What Role Does Data Play in Fighting Healthcare Fraud, Waste and Abuse?

Transcript

[Tyler Daniels] How can data help fight healthcare fraud, waste and abuse? I'm Tyler Daniels and welcome to the HHS IG's podcast. We'll start with an example of healthcare fraud: Dr. David Pon. Dr. Pon, of Florida, intentionally misdiagnosed more than 500 patients as suffering from a degenerative and incurable disease, known as wet macular degeneration. Pon then used his false diagnoses to bill the government for unnecessary diagnostic testing and unwarranted treatments.

[Caryl Brzymialkiewicz] So the scary thing is, in connection with that unnecessary testing, he was injecting victims with various dyes, that also posed a potential serious health risk, including cardiac arrest.

[Tyler Daniels] That's Caryl Brzymialkiewicz, the Chief Data Officer for the HHS Office of Inspector General.

[Caryl Brzymialkiewicz] He was just convicted on all 20 counts, it's one of our best examples from our team looking at our peer comparison analysis that involved patient care.

[Tyler Daniels] I spoke with Caryl about the role data plays in investigations, like the Dr. Pon case, as well as the larger role data plays in combating fraud, waste and abuse in the healthcare system.

[Tyler Daniels] Caryl, do you want to tell us, first of all, about the mission of your office?

[Caryl Brzymialkiewicz] Sure, so our office is responsible for providing more and better access to data and analytics to support OIG's mission. So we have three functions within our office. We have advanced analytics, which if you think about the data scientists of the world, data analysts, that's that team. We have a data operations team, which is the nitty-gritty, behind the scenes data governance, data quality. And we have strategic planning and organizational performance management, which helps with thinking about strategic plans for the organization.

[Tyler Daniels] What does actionable advanced analytics mean?

[Caryl Brzymialkiewicz] Great question. So what it really means is having high quality lead-generation for either our investigators, our auditors, our evaluators or for compliance oversight. One of two things can happen with our advanced analytics. Either the data can lead us to somebody that is potentially committing fraudulent activity or our investigators can have a hotline call where they can have a witness or a whistleblower come tell them that they suspect criminal activities happening, and we can bounce that against the data. So it's a really a combination of the data analytics and the data scientists and our statisticians and computer programmers with that field intelligence of our law enforcement agents working in the field- that combination is very powerful.

[Tyler Daniels] How long as the office of the Chief Data Officer been around and how has it changed recently?

[Caryl Brzymialkiewicz] We officially became into existence last year. So we've been around about a year. The advanced analytic team has been around for several years. What we're trying to do is accelerate the work of that team. So some of the things that the Chief Data Office is doing now is really focusing on, how do we help the OIG become even more effective and efficient in what it's doing- which includes improving our access to data. So to not only support our investigators and their cases but also with audits and evaluations. As they're working on issues, what other datasets do they need? How do we make sure that the data we have is high quality data? For example, right now Medicaid data throughout the country, having a national dataset, that's a huge issue. And so we're working closely with the centers of Medicaid and Medicare to understand, as they're implementing their new system, what does that mean, how can we potentially tap into that environment. In the meantime, we're still working on access to other systems that we know we still need for some of that work. Our office is the advocate for that, and the proponent for that.

In the predictive analytics space, we do a lot of modeling and generate risk scores. And so what that means is based on various statistical methods behind the scenes, what are those parameters that suggest we have somebody that, either a pharmacy or provider, that is of a high risk, "high risk". And they can be high, medium or low. High risk is basically flags for people. We might want to apply a little bit more scrutiny and take a closer look and understand what's going on in that case. So we're trying to accelerate our predictive analytic work. We're also looking at news ways we can visualize our data. So a lot of people, I joke that as children we always wanted to know do we have to go to school tomorrow. And so we were really good at looking at the weather map. So if I were to show you an excel spreadsheet with a bunch of numbers on it, it may or may not make sense to you. I joke about speaking 'chart'. But if I plot it on a map and show you hotspots of the country of where I think fraudulent activity is happening or where we have an issue that might warrant an audit or an evaluation that's an issue we should consider, putting it in a geospatial manner, putting it on a map is another way of just visualizing and exploring the data and highlighting things. People see things-see information in different ways. So we're responsible. We're trying to drive, what are the new capabilities and tools we want to bring into the organization to help enable that even more.

[Tyler Daniels] We've heard all about the work you do externally, but you also do internal work as well for the organization. What is that?

[Caryl Brzymialkiewicz] So we're also working to support senior leadership, to help inform their decisions, to facilitate the right conversations about where we're allocating resources, to help us position ourselves well as we're competing for increasingly scarce resources across the entire government. How do we give our executives the visibility on everything wonderful that's happening throughout the organization so that they can align to priorities and align the right skillsets and people with those priorities and then if we need additional resources, we're standing on some very solid ground in terms of our logic of what we need when we go back and ask people for additional money.

[Tyler Daniels] What can we expect for the future, where are we going?

[Caryl Brzymialkiewicz] Our office is working very heavily on investing in additional tools for our office to really focus our efforts against fraud, waste and abuse, even more. We're trying to get, we're trying to democratize data, you'll hear that term a lot with the big data. So rather than having the whole organization having to understand any particular programing language, some of our analysts are really thinking, very cleverly, about how to create tools to put capabilities in our organization's hands. Tools and products at OIG's fingertips is what we talk about. So we've got the peer comparison generator. So as I mentioned the two examples up front, we're looking at ways, how do you really spot those doctors that are outliers. How do you make that really easy for the organization? If you're trying to understand trends, what's happening, are the trends lines going up or down for payments in a specific area. So we have a lot of work looking at the costs associated with ambulance providers and figuring out if they were actually going to where they should be going. And so how do you take-you know you've taken some action-how do you see whether that money-is the overall expenditures going down where you look at it from a trend tool. We're really thinking about link analysis so as well...

[Tyler Daniels] What's that?

[Caryl Brzymialkiewicz] So link analysis is thinking about the connectedness between physicians and providers. So if we're looking at risk scores, potentially of high risk pharmacies and all of a sudden, they're connected to each other-does that provide you additional insight? How are people connected? What I'm really thinking about as well, now that we're going from fee-for-service to value-based care, inherently there are connections between organizations. A lot of them are very good. So when we're trying to find the people that are potentially committing fraud, waste and abuse, how do we need to think about our data in different way-or how do we need to bring a different approach to that problem to see and understand where we might need to look, even further.

[Tyler Daniels] Caryl Brzymialkiewicz, thank you very much for your time.

[Caryl Brzymialkiewicz]: Thank you very much, appreciate the opportunity to talk to you today.

[Tyler Daniels] Caryl is the Chief Data Officer for the HHS Office of Inspector General. I'm Tyler Daniels and thanks for listening.