Believe in the Numbers

Analytics in a nutshell: Lots of data, crunched a lot.

By now, pretty much everyone knows that the term “information technology” has something to do with computers gathering and analyzing data, or automating processes that had been manual. How about the term “analytics?” It’s new, it’s sexy, but what does it mean? More important, how much do contracting professionals need to know about it?

If JHC readers think of the difference between the first production Boeing 707 and the upcoming 787 Dreamliner, they may have an idea of how analytics compares to traditional information technology. It’s a question of scale (in the case of analytics, more data being crunched than ever before), speed, and depth of knowledge that can be gleaned. Even better, because of the Web and various shared services organizations, analytics is coming to an IDN near you.

Dan DeLay understands this more than a lot of people. He should. After all, he’s senior vice president of the supply chain analytics division for VHA in Irving, Texas. What’s more, he can explain it better than a lot of people too, perhaps because he was raised in Iowa and schooled in Chicago, the city of plain speakers.

It was at the University of Chicago 25 years ago that DeLay got turned on to the power of information. A public policy major, he studied plenty of economics at the school. “I was introduced to the Chicago School of Economics, which [envisions] a clear-eyed view of the world, based on data, statistics and information, as opposed to the way we wish the world would be, or our biases,” he says. “Based on information, you can make decisions about what’s really best for business, social policy, etc.”

After spending a few years in the Army, DeLay went to law school at Loyola University in Chicago. Upon graduation, he worked for a firm that got involved in a fair amount of healthcare litigation. DeLay often found himself representing plaintiffs who were suing insurance carriers to pay for what the latter considered to be experimental procedures. In the course of that work, he saw that data was extremely important to the outcomes of such cases.

After seven years, he decided to leave law, at least for awhile, and pursue a master’s degree in business administration. He earned his MBA – with a focus on finance and health services administration – from Northwestern University in north suburban Chicago. From there, he went to work for Computer Sciences Corp., a consulting firm with a heavy emphasis on information technology. For CSC, DeLay did some work in the area of Medicare and Medicaid fraud, with an emphasis on examining data and identifying patterns that might indicate the likelihood of fraud. He also found himself working on ERP (enterprise resource planning) installations at healthcare facilities.

At CSC, he worked with VHA, University HealthSystem Consortium and Novation as they examined their relationship with Neoforma. (In October 2005, the health systems chose to sell Neoforma to GHX.) At the close of that project, he threw his hat in the ring for the newly created supply chain analytics post, which he was awarded in March 2006.

Though the post was new, VHA’s interest in analytics was not. Prior to 2006, VHA and Novation had developed tools to help their members analyze their spending. When DeLay joined VHA, he was charged with developing a prototype of an online tool called Global Spend Analysis. The first version was released (“happily, on time”) in August 2006. Today, VHA offers its members a variety of tools to help them analyze their spending; benchmark their spending, product usage and outcomes with other facilities; and maximize contract utilization.

Journal of Healthcare Contracting: What is analytics and how does it differ from traditional data collection and analysis?

DeLay: The thing that distinguishes analytics from regular business analysis is that it involves massive amounts of data and sophisticated technology. We’re talking about terabytes of data, which are being crunched at high speeds by pretty sophisticated technology. What that allows is some really interesting insight into business questions.

Although analytics is new to healthcare providers, we’ve seen its use in other industries for years. It started out in the financial community, then migrated to other businesses. A great example is the airline industry, where analytics is used for scheduling and rate management.

People ask, “Why is a GPO involved in analytics?” The way I explain it is this: The GPO traditionally operated by pooling spend from its members, then going to the market to get a better deal. It’s the same principle with analytics. We pool our members’ data, then apply very sophisticated technology to its analysis. Then we push that information out to our members. Most hospitals – I don’t care how big they are – don’t have the budget to acquire, install and maintain the type of technology required to do this. In 2007, we processed $22 or $23 billion of spend, which gives us tremendous insight into what’s going on in the market.

JHC: To what extent does analytics replace intuition in the healthcare contracting process?

DeLay: It does replace intuition, especially in areas like supply chain management, which is so data-driven. One thing should be pointed out: The reason people use intuition in heavy-data-laden environments is that they haven’t been able to manage the information; they haven’t had the tools. Contracting professionals know about all the different factors that can affect contracts. However, until now, they haven’t had the staff or the technology to do the kind of analysis they want.

A good example is the use of vendor scorecards to figure out [an IDN’s] best suppliers and why. You need to marry that up with multiple contract choices, then do the financials around that. Then you need to look at substitutions and market trends, such as, “How are the physicians’ practice patterns changing?” This is not a trivial task. While a lot of information is out there today, the problem has been harnessing it and making sense of it. We’re not there by any means. We’re at the dawn of this age.

JHC: It has been said that when ranking potential recruits, the New England Patriots [football team] consider such intangibles as intelligence, commitment, coachability and a willingness to subordinate individual ego to the goals of the team. To what extent can intuition be incorporated into analytics, particularly in healthcare contracting?

DeLay: Intangibles can be turned into numbers. Using the football example, a great coach can look at how a player plays on the field, performs on the sidelines, etc., and rate him from 1 to 10 in terms of, “This guy is a team player.” A great contracting professional can look at fill rates, responsiveness, and things like that, and then assign a numerical value to a particular vendor. Over time, [he or she] can compare the results of their decisions with the model. One of the great strengths and beauties of analytics is that it allows you to run experiments. I can rate a vendor on responsiveness, then, over time, check the reliability of my model. If it’s off, I can tweak it.

