Three Health Catalyst Clients Recognized for Data-Driven Quality Success (HFMA)
Data Key to Quest for Quality
THREE HEALTH SYSTEMS USED A FORM OF DATA WAREHOUSING UNIQUE TO HEALTH CARE TO GAIN GREATER INSIGHT INTO CLINICAL WORKFLOWS AND IMPROVE PERFORMANCE.
At a Glance
- Late-binding data warehousing reduces the time it takes to obtain data needed to make crucial decisions.
- Late binding refers to when and how tightly data from the source applications are bound to the rules and vocabularies that make it useful. In some cases, data can be seen in real time.
- In historically paper-driven environments where data-driven decisions may be a new concept, buy-in from clinicians, physicians, and hospital leaders is key to success in using data to improve outcomes.
Rapid advancements in digitizing, integrating, and exchanging health information have given healthcare leaders access to an unprecedented volume of data to drive decisions. But with the amount of raw data bordering on information overload, the challenge these leaders face is finding the best way to use these data to improve operations and patient care.
Three healthcare organizations—Texas Children’s Hospital, MultiCare Health System, and North Memorial Health Care—recently took on this challenge, using a form of data warehousing specific to health care to capture data, and then forming teams of specialists from throughout their organizations to find ways to use the data to improve outcomes. The lessons these organizations learned could help other hospitals and health systems unlock the potential of the data captured through their data warehouses and make gains in clinical and financial performance.
A Healthcare-Specific Approach to Data Warehousing
All three organizations first implemented an enterprise data warehouse. Although commonplace in other industries, enterprise data warehouses have presented a challenge for hospitals and health systems because so much clinical information is in the form of “unstructured data,” such as nurse’s notes and physician progress notes. Yet healthcare analytics and quality improvement programs require integrated clinical, financial, and patient satisfaction data.
To solve this puzzle, all three organizations chose a form of data warehousing designed specifically for health care called a late binding data warehouse. While early binding of data can take months and even years to deliver value and is hard to adapt to evolving uses, a late-binding model assembles data from source applications just in time to address new analytic uses.
Proponents of this approach cite its agility as a principle benefit.
Although establishing a late-binding data warehouse was the engine behind each organization’s quality improvement efforts, sustained improvement would not have been possible if the insights gleaned from the data had not been integrated into clinician workflows. To accomplish this goal, all three organizations developed teams of clinicians, technologists, analysts, and quality personnel to turn new insights from the data into better clinical processes.
Although each organization applied the strategies to very different quality initiatives, all three achieved quick success.
Case Study: Texas Children’s Hospital
Texas Children’s Hospital is one of the premier children’s hospitals in the United States: US News & World Report recently ranked it the nation’s fourth best children’s hospital and the best in Texas. In 2006, with value-based payment on the horizon, Texas Children’s evaluated its quality-improvement program and data management capabilities. Knowing they needed hard data to identify inefficiency and to manage high-risk populations of patients, hospital leaders decided to install an enterprisewide electronic health record (EHR).
These leaders quickly found they had a wealth of data reports that, in most cases, clinicians did not find useful. Efforts to analyze the data and track outcomes were slow and inefficient, often taking as long as six months to generate actionable information.
But in 2011, Texas Children’s implemented a late-binding enterprise data warehouse platform—a process completed in just 90 days. The hospital then used the data to perform a financial and clinical assessment across the enterprise that examined variability of care and resource consumption.
The assessment revealed many opportunities, including a significant quality-improvement opportunity related to asthma care. The data showed that physicians were ordering a high volume of chest X-rays for asthma patients (in 70 percent of the cases) when only 5 to 35 percent of these X-rays were indicated. Decreasing the volume of unnecessary chest X-rays became the hospital’s first target.
Following a protocol developed by its data warehouse vendor, Texas Children’s put together an integrated team of clinicians, technologists, analysts, and quality personnel to translate its care guidelines into the X-ray workflow. In the past, physicians had been quick to deny the validity of data that questioned their clinical decisions. This time, the team convinced clinicians to change their ordering behaviors by drilling down into near-real-time data.
Change happened quickly: The team produced a 15 percent reduction in unnecessary chest X-rays in just 45 days. The ordering rate subsequently dropped by 30 percent in the months that followed.
Texas Children’s clinical teams are actively using data to improve care for patients by asking better questions and uncovering the root causes of variation in care. Near-real-time access to data has accelerated the process. Rather than waiting six months for useful data about patient outcomes, the team receives actionable information updated within the last 24 hours.
