Long-term outcomes
There were eight remote deaths in the elderly group and three in the control group during follow up (Table 4). Median follow-up time was 49 months (19.5–90.5) for all patients; 42.5 months (18.8–75.8) for patients in the elderly group and 53 months (22.5–99) for patients in the control group. Survival rate at 1, 3, and 5 years from initial surgical repair, excluding hospital deaths within 30 days, was 92% (95% CI: 0.801–0.969), 84.7% (95% CI: 0.703–0.924); and 84.7% (95% CI: 0.703-0.924), respectively, in the elderly group compared with 97.8% (95% CI: 0.853–0.997), 97.8% (95% CI: 0.853–0.997), and 94.2% (95% CI: 0.778–0.986), respectively, in the control group. A log-rank test comparing the two survival curves did not indicate the significant difference between the groups (P = 0.117) (Figure 3).
Causes of remote deaths included heart failure (1), stroke (2), rupture of the downstream aorta (1), and pneumonia (4) in the elderly patients. There were eleven late reinterventions including repair for aortic root (1), aortic arch (1), descending aorta (6), and abdominal aorta (3). There were no significant differences in late reinterventions between the groups.
To assess how postoperative physical activities after the surgery impacted late mortality, patients with postoperative walking difficulty, excluding hospital deaths within 30 days, were compared with patients without postoperative walking difficulty. The patients with postoperative walking difficulty were found to have significantly worse late mortality compared with the patients without it (Figure 4A) (P < 0.001).
Survival rate at 1, 3, and 5 years from initial surgical repair was 89.7% (95% CI: 0.749–0.960), 80.6% (95% CI: 0.635–0.903), and 76.4% (95% CI: 0.576–0.877), respectively, for patients with postoperative walking difficulty vs 98.2% (95% CI: 0.878–0.997), 98.2% (95% CI: 0.878–0.997), and 98.2% (95% CI: 0.878–0.997), respectively, for patients without postoperative walking difficulty. Cox proportional hazard analysis showed that postoperative walking difficulty was a significant risk factor for late mortality according to multivariable analysis, whereas age≧70 and postoperative stroke were not (Table 5) (HR 12.65; P = 0.017).
We also assessed the risk factors for late mortality in the patients who regained ambulatory autonomy at discharge. Patients with postoperative walking difficulty were found to have significantly worse late mortality compared with the patients without it difficulty (Figure 4B) (P < 0.001). It was also found in Cox proportional hazard analysis that postoperative walking difficulty was a significant risk factor for late mortality according to multivariable analysis, whereas age≧70, re-exploration for bleeding, and long-term intubation (>3 weeks) were not (Table 6) (HR 15.54; P = 0.026).