Progressive health systems are relying on data analysts to improve clinical and financial performance and help prepare their organizations for value-based care and population health management.
At a Glance
Data analysts can help organizations identify opportunities to improve clinical and financial performance. Hospitals should seek data analysts who possess:
- Knowledge of multiple data sets
- A math, science,computer science, or clinical background
- Healthcare industry experience
- Critical and systems thinking skills
The next priority position for hospitals and health systems to fill may be that of a talented data analyst. Across the country, data analysts are helping healthcare organizations apply new business intelligence tools to identify and prioritize opportunities to improve clinical and financial performance. For example, many analysts are supporting organizational initiatives to reduce variation among clinicians, which could help improve care and control costs.
“Reduction in practice variation is a huge competitive advantage,” says Steven Heck, president of MedSys Group, Plano, Texas. “Healthcare workflow and technology can no longer be separated. In a market where a growing percentage of fixed assets are spent on technology, data analytics becomes a more significant competitive factor.”
Although many healthcare organizations are still struggling to implement electronic health records (EHRs) and meet meaningful use requirements, some leading systems are enhancing their analytics functions to prepare for the changing healthcare environment.
Transforming Care at Dignity Health
The influence of data analysts in healthcare organizations is growing, according to Steven Zoner, senior director of decision support and business intelligence at San Francisco-based Dignity Health. “I used to work primarily for people in finance,” says Zoner, a 30-year veteran in healthcare analytics. “But now, I spend just as much time with people in quality, strategic planning, and human resources. And my job is about much more than just cost accounting. It’s now about solving information problems.”
Dignity Health, the fifth-largest health system in the nation, employs data analysts at the corporate level to support budgeting, strategic planning, quality, and clinical efficiency initiatives. These corporate data analysts work in tandem with local data analysts at Dignity Health’s 37 hospitals, spread across three states.
In recent years, Dignity Health has made significant investments in back-end performance management software that integrates clinical, financial, and operational data. Clinical efficiency analysts—known as “transformational care analysts”—use that software to identify variations in outcomes and identify which clinicians to target for outreach. The transformational care analysts report to the chief operating officer and are supported by a team of physicians who help translate the data for other clinicians. Recently, these analysts and physicians collaborated on an initiative to reduce inappropriate ICU utilization across the system. In FY13, Dignity Health reduced its systemwide ICU length of stay (LOS) by 9 percent, which saved more than $10 million across the organization.
Analytics software also has helped Dignity Health identify opportunities for savings in blood product use during surgery. By reducing variation among hospitals, Dignity Health was able to save almost $4 million in FY13. “Once we had the data to show physicians, it was relatively easy to influence their behavior,” Zoner says.
Dignity Health also is starting to dig deeper into outpatient data. Currently, the analytics team can “link” patient encounters to identify high utilizers of emergency services and share the data with clinicians, who then help patients access more appropriate levels of care.
To support its analytics efforts, Dignity Health is designing an enterprise data warehouse that will store the system’s vast clinical and financial data in one place. The data warehouse will allow analysts to access more granular data from the system’s new EHR, such as lab results, surgical times, and other data that are currently not accessible through the performance improvement software, Zoner says. Once the data warehouse is operational, Dignity Health will test its capabilities by focusing on an initiative to reduce heart failure readmissions and the associated costs across the system by approximately 4 percent. Analysts will pull data from the warehouse to compare costs, LOS, and inpatient mortality outcomes among both utilizing and nonutilizing physicians of the system’s pathway for heart failure patients. New software also will make it possible to attribute expenses to individual providers, instead of attributing all cost reports to the attending physician or hospitalist group, as is currently done.
The health system also plans additional clinical improvement efforts focused on reducing sepsis across the organization. Zoner says the data warehouse will help Dignity Health develop interventions for at-risk patients that could significantly reduce morbidity and save $6 million to $10 million of associated cost each year.
Despite this progress, Zoner concedes that a continuing obstacle to better performance through the use of analytics is its hospitals’ use of different patient accounting, admission/discharge/transfer (ADT), and medical records platforms. “Data governance is a struggle,” Zoner says. “We are challenged to find that single source of truth.” To minimize data integration obstacles, Dignity Health developed a data governance council to standardize data definitions for financial, clinical, and patient encounter data for use across the system.
Using Dashboards at CHRISTUS
George Conklin, senior vice president and chief information officer (CIO) at CHRISTUS Health, Irving, Texas, agrees with Zoner that a key challenge for data analysts is pooling data from a variety of inpatient and outpatient systems. Like Dignity, CHRISTUS aims to reduce obstacles to compiling data, in part through its ongoing data standardization efforts. “To some degree, we still have dueling databases that produce different numbers on the same things, but this happens much less often than in the past,” Conklin says.
Since adding a new enterprise data warehouse and a health intelligence tool in 2013, the Catholic health system, which has more than 30 hospitals in the United States and Mexico and a joint venture in Chile, has used new resources primarily to improve clinical quality rather than financial outcomes. “To talk to clinicians about abstruse financial indicators without any association to quality doesn’t get you anywhere, Conklin says. “We think there is more mileage in addressing clinical behaviors first, because clinical behavior is ultimately what is going to ensure our delivery on our mission and vision and drive cost.
