Data Detectives (Health Data Management)
[Published 01/01/2015 by Elizabeth Gardner]
Stephanie Lenzner’s favorite kind of analyst has a touch of the gumshoe-someone who’s willing to knock on doors throughout the organization, ask questions, and sift through the answers to solve mysteries.
“Analysts historically have had a tendency to hoard their databases and their insight,” she says. “It becomes ‘my database’ and ‘my analysis,’ not the organization’s. We want to tap into their private investigator nature.”
Some may be reluctant to give up the “god-like role” of being the keeper of the data, she says, but they have the potential to become even more god-like by mastering the big picture.
Lenzner, vice president of information management services at Children’s Hospital Wisconsin (CHW) in Milwaukee, is developing an analytics community, also known as a “center of excellence,” to break down information silos and position the organization for the era of accountable care. Children’s has 40 locations in Wisconsin, including a Level 1 trauma center, two hospitals, and urgent, primary and specialty care clinics. The organization has 5,000 employees and 1,000 affiliated providers. It’s pulling together a population health management strategy across its entire network, which includes a social work component in addition to clinical care.
“We’re different from other health systems in the information we collect because we play a more extensive role in a child’s life,” Lenzner says.
The analytics center of excellence is built on an enterprise data warehouse that pulls data from a variety of clinical, operational and financial sources. And CHW has five- and 10-year strategic plans for developing advanced analytics capabilities.
The question becomes: Where do CHW and other healthcare organizations find the data gumshoes they need to make their vision a reality? And how do they organize these specialists to work together, among themselves and within the larger organization?
Today’s new analytics environment-enterprise data warehouses, quality reporting, tracking patients across the continuum of care, value-based reimbursement-demands an analytics team that can go way beyond report writing and dashboard design to generate real-time understanding of how an organization is doing, both clinically and financially, and where it needs to improve.
In the rapidly changing healthcare environment, the team also must be able to perform predictive analytics, to identify which actions are most likely to get an organization where it needs to go and how different decisions might affect the outcome.
“These are among the hottest jobs everywhere in the country,” says Alistair Erskine, M.D., chief clinical informatics officer at Geisinger Health System, Danville, Pa.
Putting together these groups is no easy task. But Geisinger, CHW and other healthcare providers-including the Children’s Hospital of Omaha and the Cleveland Clinic-are finding new ways to recruit, retain and deploy analytics teams.
Team first approach
CHW’s Lenzner has a team of 24 full-time analysts attached to its center of excellence, plus 75 “pseudo-analysts” who have primary responsibilities in other departments but were also hired to contribute to the analytics efforts that touch their area. “We didn’t want to have pockets of analytics,” Lenzner says. “We’re in the process of wrapping these folks in [to the analytics community] and finding out what they know. It’s mutually beneficial-they will learn more technically and have earlier access to analytics tools, and we’re serving the organization better by presenting things in a standard way.”
The analyst community has a matrix structure: Some analysts report up to the leaders of the center of excellence, and others report to multiple departments, but they all use the same analytics tools and present data in consistent ways. The community meets regularly to pool knowledge.
“We don’t have data silos anymore,” Lenzner says. “There’s no longer any such thing as ‘my database,’ and that’s a big deal.”
Lenzner casts a wide net to find people who can fit into CHW’s environment.
“The biggest competency we hire for now is business acumen, in addition to strong analytical and technical skills,” Lenzner says. “Business acumen is very difficult to teach. Either you have it or you don’t.”
CHW is developing a program to mentor students at local universities and groom them for healthcare analytics, and it recruits heavily from industries with advanced analytics capabilities, such as retail, banking, and even gambling. “We’re trying to create a high-performance team of people from all backgrounds.”
Geisinger’s Erskine, like Lenzner, looks to the college ranks. He seeks out recent graduates who are interested in the field.
Geisinger’s analytics staff includes data engineers, who create a “unified layer” out of multiple data sources, and data scientists, who take the information from the unified layer and use it to answer questions. They’re aided by domain experts who figure out the right questions to ask.
Geisinger presents quite a different picture from what was considered the healthcare-analytics norm just five or 10 years ago.
Back in the day…
Being an analyst in a healthcare organization used to be a straightforward number-crunching task. Someone needed a report, so the analyst assembled the data, did the calculations and printed the report-daily, weekly or monthly. Each new question generated a new report. The list of available reports could number in the thousands, with many reports sticking around long beyond their usefulness. More advanced analytics teams worked on dashboards, assembling data sets that users could manipulate themselves to some extent.
Many organizations’ analytics functions are still rooted in reports and dashboards. “That approach is baked into most organizations, but it doesn’t scale,” says Nicholas Marko, M.D., chief data officer at Geisinger Health System, and an associate of Erskine. “Reports and dashboards only give you the information someone has spooned to you.”
Geisinger is considered a leader in integrated, value-based care with a sophisticated analytics operation that employs between 100 and 200 analysts distributed across an organization of 20,000.
And the organization is planning to take analytics to the next level. Geisinger’s goal is to give each person in the organization self-service access to any information they need, in real time.
Cornerstone Health Care, a multi-specialty group practice based in High Point, N.C., is a Medicare Shared Savings accountable care organization, and depends on its analytics group for the information it needs to stay afloat. CMIO Dale Eric Green, M.D., decided four years ago to develop advanced analytics capabilities, and he now has six full-time employees in the analytics group: two to whip the data into shape for analysis, two to do data visualization, and two to do the heavy-duty analysis of the data to fulfill requests from the end users.
But he agrees with Geisinger’s Marko that self-serve analytics are preferable to generating report after report.
“The most complex problem I still see for our organization is the ‘sand-painting’ problem,” Green says, referring to the practice of Tibetan monks who draw elaborate mandalas in the sand and then sweep them away: a profound meditation on impermanence but not a good way to run a healthcare organization. “That’s the problem in analytics-we spend a huge amount of time doing a report, and then we say, ‘That’s nice,’ and we sweep it away. I’ve gotten more and more fond of this idea where the technology enables the people.”
To achieve that goal, he makes sure the analytics group is closely aligned with the shared vision of the organization. “What makes people happy at their jobs is feeling like they matter,” Green says. “If you can connect their work to the mission and vision of the organization and demonstrate how they drive it, the job satisfaction takes care of itself.”
And that may mean embedding a team member, or an entire team, into the day-to-day operations of the organization.
Indeed, the analytics team of the future will be characterized by analysts who get out and about rather than huddling in their cubicles. At Children’s Hospital of Omaha, data pros sit with the staff and help them figure out what problems they’re trying to solve and what information they need to solve them. The people who need the data play a key role in designing their own dashboards. The hospital is trying to replace endless report writing with a self-serve mindset, where analysts are a resource for understanding how data can be used, rather than gatekeepers, according to CIO George Reynolds, M.D. … Click to view full article