Risk Prediction Model for Major Adverse Outcome in Proximal Thoracic Aortic Surgery.

Published

Journal Article

BACKGROUND:Proximal thoracic aortic surgery utilizing hypothermic circulatory arrest carries risks of mortality and major morbidity; however, these risks are not the same for every patient. The goal of the current study was to establish a risk prediction model for risk-stratifying patients undergoing proximal thoracic aortic surgery with hypothermic circulatory arrest for degenerative pathology, to facilitate preoperative physician-patient counseling. METHODS:A retrospective analysis was conducted on 489 patients who underwent proximal thoracic aortic surgery with hypothermic circulatory arrest for degenerative pathology between July 2005 and August 2014 at a single referral institution; patients with acute dissection (n = 139) were excluded. Multivariable logistic regression was used to build a risk prediction model and identify preoperative predictors of major adverse outcome-the composite endpoint of 30-day/inhospital mortality, stroke, acute renal failure, prolonged ventilation, or discharge to a location other than home. The results were validated using an independent cohort of 120 patients operated on from September 2014 to September 2016. RESULTS:Multivariable analysis identified age (p = 0.0002, odds ratio [OR] 2.01), total arch replacement (p ≤ 0.0001, OR 6.75), and procedure status (p = 0.0028; OR 2.73 for urgent, OR 43.58 for emergent) as independent predictors associated with major adverse outcome. The calibration curve for probability of major adverse outcome showed excellent agreement between the model and observations. The concordance index was 0.93 in external validation. CONCLUSIONS:The current study identified risk factors for major adverse outcome after proximal thoracic aortic surgery with hypothermic circulatory arrest for degenerative pathology. The proposed simple, accurate model can quantify risk and facilitate physician-patient counseling before possible surgical intervention.

Full Text

Duke Authors

Cited Authors

  • Wagner, MA; Wang, H; Benrashid, E; Keenan, JE; Ganapathi, AM; Englum, BR; Hughes, GC

Published Date

  • March 2019

Published In

Volume / Issue

  • 107 / 3

Start / End Page

  • 795 - 801

PubMed ID

  • 30414833

Pubmed Central ID

  • 30414833

Electronic International Standard Serial Number (EISSN)

  • 1552-6259

International Standard Serial Number (ISSN)

  • 0003-4975

Digital Object Identifier (DOI)

  • 10.1016/j.athoracsur.2018.09.052

Language

  • eng