By Jeff Craven
For 30 years, cardiac surgeons have tracked their outcomes data in the Society for Thoracic Surgeons (STS) National Database. With three decades of longitudinal data to look through, cardiac surgeons have used the information in the STS database to improve workflows, develop protocols and create clinical pathways.
“Cardiac surgeons have always been extremely focused on their outcomes,” says Paul A. Kurlansky, MD, Associate Professor of Surgery and Director of Research, Recruitment and Continuous Quality Improvement at Columbia. “But in all fairness, it's an area which is very readily susceptible to outcome analysis, because you have a discrete event.”
In other specialties, data is more challenging to sift through and not as straightforward. If a physician is managing diabetes or hypertension, the reasons for a given patient condition can be multifactorial, whereas in cardiac surgery, data collection has a higher impact, says Dr. Kurlansky.
“Cardiac surgery is very unique because it's such a common and high impact and high expense type procedure,” says Dr. Kurlansky. “It's actually been under scrutiny for a very long time and it's been under scrutiny in New York state for a very long time and cardiac surgeons have been very proactive in this whole process.”
A gold standard for cardiac outcomes tracking
In 1988, the New York State Commissioner of Health, David Axelrod, was concerned about the outcomes for coronary artery bypass graft (CABG) surgery in his state. When the Cardiac Advisory Committee in New York produced a paper for JAMA in 1990 on independent risk factors for CABG and valve surgery in-hospital mortality, Axelrod sent the names of the hospitals in the JAMA paper to the New York Times that same day.
Following the Times story, the Long Island newspaper Newsday subsequently sued and won a lawsuit against the U.S. Department of Health for the right to release surgeon-specific hospital mortality rates to the public.
Axelrod then created the New York State Cardiac Surgery Reporting System, bringing together cardiologists and statisticians to create a registry that has since become the gold standard for reporting physician and hospital quality.
Outcomes are better for patients
Tracking cardiac surgery outcomes can improve patient satisfaction, identifying areas of the patient experience that could be innovated on—everything from what information a patient fills out on their preoperative questionnaire, to how a certain patient will fare during surgery, to their reactions to anesthesia, pain medication, and postoperative complications such as wound infection and readmissions.
From a national standpoint, outcome measurements will traditionally help centers identify where they are in comparison with other similar centers. But even if your rates are better than the national average, there is always room to be better, says Dr. Kurlansky. For example, of the hundreds of thousands of people who undergo coronary artery bypass grafting (CABG) each year, the risk of stroke is between 0% and 5.2% across studies with a national rate of 1.3%.
“Stroke is a very morbid event. Some patients would rather die than have a stroke,” says Dr. Kurlansky. “Sometimes, you [need to] look carefully at these things and try to figure out what possible strategies you could use to try to improve your results—even if they're pretty good.”
Outcomes promote interdisciplinary care
Cardiac outcomes are also better for surgeons because it not only helps them see their own results, but lets them identify interdisciplinary areas of improvement, says Dr. Kurlansky. The cardiac surgery group at Columbia not only meets on a weekly basis, but also meets with the Department of Anesthesia weekly to discuss problematic cases and how to improve outcomes.
Columbia extends this interdisciplinary approach even further with their HeartSource program, where cardiac surgeons work with their colleagues in cardiology to consult with outside institutions, helping them build or turnaround heart programs around the world.
“One of the things that we look at very carefully in these institutions is their data,” says Dr. Kurlansky.
At the Center for Innovation Outcomes Research, surgeons partner with statisticians to filter STS data by time, procedure and provider to run comparisons and percentages, but also more sophisticated analyses to answer bigger, more complicated questions.
Columbia’s question was wound infections: “We had an issue with wound infections because we use a lot of bilateral internal mammary arteries—which are probably lifesaving in the long term, but they also put you at increased risk for wound infections,” says Dr. Kurlansky. “So, we did an analysis over 4 years of what the statistically significant independent predictors of wound infection were.” They found diabetes had less of an impact on wound infection than originally expected, while kidney disfunction had more.
Four wishes for cardiac surgery outcomes measurement
While tracking cardiac surgery outcomes improves quality of care, the outcomes process itself has room for improvement. Dr. Kurlansky outlined four main areas for improvement in cardiac outcomes, beginning with a way to link or merge data with the National Death Index at the CDC for tracking long-term follow-up of cardiac surgery patients.
Interoperability of electronic health records is another area for improvement, says Dr. Kurlansky, because there can be a discrepancy between what the hospital sends to an insurance company or Medicare and what is collected for databases like the New York State Cardiac Registries and the STS National Database.
“The clinical data tends to be much more robust in the databases. The problem is, it's very time intensive and laborious to collect all this data,” says Dr. Kurlansky.
Having a better handle on what cardiac surgery costs per patient would also be beneficial, as hospitals tend to use proprietary accounting systems that are non-transparent for data analysis, he added. In addition, cost metrics like readmissions are controversial because entities like CMS have penalized centers with excess readmissions for certain diagnoses, complicating the matter.
“As the readmission rate went down for congestive heart failure, the mortality went up. Sometimes, patients need to come into the hospital and need to be taken care of,” says Dr. Kurlansky.
“In some ways, quality care costs more and in some ways, quality care costs less, he added. “Everybody talks about [how] quality care is cheaper. It's not, necessarily.”
Another opportunity for improvement is finding the correlation between a patient’s perception of quality and clinical/professional assessment of quality.
“We have devoted a lot of energy into what we feel professionally/clinically is important. And we've got a good job of it,” says Dr. Kurlansky. “[It] would be nice to develop what is, professionally, clinically important to patients [to] to figure out what kind of correlations there are between patients' perception of quality and professional perceptions of quality.”
There are some elements of patient care that have yet to be captured with statistics, admitted Dr. Kurlansky, such as the effect of a long recovery from cardiac surgery on a patient and his or her family.
“Sometimes, the patients that we operate [on] are very, very sick and sometimes they survive, but they spend a long time in the ICU. They have impairment, [it] takes them a long time [to recover], and sometimes they don’t always come back so well—they're older, or they develop cognitive dysfunction,” says Dr. Kurlansky. “Those types of things don't always come through so clearly in the statistics. But for a family—or for a person—may be a major, major outcome.”
“It's very important to help define, to better understand for everybody—for the patient, for everybody—what the patient's goals are,” says Dr. Kurlansky. “What do you expect to get out of this, and what's important to you?
Measurement and tracking of cardiac surgery outcomes has gotten “tremendously better” and “continues to get better” over the years. As a profession, cardiac surgery has been more proactive than most specialties in tracking surgical data and determining professional quality of care.
“Where we are now is light years ahead of where we have been in the past.” says Dr. Kurlansky. “But we have a long way to go.”