Background: We evaluated the predictive value of immature granulocyte (IG) percentage in comparison with white blood cell counts (WBC) and C-reactive protein (CRP), for infection, its invasiveness, and severity in critically ill patients. Methods: In 46 consecutive patients, blood samples were collected at the day (0) of a clinical suspicion of microbial infection and at days 1 and 3 thereafter. We defined infections, bloodstream infection, and septic shock within 7 days after enrollment. Results: Of the 46 patients, 31 patients had infection, 15 patients developed bloodstream infection, and 13 patients septic shock. C-reactive protein and IG percentage increased with increasing invasiveness and severity of infection, from day 0 onwards. Receiver operating characteristic analysis to predict infection showed an area under the curve of 0.66 (P = .10) for WBC vs 0.74 (P = .01) for CRP and 0.73 (P = .02) for IG percentage on day 0. Comparing WBC and CRP to WBC and IG percentage results in comparable prediction of microbial infection. Comparing WBC and CRP with WBC, CRP, and IG percentage suggests an additional early value of IG percentage, when not elevated, in ruling out infection. Conclusion: Immature granulocyte percentage is a useful marker, as CRP, to predict infection, its invasiveness, and severity, in critically ill patients. However, the IG percentage adds to WBC and CRP in the early exclusion of infection and can be obtained routinely without extra blood sampling or costs.

, , ,
doi.org/10.1016/j.jcrc.2014.03.033, hdl.handle.net/1765/63569
Journal of Critical Care: improving patient care by integrating critical care systems knowledge into practice behavior
Department of Clinical Chemistry

van der Geest, P., Mohseni, M., Brouwer, R., van der Hoven, B., Steyerberg, E., & Groeneveld, J. (2014). Immature granulocytes predict microbial infection and its adverse sequelae in the intensive care unit. Journal of Critical Care: improving patient care by integrating critical care systems knowledge into practice behavior, 29(4), 523–527. doi:10.1016/j.jcrc.2014.03.033