Data Warehouses Provide Foundation For New Business Intelligence
Meaningful use and an increasingly competitive healthcare market is driving the need for the capability of 360° data analysis. Electronic data warehouses may provide the means.
Northeast Georgia Medical Center (NGMC) in Gainesville, Ga., created a virtual data warehouse because they found that a traditional data warehouse was “a little too cumbersome to respond to some of the frontline needs,” says James Bailey, MD, MPH, CMIO at NGMC.
There was a general feeling that the IT team was overloaded with report requests, he says. The goal was to establish a system that allowed staff to run their own queries. Today, “my data services team tries to manage the extraction and data quality and provide tools that allow them to run their own queries,” says Zan Miller, MS, chief data architect at NGMC. Providing querying tools allows a much more rapid turnaround of requests.
A big plus has been the ability of users to see their data quality, says Miller, and determine whether the data are complete and accurate. “We can quickly put processes in place to improve data quality,” rather than waste time analyzing inaccurate data. He’s also improved his efficiency by avoiding situations in which “we get three requests and build the same things using different tools.”
The warehouse has been very helpful in the computerized physician order entry (CPOE) rollout, says Bailey, since they could drill down to individual physician utilization.
Another use is finding out where better data collection is needed. For example, Miller says they’ve been looking at medication variances requiring pharmacist intervention. “We have found that the collection of a couple more numbers, such as order number, enables us to leverage a bigger CPOE tool. Now, we can determine where those variances are and the cause, whether it’s computer entry, written or a verbal order.”
Healthcare also is underinvested in IT. In fact, Biehl compares the current health IT with corporate IT of the 1980s—a perspective that has served him well in working with healthcare organizations.
Another difference between corporate and health IT is data status. Less than one-tenth of corporate data are bad, Biehl says, but in healthcare, closer to 90 percent of data are inaccurate and there is no one to clean it up. “We had to invent ways to build better warehouses, and be more flexible in bringing in data no matter how unclean they are.”
Hospitals undergoing EMR and CPOE implementations will “give them a level of IT that will make it meaningful to talk about combining all of those data in a warehouse.” Vendor warehouses aren’t delivering and, despite big investments, users still can’t perform some of the analytics they want to do.
Biehl’s first healthcare data warehouse was for a New York hospital six years ago and, going in with no healthcare experience, he found the experience “very humbling. I thought it would be a piece of cake, but healthcare data was clearly different from corporate data.” In fact, most of his assumptions weren’t true: Not every drug ordered had a valid code in the pharmacy system and not every medication administered had a valid order.
While data clean-up is a big challenge for IT, the implementation of a data warehouse shouldn’t have much impact on physicians, Biehl says. It should be used to analyze charts in an automated way, such as reviewing the top 10 DRGs by clinical specialty or utilization of a surgical room by DRG. Physicians can compare their own practices with best practices across the warehouse.
Eric Widen, MHA, most recently director of performance improvement for Lucile Packard Children’s Hospital and now managing director of HBI Solutions, both in Palo Alto, Calif., founded new business specifically because of the new data warehouse market. “The market is rapidly growing, and we’re developing leading solutions for the space,” he says.
Providers have their hands full implementing EMRs, Widen says, while leading institutions are thinking about how to put together a solid business intelligence and data management program. Federal stimulus money for EMRs will continue and in the next five to seven years, “we’ll see more electronic data warehouse work. The logical next step is providing that infrastructure.”
Business intelligence seeks to help organizations use data to drive business forward, he says. Many facilities install an enterprise data warehouse for the wrong reasons, because they haven’t considered how the business intelligence/enterprise data warehouse connects to their enterprise strategy.
Widen cites two problems. First, users must realize that enterprise data management should be the underlying program while the data warehouse is just a tool. The program involves integrated, trustworthy data that people are educated about and buy into. Those trustworthy data provide a platform to perform analytics to find opportunities, study trends, write reports and produce key performance indicators.
Another problem is that most people expect someone else to perform data analytics for them. However, “people on the frontlines are most in tune with the data they need,” Widen says. To work effectively, he says organizations need business analyst staff who know how to connect the dots to solve operational problems. Healthcare hasn’t invested in those people to the degree required. “These [specialists] are worth their weight in gold. They know how to put together analytics to identify and solve problems.”
Meaningful use requirements will drive data warehouse efforts in the future, Biehl says. “I’m seeing a push and anxiety among IT people because they need to get a warehouse in place or they will never be able to consolidate data.” Meanwhile, data haven’t excited executives in the past, he says, but meaningful use has money attached to it, and “now executives are listening to IT about consolidating data. Meaningful use will drive business intelligence.”
Northeast Georgia Medical Center (NGMC) in Gainesville, Ga., created a virtual data warehouse because they found that a traditional data warehouse was “a little too cumbersome to respond to some of the frontline needs,” says James Bailey, MD, MPH, CMIO at NGMC.
There was a general feeling that the IT team was overloaded with report requests, he says. The goal was to establish a system that allowed staff to run their own queries. Today, “my data services team tries to manage the extraction and data quality and provide tools that allow them to run their own queries,” says Zan Miller, MS, chief data architect at NGMC. Providing querying tools allows a much more rapid turnaround of requests.
