Today’s data avalanche has left no business unaffected by the free flow of information, and that includes health...
care organizations striving to meet new government regulations.
While Danville, Pa.-based Geisinger Health System was ahead of the curve, having introduced electronic medical records (EMRs) 15 years ago, finding ways to manage a growing avalanche of data continued. Recently, Geisinger, a health care system serving more than 2 million people, focused its efforts on the pathology department, which performs tests on tissue and fluid samples to help diagnose disease. Its three departmental developers used Flash, Flex and ColdFusion to present data to medical professionals, but Geisinger realized the methodology was too laborious to match the growing demand and aimed to seek out new approaches.
“The requests were coming fast and furious,” said Al Shulski, Geisinger’s laboratory information systems (LIS) director. “It just took us too long to programmatically present that information using those applications.”
Data demands, government regulations
Geisinger’s pathology department has two major components: The clinical lab analyzes fluid samples like blood and urine, while the anatomic lab analyzes tissue samples. Geisinger has stayed on top of anatomic requests by using data visualization to program dashboards, but the clinical side posed more challenges.
Volume between the two, for example, varies greatly: The anatomic lab processes roughly 150,000 cases annually, whereas the clinical side can perform from 4 to 5 million lab tests a year.
“But volume is not the only factor,” Shulski said. “The tests, procedures and processes on the clinical side are far more varied and involve a good deal more resources -- expendable and human.”
Geisinger, listed as one of the nation’s most wired hospitals and health systems by Hospitals & Health Networks magazine for nine years in a row, knew it could hire more employees to meet the rising demand or find better tools to perform the work. It decided on the latter approach, spurred on by new government regulations and the introduction of penalties for underperforming institutions.
“Even on a good day, we’re not reimbursed enough to cover our costs from groups like Medicare,” Shulski said. “So any ratcheting back of what we’re getting is going to be even more of an impact. In order to keep up, you either have to find ways to be more efficient, keep your costs down and meet the regulations or you’re just going to fall behind.”
Testing out the market
When seeking out an informatics or business intelligence (BI) platform vendor, certain qualities topped the list of must-haves, such as solid efficiency in meeting requests and the ability to pull data from different systems within the organization.
“For years and years, Geisinger had a best-of-breed environment, where every department had their own system,” Shulski said. “A lot of that is now being integrated into the EMR system, but there’s still a lot of these independent systems and databases out there where data is housed.”
All data has value, Shulski said, but removing departmental silos so that data can be combined across the organization adds to its worth. In addition to data integration, Geisinger sought easy-to-create dashboards so end users could build their own instead of relying on technical staff.
“They would get the best results if they could design their dashboards to look the way they wanted them to look as opposed to trying to explain it to us and having us take an arbitrary crack at what they wanted,” Shulski said.
In the latter half of 2010, Geisinger searched the Internet for potential vendors, which turned up three or four top-of-the line BI systems (Shulski wouldn’t say which ones). In January, the health care organization held its first training session with Media, Pa.-based Altosoft Corp., an end-to-end BI vendor with a product designed specifically for pathology departments.
Altosoft’s Insight for Pathology met Geisinger’s data integration requirement without the need for a separate extract, transform and load (ETL) tool.
“It’s given us the ability to expand the information we provide to our folks because we can now dig into the information in indirect databases,” said Shulski. “That information is vital but was a little out of our reach.”
It also supported both the anatomic and clinical lab components with a set of prebuilt dashboards, helping to more quickly turn information around. Ideally, Geisinger’s highly technical workers will tend to the data integration aspects, and end users will customize the prebuilt dashboards to fit their own requirements.
While end users have not built their own dashboards to date, Shulski said training is under way.
“I believe there are quite a few people in our lab that have the ability and that, in turn, would take some of the heat off of us as far as utilization of our resources to [build dashboards],” he said.
Finally, Shulski said, the system has helped keep tabs on staffing levels and turnaround times for lab tests, flagging those that aren’t meeting the prescribed turnaround times.
The final hurdle: Talent
Recently, the one full-time employee in charge of extracting information and putting the metrics together left the pathology department for another position. Replacing that loss has been a challenge for Shulski.
“It’s like people have been starved for this kind of information for years and years, and when we finally gave it to them, the floodgates opened up and they saw a new array of things they could ask,” he said.
Shulski said he wants someone with a basic data analytics and analysis background, with the skills to visualize how data can be presented and with laboratory experience to help translate what the lab wants.
“They are not very plentiful out there,” he said. “It is on Geisinger’s list as one of the hard-to-recruit positions.”