By using website you agree to our use of cookies as described in our cookie policy. Learn More


Cardiac Patient Data Helps Small Hospital Compete

By Linda L. Briggs

With healthcare in the national spotlight these days, a small San Francisco hospital has shown that good data and analytics can improve outcomes for cardiovascular patients. In the past five years, the hospital has reduced the mortality rate in its cardiac surgery by more than 50 percent.

Sequoia Hospital is nestled on the San Francisco peninsula, close to nationally renowned hospitals and medical centers such as Stanford University and the University of California San Francisco. To survive in that climate, the hospital has developed a world-class cardiovascular program with a backbone of rich stores of patient data and powerful analytics software.

Thanks to more than 1,000 employees and nearly 400 staff positions, Sequoia Hospital handled over 250,000 patient visits in 2010. The hospital is part of a larger health system, Catholic Healthcare West, a not-for-profit with 42 hospitals and medical centers in California, Arizona, and Nevada. The hospital performs procedures ranging from stents and catheterizations to valve replacements, angioplasties, and coronary bypass surgeries.

According to Audrey Fisher, director of cardiovascular services at Sequoia, IBM SPSS analytics software provides a crucial competitive edge to the small community hospital. Its cardiovascular program is built on referrals from all over the western U.S. “This is one way we’ve maintained our competitive edge,” says Fisher. “We have to have a compelling reason for physicians to send patients to us.”

Surgeons at Sequoia use predictive analytics to help determine if a patient is a suitable candidate for surgery, and how to best manage the case. The software compiles patient factors such as age, weight, current health, and previous surgeries, then analyzes and compares the data to similar national and local cases. The resulting prediction of outcome and risk helps doctors provide the best recommendation for patients and families.

Rich in Data, Poor in Analytics

When Fisher joined the hospital in 1998, software was in place for collecting information, but there were limited ways to access and analyze the data. “We had this great, rich repository of information,” she says, “but we did not have easy and timely access to it.”

If a doctor needed a statistic such as the five-year mortality rate in diabetic patients under a certain type of care, Fisher would call the software company, request a custom report on disk, and wait for it to arrive in the mail. “Maybe three weeks later, I would have an answer for the physician,” she says. “By then, they’d forgotten why they’d even asked me in the first place.”

Now, when doctors come to Fisher with requests for information, she can provide it quickly enough that a doctor in consultation with a patient can get specific information quickly on a question such as “How many double-valve operations on patients over the age of 80 have we done in the last five years, and what are our results so I can tell Mr. Smith, who’s sitting in front of me?”

SPSS, an IBM product since 2009, was first installed around 2000, when Fisher was still relatively new to the hospital. In over 10 years of use, the product and its application at Sequoia have evolved, she explains, but the results continue to speak for themselves.

A so-called referral center, Sequoia Hospital must give physicians a strong reason to refer patients to a hospital outside their local communities, since doing so often has both a financial and political cost for the doctor. “We have to demonstrate that there is some compelling reason for them to send their patients to us,” Fisher says. For Sequoia, that reason has become the proven outcomes the hospital has achieved in cardiac surgery—outcomes that it has been able to continue demonstrating to referring physicians and to the community.

Sequoia’s sophisticated use of its data has caught the attention and interest of other hospitals and medical centers. One example of how the hospital stays competitive by making good use of data on cardiac patient outcomes is a “results brochure” that is produced every other year and sent to patients and referring physicians. The hospital has been distributing the brochure for six or seven years now, and Fisher hears frequently from other hospitals that would like to do the same thing. “It’s another thing we’ve been able to do because we have SPSS, and we’re able to analyze all of our information.” The brochure, she says, “has really helped us keep our competitive edge.”

The hospital’s smart use of data and analytics is just one link in a long chain of factors that contribute to the success of its cardiovascular program, but Fisher calls the analytics software the program’s headlights. “It illuminates the path for us and lets us know what’s out there, through our data, and it lets us demonstrate clear results to ourselves and to others.”

For example, she cites high-risk patients who have what’s known as co-morbidity conditions, such as diabetes and lung issues. Managing those conditions carefully before patients come to the hospital for surgery so that they are in optimal health before the operation has meant much better outcomes post surgery.

“That’s been one of our big epiphanies over the last few years,” she says. “It’s really been guided by our data.” Those patient populations that are extremely high-risk, she explains, tend to do poorly after surgery, but that outcome can be changed if patients are in better condition before they enter the operating room.

“It’s one of those things where you might think, no kidding, right?” Fisher says—but nothing similar has been widely done or widely publicized in cardiac medicine.

How Data Is Stored and Accessed

The hospital’s local database system, called Heartbase, resides on a dedicated onsite server. That software, installed at the same time as SPSS, delivers its own internal reporting tools and other functionality. Data is entered manually into Heartbase from each patient’s electronic medical records. There is also an electronic link to the hospital’s admissions demographic system for tracking demographic data on patients. From Heartbase, data is exported directly to SPSS. “That’s the nice thing with SPSS,” Fisher says. “You can use it as its own repository for data, and enter data into it, or you can pull data out of any other system, and analyze it in SPSS.”

The two systems work together quickly and virtually seamlessly in either direction. For example, Fisher can also run a report in the Heartbase system that pulls data from SPSS. “I can have 10,000 patients in a file on my [computer] in a matter of minutes,” she says. “It’s extremely quick.”

Even with Heartbase in place, SPSS provided immediate and valuable ways to manipulate and analyze data. “I latched onto SPSS as soon as I got training on it,” Fisher says. “I was off and running. I’ve never looked back. ... It has opened up a whole new world. It’s really been tremendous for us.”

Additional data on the cardiac program is collected and added to the Heartbase database in other ways. For example, phone calls are made to patients after they have returned home after surgery to learn whether the patient is feeling better, the same, or worse than before surgery. Questions are formulated in the Heartbase software and information is entered there, then imported to SPSS for analysis of the results.

Steep Learning Curve

If there’s a drawback to SPSS, Fisher admits, it’s the learning curve. “You need to make a time investment in training and learning how to use the product,” she says. “It’s a heavy-duty statistical software product, so it’s not for the faint of heart, and your average computer user isn’t going to be able to latch onto it quickly or easily.” Assign users who have some analytical proclivity, she advises, and definitely plan on training. The hospital has a license with SPSS that includes unlimited training over an extended period—training that Fisher describes in hindsight as “phenomenal.” She attended a total of some four to five weeks of training over a six-month period, enough to get her up and running with the software.

That training, she stresses, is essential. She has spoken with other hospitals that have tried to implement SPSS; she found that if training is rushed or is conducted with the wrong person, it simply doesn’t work. A challenge in healthcare can be finding someone with analytic tendencies to master the software and run reports. Fisher is currently overseeing SPSS training for a coordinator with a background in research whom Fisher describes as a perfect fit for the analytics software.

In the future, Fisher would like to use IBM SPSS more for physician scorecards and reporting—the type of physician-level information and performance she says Medicare is increasingly looking for from hospitals. “I think that SPSS could be extremely instrumental in that.”

Linda L. Briggs writes about technology in corporate, education, and government markets. She is based in San Diego. Contact her at

This article originally appeared in the issue of .

TDWI Membership

Get immediate access to training discounts, video library, BI Teams, Skills, Budget Report, and more

Individual, Student, and Team memberships available.