Unity Health Care, Inc. is a federally qualified health center (FQHC) in Washington, D.C., with over 25 care sites in the District. Unity provides ambulatory care, including primary care, specialty care, and dental care, as well as school-based healthcare at five sites, and services to the District’s two jails. It also has 11 care sites in homeless shelters. The FQHC has about 250 salaried physicians and 1,000 employees altogether.
Unity Health Care, which was founded in 1985 as Healthcare for the Homeless, is celebrating its 30-year anniversary this year. Last year, the organization saw 105,000 unique patients, and provided care via a half-million visits altogether. Its population is three-quarters African-American and 15 percent Latino, and serves all wards of the District. The organization operates on an annual budget of about $100 million, with about 85 percent of its revenues deriving from Medicaid payments, about 8 percent from grants, about 6 percent from private health insurers, and a small percentage from Medicare. It receives federal funding compensation for uncompensated care. About 11 percent of its patients are homeless.
Not only is Unity Health Care certified as a level 3 patient-centered medical home; it also became a Davies Award winning organization in 2012, recognizing its leveraging of healthcare IT.
Angela Duncan Diop, ND, vice president of information systems at Unity Health Care, spoke recently with HCI Editor-in-Chief Mark Hagland, with regard to the topic of population health. Diop will be moderating a panel on the related topic of “Driving Organizational Performance with Predictive Analytics,” at the Health IT Summit in Washington, to be held June 16-17 at the Ritz-Carlton Tyson’s Corner, in McLean, Virginia. Below are excerpts from that interview.
What are some of the biggest challenges in doing population health right now? Is transience one problem, in terms of harnessing data?
Yes, transience is a problem—not only with homeless patients, obviously, but even with our non-homeless patients—coming up with addresses and phone numbers to do outreach with patients, is a challenge. Especially as we become more automated. We do phone call reminders. We have a patient portal. We’re actually trying to meet meaningful use Stage 2 under Medicaid. And one of the things you have to do is to have your patients link to you through a portal.
And it obviously is difficult to achieve that, with a large percentage of homeless and near-homeless patients, correct?
Yes. We have a grant through the Aetna Foundation that provides an umbrella including an app and a patient portal. And while these technologies don’t really fit with all of our patients, most of our patients actually have smartphones. They might have a smartphone even before they have a home to live in. So that is challenging. But we ask for demographics every time when they come in. We ask, is your phone number the same? Is your address the same? I always say… as an FQHC, in many ways, we’re ahead of some other types of organizations, because we’ve been having to do this for years. Before, we were doing chart audits. And we’ve always been required to collect demographics. So we’re ahead in that way.
Are you actively using “big data” yet? Perhaps only “medium data” so far?
Yes, we actually call it “medium data” here [laughs]. But we’re a big provider customer of [the Westborough, Mass.-based] eClinicalWorks. To have 250 providers, that’s a lot of doctors, so we have a pretty good-sized database. And we’re one of the largest FQHCs in the country.
What have you been doing in that area?
We’ve been doing a lot. It’s really kind of exciting what we’ve been doing with the data. We’re going into our sixth year live with our EMR. And there was a period of stabilization following our go-live years ago, but at the same time, we really had to get data out of our EMR right away, because our government grantors weren’t going to give us extra time to share that data with them. And all the FQHCs, we all have to annual report unified data sets (UDS) to the government. It’s a big, huge report that tells us about our patients—their zip codes, where they live, their ages, their demographics, and then we report on quality measures for them… top diagnoses, etc. UDS reports are public information, you can Google them. So out of the gate, we were having to get these quality reports. So we spent a couple of years just making sure we were getting reports to everyone who required them from us. Now, we’ve got the bandwidth to look at data and use data in ways we want to use data. And we want to become data-driven, that’s actually our goal. We actually have data governance here in our organization, and have had it in small ways for a long time, but in the past two years, we’ve created a more formal data governance structure. And we’ve always had a formal data team that meets twice a month, and includes senior leaders; at the beginning, we were mostly focused on data integrity. But now we’re looking at data as a strategic resource—who are we using the data, how are we prioritizing it, what should our long-term partnerships be?
