• Ville Pettilä, Marja Hynninen, Olli Takkunen, Pentti Kuusela, and Matti Valtonen.
• Department of Surgery, Division of Anesthesiology and Intensive Care Medicine, Helsinki University Hospital, P.O. Box 340, 00029/HUS, Helsinki, Finland. ville.pettila@hus.fi
• Intensive Care Med. 2002 Sep 1;28(9):1220-5.

ObjectiveTo evaluate the performance of procalcitonin (PCT), interleukin-6 (IL-6), C-reactive protein, leukocyte count, D-dimer, and antithrombin III at onset of septic episode and 24 h later in prediction of hospital mortality in critically ill patients with suspected sepsis.Design And SettingProspective, cohort study in two university hospital intensive care units.Patients61 critically ill patients with suspected sepsis.Measurements And ResultsThe outcome measure was hospital mortality. Hospital survivors ( n=41) and nonsurvivors ( n=20) differed statistically significantly on day 1 (admission) in PCT, IL-6, SOFA score, and APACHE II score, and 24 h later in PCT, IL-6, and D-dimer values. AT III, CRP, and leukocyte count did not differ. The areas under receiver operating curves showed reasonable discriminative power (>0.75) in predicting hospital mortality only for day 2 IL-6 (0.799) and day 2 PCT (0.777) values which were comparable to that of APACHE II (0.786), and which remained the only independent predictor of mortality.ConclusionsAdmission and day 2 IL-6, and day 2 PCT, and day 2 D-dimer values differed significantly between hospital survivors and nonsurvivors among critically ill patients with suspected sepsis. However, in prediction of hospital mortality, only the discriminative power of day 2 PCT and IL-6 values, and APACHE II was reasonable as judged by AUC analysis (>0.75).

30 keywords...

hide…