Patients who undergo repair of acute aortic dissection by surgeons who average less than one case a year are almost twice as likely to die in the hospital than patients whose surgeons average five or more cases annually, according to an analysis published in the May issue of the Annals of Thoracic Surgery. Hospital volume also predicted mortality in the study.
Joanna Chikwe, MD, of the department of cardiothoracic surgery at Mount Sinai Medical Center in New York City, and colleagues used the Nationwide Inpatient Sample to identify patients with a discharge diagnosis code of thoracic or thoracoabdominal dissection between 2003 and 2008 who underwent surgical repair during their admission. The database is maintained by the Agency for Healthcare Research and Quality with a 20 percent stratified sample of hospitals in 45 states.
The primary outcome was in-hospital mortality. The analysis covered 11 states that included unique physician identifiers and outcomes of 5,184 patients who were diagnosed with acute aortic dissection. The researchers calculated surgeon and hospital volumes separately and risk adjusted for preoperative comorbidity. To assess trends over time, they also analyzed a weighted total of almost 25,000 patients across all states in the dataset.
The overall mortality rate was 21.6 percent; 71.1 percent of patients experienced major postoperative complications. Operative mortality decreased over time, from 23 percent in 1998 to 2000 to 19.1 percent from 2005 to 2008. The number of procedures was mostly constant and the patient risk profile did not change.
The multivariate analysis showed that institution and surgeon volume were strong predictors of mortality. Hospitals with an institutional volume of less than three acute dissection surgeries annually had a mortality rate of 27.4 percent; facilities with 13 or more cases had a mortality rate of 16.4 percent.
The mortality rate for surgeons with an annual volume of less than one acute dissection case totaled 27.5 percent compared with 17 percent for surgeons with five or more cases a year. Patients whose surgeons performed 60 or fewer total cardiac surgeries annually were two to three times more likely to die in the hospital compared with patients whose surgeons performed more than 170 cardiac surgery procedures a year.
Surgeon and institutional volume had a greater impact on outcomes than did preoperative morbidity.
“The most significant improvement in outcomes occurred at a cutoff of roughly two repairs of aortic dissections a year: below that number, operative mortality was more than 23 percent, above that number, operative mortality dropped to 19 percent. This difference remained highly significant after adjusting for patient factors,” Chikwe et al wrote.
They observed that it may be difficult for some surgical teams to build the volume that may be needed to optimize outcomes. “Given declining duration and intensity of cardiac surgical training, and rapid access in most urban centers to surgical teams with focused expertise in management of aortic dissection, it may no longer be appropriate to emphasize the speed of surgical transfer beyond speed of transfer to an experienced center.” They suggested looking at a model used in heart transplant surgery, where experience and expertise is consolidated.
The Nationwide Inpatient Sample is primarily an administrative dataset that does not include distinctions such as type A and type B aortic dissection, nor does it track outcomes after discharge. As an observational study, the results may be biased by possible confounding.