Abstract 14076: Mortality Risk Spectrum and Time-Varying Prognostication After Acute Type A Aortic Dissection Repair
Introduction: Traditional mortality risk-assessments after acute type A dissection(AAAD) repair assume that predictors have a constant risk during the postoperative period. Such analysis can over/underestimate their effects if the risk associated with baseline variables significantly changes over time.
Hypothesis: To ascertain the spectrum and time-varying effects of risk-factors in a large cohort of AAAD patients over the first year (1yM) after repair.
Methods: AAAD patients from 1999-2014 were reviewed. Prognostic models were developed from univariate and multivariable analyses using multiple logistic and generalized linear model regression. Individual risks were derived from regression estimates.
Results: The 113/523(22%) repaired AAADs who died within 1year were older (65±16vs57±14years; p<.001), males (41% vs 26%);p=0.006), had longer bypass time (186±67 vs 155±46, P<.001). Malperfusion (MPS) of ≥2 vascular beds significantly increased (32% vs 20%, OR 1.9, P=.022) 1yM risk and most of early deaths occurred in those who developed multisystem dysfunction (62% vs 38%, p<.001). 23% (26/113) of 1yM occurred within 24h, 71% at 30d (greatest hazard during first 10days); and 88% at 6mo after AAAD repair. The prognostic factors of mortality and individual risk estimates are shown in Table. Rupture AAAD was a major predictor associated with a consistent 2-fold increased 1yM risk at each interval. Other significant predictors included coronary artery disease (CAD), age>70, and visceral MPS. After initial drop to 78% in 1year, actuarial survival after AAAD plateaued over a median follow-up of 3.8yrs (73% at 5yrs) and gradually declined to 56% at 10years.
Conclusion: In AAAD, the predictors of mortality (age, CAD, MPS, and rupture) remain consistent throughout the first year. Although surgical results are conventionally measured as early and late mortality, risk prognostication should account for varying effects of predictors over time after AAAD repair.
Author Disclosures: H.K. Sandhu: None. A.K. Tanaka: None. K.M. Charlton-Ouw: None. E. Habrel: None. C.C. Miller: None. H.J. Safi: None. A.L. Estrera: None.
- © 2016 by American Heart Association, Inc.