Volume Does Not Correlate with Outcomes in Adult Cardiac Operations

Hospital or surgeon volume does not impact observed-to-expected operative mortality or 30-day readmission, according to a study presented at the Southern Thoracic Surgical Association’s 65^th annual meeting.

For the study, researchers utilized data from the Pennsylvania Health Care Cost Containment Council (PHC4). They included isolated coronary artery bypass grafting (CABG), isolated valve surgery, and CABG plus valve surgery performed between 2014 and 2016. Operative mortality and 30-day readmission were the primary outcomes. PHC4’s own risk models were used to determine expected operative mortality and 30-day readmission. Researchers correlated observed-to-expected (OE) ratios with procedural volume with linear regression analysis.

Final analyses included 29,578 surgeries, of which 16,641 were isolated CABGs, 8,618 were isolated valves, and 4,319 were CABG plus valves. Procedures were performed by 182 surgeons at 60 hospitals. The predicted risk of operative mortality for surgeons was as follows: 1.5% for isolated CABGs, 1.8% for isolated valves, and 4.3% for CABG plus valves; for the same operations, the predicted risk of operative mortality for hospitals was 1.5%, 1.7%, and 4.3%, respectively. Predicted 30-day readmission for surgeons was as follows: 10.3% for isolated CABGs, 13.4% for isolated valves, and 14.4% for CABGs plus valves; for the same operations, hospitals’ 30-day readmission predicted rates were 10.2%, 13.2%, and 14.3%, respectively.

“There was no correlation between surgeon or hospital volume and either predicted risk or observed-to-expected operative mortality or 30-day readmission for any of the index operations,” the study authors wrote.

Study author Valentino Bianco, a cardiothoracic surgery resident at the University of Pittsburgh Medical Center, presented the findings. Her team also performed a secondary analysis.

“We wanted to see if there was an association between surgeon and hospital volume and the expected risk, or the expected mortality, of these patients,” Bianco said. “In other words, we wanted to answer the question, do high-volume surgeons and high volumes in fact operate on higher-risk patients? And based upon the results of this linear regression analysis, the answer to that question is no, there is no direct correlation between surgeon and hospital volume in respect to patient risk.”

The data were limited to just one state, which Bianco and colleagues cited as a study limitation. Data entry also carries a potential for error.

“There is also the potential that the risk models advised by the PHC4 could fail to accurately predict the risk of the patients,” Bianco added.

The data presented contradicted results seen in previous studies, several audience members noted during discussion.

“This is a timely and most provocative topic,” said discussant Alan Speir, MD, of Inova Medical Group in in Falls Group, Va., who praised the efforts of Bianco and his co-authors. “But despite their diligence, however, I am not convinced that this conundrum can be solved from the data that have been presented. It’s hard to believe that the relationship between volumes and surgical outcomes, which were first reported over 40 years ago in the New England Journal of Medicine, have demonstrated that hospitals with larger surgical volumes demonstrated better outcomes.”

When asked why the research team chose to study state data rather than STS data, Bianco explained, “This is something that surgeons are going to be held to. They are going be judged by the methods used by the states, not STS.”

Another audience member suggested that the study perhaps could have looked at more complex cases.

“I think one of the issues with your analysis is that you picked CABG valve replacement,”
the attendee said. “I think if you looked at the more complex cases, you might find a difference. If you look at the congenital analysis from the STS database, you’ll find exactly the opposite of what you talked about.”

“Are you saying that the belief ‘the more you do, the better you are’ is not true?” another attendee asked. “You’re saying that all of our studies in cardiac surgery that say volume equals a better outcome are not true?”

“It does not necessarily mean that that association doesn’t exist, but it certainly should be called into question,” Bianco responded. “For these basic relatively basic operations, I think that’s what we’re getting at.”