JHC: Can you talk about the traditional parameters, or data points, that contracting professionals have used in their analyses, and then talk about some of the newer ones, which may yield even more usable information?

DeLay: It’s not so much that there are new parameters, but that we now have the ability to process the parameters we had before. An example is clinical outcomes. Today, we talk a lot about this product vs. that one, and its reliability and effectiveness. But how many hospitals have the ability to bring those factors to bear in a contracting analysis? Another factor – particularly if I’m a big health system – might be, “How much of a particular vendor’s business do I represent?” There are all these factors that sophisticated supply chain management professionals have known about for a long time, but they have looked at them on an isolated basis. Now we’re talking about bringing them together in a unified model. So it’s not so much new inputs as outputs. Some non-intuitive outcomes might emerge from this analysis. That’s what we keep our eyes out for.

JHC: How do today’s information systems differ from those of, say, 10 years ago, in terms of helping providers take advantage of analytics?

DeLay: Today’s ERP systems were designed to be transactional, not analytical. It’s still very, very difficult for contracting and purchasing departments to be able to tease the information they need out of these systems. The other reality is, if you look at health systems made up of multiple facilities, quite often you’ll find that [each facility] has a different information system or different version. So everything turns into a consulting engagement when you try to find out what’s going on. One of the great things about our program is that instead of managing the supply chain by initiative, you manage it by data and information – which is ongoing and repeatable, and which doesn’t require consulting engagements.

JHC: Explain what you mean when you say “managing by data instead of by initiative.”

DeLay: What often occurs is that a hospital might do an initiative around cardiology or orthopedics. They call in the consultants, who … gather data for a few months, analyze it, then go to the physicians and try to achieve a specific cost-reduction result. That process is often a worthwhile thing. The problem is, what happens next week? Without consistently and constantly managing it, the benefits quite often evaporate over the course of two or three years. So when people ask, “What’s the difference between the healthcare supply chain and that of other industries?” the answer is, healthcare typically hasn’t had the availability and usability of information over time.

JHC: What kind of internal structure does the IDN need to transform itself into an analytics-driven organization? Must it hire full-timers to participate or drive the process, or can that function be delegated to existing employees?

DeLay: For a long time, folks in the supply chain field have complained that [supply chain] doesn’t receive more recognition from the powers-that-be in their hospitals. “Where’s my board representation?” “Why aren’t we talking about these issues at executive committee meetings?”

I think that those complaints have been legitimate. However, without bringing something to the table that people can get involved in, it’s difficult to get traction, especially when the supply chain isn’t your core business [which it isn’t for hospitals, whose core business is clinical care]. This analytics capability is scientific in nature, so it’s in line with the training of a lot of people who have influence over these institutions. This is a great opportunity for materials management and contracting [professionals] to take on larger and more visible roles. There’s nothing like data and numbers to get somebody’s attention. As that proceeds, it will become clear and obvious to senior leadership that there’s “gold in them thar hills.” Given the size of the expense associated with supply chain management. I think it’s an iterative process, meaning that as more opportunity is identified, the more traction you’re likely to get.

JHC: What kind of people are needed in order to bridge the gap between collecting and analyzing data, and then discussing its implications with decision-makers (such as administrators, contracting professionals, clinicians)?

DeLay: It’s hard to say “You need to do X to be successful.” I think the primary ingredient is an attitude and a belief in the validity of the process. I went to a presentation by one of the MIT guys [who were the subjects of the 2004 book Bringing Down the House, which described the students’ successful attempt to beat the Las Vegas blackjack tables using math]. He said that when they started, one of the first things they realized was that they had to believe in the numbers. At one point, he was down $75,000. He went to his hotel room, recommitted himself to the numbers, then went back downstairs and walked away with $300,000. If you’re going to go with a program like this, you have to leave intuition and gut feelings aside, and really begin to rely on the numbers. That means you’ll have to put some framework of a program in place where you review the information and then act on it …. And you have to have the people and business skills to communicate to physicians, clinical staff and hospital administration what those numbers are saying.

JHC: To what extent can analytics be used not merely to examine what’s going on now, but to predict what will happen in the future? Is there a role for so-called “predictive modeling” in the healthcare supply chain?

DeLay: Yes. It’s not something we’re doing now; but it is on our drawing board. [An example would be], “Given the introduction of a new technology or a new physician group in my facility, and combining that with what I know about disease states, pricing and contracting, can I begin to predict which products will be utilized, by whom, when, and at what cost, so I can do a better job of budget management?” You can also look forward and ask, “Am I setting myself up for errors or mistakes, which I can avoid?”

Then you can get to the Holy Grail – optimization. Say someone has been admitted for stroke. I can look at the optimized course of care for that individual, and the optimized bill of materials from a supply chain point of view, which will deliver the highest quality care at a reasonable cost. With an optimized course of care, I can go to the physicians and say, “There’s a delta between the way you’re doing this, and what we see as the optimized way.” That’s where we’d really like to get.

JHC: How can an organization keep the demand (and hence, use) of analytics alive over a prolonged period? Some organizations employ “business advisors” to show business units how useful analytics can be to them.

DeLay: I think having those types of advocates or salespeople in an institution will be important. But in my experience, once you whet people’s appetite for this, I don’t think you need to worry about keeping it going. My biggest problem is meeting demand. The more information I deliver, the more they want. That has already started among our members.