Population health management is becoming increasingly more important as the Baby Boomer generation has started enrolling in Medicare, information and data are progressively available, chronic disease is on the rise, and health reform has set a timeline for change. The leading patient care organizations are adjusting to the change, with numerous IT initiatives to aid them along the way.
To this end, about a year ago, Baylor Health Care System in Dallas, Tex., and Scott & White Healthcare in Temple, Tex. became one. The merger of the two organizations has resulted in the largest not-for-profit healthcare system in the state, serving adjacent regions of Texas. Baylor Scott & White Health includes 43 hospitals, more than 500 patient care sites, more than 6,000 affiliated physicians, 34,000 employees and the Scott & White health plan. In the past year, the health system created the Baylor Scott & White Quality Alliance (BSWQA), a 3,700 physician-strong network that is one of the largest accountable care organizations (ACOs) in the country.
According to Nick Reddy, the organization’s senior vice president of information system investments, Baylor Scott & White Health is a “baby Geisinger,” referring to the Danville, Pa.-based Geisinger Health System, a known industry leader. On Dec. 10 at the Institute for Health Technology Transformation’s (iHT2) Health IT Summit in Houston, Reddy will be part of a panel discussion that discuss why data is the key to population health management. The session will address successful ACO models and how to best utilize data to take care of patient populations. Click here to register for the Houston Health IT Summit to see this panel and many others. (iHT2 is a sister organization with Healthcare Informatics under the corporate umbrella of the Vendome Group). In a recent interview with HCI Associate Editor Rajiv Leventhal, Reddy previewed the panel discussion and dove deeper into the IT initiatives at Baylor Scott & White Health.
What are some new ways you are using data for better population health management?
What we have found that’s interesting, is that with data, we have more than 5,000 care sites where people can receive care, and more than 400 different electronic medical records (EMRs), so we’re under a large connection mechanism. We are using dbMotion (Pittsburgh, Pa.) to connect and interoperate. If I showed up to any of those sites as physician, you will know everything about me. Unfortunately, today you don’t. It’s what I call "data for care,” and it’s irrelevant to analytics. If a patient shows up and the clinician doesn’t know everything about him or her at the point in which the patient is there in the site, then analytics are irrelevant. So we’re trying to provide data for care at the point of care. That’s a big part of our initiative.
The second aspect is on the analytics side. We are spending a bit of money on consolidation, for one, going pretty far with one of our vendors, building a warehouse where payer and provider information is merged. We are taking that information and we have analytical tools that give us patient stratification and micro segmentation, things of that nature. So I could tell you, even though it’s common sense in healthcare that 1 percent of patients account for 40 percent of the costs, what you can realize through our tools is that 1 percent is almost insignificant because it’s too late. You can’t bend the cost curve, as those patients are normally terminally ill, in hospice. You can talk about it and be irritated by that, but you can’t move that cost curve. The next criteria are who will be the next 15 percent who will become the 1 percent that I need to prevent. So we’re able to do that through tools and strong case management, with consolidated case management across institutions, starting with very basic things such as readmissions.
The third aspect, and we’re mature in this, is what we’re calling a digital experience of our patients to give them a platinum experience. If you’re a diabetic patient, for example, how do I treat you differently, if you’re not doing something? How do I get triggered and have my case managers get to you? How do you get to telemedicine? It’s a little more exhaustive—there is nothing more inconvenient in this country than healthcare. We just make it so difficult for our patients, and most of us won’t do it unless we have to do it. We want to go after that change paradigm; convenience is a big issue we’re driving.
How are you progressing with your ACO?
We’re applying for the Medicare Shared Savings Program (MSSP), and the short story is that we have more than 300,000 patients we care for under risk. We have always had a health plan, so for us, this isn’t something we’re doing because the market is telling us to. Our department chairs will tell you this is how we ran for the last 60 years. We have also gotten paid that way for 60 years. We have taken our 30,000 plus employees and we’ve seen quite a bit of improvement in first year of savings—about $7 million in savings within our patients, and we’re past the low-hanging fruit here. A lot of it comes from better case management, and we feel good about continuing to bend the cost curve through this model, truly having a population health framework. There are three or four assets you need to do ACO. One is the financing arm, and we believe we have the health plan and the MSSP. The second is the technology, as you need the IT stuff that give you things providers have usually never been good at. The third is the operational framework around case managers. Obviously the final thing, you need to control physician behavior—all of this is just Kool-Aid that we’re serving if you don’t do that.
Does this require a cultural shift?
It takes a change in mindset, and most of our physicians get it. You have to realize 2,200 of them are employed, so it gives you a lot of control. They know it’s the right thing to do, and they love reacting to intelligence and data. It is positive reinforcement, when people know that you’re tracking them doing well. I feel good about physician behavior, but maybe not great. Most physicians I speak to get what’s happening, and there are conversations about reimbursement and personal compensation, obviously.
Your organization recently went through an extensive IT initiative called Project Phoenix. Can you explain this further?
We had one of the most aggressive Epic installs in the country relative to speed and time. We really wanted to connect the various aspects of our business and this project was successful in doing that. Led by our CEO, we measured 33 clinical and operational metrics; they’ve all increased after install. We didn’t think about it as IT, but rather as patient flow. So the patient comes in, checks into the ER, goes into surgery, comes to clinic three weeks later somewhere in Texas, and we have taken that flow and implemented Epic relative to patient flow where its certain modules. We’re talking about clinical and financial, acute and ambulatory—anywhere a patient goes we’re able to know a lot about him or her.
For example, we’d have a flag that would say when you open a patient chart, this is a BSWQA patient, so we know it fits in our quality alliance, and it would tell us if it fits in our 15 percent [category]. It seems like common sense, but it’s the things we haven’t done. I use banking as an analogy—if I go to the bank, I get a teller at Bank of America, and they send me home. But if Warren Buffet walks in, he’s treated differently because he might have $8 million in that bank. It’s flagged. You can’t treat everyone in healthcare evenly either. You at least need to be conscious of the people who need the most amount of care, the high-risk patients. There are about 18 transactions you can do on a bank’s mobile site. Clinics will be same way. If you don’t do it, someone else will.
To learn more about using data for population health management, please check out the Health IT Summit in Houston, December 10-11, 2014 sponsored by the Institute for Health Technology Transformation.