Vasopressor use during emergency trauma surgery

Robert M. Van Haren, Chad M. Thorson, Evan J. Valle, Gerardo A. Guarch, Jassin M. Jouria, Alexander M. Busko, Nicholas Namias, Alan S. Livingstone, Kenneth G. Proctor

Research output: Contribution to journalArticlepeer-review

10 Scopus citations


Most evidence suggests early vasopressor use is associated with death after trauma, but no previous study has focused on patients requiring emergency operative intervention (OR). We test the hypothesis that vasopressors are harmful in this population. Records from 746 patients requiring OR from July 2009 to March 2013 were retrospectively reviewed and stratified based on vasopressor use (epinephrine [EPI], phenylephrine, ephedrine, norepinephrine, dobutamine, vasopressin) or no vasopressor use. Vasopressors were administered to 225 patients (30%) during OR; 59 patients (8%) received multiple vasopressors. Patients who received vasopressors were older, more severely injured, had worse vital signs, and increased mortality rate (all P < 0.001). EPI was independently associated with mortality (odds ratio, 6.88; P = 0.001). If patients who received EPI were excluded, there was no difference in mortality between those who received vasopressors alone or in combination and those that did not (5 vs 6%, P = 0.523), although multiple markers of injury severity were worse. We conclude that vasopressor use is relatively common in the most severely injured patients requiring OR and is associated with mortality. EPI is most often used for cardiac arrest, whereas other vasopressors are used for their vasoconstrictive properties. This suggests that, except for EPI, vasopressors during OR are not independently associated with mortality.

Original languageEnglish (US)
Pages (from-to)472-478
Number of pages7
JournalAmerican Surgeon
Issue number5
StatePublished - May 1 2014

ASJC Scopus subject areas

  • Surgery


Dive into the research topics of 'Vasopressor use during emergency trauma surgery'. Together they form a unique fingerprint.

Cite this