Highlights from an interview with Dr. Steven Udvarhelyi, President and CEO of Blue Cross and Blue Shield of Louisiana.

‘Fail fast’ isn’t just a mantra for high-tech companies. It’s how innovative health plans are driving better results, too. And that takes a culture that embraces analytics as a way to make better decisions.

Dr. Steven Udvarhelyi, President and CEO of Blue Cross and Blue Shield of Louisiana, shared his insights into building a culture of analytics with Eric Grossman, CEO of NextHealth Technologies at the 2018 Health Evolution Summit. Here are the 10 lessons he’s applying as he leads his organization to top results.


Lesson #1: Teach people to ask the right questions

In a lot of planning, you ask your team to build a financial plan without requiring an analytics plan for every initiative. Yet, it’s important to build a culture of analytics where people ask, “how do you know it’s going to work?”

People need to think about collecting data and information to make a good decision from the start. The evolution is to get people to stop saying, “I need a database or report.” Instead, we want them to come to a center of excellence to say: “I have a business problem; I need to understand how to build a business plan; I need to reach a population of members; I want to optimize a marketing campaign.” Very often, this means you have to change operations to collect information that’s not being captured today.


Lesson #2: Bake analytics right into the business plan and operations

At Blue Cross and Blue Shield of Louisiana, we won’t fund any work without an analytics plan. And, we’re doing that enterprise-wide, so that we’re not working in a siloed fashion. We’re working to try new things and fail fast, at scale. For example, we staffed a care management area with an analytics team recently. Now, they can test within eight weeks whether something is working or not. Then, they can decide whether to scale. Organizationally, it’s a way to have a high level of confidence that you’ll know what’s going to work and why. It’s a way to see what’s going to make a breakthrough in results.


Lesson #3: Get to one set of numbers for the truth

When working with analytics, the way to avoid tension between an analytics team and actuaries is to make everyone a part of the effort. There’s one set of numbers that we all use. How it works is that the analytics team collaborates with actuaries to estimate the impact of a program—such as determining the impact of a vendor that we’re evaluating. That’s how we eliminate churning about whose numbers are right.


Lesson #4: Be open to using non-traditional data sources

We’re finding that you have to look at non-traditional data sources. In Louisiana, we have a lot of people on Medicaid. It turns out that the number one variable that helps predict adverse health events is not in the claims system. It’s the credit score.

The FICO credit score is a powerful predictor of health outcomes. It lets you know who is vulnerable financially. When you’re vulnerable financially, as hourly employees often are, you tend not to have time to take off of work to visit a doctor. You can’t afford the loss of wages. You also tend to struggle with out-of-pocket expenses. Cost becomes a big factor in deciding to see a doctor. So, we experimented with eliminating cost-sharing to see if it changed outcomes, and it did. Now, we’re seeing who can benefit from relaxed cost-sharing where it optimizes health outcomes – for example, with members who have chronic diseases.


Lesson #5: Analytics is a constant journey

Our view is that we’ll never be advanced. The state of the art is evolving too rapidly. Analytics is a continual race. The data and analytics platform and our infrastructure are the largest capital investments that we’re making as a company. We completely restructured the organization around it. Since it’s the second time I’ve built a culture of analytics, we’re moving three times faster than the last time.


Lesson #6: Traditional reimbursement relationships may not matter as much as you think

To look at impacting payer and physician relationships, we partnered with a large health system to create a laboratory to test reimbursement arrangements. The interesting thing we found is that the reimbursement relationship between the payer / provider entity may not be what matters. It’s how the provider entity pays their individual clinicians that matters the most. So, now we’re looking at how we engage differently and downstream savings to the individual clinicians.

The other interesting thing is that the provider entities kept asking us for claims data. Thanks to our predictive models, we changed what data and tools we’re both using. Now, we’re working to get information into the actual workflow to reach the clinician directly.


Lesson #7: (Almost) everyone is in analytics

From a governance stand point, we don’t focus on individual initiatives. We focus on shared roles and responsibilities around data collection. A simple way to look at it is this: If your job is sourcing data, you’re moving into analytics. If your job is programming to allow us to access data (unless you’re the actuary), your job is moving into analytics. If your job is to maintain a repository of data, then your job is moving into analytics. There is no place to go in the organization now unless you’re tied to analytics.


Lesson #8: Make it impossible for the status quo to continue

Our lesson learned early on was that you cannot get an organization to change unless you eliminate the ability for the status quo to continue. Our first step was to eliminate the status quo. No one can go to the IT organization and say, “I want a database.” They can’t go to a vendor and ask for a new data source. It has to come through the Chief Analytics Officer so that we can support business needs and turn off the process of every business area creating their own arrangements.


Lesson #9: Simple changes can impact healthcare the most

In practice, change can be simple. For example, let’s look at open enrollment. Why not change open enrollment to ask the right three questions that end up being a better predictor of who’s going to have problems than any claims data?

If you’re going to issue an ID card, why not ask: Do you live alone? Do you have someone who goes to doctor appointments with you? Can you get to the doctor and pharmacy when you need to? Those three questions will end up predicting who is going to have problems far more than any claims data. But we don’t design our open enrollment process as a data-collection instrument. Well, we are now. And it’s not that much harder to ask three more questions. It’s easiest to test and scale this kind of change. We need the feedback to get the results we want.


Lesson #10: How do you accelerate the pace of change?

We saw that everyone wanted to change. The company, the providers, and the brokers all wanted to change. Maybe not the same change, but change was a goal. If the change is coming from the top down, it’s easier to get the company to buy in. There was some resistance to how we implemented change. But we created early wins by testing people’s business hunches to see what worked. It helped build trust. People could see that the results were better.


Steven Udvarhelyi, CEO, Blue Cross & Blue Shield Louisiana

Dr. Steven Udvarhelyi joined Blue Cross and Blue Shield of Louisiana as president and chief executive officer in 2016. He is a board-certified internist and has more than 25 years of experience in the health insurance industry.

Prior to joining Blue Cross and Blue Shield of Louisiana, Dr. Udvarhelyi was with Independence Blue Cross (IBC) in Philadelphia for almost 20 years, most recently serving as executive vice president, health services and chief strategy officer. Before IBC, he worked for Prudential Health Care in a variety of roles, including vice president, operations for Florida and vice president, medical services.

Dr. Udvarhelyi received an A.B. degree from Harvard College, an M.D. degree from the Johns Hopkins University School of Medicine and a Master of Science degree in Health Services Administration from the Harvard School of Public Health. Prior to his career in the managed care and health insurance industry, Dr. Udvarhelyi was a faculty member at Harvard Medical School with a focus on health services research.

Dr. Udvarhelyi currently serves on the National Board of Trustees for the Devereux Foundation. He has also served on the Board of Directors of NaviNet, which he chaired, the Board of Trustees of the Franklin Institute in Philadelphia, the Board of Managers for Tandigm Health,  the Board of Directors of NCQA, the Board of America’s Health Insurance Plans, the Institute of Medicine (IOM) Roundtable on Evidence-based Medicine and the IOM Committee on Comparative Effectiveness Research Priorities.