Four geographically disparate health systems are banding together to redesign clinical care based on data to solve population health challenges such as preventable readmissions.
Name the technology, and unless it's a da Vinci robot or some other clinical whiz-bang item, hospitals and health systems are constantly criticized for being behind the curve. That's partially because it's been true. Hospitals and health systems have historically resisted investment in technology that would help make coordinate care better.
They've never had to, because their reimbursement didn't depend on it. But that's changing. Rapidly.
Until several years ago, medical records were still largely based on paper—never mind something as innovative as redesigning clinical care based on data. Real-time interventions based on predictive analytics seemed as unattainable as world peace.
But four members of the Premier healthcare alliance, with the help of IBM, are looking to change that with Wednesday's announcement of the formation of the Data Alliance Collaborative.
Texas Health, Fairview Health Services, Catholic Health Partners, and Carolinas HealthCare System, along with IBM, are the founding partners of the data collaborative, which is billed as the first of its kind in healthcare.
It seeks to solve long-term challenges in healthcare, such as preventable readmissions and pharmacy compliance, as a launching point to redesign clinical care so that it is more cost-effective and patient-centric. In other words, the alliance seeks to put some meat on the bones of what's called population health.
Through other previous work together, "we realized the supporting data infrastructure for coordinated care was integral piece of the puzzle," says Sean Cassidy, Premier's vice president and general manager of the Premier Data Alliance. "We had massive problems with our infrastructure."
See Also: Preventing Hospital Readmissions Presents Financial Paradox
Quietly testing its modules and recruiting participants for the past 18 months, the collaborative now boasts nearly 100 hospitals and more than 1,600 non-acute sites that care for 28 million people. The widely disparate geographical reach of the four founding organizations, as well as their ability to invest in so-called "big data," attempts to leverage four different budgets to create a data analytics package that can be used by all collaborators. Other partners are expected to come aboard within the next six months.
"Two of us were Pioneer ACOs, so the notion of improving required us to think about how we would achieve that capability when we knew we didn't have the capital alone," says Terry Carroll, chair of the collaborative and senior VP and chief information officer at Fairview Health Services in Minneapolis."
Allen Naidoo, vice president of Carolinas Healthcare Systems Dickson Advanced Analytics unit, says EMRs are only a first step, and that the chief problem with attacking most of the waste in healthcare is that so-called off-the-shelf technology solutions in healthcare that are aimed at solving short-term challenges in healthcare.
In other words, healthcare organizations have been reactive rather than proactive in their attempts to leverage technology to improve care and cut waste. "Now we are asked to merge clinical data with payer claims data to get a total view of the patient," he says. "With this collaborative, we can learn from each other and leverage the same technologies. I call it 'better together.'"
Indeed, such investments are expensive—one reason the four systems are banding together to address the issue. Carroll says the time for urgency on such big data solutions is now, because healthcare business cycles are lasting 18—24 months instead of a decade, as in the past.
"I have hundreds of people doing data management," says Carroll. "But they don't get to the insight/foresight aspect of what we're doing because they don't have the tools. We don't have enough gas left in the tank in our clinical and business models to go further. Put simply, we can't improve fast enough without this technology."
The collaborative seeks not only to solve the problems that exist today in healthcare, but to anticipate.
For example, hospitals are now penalized for certain "preventable" readmissions—yet the collaborative seeks to prevent "all-cause" readmissions, anticipating that hospitals and health systems will have to transform even further the way they provide clinical care.
Often, patients are readmitted because they've failed to follow their drug regimens post discharge, or because they don't have the support to take care of themselves after they are discharged. This is a huge source of waste in healthcare because of lack of capability.
For example, the 30-day readmission rate for Medicare beneficiaries is around 20%, which equates to about 2 million people a year. Those dollars quickly add up. The collaborative aims to change that by being able first to predict based on a variety of disparate sources of data, who is likely to be readmitted, so that hospital can address issues that may cause a readmission proactively.
The DAC pharmacy compliance model is designed to remind patients and their caregivers about the importance of follow-up medications. Analytics in the module will work to notify providers within 24 hours who has not filled their prescriptions, for example, and immediately intervene.
With its readmissions module, the DAC model will analyze both EMR and administrative data to identify patients who are most likely to be readmitted before they are discharged, and will identify potential risk factors for particular patients, tying patients to evidence-based checklists based on their condition.
Cassidy says the intent of the DAC is to work on some problems that are immediately critical for hospitals' population health strategies, but also to work on tomorrow's problems.
"Obtaining the real value from aggregation is really hard to do," he says. "You have to be looking over the horizon to understand what's coming—for example, doing things relative to performance based on payment models other than the government. Populations can be segmented, with different requirements based on the business relationship with whoever's financing the care."
Cassidy uses an aviation analogy to describe the approach most hospitals and health systems have toward data utilization. Most organizations have spent amply to acquire their shiny Piper Cubs and are doing well flying them, but their systems are outdated and can't go the distance. They don't provide real-time information that can prevent expensive interventions.
"They're going to have to fly 747s," he says. "Through this collaborative, we're hoping each of them doesn't have to build their own."
Philip Betbeze is the senior leadership editor at HealthLeaders.