial intelligence] and machine learning, they were able to reduce the amount of time it took to assign beds by 30 percent.

And they were able to cut the wait time for those members who came in through the emergency room or those patients, I should say, through the emergency room, by 20 percent. So that’s one example of using technology to focus on those areas that are cumbersome and improve inefficient processes.

David: It sounds like, one of the most fundamental issues is the alignment of the economic incentives. And so, you’ve talked a little bit about fee-for-service environment, and then the types of operating models that have evolved over time. Are there ways that you see today, steps or progress being made, where we can actually start to make some of these changes?

Allen: There are still barriers that exist today. However, I have seen significant change which I believe will improve productivity significantly as we move forward. As we change the economic model and we align the incentives with hospitals and physicians, the administrative processes that we have in place today, such as pre-authorization or concurrent review, are either reduced or they go away.

Because now by moving to a value-based [model], we’re passing some of the financial risk to those providers along with the tools necessary to manage that risk. And therefore, we can eliminate some of those processes, which will lower cost and provide a much more efficient flow for the patients and for our members.

The other area would be on our side, being a payer using the AI and machine-learning technology to be able to identify those pre-authorizations, for example, that are overturned a significant portion of the time. So, we could teach the machine how to look for those, and then when a provider calls in for a pre-authorization, they would be able to get that pre-authorization on an automated basis.

And that allows our nurses to focus on those clinical conditions that are either in a gray area, or not yet supported by the evidence. And makes the whole process much more efficient, faster, [and] less paperwork for physicians and hospitals.

David: [Perhaps you would] be able to talk a little bit about steps that Horizon Blue Cross Blue Shield is taking, where you’re trying to move forward today to capture some of these improvements?

Allen: Sure. We have built relationships with our hospital systems and providers in New Jersey. We have about 4,500 primary care physicians that are under a value-based reimbursement system. So, we pay them for their ability to provide more affordable costs for our members. And also hit certain quality screens. It always starts with quality.

We then give them the data and analytics capability by providing insights to our providers so that they can take action for those areas that are creating higher costs.

For those services that are provided in the market like radiology, for example, where there may be too much testing. Or another example would be where a patient can go to a freestanding outpatient center for care and not only get a much more efficient and better experience, but also be able to reduce costs because of the cost structure and the processes that are in place in those outpatient centers.

David: It does sound like the direction is really trying to align the delivery system around evidence-based medicine. So, these are clinically appropriate interventions. And the interventions themselves are happening in the appropriate set of care. And make sure that it’s fit for purpose.

When you think about some of the shifts inside of care, does that fit for purpose you mentioned, the outpatient ambulatory infrastructure, exist in as robust a form as is necessary? Or do you see, over coming years, as perhaps a more defined set of services are provided in an inpatient setting, [that] there’s a real emphasis on expanding some of the capacity to accommodate more that can be done in an outpatient setting?

Allen: It’s a journey. And it’s a marathon, not a sprint. Because what we’re asking the providers to do is think about how they manage their patients in a different way.

We have a significant number of primary physicians under [a] value-based [program], the six largest hospital systems in New Jersey, as well as about a thousand physicians who are under an episode-of-care model, which is a total cost-of-care model built specifically for those specialties.

The episode-of-care model would be more focused on a particular group of specialties and using data and the relationships that we have with the specialists to drive towards the best outcomes [and] lowest costs.

For the hospitals, it’s more [about the] total cost-of-care budget. So, “This is what we’re spending for these patients,” and providing the insights for the health system, and particularly the physicians who lead those health system physician groups, to understand where opportunities lie.

David: You know, some in the industry would say that the health systems have the most entrenched interests around the ongoing prevalence of the inpatient setting as a place for care delivery. And that it’s harder to have health systems take the steps that shift more of the activity and care into outpatient settings.

What has your experience been about finding health system partners that see how healthcare is unfolding in America, and what we need to do together to get to affordable quality, and the steps that they can take to evolve their footprint and their mix of services?

