Contemporary outcomes of open thoracic aortic surgery in a veteran population: do risk models exaggerate mortality?

Background: We evaluated contemporary outcomes of open thoracic aortic surgery at a Veterans Affairs (VA) medical center affiliated with a major academic aortic program and examined the predictive value of 2 established cardiac risk models.
Methods: We retrospectively reviewed all open thoracic aortic operations performed between April 1998 and April 2008 (n = 100). Both the EuroSCORE and the VA Continuous Improvement in Cardiac Surgery Program (CICSP) scores were evaluated.
Results: Procedures included ascending aortic repair (n = 74, 15 with arch repair), descending thoracic repair (n = 11, 1 with arch repair), and thoracoabdominal aortic repair (n = 15). Emergency surgery was necessary in 15 cases, and 19 procedures were reoperations. The patients' logistic EuroSCORE and the CICSP scores were similar (18.7% and 18.2%, respectively), but both scores significantly exceeded the observed operative mortality rate (8.0%, P = .008).
Conclusions: Good outcomes can be achieved when thoracic aortic surgery is performed at an experienced VA center. The cardiac risk models we examined overpredicted operative mortality.