Wednesday, November 06, 2013

Efficient Big Data Capabilities Help Cerner Drive Needed Improvements into Healthcare Outcomes

Transcript of a Briefings Direct podcast on how a large provider of healthcare services is providing insight into patient outcomes as well as EMR system performance.

Listen to the podcast. Find it on iTunes. Download the transcript. Sponsor: HP.

Dana Gardner: Hello, and welcome to the next edition of the HP Discover Podcast Series. I'm Dana Gardner, Principal Analyst at Interarbor Solutions, your moderator for this ongoing discussion of IT innovation and how it’s making an impact on people’s lives.

Gardner
Once again, we’re focusing on how IT leaders are improving their business performance for better access, use and analysis of their data and information. This time we’re coming to you directly from the recent HP Vertica Big Data Conference in Boston.

Our next innovation case study highlights how a healthcare solutions provider leverages big-data capabilities. We’ll see how they deployed the HP Vertica Analytics platform to help their customers better understand population healthcare trends, as well as to help them run their own systems internally.

To learn more about how high performing and cost effective big data processing forms a foundational element to improving healthcare quality and efficiency, please join me now in welcoming our guest, Dan Woicke, Director of Enterprise Systems Management at Cerner Corp. based in Kansas City, Missouri. Welcome, Dan. [Disclosure: HP is a sponsor of BriefingsDirect podcasts.]

Dan Woicke: First of all, thank you very much for having me. It’s my first time in Boston. So I'm having a blast here.

Gardner: Terrific. Let’s start at a high level and talk a little bit about why, in the healthcare industry in particular, big data is super important. We're going through some major transitions in how payments are going to be made and how even the definition of good care is defined. We're moving from pay for procedures to more pay for outcomes. So tell me about Cerner, and why big data is such a big deal.

Woicke: Obviously, you hit the nail on the head. The key element here is that the payment structure is changing to more of an outcome model. In order for that to happen, we need to get all the sources of data from many, many disparate systems, bring them in, and let our analysts work on what the right trends are and predict quality outcomes, so that you can repeat those and stay profitable in the new system.

Gardner: It’s interesting that, on one side of the coin, you're looking to bring large data sets together to analyze what’s going on in the field, but in order to allow you to better serve those needs, you also have big IT systems. They're putting out a lot of data, and you need to analyze them. Tell us a little bit about two ways in which big data is being employed there at Cerner.

Woicke: We’ll touch on more of the clinical side of how we are processing this data in the new model. My direct responsibility is to bring in massive amounts of performance data. This is how our Cerner Millennium systems are running.

We have hundreds of clients, both in the data center and those that manage their own systems with their own database administrators (DBAs). The challenge is just to have a huge system like that running with tens of thousands of clinicians on the system.

We need to make sure that we have the right data in place in order to measure how systems are running and then be able to predict how those systems will run in the future. If things are happening that might be going negative, how can we take the massive amounts of data that are coming into our new analytical platform, correlate those parameters, predict what’s going to happen, and then take action before there is a negative?

Effect change

We want to be able to predict what’s happening, so that we can effect change before there is a negative impact on the system.

Gardner: Everybody, almost across any business you talk to, wants to be more proactive and get out in front of these issues. Tell me how big data and the ability to manage big data gets you closer to the real time and then, ultimately, proactive.

Woicke: Since January of this year, we've started to bring in what we call Response Time Measurement System (RTMS) records. For example, when a doctor or a nurse is in our electronic medical record (EMR) system is signing an order, I can tell you how long it took to log into the system. I can tell you how long you were in the charting module.

Woicke
All those transactions produce 10 billion timers, per month, across all of our clients. We bring those all into our HP Vertica Data Warehouse. Right now, it’s about a two-hour response time, but my goal, within the next 12 months, is to get it down to 10 minutes.

I can see in real time when trends are happening, either positive or negative, and be able to take action before there is an issue.

Gardner: That’s impressive. Tell us a little bit about Cerner and describe the company -- what they do, and this idea that you have not just your own systems, but you're managing systems that other people use as well.

Woicke: We have two data centers in Kansas City, Missouri and we host more than half for our clients in those data centers. The trend is moving toward being remote-hosted managed like that. We still have a couple of hundred clients that are managing their own Millennium domains. As I said before, we need to make sure that we provide the same quality of service to both those sets of clients.

