Background: Deranged physiology in trauma complicates the clinical identification of sepsis, resulting in overscreening for bacteremia. No clinical signs or biomarkers accurately diagnose sepsis in this population. Our objective was to evaluate the accuracy of the current criteria used to prompt screening for bacteremia in trauma patients and determine independent predictors of bacteremia.
Materials and methods: Adult trauma patients admitted to our level I academic trauma center who had blood cultures (BCs) drawn were identified. Those with positive BCs were compared to those with negative or false positive BCs. False positive was defined as a BC deemed contaminated and not treated at the discretion of the attending physician.
Results: Over a 2-year period, 366 trauma patients had BCs drawn. After excluding surveillance cultures (those drawn to demonstrate bacteremia clearance), 492 unique BC sets were evaluated; 104 (21.1%) BC sets were positive; 30 (28.8%) of these were falsely positive, resulting in a true-positive rate of 15% in the screened population. Univariate analysis suggested temperature and heart rate were associated with positive BC, while multivariable analysis found only the presence of a central line and lactic acid to be predictive. Procalcitonin (PCT) was poorly predictive, with a positive predictive value of 18% and a negative predictive value of 91%.
Conclusion: Current tools for identifying bacteremia in trauma patients result in overscreening. PCT may have a limited role as a negative predictor for bacteremia. Given that false-positive BCs have negative patient and economic consequences, future study should focus on development of alternative screening modalities.
Keywords: bacteremia; procalcitonin; trauma.