For example, CHRISTUS has used analytics to help “connect the dots” between the system’s implementation of what CHRISTUS calls computerized patient order management (CPOM)—also known as computerized provider order entry—and outcomes, such as reduced mortality, LOS, sepsis, and readmissions. To help clinicians “visualize” the impact of their CPOM usage on these quality measures, CHRISTUS purchased a tablet-based app that sits on top of their data repository and creates a dashboard of key metrics. “What we have been able to show is a very clear connection between higher levels of CPOM adoption and lower mortality, fewer readmissions, and other positive outcomes,” Conklin says.
Although CHRISTUS has no current plans to form an accountable care organization (ACO), the system plans to accept more financial risk from payers, Conklin says. To accommodate more risk, CHRISTUS is adding a population health management tool to help its data analysts identify physician behaviors in the community that could affect the cost and quality of care. For example, the population health management tool, combined with the other business intelligence tools, will allow CHRISTUS to determine highest quality and lowest cost service plans for members/patients and to promote those across the system. The tools also will span the entire service continuum, from healthcare services provided through wellness services delivered in people’s homes, to ensure optimum outcomes. Last, the tools will provide CHRISTUS with the capability to more effectively stratify and manage risk in the increasingly complicated—and competitive—healthcare marketplace, Conklin says.
CHRISTUS employs approximately 65 data analysts to support its performance improvement initiatives. These analysts serve in one of three main roles:
- Business intelligence (15 analysts)
- Health informatics (40 nurse informaticists, pharmacists, and other clinicians)
- Strategy, such as service line expansions in specific markets (10 data analysts at the corporate and regional level)
Although these professionals are dedicated to analytics, Conklin says that his organization increasingly will rely on clinicians and staff at every level to use data to improve performance. “Our philosophy is that we want to keep the tools as close as we can to the users and decision makers,” Conklin says. “In some ways, we have 15,000 data analysts, because everyone in the organization has access to tools to produce their own analytics.”
Consolidating Analytics at UAB
Many health systems’ data analytics expertise is spread across the organization, with business analysts and quality analysts focused on different pieces of the puzzle. Until recently, diffuse analytics capabilities were prevalent at the University of Alabama at Birmingham (UAB) Health System, Birmingham, Ala., one of the five largest academic medical centers in the United States. But UAB is moving its data analytics resources under one umbrella, says Joan Hicks, CIO.
By consolidating UAB’s analytics functions, the health system’s leaders aim to allow its data analysts to share information and resources across departments, Hicks says. Until recently, UAB’s data analysts handled data requests independently, which meant data analysts for the cost accounting system might not have shared information with analysts in health information services. “Many times, we pull data for the same reason and get different answers. Now, we’re undertaking an initiative to make sure we understand the sources of data and are able to get the same answer, regardless of who is performing the analysis,” Hicks says.
As part of the information-sharing initiative, UAB will cross-train its data analysts to use different tools and data sets from across the organization. “We hope to take the individual problem-solving skills of each group and start applying them more broadly across the enterprise,” says Geoff Gordon, IT director. UAB officials also hope that the common training will help analysts connect the clinical and financial data sets, which provide two different views of the same patient encounter. “To improve quality and financial efficiency, you want to synthesize these two views so you get a whole view of that encounter,” Gordon says.
The health system also employs nurse informaticists and other clinicians to work in tandem with data analysts and serve as operational process analysts. “They are really good at optimizing and understanding clinical workflows,” Gordon says, “but they also understand the IT side and can help use the data to propose solutions for changing behavior on the units and in the clinics.” The clinical informaticists are stationed in UAB’s nursing, pharmacy, radiology, laboratory, and ambulatory clinics, rather than in IT. The department-based distribution helps enhance the collaboration between the clinicians and informaticists and fosters a sense of ownership for change within each unit, he says.
As in other organizations, data integrity is one of the biggest challenges for data analysts at UAB, says Suzanne E. Daniel, director of health information. To meet the data integrity challenge, UAB’s health information management (HIM) department has added a manager of data integrity and quality to bridge IT and finance. UAB also added several HIM leaders with master’s degrees in informatics. “Moving forward, there will need to be more cohesiveness between IT and HIM because this can have a tremendous impact on organizational success,” Daniel says.
The Ideal Data Analyst
The data analyst characteristics most valued by hospitals and health systems today reflect a broader set of skills and industry knowledge than these organizations might have sought in these individuals a few years ago, experts say. Broadly, healthcare leaders are looking for the following critical qualities in potential analysts.
Knowledge of multiple data sets. “Although data analysts in the past needed to be experts in only one data set, today’s analysts need to be able to understand multiple data sets from different systems, venues, and stakeholders,” says Asif Dhar, principal, Deloitte Consulting LLP, and managing director, Deloitte Health Informatics LLC, McLean, Va. “Ultimately, the goal of analytics is not to provide insight into how to do more with less, but how to do more at the right place and at the right time.”
A math, science, computer science, or clinical background. Analysts with a strong statistics background are especially valuable to healthcare organizations because these individuals are comfortable mining large data sets to find correlations, Dhar says.
Healthcare industry experience. “We look for analysts with a background in clinical informatics or healthcare finance and cost accounting,” says Zoner of Dignity Health. “It’s almost impossible to find someone with both, such as a clinician who knows cost accounting. If we can find someone with one background or the other, we can teach them what they don’t know. But it’s really hard for someone to come in from another industry.”
Critical and systems thinking skills. “Data analysts need to be able to understand how to pull data from different entities into a meaningful whole,” says Conklin of CHRISTUS. “And those who can will prove to be extremely valuable to their organizations.”
Laura Ramos Hegwer is a freelance healthcare writer and editor based in Lake Bluff, Ill.
Publication Date: Monday, February 03, 2014