Abstract 16237: Updated Risk Prediction Model for Acute Type B Aortic Dissection

Background:Independent predictors of in-hospital mortality for Type B acute aortic dissection was previously published in 2003 using the IRAD database. In this cohort of 384 cases three significant predictors of death were revealed, the deadly triad of hypotension/shock, absence of chest/back pain on presentation, and branch vessel involvement. A 2014 updated analysis in 1034 cases identified mesenteric ischemia, hypotension/shock, descending diameter ?5.5 cm, renal failure, periaortic hematoma, acute limb ischemia, and age to be predictive of in-hospital death. An updated risk prediction model of type B aortic dissection that further reflects current practice is warranted.Methods:TBAAD patients enrolled in IRAD in the past decade from 2009 to 2019 (n = 1,028) were queried. The 2003 model of the deadly triad of hypotension/shock, absence of chest/back pain on presentation and branch vessel involvement was then applied to this later cohort and fitness was evaluated. A new model was generated for the 2009 to 2019 population.Results:The previous 2003 model of the deadly triad strongly predicted mortality in that cohort (c = 0.82) but was weaker when applied to the latter (2009-2019) cohort (c=0.643). Updating the risk model yielded a more robust c-statistic (c = 0.78). Factors common to both models were age, gender, hypotension and mesenteric ischemia/infarction. However, additional factors such as intramural hematoma, coma/spinal cord ischemia, periaortic hematoma and cardiac tamponade also influenced in-hospital mortality in the past decade.Conclusions:Factors which previously predicted mortality are no longer reflective of risk for Type B patients in the face of changing patient characteristics and management. IRAD provides an updated risk prediction model which strongly predicts mortality in today?s patients and may be a more accurate decision making tool.