Grand Junction (population 58,704), in the high desert country (elevation 4,597 feet) of western Colorado, is typical of certain types of small cities in the American West. Compared to the large metropolises of the East, it is relatively tiny; yet out here, in the sparsely populated lands west of the Continental Divide, it is a major regional center of commerce and connectivity for western Coloradans.
Not surprisingly, it is also the home base for the Colorado Beacon Consortium, a not-for-profit collaborative of four healthcare organizations (the Quality Health Network, Mesa County IPA, St. Mary’s Hospital & Regional Medical Center, and Rocky Mountain Health Plans) that came together in 2010, with the aim of optimizing the “efficiency, quality, and performance of our health care system, and integrat[ing] the delivery of care and use of clinical information to improve community health” for the counties of Mesa, Delta, Montrose, Garfield, Gunnison, Pitkin, and Rio Blanco.
The fact of those four healthcare organizations—a regional health information exchange (HIE), a large independent physician association (IPA), the largest hospital in the area, and the regional health plan—coming together collaboratively, speaks to the nature of what’s being shared here: data and information, in order to improve care management and care communications across this far-flung region of mountains and plains, canyons and desert.
And what’s excitingly innovative about what the Colorado Beacon Consortium (CBC) is doing is also the reason that the organization is being recognized as a co-second place winner in the Healthcare Informatics Innovator Awards program this year. What the leaders of the program are doing is using the advanced HIE architecture of the Quality Health Network to deploy a state-of-the-art decision support tool called Archimedes IndiGO (from the San Francisco-based Archimedes) across multiple, independent primary care sites, operating on multiple, independent electronic health records (EHRs) throughout western Colorado, in order to perform data analytics across a base of more than 55,000 patients by the end of the first quarter of this year.
Using individual personal health information, de-identified from the EHR of individual patients, CBC leaders are calculating the risks of different diseases, and the overall health impact of different medication and lifestyle interventions on thousands of patients in the system, in effect creating individualized guidelines designed to help the physician and patient to make optimal preventive care decisions. What’s more, IndiGO then displays this information in a user-friendly graphical interface, allowing the physician and patient to compare the effects of different treatment choices.
The goal is to implement IndiGO efficiently into existing, EHR-based practice workflows, with single sign-on support provided via health information exchange. At the highest strategic level, what he and his colleagues are focused on, says Patrick Gordon, executive director of the CBC and associate vice president for community integration at Rocky Mountain Health Plan, is that “We’ve worked to do three things: first, promote a broad-based practice transformation and quality improvement program that is self-sustaining, through learning collaborative and other techniques. The second,” he says, “is the creation of a broad-based healthcare transformation effort that is self-sustaining, and will continue after Beacon ends. And the third thing is the health information exchange architecture, which now is sufficient to support population health management and improvement functionality.”
Gordon, and Marc Lassaux, technical director of the CBC and director of new technical initiatives at Quality Health Network, and their colleagues, have been delighted with the results so far of their hand-in-glove collaboration with the Archimedes people; the CBC initiative is certainly Archimedes’ showcase case study to date. And on one level it certainly was gutsy to start out partnering with a small start-up vendor when laying the IT and data foundations for such an ambitious regional population health initiative. But Gordon and Lassaux never had any doubt over their partnership choice. “At the 50,000-foot level, the basis of the collaboration was philosophy and insight,” Gordon says. “At the end of the day, their tool, technology, model, is simply superior; it’s disruptive. And we were attracted because not only did it have a fundamentally different approach to predictive modeling and measurement, but the tool itself could be deployed in simple ambulatory and patient care settings; it’s active and actionable in clinical practices of all sizes, and is patient-facing; and yet the model itself is transformative.”
As Gordon and Lassaux wrote in their Innovator Awards submission, “In this project, we are actively bridging advance population health analytics, risk stratification, targeted clinical interventions and patient activation in a single architecture. There are currently few (if any) working examples of multiple sources of disparate, clinical and payer data being aggregated and normalized within HIE to support deployment of a shared analytic tool in multiple, independent, primary care settings. The patient-facing component of IndiGO aligns well with the active promotion new skills, such as motivational interviewing, and new measures, such as the Patient Activation Measure (PAM), in comprehensive primary care models.”
And, they added, “The project will enable adopting practices to fulfill the practice-based risk stratification milestone set forth by CMS in the Comprehensive Primary Care Initiative (CPCi). In addition, it will enable private payers (Rocky Mountain Health Plans) to adopt practice-wide IndiGO values as the basis of risk-adjusted payment and gainsharing arrangements for the purpose of CPCi and other payment reform initiatives—and begin retiring plan-based, claims-oriented risk adjustment methods that are less clinically-pertinent, less predictive and segmented by payer source.”
BECOMING A LEARNING
Gordon compares the early stages of any data-driven initiative like this to “a hockey stick—it’s a long, slow process of getting to that critical tipping point; the health plan and physicians have been working on some of this stuff for a decade now.” But, he says, at some point, “You get to sort of a critical tipping point of engagement, and the learning process and the pace of change accelerate. And we’ve seen that in terms of the quality improvement and the competence of use of data measurement and methodologies.” At that point, he says, “you really get to what Don Berwick had envisioned in terms of a learning collaborative, in terms of active, collaborative learning. And that’s how we’ve gotten community engagement. QHN has been doing that for years now with health information exchange. We’re just beginning to get to a critical tipping point now, after eight years of work.” Among the initial results, Gordon reports, are improved patient medication adherence, behavior change, and shared decision-making, following physicians’ use of the IndiGO capabilities during patient visits.
Given that what Gordon, Lassaux and their colleagues have been doing, here on the leading edge of data-driven, region-wide population health management, do they have any advice for the CIOs, CMIOs, and other IT leaders in other organizations who might try to follow their example?
“One of the phrases we’ve been using,” Gordon says, “is, ‘waiting is wasting.’ Waiting for a perfect IT infrastructure architecture, for perfect health information exchange, and so on, is unrealistic. Every day, you miss opportunities to develop your own insights and develop new solutions. Data production and analysis remain insufficient, so we’ve had to create our own models. It’s costly and takes time, but had we not rolled up our sleeves and done the hard work of data aggregation, and developed policy insights from that work, we would not be where we are now.”