Gardner: So you're used primarily by healthcare organizations. Tell us how you actually function within healthcare and the services that you provide to these organizations.

Woicke: We run the largest EMR in the world. We have well over 400 domains to manage  -- we call them domains -- which allows us to hook up multiple facilities to those domains. Once we have multiple facilities connecting into those domains, at any given time, there are tens of thousands clinicians  on the system at one time.

Gardner: I'm still trying to tease out a little bit of more understanding of the function that you provide to these health providers. Are you doing their medical records inventory for them or do you have a set of applications in addition to that? Help us understand better what services you provide.

Single database

Woicke: Cerner Millennium is a suite of products or solutions. Millennium is a platform where the EMR is placed into a single database. Then, we have about 55 different solutions that go on top of that platform, starting with ambulatory solutions. This year was really neat. We were able to launch our first ambulatory iPad application.

There are about 55 different solutions, and it's growing all the time with surgery and lab that fit into the Cerner Millennium system. So we do have a cohesive set of data all within one database, which makes us unique.

Gardner: Before we go to some more insights about the healthcare industry, population health, and some of the great analytics that can be brought there, let’s drill down a little bit into what you're doing on site. Where does the data come from primarily? Is this log information. Do you have a set of management systems of your own, and how much data we are talking about?

Woicke: We're talking about quite a bit of data, and that’s why we had to transform something away from a traditional OLTP database into an MPP type database, because those systems that are now sending data to Cerner. 

We have claims data, and HL7 messages. We're going to get all our continuous care records from Millenium. We have other EMRs. So that’s pretty much the first time that we're bringing in other EMR records.
What that's going to do is bring the total cost of your healthcare down, which is really the goal.

We have health-plan enrollments, and then of course, within Millennium, we're going to drill down into outcomes, re-admissions, diagnosis, and allergies. That’s the data that we need to be able to predict what kind of care we are going to have in the future.

Gardner: Now, you're also looking to how you can better understand the marketplace and provide insights, so that people can literally change on a dime, change the wings on the airplane while it’s still in the air, if you will, in healthcare and population health. What are the insights that you can get there and what are the data sets that you need in order to do that?

Woicke: The data sets are similar to what we just discussed. You’ll have that claim data that comes in from multiple sources, multiple EMRs, but the whole goal of population health is to get a population to manage their own health. That means that we need to give them the tools in their hands. And they need to be accurate, so that they can make the right decisions in the future. What that's going to do is bring the total cost of your healthcare down, which is really the goal.

Gardner: So it seems to me that we talk about "Internet of things." We're also going to the "Internet of people." More information from them about their health comes back and benefits you and benefits the healthcare providers. But ultimately, they can also provide great insights to the patients themselves.

Do you see, in the not too distant future, applications where certain data -- well-protected and governed of course -- is made into services and insights that allow for a better proactive approach to health.

Proactive approach

Woicke: Without a doubt. We're actually endorsing this internally within the company by launching our own weight-loss challenges, where we're taking our medical records and putting them on the web, so that we have access to them from home.

I can go on the site right now and manage my own health. I can track the number of steps I'm doing. Those are the types of tools that we need to launch to the population, so that they endorse that good behavior, which will ultimately change their quality of life.

Gardner: Then, there is also this notion of anonymized patient information, where you can take an aggregate and find out what works and what doesn’t work when it comes to behavior, patterns of fruition when it comes to things like weight loss. Tell me how that grander view, the holistic view of the data, comes to bear as well. 

Woicke: Right now, we're in production with the operation side that we talked about a little bit about earlier. Then, we are in production with what we call Health Facts, a huge set of blinded data. We hire a team of analysts and scientists to go through this data and look for trends.
You can see what that’s going to do for the speed of the amount of analysis we could do on the same amount of data. It’s game changing.

It’s something we haven’t been able to do until recently, until we got HP Vertica. I am going to give you a good example. We had analysts log a SQL query to do an exploratory type of analysis on the data. They would log that at 5 p.m., then issue it, and hopefully, by the time they came back at 8 a.m. the next day, that query would be done.

