The financial world buzzes with banks crunching numbers in real time to reduce risk and find opportunities. SAS expects that soon hospitals will do that too.

The Cary business intelligence and analytics software company is moving to bolster its position in the health care space with new tools that allow hospitals to analyze data “big data” in real time and use that analysis to more readily address problems.

Graham Hughes, chief medical officer for the Cary business intelligence and analytics software company, says hospitals today take a “rear-view mirror” approach to analytics, using data to look back at how they’ve performed. But hospitals can act more quickly if they take a forward looking, predictive approach to their data. Predictive analysis of the data will help hospitals improve on measures such as medical errors and patient readmission rates.

“Those are the things that you can do if you’re running large quantities of data in near real time,” Hughes told WRALTechWire. “You see patterns quickly, you can predict quickly and you can intervene quickly. That’s the kind of nimble environment that gets created.”

SAS executives announced the health care effort Wednesday during the company’s annual Health Care & Life Sciences Executive Conference. The conference drew an audience of about 200 to SAS’s campus for presentations focused on health analytics.

Predictive analytics leader

The new SAS health care analytics tools could help the company keep its lead in the market for predictive analytics. SAS holds 35 percent market share in predictive analytics, said SAS Chief Marketing Officer Jim Davis. IBM has 14 percent; Microsoft has just 4 percent.

SAS and IBM are the “unshakable leaders” in predictive analytics, Forrester Research analyst Mike Gualtieri said in a recent report. Forrester ranks SAS ahead of IBM, in part due to the performance of its software across a wide range of categories. Forrester estimates that SAS has 3,000 customers using its predictive analytics tools. By comparison, IBM has an estimated 1,500 predictive analytics customers.

Davis characterizes the new health care analytics tools as an evolution from SAS’ business intelligence work, which yielded actionable insights by through analysis of a company’s data. As companies produced more data, business intelligence became “big analytics.” But with the amount of data produced now growing exponentially every year, what is now commonly called “big data” has become a big obstacle, slowing many companies as they are forced to devote resources and time to analyze it.

Predictive analytics is different from the business intelligence tools that companies use to report on their business, Forrester’s Gualtieri said in the report. The deeper analysis from predictive analytics finds patterns that business intelligence might not reveal. And predictive analytics is an ongoing process; companies analyze data weekly or even continuously.

“Big data is the fuel and predictive analytics the engine that firms need to discover, deploy and profit from the knowledge they gain,” Gualtieri said.

Roots in financial services

The health care applications of SAS’ new analytics tools evolved from work that SAS was doing in financial services. SAS CEO Jim Goodnight told conference attendees that about four years ago, a banker in Singapore lamented to him about the 18 hours it took the bank’s software to calculate its reserves.

Goodnight recalled that another bank would use analytics to gauge the likelihood of defaults in its 2 million loans. The bank would analyze just a small sample, and even that would take up to three days. SAS can now analyze all of that bank’s loans, not just a sample, resulting in better understanding of default risks. Just as important, Goodnight noted, SAS has cut the analysis time down to 20 seconds.

Speed matters in health care, too. Analysis of health care data that takes days, even weeks, slows a hospital’s ability to make decisions about its operations. Hughes said that analytics will enable hospitals to look for patterns. From the patterns, it can do predictive modeling. For example, when a hospital admits a patient with a particular condition, it can run an analysis of what happened with patients that the hospital has admitted with the same condition. The model could guide patient care, reducing the likelihood of readmission.

“Predictive modeling is a a guide to prevent a future you don’t want to see,” Hughes said.

[SAS ARCHIVE: Check out more than a decade of SAS stories as reported in WRALTechWire.]