Testing Low-Risk Patients for Suspected Pulmonary Embolism: A Decision Analysis
Presented at the Society for Academic Emergency Medicine annual meeting, May 2009, New Orleans, LA.
Received 28 June 2009; received in revised form 1 November 2009 and 22 November 2009; accepted 2 December 2009. published online 11 January 2010. Corrected Proof
Study objective
The Pulmonary Embolism Rule-out Criteria (PERC) identifies low-risk patients who are treated in the emergency department for suspected pulmonary embolism and for whom testing may be deferred. The purpose of this study is to develop a decision model to determine whether certain elements not included in the PERC methodology could better estimate the testing threshold for pulmonary embolism (ie, the pretest probability below which a patient should not be tested for pulmonary embolism). In addition, we determine which risks and benefits of pulmonary embolism evaluation and treatment have the greatest effect on the testing threshold.
Methods
We built decision models of low-risk patients with suspected pulmonary embolism, as determined by the PERC. We obtained model inputs from the literature or by using clinical judgment when data were unavailable. One-way sensitivity analysis derived the testing threshold, and 2-way sensitivity analysis was used to determine the main drivers of the testing threshold.
Results
We found an average testing threshold of 1.4% across all age and sex cohorts. Two-way sensitivity analysis demonstrated that risk of major bleeding from anticoagulation, mortality from contrast-induced renal failure, risk of cancer from computed tomography scan, and mortality from both treated and untreated pulmonary embolism had the greatest effects on the testing threshold.
Conclusion
We found a testing threshold for the PERC similar to that calculated by the Pauker and Kassirer method, using somewhat different assumptions. The 5 major drivers for the testing threshold are variables for which there is a paucity of literature to assess accurately for low-risk patients.
aDepartment of Emergency Medicine, University of Pennsylvania, Philadelphia, PA
bDepartment of Radiology, University of Pennsylvania, Philadelphia, PA
cAllergy and Critical Care Division, University of Pennsylvania, Philadelphia, PA
dUniversity of Pennsylvania School of Medicine, Philadelphia, PA
eDepartment of Emergency Medicine, George Washington University School of Medicine and the Department of Health Policy, and School of Public Health and Health Services, Washington, DC
Address for correspondence: Adam L. Lessler, 917-612-9620
Supervising editor: Steven M. Green, MD
Author contributions: ALL and JMP conceived the study, designed the decision model, and performed the data analysis. RA and HIP provided content advice on the radiology- and pulmonary embolism-specific components, respectively. JAI assisted with the research methodology. ALL drafted the article, and all authors contributed substantially to its revision. ALL takes responsibility for the paper as a whole.
Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article that might create any potential conflict of interest. The authors have stated that no such relationships exist. See the Manuscript Submission Agreement in this issue for examples of specific conflicts covered by this statement.
ABSTRACT