A big plus has been the ability of users to see their data quality, says Miller, and determine whether the data are complete and accurate. “We can quickly put processes in place to improve data quality,” rather than waste time analyzing inaccurate data. He’s also improved his efficiency by avoiding situations in which “we get three requests and build the same things using different tools.”
The warehouse has been very helpful in the computerized physician order entry (CPOE) rollout, says Bailey, since they could drill down to individual physician utilization.
Another use is finding out where better data collection is needed. For example, Miller says they’ve been looking at medication variances requiring pharmacist intervention. “We have found that the collection of a couple more numbers, such as order number, enables us to leverage a bigger CPOE tool. Now, we can determine where those variances are and the cause, whether it’s computer entry, written or a verbal order.”
Transition from corporate world
Electronic data warehouses got their start in banking and other industries. “Data warehousing caught on in the corporate world because, by definition, it’s designed to pull data in from lots of sources,” explains Richard E. Biehl, PhD, University of Central Florida’s department of health management and informatics and owner of Data-Oriented Quality Solutions in Orlando, Fla. However, the proprietary nature of most health IT products means most healthcare systems cannot bring outside data into their EMR.Healthcare also is underinvested in IT. In fact, Biehl compares the current health IT with corporate IT of the 1980s—a perspective that has served him well in working with healthcare organizations.
Another difference between corporate and health IT is data status. Less than one-tenth of corporate data are bad, Biehl says, but in healthcare, closer to 90 percent of data are inaccurate and there is no one to clean it up. “We had to invent ways to build better warehouses, and be more flexible in bringing in data no matter how unclean they are.”
Hospitals undergoing EMR and CPOE implementations will “give them a level of IT that will make it meaningful to talk about combining all of those data in a warehouse.” Vendor warehouses aren’t delivering and, despite big investments, users still can’t perform some of the analytics they want to do.
Biehl’s first healthcare data warehouse was for a New York hospital six years ago and, going in with no healthcare experience, he found the experience “very humbling. I thought it would be a piece of cake, but healthcare data was clearly different from corporate data.” In fact, most of his assumptions weren’t true: Not every drug ordered had a valid code in the pharmacy system and not every medication administered had a valid order.
While data clean-up is a big challenge for IT, the implementation of a data warehouse shouldn’t have much impact on physicians, Biehl says. It should be used to analyze charts in an automated way, such as reviewing the top 10 DRGs by clinical specialty or utilization of a surgical room by DRG. Physicians can compare their own practices with best practices across the warehouse.
Increasing interest
That kind of analysis only will grow in the future, and Bailey anticipates an increased interest in data. “Physicians want to know how they’re performing.” The data warehouse will help answer their detailed performance questions. It goes way beyond core measure compliance, he says. “Surgeons have told me they want to know about their length of stay on certain procedures. Business intelligence actually will be clinical medical intelligence.”Eric Widen, MHA, most recently director of performance improvement for Lucile Packard Children’s Hospital and now managing director of HBI Solutions, both in Palo Alto, Calif., founded new business specifically because of the new data warehouse market. “The market is rapidly growing, and we’re developing leading solutions for the space,” he says.
Providers have their hands full implementing EMRs, Widen says, while leading institutions are thinking about how to put together a solid business intelligence and data management program. Federal stimulus money for EMRs will continue and in the next five to seven years, “we’ll see more electronic data warehouse work. The logical next step is providing that infrastructure.”
Business intelligence seeks to help organizations use data to drive business forward, he says. Many facilities install an enterprise data warehouse for the wrong reasons, because they haven’t considered how the business intelligence/enterprise data warehouse connects to their enterprise strategy.
Widen cites two problems. First, users must realize that enterprise data management should be the underlying program while the data warehouse is just a tool. The program involves integrated, trustworthy data that people are educated about and buy into. Those trustworthy data provide a platform to perform analytics to find opportunities, study trends, write reports and produce key performance indicators.
Another problem is that most people expect someone else to perform data analytics for them. However, “people on the frontlines are most in tune with the data they need,” Widen says. To work effectively, he says organizations need business analyst staff who know how to connect the dots to solve operational problems. Healthcare hasn’t invested in those people to the degree required. “These [specialists] are worth their weight in gold. They know how to put together analytics to identify and solve problems.”
Meaningful use requirements will drive data warehouse efforts in the future, Biehl says. “I’m seeing a push and anxiety among IT people because they need to get a warehouse in place or they will never be able to consolidate data.” Meanwhile, data haven’t excited executives in the past, he says, but meaningful use has money attached to it, and “now executives are listening to IT about consolidating data. Meaningful use will drive business intelligence.”
Beth Walsh,
Editor
Editor Beth earned a bachelor’s degree in journalism and master’s in health communication. She has worked in hospital, academic and publishing settings over the past 20 years. Beth joined TriMed in 2005, as editor of CMIO and Clinical Innovation + Technology. When not covering all things related to health IT, she spends time with her husband and three children.