In Vertica, we've timed those queries at between two and five seconds. So you can see what that’s going to do for the speed of the amount of analysis we could do on the same amount of data. It’s game changing.

Gardner: Let me ask you, Dan, about that process through which you acquired Vertica. How did you adopt it? What were some of the requirements, and why didn’t some of the other alternatives work out?

Woicke: There were a lot of competitors that would have worked out, but we had a set of criteria that we drilled down on. We were trying to make it as scientific as possible and very, very thorough. So we built a score sheet, and each of us from the operation side and Health Facts side graded and weighted each of those categories that we were going to judge during the proof of concept (POC). We ended up doing six POCs.

We got down to two, and it was a hard choice. But with the throughput that we got from Vertica, their performance, and the number of simultaneous users on the system at a given period of time, it was the right choice for us.

Gardner: And because we're talking about healthcare, costs are super important. Was there a return on investment (ROI) or cost benefit involved as well?

Extremely competitive

Woicke: Absolutely. You could imagine that this would be the one or two top categories weighted on our score sheet, but certainly HP Vertica is extremely competitive, compared to some of the others that we looked at.

Gardner: Dan, looking to the future, what do you expect your requirements to be, say, two years from now? Is there a trajectory that you need to take as an organization, and how does that compare to where you see Vertica going?

Woicke: Having Vertica as a partner, we navigate that together. They invited me here to Boston to sit on the user board. It was really neat to sit right there with [HP Vertica General Manager] Colin Mahony at the same table and be able to say, "This is what we need. These are our needs coming around the corner," and have him listen and be able to take action on that. That was pretty impressive.

To answer your question though, it’s more and more data. I was describing the operations side, where we bring in 10 billion RTMS records. There's going to be another 10 billion type of records coming in from other sources, CPU, Memory, Disk I/O, everything can be measured.

We want to bring it into Vertica, because I'm going to be able to do some correlation against something we were talking about. If I know that the RTMS records show a negative performance that's going to happen within the next 10-15 minutes, I can figure out which one of those operational parameters is most affecting that outcome of that performance, and then can send the analyst directly in to mitigate that problem.
By bringing in more and more data and being able to correlate it, we're going to show all the clients, as well as the providers, how their system is doing.

On the EMR side, it’s more data as well. On the operations side, we're going to apply this to other enterprises to bring in more data to connect to the experts. So there is always somebody out there. That’s the expert. What we're going to do is connect the provider with the payers and the patient to complete that triangle in population health. That’s where we're going in the next few months.

Gardner: I certainly think that managing data effectively is a huge component of our healthcare challenge here in the United States, and of course, you're operating in about 19 countries. So this is something that will be a benefit to almost any market where efficiency, productivity, quality of care come to bear.

Woicke: At Cerner Corp., we're really big on transparency. We have a system right now called the Lights On Network, where we are taking these parameters and bringing them into a website. We show everything to the client, how they're performing and how the system is doing. By bringing in more and more data and being able to correlate it, we're going to show all the clients, as well as the providers, how their system is doing.

Gardner: Well, great. I'm afraid we’ll have to leave it there. We've been learning about how a healthcare solutions provider has been leveraging big-data capabilities, and we've seen how at Cerner Corp. they've deployed HP Vertica Analytics Platform to help their customers better understand population health trends, as well as to gain terrific insights into their own systems and the systems that they host for others.

So, a big thank you to our guest, Dan Woicke, Director of Enterprise Systems Management at Cerner Corp. Thanks so much, Dan.

Woicke: Thank you for having me.

Gardner: And thank you also to our audience for joining this special HP Discover podcast coming to you directly from the recent HP Vertica Big Data Conference in Boston.

I'm Dana Gardner; Principal Analyst at Interarbor Solutions, your host for this ongoing series of HP sponsored discussions. Thanks again for joining, and don’t forget to come back next time.

Listen to the podcast. Find it on iTunes. Download the transcript. Sponsor: HP.

Transcript of a Briefings Direct podcast on how a large provider of healthcare services is providing insight into patient outcomes as well as EMR system performance. Copyright Interarbor Solutions, LLC, 2005-2013. All rights reserved.

You may also be interested in:

No comments:

Post a Comment