Pilot study of cytokines in emergency department patients with systemic inflammatory response syndrome

Objective: To determine the potential utility of cytokine and arachidonic acid metabolite levels in ED patients with systemic inflammatory response syndrome (SIRS) as a predictor of progression to severe sepsis. Methods: A prospective, observational study of test performance was performed using convenience samples of adult control subjects and admitted patients. The latter patients were identified in the ED as having signs of SIRS. Levels of cytokines and arachidonic acid metabolites measured from specimens obtained in the ED were compared between groups and associated with the progression of sepsis within 24 hours in the SIRS patients. Results: There were 30 control patients and 29 SIRS patients. There were 8 SIRS subjects who progressed to severe sepsis within 24 hours using the following criteria (hypotension, n = 5; hypoperfusion, n = 1; and organ dysfunction, n = 2). Of the 21 SIRS subjects who did not progress to severe sepsis, 11 had resolution of SIRS criteria at 24 hours. There were no significant differences in mean mediator levels between the SIRS patients who progressed to severe sepsis and those who did not. Of the 1l patients with resolution of SIRS criteria at 24 hours, the mean interleukin-6 (IL-6) level was significantly lower than that for the patients who did not recover or who progressed at 24 hours (n = 18); 65.4 ± 49.1 vs 230 ± 112 pg/mL, p = (0.001). Six of 15 subjects with IL-6 >150 pg/mL progressed to severe sepsis (p = NS). Using threshold values based on the range of levels for normals, the sensitivity of an abnormal marker for the development of severe sepsis within 24 hours varied from 50% to 87%, while the specificity varied from 11% to 84%. Conclusion: While mean levels were significantly elevated when compared with those of normal control subjects, they had limited ability to predict the subset of patients likely to progress to severe sepsis. However, initial low levels of cytokines may have exclusionary prognostic value. Prospective validation of the latter finding is warranted.