Allen: I think the changes in the ecosystem are driving a lot of the interest in changing the model and understanding the value-based care model. For instance, we have competitors, who are acquiring physicians and acquiring outpatient capacity, [and] acquiring [surgical] centers, with the intent of providing more of that care in that setting versus the hospital setting.

So, that has real implications if you think about the existing fee-for-service models that are in place. However, if you move to a value-based care model, it behooves the health systems and the physicians to invest their capital in population health management, in people, processes, and technology, that support that.

The future marketplace, and what’s happening right now, is that the lines between payers and providers are blurred. So, some of the functions that payers perform today, if the marketplace moves towards a value-based economic model, the payers’ strength will be to bring analytic capability, operational capability, technology, and customer relationships to the table.

But the care will be delivered and managed at the physician level, at that point of service, as opposed to in an administrative office. So, I think the changes in the marketplace are driving health system CEOs and physician leaders to understand that they’ve got to make more investment in technology, to be able to support the new models, and the new way care will be delivered than building buildings or adding beds.

David: What I hear you saying is the shift towards value-based payment in the case of hospitals, the global budget allows them to think more about the continuum of care, and then, right care, right place, [and] right time across that continuum. And then, by aligning the care in that way, they’re able to make mission-sustaining margins and be able to continue investing in infrastructure.

Allen: That’s exactly right. I think historically hospitals have been focused, as have physicians, on revenue and not necessarily margin. And the cost for a provider in a value-based model, and the margins they could make in that model, if they perform, is substantially higher than in a clinical model.

Because you don’t have to hire more people. I said earlier that labor is the biggest cost within the healthcare delivery system. You don’t have to fill more beds. You don’t have to build more buildings. So, you don’t have to generate as much clinical revenue to drive the same or better margins than you’re doing today. I think that learning is beginning to catch on.

David: What is your point of view around the future models in healthcare? On the delivery system side, does the system get optimized through vertically integrated delivery systems that go the complete continuum of care and can move patients appropriately across that continuum?

Or do you see mixed models, where the health systems, as an example, would own some portion of that continuum and then other entities may be aggregate physician supply and provide the enabling technology, and then provide more of the integration and coordination of them?

Allen: What I see health systems doing is not employing physicians, but recruiting physicians into their clinically integrated networks, where they’re managing the relationships with the payers [and] the contract providing the services, that a small physician office can’t do on their own. They can’t afford it.

And there are aggregators in the market that are aggregating groups of physicians. The medical home model, where they have large enough contingents of primary care physicians. And I’m starting to see specialists beginning to be organized in the model as well.

So, if I put my health system hat on, I want to make sure that I have enough physician capacity that I have some control over where those services are being delivered. But, more importantly, I think, is helping me manage total cost of care under these new arrangements.

So, for example, if there’s a risk-share arrangement, or a shared-savings arrangement, upside only, it would be a tri-party agreement between the payer, the physician, and the hospital.

That allows them to get the assistance that they need at the physician level. The physicians are the ones who are the drivers of both cost and quality. There’s also private equity money now in the healthcare space where private equity firms are buying interest in physician groups as well.

Would you like to learn more about our Healthcare Systems & Services Practice?

David: When you think about that mixed model, what would be examples of the type of enabling infrastructure, valuated services that Horizon would offer to your provider partners across that continuum of care?

Allen: So, we have built private HIE [health Information exchange], which we call Health Sphere, which allows us to take our claims data, [and] marry it with the clinical data that the providers have. [We] then use our analytic capability to lay on top of that and provide insights to the health systems and the physicians about what’s going on with their patients.

We have implemented other tools that allow physicians, at the point of service, to identify prescription drugs that are generic versus brand cost differentials. And then make a decision, at the point of service, for that patient to provide the most efficient drug for that particular condition.

We also provide a front-end system that runs through our portal that allows our members to be able to connect with the providers and schedule appointments on their phone or on their iPad, much like you would in an OpenTable scenario.

And that requires our physicians to work with us to give us the open spots on their schedule. We also provide telemedicine, where it’s required, and care coordination. So, we can help direct that patient whose child might have an earache at 11:00 p.m. in the evening to the appropriate site of care.

