• K J Brasel and J A Weigelt.
• Department of Surgery, St Paul-Ramsey Medical Center, University of Minnesota, USA.
• Arch Surg Chicago. 1996 Jun 1;131(6):619-25; discussion 625-6.

ObjectiveTo evaluate the influences of patient preference and treatment costs on the diagnostic approach to blunt aortic trauma.MethodsDecision and cost-utility analysis.Data SourcesA MEDLINE search of all literature dealing with the diagnosis and management of blunt aortic injury was used to establish assumptions and assign baseline probability estimates. Utility assignments were made from published data and our own assignments. We obtained institution-specific cost data.Study SelectionOnly randomized, prospective trials that used aortography as the gold standard test were used to assign baseline accuracy of transesophageal echocardiography and dynamic chest computed tomography. Other baseline estimates were taken from class II and class III published data.Data SynthesisA decision tree compared 4 diagnostic approaches for blunt chest trauma after an initial normal chest radiograph: observation with follow-up chest radiography, aortography, transesophageal echocardiography, and dynamic chest computed tomography. Utility (a quality-of-life measure) was assigned to ultimate health states to incorporate patient preference. Chest radiography and aortography had similar utility. Aortography gained 1 quality-adjusted life year for minimal cost. Transesophageal echocardiography and dynamic chest computed tomography lose quality-adjusted life-years at increased cost. No variable changed the relative cost-utility of the screening methods in 2-way sensitivity analyses.ConclusionsAortography gains additional quality life at minimal cost when used as a screening method for all patients with blunt chest trauma regardless of the results of the initial chest radiograph. With a normal initial chest radiograph, transesophageal echocardiography and dynamic chest computed tomography are associated with increased cost and loss of quality-adjusted life.

32 keywords...

hide…