Log in

By JULIE MINDA

Dr. Jim Reichert believes that people should be able to go to any health care provider and know that they are getting high-quality care, and they should be able to expect an optimal outcome.

Reichert

Reichert, who is vice president of clinical analytics for Catholic Health Initiatives, knows U.S. health care is still a long way from being able to deliver on the promise of consistent high-quality care. But he believes data analytics holds promise for moving the industry closer to that aspirational goal.

He explained to Catholic Health World how Englewood, Colo.-based CHI is using data to improve care quality in its hospitals and guide providers to reduce variance in health care delivery. The initiative soon will be expanded to CHI primary care clinics.

What is the problem you think analytics can help solve?
We haven't really lived up to the promise of high-quality care for everyone, in every health care facility. It makes a huge difference where you go for care, who you see, and also your individual characteristics, such as your age. All impact what your outcome may end up actually being, and that's unfortunate.

How could analytics address this problem?
Through transparent reporting, through standards, through risk-adjusted measures, we can start to measure across not only organizations, but also across the continuum, to ensure that everybody is getting the same high quality of clinical care.

What are the goals of CHI's systemwide data analysis work?
Our work is about using the data that we have to impact the clinical care of our patients for their experience, the efficiency of their care and the clinical outcomes.
We want to reduce the variance in all three of these areas and be able to monitor and impact the efficiency of care that's delivered so that the care is consistent and efficient wherever you are.
Right now we have a lot of inefficiencies that are opportunities. We now have data and can understand those inefficiencies, such as inefficiencies that lead to unnecessarily long lengths of stay and inefficiencies related to supply utilization.

What are you measuring?
We are focused on seven quality measures, eight safety measures and eight patient experience measures. All of that was our focus, and that kicked off about two years ago. We've been reporting it internally for a couple of years now, and show some dramatic improvements in those scores. (See sidebars.)

How does the analytics process work?
CHI collects and analyzes data on all 23 measures for all of our wholly owned hospitals. We then do statistical analysis on all of those measures so that we can show individual facilities how they are doing. This way, they can know where they need to improve.

How does the information flow between CHI and the hospitals?
We disseminate all our analytic reports through our analytic portal, which is a one-stop shop. Virtually everybody in the markets who is inside CHI and its facilities has a one-click access to these reports.

If you tell everyone: "You have to do it this way," people may or may not accept that completely; they may or may not adopt it completely.

The approach that we've taken is more of an organic or bottom-up approach. We provide the data analysis to each facility, and when a hospital underperforms in a measure, we talk about it, and position it as an opportunity to improve. And when you get agreement at that level, then all of a sudden there's the ownership. It shifts from the national headquarters pushing something to a market pulling.

And I think this has more success in moving the needle because they're driving toward a goal and a given metric that they're working on as opposed to some standard of practice that national has pushed on them.

Do you work with hospitals to fix problems, or are they on their own to find solutions?
When it comes to actually getting at the root cause when our hospitals' outcomes are off, and figuring out what needs to be done, CHI analysts — usually at the hospital level, with support from the national level — will go in and they'll actually pull the records for the patients in that facility and do the root cause analysis to understand what's going on there.

And it's been fascinating because with surgical site infections, for example, we've made a lot of progress within a short period of time. Sometimes the analysts will identify that a particular surgeon is responsible for an increased number of infections occurring with patients in his or her care. Sometimes the analysts will identify that the way a surgical team has been doing surgery prep for years isn't up to a key standard. And so the analysts and clinicians learn from other markets: This is the standard and here's the best way to reduce variances.

What's next in this effort to apply analytics to improve clinical practice?
We plan to move into ambulatory care this year and next year. We're planning to use forecasting and predictive analytics so we actually move the analytic engine closer to the bedside. This way, we can support the provider in near real time, or real time. For example, we may identify through analytics that a patient who has had heart failure is a high mortality risk, and we may use that information to manage that patient with more progressive outpatient care.

As we move in this direction, the data requirement will be to get more granular, operational, real-time data so we can start acting on that information quickly.

How do your teams share successful approaches they develop?
CHI has an operational performance review with each division. Taking part are the senior vice presidents of the divisions, the chief medical officers, the chief nursing officers, sometimes the presidents of the facilities, and definitely other leaders whose sole area of responsibility is quality improvement. And they go through, measure by measure, what they're doing. They identify areas where improvements have been achieved and identify gaps that require attention.

Then there's also a meeting of the quality leaders, directors and other individuals who are actually doing the real work to move this needle. It's in those forums that best practices are shared.

Is data analysis expensive?
It's fairly low cost right now, and yet there's a potential for significant dollar returns and reductions in value-based payment penalties and other incentives that the federal government is offering.

Our (system headquarters) team is five people. And we produce upward of over a thousand analytics reports a month right now. We've been able to leverage technology, and we're able to standardize the reports. Once we do that, we can automate the reports. With automation, a lot of these analytic engines will be running on their own with very little time required for maintenance or upkeep.

What I hear from people frequently, is: "We spend less time doing hunting and gathering of data, which is what we used to do; that was our career." And now what they spend more of their time doing is on the root cause analysis and the performance improvements.

Do payers reward you when your analytics work improves outcomes?
I don't think we've moved to that level yet within our organization to develop a strategy around how improvements in our clinical outcomes can be shared more broadly with the payers that we have contracts with.

What are some of the main challenges of conducting this analytics work?
What we're wrestling with now… is that we have the analytic information, and we have a lot of insight into opportunities to improve care and to address variances across the organization. But we want to determine how to set up our clinical care work throughout our facilities with the structure and processes that are needed to use the data. We want to be data-driven and actually drive improved efficiency and clinical outcomes for any procedure and any care process we wish to implement. And I think that's the opportunity right now that exists for us.

We've been able to make such improvements in our hospital environments, but now how do we do it when we move it to a service line level? And the other area where we want to use analytics in this way is with our ambulatory work. We're putting in place in our ambulatory care areas analytics work that is similar to what we have used for inpatient care. We're just beginning to roll that out in CHI's ambulatory care facilities.

But what's been clear is that the governance and structure used to advance the analytic work need to be thought out very clearly, and need to be organized in a way that previously we haven't had to do.

How does CHI's work differ from what many others are doing with analytics?
What you often see (elsewhere) is two tracks. You see a track of the executive leadership team moving along one rail, and then you see a parallel track of analytic developers moving along the other rail. The analytic developers will build solutions, but those solutions aren't finding their way into the C-suite. They're not finding their way where the executives are actually using it to make and drive improvements. Then on the flipside, if you talk to those executives in those other organizations, they complain of a data gap.

So what we've done here at CHI is: We've had data and technology as a part of our work, but we've really aimed to focus on key elements like governance, structure, culture, people and process to hold people accountable and drive transformation.

How can analytics experts help advance the work?
I think in general in health care, leaders haven't necessarily gone through formal analytics training and education. I think our job is to provide that education. If leaders are willing to take the time to do that formal training, then I think they're also willing to make it a priority in their organization to become more data-driven.

What happens to this analytics work, if the Affordable Care Act is dismantled?
From my perspective, our work does not change significantly because we have a commitment to our patients to use data to make sure that wherever they go for care, they get a consistent, high-level experience, and optimal outcome.