Skip to content Skip to navigation

Leveraging Data

February 21, 2011
by John Degaspari
| Reprints
Providers Are Using Data to Drive Performance Improvement in Their Organizations

Hospitals, medical groups, and health systems have taken a page from the playbook of manufacturing and other industries to achieve measurable improvements in care quality and the bottom line

The manufacturing industry has long used concepts of six sigma and lean management to drive improvements in quality. And the idea of applying data to improve performance-which years ago spread to other industries such as air transportation, retailing, and shipping-has in recent years caught the attention of healthcare administrators, who hope to achieve measurable results in healthcare. To be sure, hospitals and factories are worlds apart in many ways, but-as leading provider organizations have already demonstrated-they share common tools when it comes to quality.


One organization that has the benefit of a broad view of data-driven performance improvement in the healthcare arena is the Premier healthcare alliance, whose membership includes 2,300 hospitals and 70,000 outpatient facilities. Premier's leaders cite as their mission the goal of improving the efficiency of its member hospitals, as well as the health of the communities that the provider organizations serve.

Richard Bankowitz, M.D., Premier's enterprise-wide chief medical officer, believes that continuous quality improvement tools have a lot to offer health providers. As evidence, he points to the Hospital Quality Incentive Demonstration (HQID) Project, which Premier has been jointly running with CMS as a six-year demonstration project focused on evidence-based care. Top-performing hospitals have been those with a data-driven culture, using data to identify opportunities for performance improvement or monitor their progress, he says.


Premier has expanded on the demonstration project, forming a collaborative of 200 hospitals across the U.S. They share five goals: reducing preventable mortality, improving patient safety, increasing the amount of evidence-based care provided, improving patient satisfaction, and reducing cost. Data-driven quality improvement is at the core of the collaborative, Bankowitz says. “We focus on defining, measuring, and determining the definitions of what we want to measure, and we focus on setting targets that are quantifiable. And we focus on transparency, so that everyone in the collaborative knows where they are; everyone knows the top performer and the bottom performer; everyone knows we have pockets of excellence.”

Premier's vast databases contain about one in five discharges in the U.S. Using them as a basis, Premier has been able to show hospitals where they stand relative to top performers, risk-adjusted mortality, and see their performance according to clinical product line or department. “They are able to set performance thresholds and use this wealth of data to see if they are providing the best care possible,” he says.

According to Bankowitz, CMIOs have noted that their participation in the collaborative has allowed them to use standardized measurements; use transparency to create a climate of healthy competition; and accelerate improvements by sharing best practices. Barriers, of course, still exist. Among them, data can be difficult and labor-intensive to obtain; and actionable data has to be timely, accurate, placed in context of relative performance, and it must be easy to find opportunities. He adds that it's important for clinicians to focus on the big picture, by concentrating on best practices and figuring out how to reach that goal without getting caught up in nuances. It's also important to realize that not everything in healthcare can be standardized. “The question is, ‘How much variation is justified?’” he says.

Bankowitz says that being a data-driven performance organization doesn't necessarily mean hiring additional staff; but staff should be employed more effectively. Sometimes it could mean investing in products for automated data collection or data analysis, but these have a return by shifting labor from manual data collection, which that then can be shifted to bedside care.

Both EHRs and computerized physician order entry (CPOE) are helpful, he says. But it is also necessary to have the means to do near real-time data analysis to aggregate data and determine where opportunities exist. Also useful are tools that look at physician practice patterns, as well as those that go beyond CPOE to provide alerts for interventions. Overall, the most successful hospitals have support from their boards of directors.

Bankowitz maintains that during its first two years the collaborative has prevented or avoided 2,000 in-hospital deaths, and reduced the average cost of a discharge by $600 on an inflation-adjusted basis, for a total of $2 billion in savings.