So, things that we do as a payer to help control medical costs, we’re now taking that information and we’re sharing it with our hospital systems and physicians. And then we tie that to the incentive system.

David: With that array of enabling support services, are any of the components offered, on an all-payer basis? Or are each of these proprietary solutions between Horizon and your provider partners?

Allen: Some of the analytic capability would be available to all payers. Our view is if we can lower the cost of care for our patients, and we have the incentive systems and we have the care coordination all built in, we’re still going to get a better result.

So, we offer that to them. There are certain things like the front-end component, which is the consumer experience piece. That is proprietary to Horizon because it’s a service that we’re offering to our members that differentiates us from our competition.

David: When you think about moving to this model and assembling this type of paradigm, how has that changed the skill sets you need specifically within the four walls of Horizon? And what has that meant in terms of the kinds of people you hire, skill sets you’re looking for?

Allen: So, we did have to look at the skill sets of all of our team. So, for example, in the provider contracting area, historically, the way we did business is we would try to strike the best unit cost deal that we can. The providers would do the same thing. There was a lot of disagreement [in] where we should end up.

That skill set is great, and we still need some of that. But in the new model, where we’re still facing off with our health systems and physicians and trying to decide collectively how we provide the best value for our customers required a different skill set, more of a collaborative skill set, a higher level of understanding of how the total cost of care worked.

And the ability to influence and convince these health system partners and physicians why this was in their best interests as well as ours. So, we had to bring in particularly, at the senior level, folks who have had a lot of years of experience in dealing with partnerships [and] in dealing with collaborating with physicians.

And we were able to do that, both on the clinical side and the network and operational side. Understanding the new model, and how the new model worked, what it meant for us as a health plan, and explaining that to the providers, we needed to get some people that had that skill set as well. Because it’s a different ballgame.

David: What sort of results have you seen from this new model?

Allen: We’ve seen tremendous results. We’ve seen reduction in readmission rates. We’ve seen reduction in inpatient utilization [and] reduction in emergency room utilization. We have been in the position, the last several years, where we have been able to create significant savings that allow us to pay out a portion of that savings back to our health system and physician partners.

Which is always positive, because they begin to believe that this model actually works. So, on the clinical side we’ve seen very good results in terms of improving gaps in care, and some of the other operational quality measures that we typically look at.

Overall, the relationship with the health systems and the physicians have given us, a major advantage because we’re now beginning to speak the same language and focus on the same things.

David: This has been rolled out in the employer-sponsored market. And what kind of reception have you received from this? How much do employers like it? What kind of questions do they have, as they see these new propositions in the market?

Allen: So, we had a lot of success in not only the employer market, but in the individual market, and the small-group employer market in particular. We’ve been able to take a product built around the health alliance and the other value-based providers.

And the first year, we were expecting around 250,000 members. We got 275,000. Now we’re up to 450,000. so clearly the market has told us that this is a product that they value. Our price points have been very competitive.

As we go upmarket, it took a couple years for those employers to see the value. But they are now seeing that value. And we are growing our business in those markets for this particular product.

So, we’re able to take these relationships that we’ve built with the provider community and explain in detail how they’re working and why we believe we’ll continue to be able to control costs and improve quality and service as we go forward. Having those relationships in place for a number of years, we could have very focused conversations with employers about the advantages that we bring to the marketplace.

David: How administrative-expense intensive is this new model, compared to the old way of providing broader PPO [preferred provider organization] tech networks?

Allen: So initially there is some administrative expense necessary to build the teams, to be able to face off with the providers, and the technology to support the providers. But over time, it’s much more efficient. We can move some of the functions out to the point of care. And as we do that, we can lower our administrative costs and be able to have those patients handled by care coordinators [or] by others within the delivery system to manage that cost.

David: When you think of how this model evolves over time, I see the shift that you talked about, the blurring of the lines between the payer and provider. Are there risks around [the] potential [that] they feel like they can just take on the insurance risks themselves [and] start their own health plans? Or do you see ways in which the model remains intertwined and interdependent over time?

