This Article Résumé de cet Article Full Text Full Text (PDF) Submit a scholarly reply Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Comfere, T. B. Articles by Warner, D. O. Search for Related Content PubMed PubMed Citation Articles by Comfere, T. B. Articles by Warner, D. O.

Predictors of mortality following symptomatic pulmonary embolism in patients undergoing noncardiac surgery

[Les indicateurs de mortalité à la suite d’embolies pulmonaires symptomatiques chez des patients subissant une chirurgie non cardiaque]

Thomas B. Comfere, MD^*, Juraj Sprung, MD PhD^*, Kimberly A. Case, SRNA^*, Paul T. Dye, SRNA^*, Jamey L. Johnson, SRNA^*, Brian A. Hall, MD^*, Darrell R. Schroeder, MS, Andrew C. Hanson, BS, Mary E. S. Marienau, CRNA MS^* and David O. Warner, MD^*

^* From the Departments of Anesthesiology and
Health Sciences Research, Mayo Clinic College of Medicine, Rochester, Minnesota, USA.

Address correspondence to: Dr. Juraj Sprung, Department of Anesthesiology, Mayo Clinic College of Medicine, 200 First Street SW, Rochester, MN 55905, USA. Phone: 507-255-3298; Fax: 507-255-6463; E-mail: sprung.juraj{at}mayo.edu

Purpose: To determine 30-day mortality and predictors of mortality following perioperative pulmonary embolism (PE).

Methods: We searched both the Mayo Clinic electronic medical records and Autopsy Registry, between January 1, 1998 and December 31, 2001, for patients who developed PE within 30 days after noncardiac surgery performed under general or neuraxial anesthesia. Medical records of all identified patients were reviewed using standardized data collection forms. The association between risk factors for PE and 30-day post-PE mortality was assessed using t tests, exact binomial tests, and logistic regression.

Results: We identified 158 patients with probable or definite perioperative PE. The overall 30-day mortality from the day of PE was 25.3%, i.e., 40 patients died. Hypotension requiring treatment, need for mechanical ventilation, and intensive care unit admission were the prominent univariate predictors of 30-day mortality (all P 0.001). Other significant factors were exact bi normal tests, and higher ASA physical status (P = 0.002), longer surgical time (P = 0.030), recent central vein cannulation (P = 0.021) and intraoperative use of either blood transfusions or other blood products (P = 0.010). Using multivariable analysis, hemodynamic instability was found to be the dominant independent risk factor associated with mortality.

Conclusions: Perioperative PE is associated with a high 30-day mortality. Patients who experience hemodynamic instability and require vasoactive treatment at presentation of PE have extremely low survival rates; therefore, for these patients the most aggressive therapeutic modalities should be considered.