Comparing the accuracy of four prognostic scoring systems in patients operated on for ruptured abdominal aortic aneurysms

J Vasc Surg. 2017 Mar;65(3):609-615. doi: 10.1016/j.jvs.2016.08.082. Epub 2016 Oct 12.

Abstract

Objective: Ruptured abdominal aortic aneurysms (rAAAs) are associated with high mortality and morbidity. Several prognostic scoring systems are available for prediction of outcome, but scarcity of external validation and evaluation of predictive value has hampered widespread implementation. The aim of this study was to examine the discriminatory value of four scores in a consecutive Norwegian cohort.

Methods: This was a retrospective study of a consecutive series of patients operated on for primary rAAA at Stavanger University Hospital from January 2000 to December 2014. The Hardman Index, Vancouver Score (VS), updated Glasgow Aneurysm Score, and Edinburgh Ruptured Aneurysm Score (ERAS) were calculated. Predictive ability in discriminating survivors and nonsurvivors was compared using receiver operating characteristics analyses and presented as area under the curve.

Results: Altogether, 177 patients underwent surgery for rAAA. Mortality at 30 days was 46.3%. In receiver operating characteristics analysis, the Hardman Index had an area under the curve of 0.674 (95% confidence interval [CI], 0.588-0.753); the VS, 0.684 (95% CI, 0.610-0.752); the Glasgow Aneurysm Score, 0.680 (95% CI, 0.605-0.749); and the ERAS, 0.586 (95% CI, 0.509-0.660). VS had a significantly better fit than ERAS (P = .022).

Conclusions: The accuracy of the available scores is limited. The findings question the clinical value of such scores for decision-making.

Publication types

  • Comparative Study
  • Observational Study

MeSH terms

  • Aged
  • Aged, 80 and over
  • Aortic Aneurysm, Abdominal / diagnostic imaging
  • Aortic Aneurysm, Abdominal / mortality
  • Aortic Aneurysm, Abdominal / surgery*
  • Aortic Rupture / diagnostic imaging
  • Aortic Rupture / mortality
  • Aortic Rupture / surgery*
  • Area Under Curve
  • Chi-Square Distribution
  • Decision Support Techniques*
  • Discriminant Analysis
  • Female
  • Health Status
  • Health Status Indicators
  • Hospitals, University
  • Humans
  • Logistic Models
  • Male
  • Middle Aged
  • Multivariate Analysis
  • Norway
  • Predictive Value of Tests
  • ROC Curve
  • Reproducibility of Results
  • Retrospective Studies
  • Risk Assessment
  • Risk Factors
  • Treatment Outcome
  • Vascular Surgical Procedures* / adverse effects
  • Vascular Surgical Procedures* / mortality