Allen: There are things that we do around benefit design and ID cards and other ways of communicating with these patients that would be very difficult for a provider to take on.

Just like it would be difficult for us to get into the hospital or health system business, right? So, I think those complementary skill sets and tools that we bring to the table will allow us to continue to collaborate over time.

David: Healthcare doesn’t happen in a vacuum. So, thinking about social determinants of health, the other factors more broadly that play a role in driving healthcare costs. Any thoughts on productivity, greater affordable quality, and how we’d incorporate some of those dimensions as well?

Allen: From a social determinants of health perspective, it’s been very interesting for me to understand how much where you live has to do with your health status versus which health plan you have, or the medical care associated with your condition.

The statistics show that 60 percent of a person’s health status has to do with the ZIP code that they live in, because of things like food deserts, transportation, the pollution levels in those areas, [and] crime. So, we did a pilot in Newark, with Robert Wood Johnson Barnabus Health around one of their hospitals, North Beth Israel.

We looked at four ZIP codes in that area where the expense was extremely high relative to what we normally see, and the age cohorts were much younger, between 35 and 50. We found that the root cause of the problem was that many of them were using the emergency room as their primary care office, as opposed to being tied into the delivery system. They did not know what their conditions were. So many of them have co-morbidities [and] have behavioral health issues.

And so, we employed a model called the University of Pennsylvania Impact Model. It’s a model that engages community health workers who are people who actually live in those communities, who are not healthcare professionals necessarily, but know these people. And that’s the key.

It’s one thing to identify them. It’s another thing to actually get in the door to be able to make an impact on their lives. And so, we hired these people [and] we trained these people.

The other benefit was, we were fully employing some people that may have not been previously employed. But we put them through training. We then hired a person called a personal care [assistant—a] nurse, generally, or a care manager who was doing additional work with those members directly to try to help them navigate through a very complex healthcare system and was providing oversight to the community health workers.

The bottom line was that our engagement rate was [over] 45 percent. Which if you look at the statistics is significantly higher. We identified a thousand members that we can reach. We ended up reaching out to 450. We found interesting reasons for why their costs were so high.

So, for example, there’s a single mom, with a couple [of] kids, who has multiple co-morbidities, [is] taking care of an elderly parent, [and who] doesn’t have a primary care physician because they can’t find one that they trust. And [she] also has depression, undiagnosed.

We were able to intervene, connect this person to a primary care physician in the delivery system, [and] connect them to a behavioral health professional. We tracked that person after six months and found that they had no emergency room visits. They were not hospitalized. Their condition was much better in terms of their health status. And that was a tremendous win for us.

We had a 20 percent reduction in our emergency room costs, and a 25 percent reduction in inpatient visits, and a significant reduction in readmissions.

David: That’s really impressive. And when you think about the insights here primarily the examples you gave here were underserved segments [and] fragile populations. Do you see this having relevance as well in other markets—commercial market, employer sponsored, [or] individual?

Allen: Yes David. This was actually a commercial group of members. But here’s an interesting fact. The disease states for commercial members who live in these ZIP codes were almost identical to the disease states of the Medicaid members.

So, COPD [chronic obstructive pulmonary disease], obesity, hypertension, [and] those types of conditions were identical. Which proves the theory that where you live has much more to do with your health. And we believe now that we can apply these same models to other parts of the State of New Jersey.

David: Any advice for listeners? Lessons you’ve learned?

Allen: When you make a systemic change, in an industry like healthcare, and particularly when you’re changing the economic model, it takes time. It takes preparation. It takes work. And it takes a lot of collaboration.

So, establishing the trust with whoever the partner is, [is] critical in being successful. I also think what I’ve learned is understand the barriers that the party that you’re dealing with sees. Don’t assume that they see things the way you do.

And understand that so that you can come up with solutions to get over those barriers. And the last thing I would say, particularly about the healthcare system, is think about administrative simplification. Because we are in an extremely complicated industry to understand, whether you’re a customer or a member. And so, the easier we can make it for our customers and members to navigate through the system, the better.

David: Allen, thank you again for joining us today. This has been a super interesting conversation.

Explore a career with us