Texas Children’s is now focused on improving outcomes for appendectomy, spine surgery, pneumonia, diabetic ketoacidosis and expanding beyond hospital-based care to include primary pediatric practices and clinic-based care. The teams also plan to implement registries for the hospital’s 38 guidelines of care by the end of the year.
Case Study: MultiCare Health System
MultiCare is an integrated health organization in the Tacoma, Wash., area composed of five hospitals, numerous outpatient specialty centers, primary and urgent care clinics, as well as a variety of other services and community outreach programs. The health system is consistently on the cutting edge of IT advancements and has been recognized as one of the nation’s Most Wired healthcare organizations by Hospitals & Health Networks.
When it discovered one of its hospitals performing below national mortality averages for septicemia, it decided to take action.
Implementing a late-binding enterprise data warehouse proved to be an important means to unify clinical, IT, and financial leaders and drive accountability for performance improvement. Like Texas Children’s Hospital, MultiCare established teams of clinicians, technologists, analysts and quality personnel to collaborate on reducing septicemia mortality. One of the first tasks was to analyze the data to refine the clinical definition of sepsis, a difficult undertaking given the complex comorbidity factors that lead to
Next, the teams established a systemwide critical care collaborative to integrate the new definition into clinical and process improvement. The collaborative focused on three tasks: establishing a systemwide standard for the care of severely septic patients, developing an early-warning dashboard to identify patients at risk of becoming clinically unstable, and ensuring timely implementation of the defined standard of care for all identified patients.
In just 12 months, MultiCare was able to reduce septicemia mortality rates by an average of 22 percent, leading to more than $1.3 million in cost savings. Hospital leaders are confident similar results can be realized in other areas, including heart failure, emergency department performance and inpatient throughput.
Case Study: North Memorial Health Care
A 518-bed, two-hospital system in the Minneapolis-St. Paul metro area, North Memorial Health Care was struggling with rising costs, stiff competition, and unpredictable payment. At the same time, difficulty collecting and analyzing data from its EHR and other IT systems left hospital leaders with an incomplete view of their financial pain points and opportunities for improvement.
Using a late-binding data warehouse, hospital leaders identified elective deliveries prior to 39 weeks of gestation as the area with the greatest opportunity for improvement and financial return. Although North Memorial’s elective delivery rate was relatively low at 1.2 percent, further incentive to act came from a payer partner that promised a bonus for cutting the rate to 0.6 percent.
North Memorial leadership established a service line guidance team comprised of OB/GYNs, primary care physicians, nurses, data architects, and outcomes analysts. Their task was to define when early term deliveries were appropriate, standardize workflow, and create improved processes for pregnant women and newborn care.
Using the data warehouse to analyze existing workflow and performance measures, the team defined a new process for managing elective pre-39-week deliveries. Nurses now must use a checklist of requirements for medical necessity to schedule early-term deliveries. Any scheduled elective deliveries that make it past the checklist are immediately referred to the medical department chair.
Gaining physician buy-in was essential. To get it, North Memorial leaders attended staff meetings to explain the reasoning behind the changes and share the data. They accelerated buy-in by appealing to their physicians’ competitive natures: each clinic posted internal provider report cards listing the number of elective pre-39-week deliveries approved by each provider.
North Memorial reduced the rate of elective early-term deliveries by 75 percent in six months. The decrease from 1.2 percent to 0.3 percent was well below the payer’s goal and earned the hospital a significant bonus payment.
Hospital leaders are now expanding the program and approaching other payers about entering into shared-savings contracts for additional care processes.
A Data-Smart Approach to Reform
Information and communication technology have intensified the problem of information overload in virtually every industry. The healthcare industry, with all of its complexity and technological variability, is a relative latecomer to their problem, but leaders must find ways to make the data work for them.
The three organizations highlighted here quickly discovered that unlocking the potential of the data captured through a data warehouse is vital to thriving in the industry’s value-based future. The key to success for each of these organizations was forming teams of people who worked together to identify ways to use the data to improve outcomes and who gained buy-in from physicians and hospital leaders in putting the data to use.
Florence S. Chang is executive vice president, MultiCare Health System, Tacoma, Wash..
Charles Macias, MD, MPH, is associate professor of pediatrics, director of the Center for Clinical Effectiveness, and director of the Evidence Based Outcomes Center at Baylor College of Medicine/Texas Children’s Hospital, Houston.
Jon Nielsen, MD, is medical director of women’s and newborn services, North Memorial Health Care, Maple Grove, Minn., and a national lecturer and physician instructor on minimally invasive surgery.
Publication Date: Friday, November 01, 2013