What are the main programmatic areas you’ve focused on so far?
Our biggest effort for the past six to nine months has been on hypertension. I’ve been here for eight years now, and we were already doing work on diabetes, but I asked, why aren’t we doing more in this area, the number-two area? So it’s exciting in terms of how we’re using data. We wanted to better utilize our work teams for hypertension so that we made sure the patients were coming back to get their blood pressure taken. The patient might see the physician and the doctor would say, your hypertension is out of control. So we’ve redesigned the workflow so that for the follow-up visit, the nurse could see the patient rather than the doctor. And we made sure the EMR could capture that. And we trained all of our clinicians. So we trained everyone, the entire team, as we’re a PCMH. And we had data on outcomes. We were able to measure the impact on the decrease in blood pressure, and we were able to measure that. We were even able to measure the use of the templates. Because sometimes we would design something in the EMR, and would find out a year later that no one was even using it, so we couldn’t measure data. So this one, we actually went in and retrained everyone.
What was the timeframe for this development?
It was a six-to-nine-month period of time. The clinicians were trained last fall, and then last month, April, they were retrained. So it’s a more holistic approach. We’re trying to look across the whole chain of information, not just looking at data, but looking at information. We want to put in good quality, put in good workflows, learn about the outcomes, and make changes based on the data.
And per templates, we do a lot already with diabetes and asthma; those programs were existing and in place. We actually created specific asthma templates, and recently we realized people weren’t actually using them that much. So we wanted to find out whether they were recording the data someplace else.
The other big population health effort we’ve had within the last six to nine months has been the rollout of the FIT Test. It’s to screen for colon cancer. The patient can just take it home and screen themselves and put the stool sample in the mail, and we have gone from abysmal rates of screening to really very good ones.
What have been your biggest lessons learned so far in all of this?
I think sometimes that if you’re a small organization, or if you’re not an academic teaching hospital, you might think, we can’t do research. But start small, find those people who are interested in data, or outcomes measurement, or change management, and start where you’re at. It would be hard to believe we’ve gotten as far as we have, but we’ve spent five years on this. And we meet twice a month, and now we have breakout meetings—we’ve created a governance group and we have subgroups.
Also, you need to be patient, because the capacity of the organization has to grow. If you’re tiny and not data-driven, you’re not going to get there over night. We had to start by explaining what kinds of data we could even give people. So that’s been a long road. And I think that it’s important to focus on data integrity first. We didn’t start with lofty governance goals; we started working on the integrity of the data, because if people don’t believe in the data, everything else goes down the chute. Also, really developing change management systems is important.
What would you advise other CIOs? You don’t have unlimited resources, obviously, and others don’t, either.
It’s important to sort of leverage yourself, because you can’t do it all yourself. In data governance, I have a counterpart whom I’m working with; she’s the vice president of strategic planning. So you can usually an ally in someone else who cares about the data. And she actually is responsible for our grants work, the grants people report for her. So she really cares about the data. Or it might be someone finance, or QI. So look for allies in those other places, and also look for allies amongst the staff; because that’s how you’re going to leverage yourself.
Do you feel optimistic going forward?
I feel incredibly optimistic! Our goal in our governance group is to move our organization to e data-driven. And there’s a couple of key questions we try to ask ourselves that five years ago we didn’t ask ourselves. And one of the things a small organization can do is to ask, how will we know if we’re successful—what’s the measure? And just asking that has helped people to think about our goals even before we start any project. And second, we’re asking ourselves more, how we know just when to quit. So those are two key questions that we’re asking and that other people can ask. Also, in terms of predictive analytics, we’re not there with predictive analytics yet, but I can see it on the horizon. And I think that is what’s really exciting. So that’s where we’re at.