Date of Award

Summer 2011

Document Type


Degree Name

Doctor of Philosophy (PhD)


Health Services Research

Committee Director

Anna H. Jeng

Committee Member

Andrew E. Balas

Committee Member

Stella B. Bondi


Pulmonary embolism (PE) is a serious, life-threatening thrombotic disease, which results in considerable health and economic consequences each year for the United States. These consequences include a toll of 83,000 deaths and an economic impact between 1,5 and 5 billion. Approaches to strategy selection by physicians and other health-care specialists are based mainly upon cost, technology availability, and cultural tolerance regarding radiation exposure. The purpose of this study was to determine the most cost-effective diagnostic strategy with patients suspected of PE among several strategies currently used by examining their detection failure rates. This objective was met by (a) assessing parameter estimates and their uncertainty using triangular and 7 distributions, (b) conducting a cost-effectiveness analysis, and (c) testing the model for errors using sensitivity analysis.

Cost-effectiveness analysis based upon a decision tree model revealed that among the investigated strategies for patients with suspected PE the most cost-effective strategy appears to be strategy 3, composed of a clinical decision rule (CDR), a D-dimer test (DD), a compression ultrasonography test (CUS), and a computed tomography pulmonary angiography (CT). Strategy 5, composed by a CDR, DD, a CT, a CUS, and an invasive pulmonary angiography (PA) appeared to be a cost-effective method, but it was more expensive than strategy 3 and included an invasive pulmonary angiography (PA).

The results of a Monte Carlo simulation sensitivity analysis were robust over a number of distributions regarding the PE diagnostic test costs, sensitivities, specificities, and strategy effectiveness. Additionally, the results of this investigation were valid over an extensive range of one-way, two-way, and three-way sensitivity analyses regarding PE diagnostic test costs. Overall, the proposed analyses identified uncertainty and eliminated error; thus, it provides a practical approach to help medical professionals estimate uncertainty in the diagnosis of PE. Although this research has broadened the ability to identify uncertainty and eliminate error, further research is needed to validate these findings in a prospective clinical trial before the delivery of a clinical recommendation.