13 May Can business intelligence really improve healthcare?
Madison, Wis. – The deployment of clinical and business information systems that generate digital information could change the delivery of healthcare in many ways, but perhaps the most profound impact will be in mining the data with analytics – especially if organizations learn to present it in meaningful ways.
An enormous flow of data is generated from patient encounters on a daily basis, and those who promote business intelligence believe this data can become useful information with the help of “BI” tools. Not just any information, but information that can help hospitals derive added value from their IT investments by sharpening their decision-making on both the business and clinical sides.
“I think the argument [for business intelligence] is intuitively obvious,” said Galen Metz, information services director of Group Health Cooperative of South Central Wisconsin. “I think most executives find themselves in a position where they wish they had the data to make a decision.
“My guess is that many times, they don’t.”
GHC-SCW, the first Madison health organization to deploy an electronic medical record, is turning its attention to the reporting mechanisms of EMRs, which serve as the foundation for business intelligence in healthcare.
For inspiration, Metz can turn to John Hansmann, a regional manager for Intermountain Healthcare. Intermountain is a Utah-based nonprofit organization that has used analytics for its clinical operations.
Hansmann, who spoke at WTN’s annual Digital Healthcare Conference, defined business intelligence as the use of data to help make better business operational decisions, “whether it’s financial or clinical.”
He provided several examples of how Intermountain uses BI to track operational performance in the emergency department, and various aspects of its staffing and clinical programs. Like many emergency departments, Intermountain “EDs” measure door-to-doctor time, in part to determine whether workflow improvements are necessary to reach its goals. The organization is in the enviable position of having a lot of patients coming through its doors, but that also places stress on hospital staff, including one whose door-to-doctor goal is 30 minutes.
The hospital in question hit the goal this year, but not before an intervention that resulted in improvements. “So we track this, and we understand how we can improve operations on a minute-by-minute basis,” Hansmann said.
Intermountain, which has 22 hospitals in Utah, also uses patient census data to plan master schedules on nursing floors and plan for other resource-utilization needs. The data helps ensure “we have the right workforce there at the times we need it, so we don’t over staff or under staff at any one given point in time,” he said.
Analytics also come into play for collaboration, he added. Intermountain partnered with GE Healthcare to develop an ED module used to track the frequency of deviation from workflow processes.
Hansmann cited BI uses in other industries, including retail for optimization based on customer needs, banking to gain a better understanding of customers, and grocery stores for forecasting product demand using customer-purchasing patterns.
“I think it’s a sad state of affairs that most of our grocery stores know more about us than our healthcare providers do,” Hansmann said.
Some organizations in the healthcare space already are moving to outpace groceries. Metz plans to leverage data captured in the EMR to optimize workflows and identify preventive care opportunities in patient populations.
The organization already has McKesson as a data warehouse vendor, and Metz characterized the data as “clean and loaded.” Since GHC-SCW has what Metz described as a “do-it-yourself culture,” it will produce much of the analytics itself and look at vendors and tools to supplement in-house work.
Metz has hired several MBAs on its BI team to bolster data analysis capability, and is reviewing offers from several vendors – including Symphony Corp. – on the basis of technical environment, cost, and training and support.
In terms of metrics, different types of organizations have different needs. Reed Eichner, vice president and CIO for the Healthcare Information Management Systems Society, recently adopted BI to analyze membership statistics for individual members, corporate members, and organizational affiliates.
These metrics, which influence the organization’s membership recruiting efforts, could not have been identified by the information technology department. Eichner, who said the goal of BI is to help executives make “actionable decisions,” said the metrics worth measuring were determined by stakeholders during departmental meetings.
For the vendor, the organization selected Business Objects, which has a data-analyzing tool that enables users to develop their own reports. In terms of the enabling technology, Eichner believes providers in this space are very good, but the choices are narrowing as vendors consolidate through mergers and acquisitions.
According to Eichner, HIMSS used its Gartner EXP account to research vendors, and performed more vendor due diligence with associations and businesses that already deployed BI technology.
Whether the information is presented on dashboards, which can be customized to present performance indicators, or scorecards, which monitor goals against targets, Eichner said they key is to present information in an understandable way, as opposed to offering a mix of data that may be confusing and complex.
Public reporting and the value chain
Internal uses may drive decision-making, but analytics also play a role in the public reporting of clinical performance data. For healthcare, the external pressure to publicly report clinical data is building. That’s especially true in Wisconsin, where organizations like the Wisconsin Health Information Exchange, which is working to share patient data among healthcare organizations in southeastern Wisconsin, intends to produce analytics on issues like Medicaid cost impacts, and where the Wisconsin Collaborative on Healthcare Quality collects and reports quality information from hospitals.
Karen Knecht, associate partner of Healthlink Solutions for IBM, said healthcare organizations will move up the “value-chain” with information systems that drive efficiency, quality improvement, and competitive differentiation that is reflected in public reporting.
“The ability to move up the value chain will require a new infrastructure and a whole new way of thinking about this,” Knect said. “You still have to do it incrementally. You can’t solve all the problems at once, but you have to think about the information value chain.”
When it comes to competitive markets, Knecht said publicly reporting organizations have to offer more than reports on, for example, the routine dispensing of aspirin and instead focus on outcomes. This results-oriented style of analytics has been tried successfully in other industries, she said, and clinical analytics should help healthcare organizations move up the value chain.
In terms of analytics, there is disagreement about how much healthcare can learn from other industries, in part because clinical analytics happen on the front end, not the back end. Attorney Erik Phelps of Michael Best & Friedrich, however, said much like clinical decision-making, purchasing decisions made by Wal-Mart involve “multi-variable analysis.”
“Retail isn’t perfect,” he said, “that’s why we have clearance sales, but I would love to have a Wal-Mart data analyst in this room because I bet there would be something for us to learn.”
Knecht, who believes data mining would be well served by a formal knowledge-management process, cited the airlines and their ability to fill more seats on individual flights. However, she indicated that healthcare could learn the most from complex analytics done by industries like finance, which have uncertainty.
“Even geology or Earth science tends to be less certain, and there have been deep analytics capabilities applied in that industry,” she said.
• Business knowledge has become a CIO’s most important skill
• CIO Leadership: Metz launches business intelligence operation
• CIO Leadership Series: Eichner tackles business intelligence for HIMSS
• Patrick Gray: The subprime mortgage mess: A case of subprime data?