Procalcitonin: analysis of diagnostic specificity and effectivity in comparison with other markers of inflammation in the critically ill

Aim

To evaluate diagnostic specificity and sensitivity of procalcitonin (PCT) in comparison with C-reactive protein (CRP) and other markers of inflammation phase reactants in relation to clinical status and microbiology examinations.

Materials and methods

A prospective study in ICU patients. One hundred and sixty-four examinations of PCT, CRP, orosomucoid, prealbumin, fibrinogen, INR, leukocyte and platelet count were evaluated in 43 patients. The APACHE II score was calculated on admission, the SOFA daily. SIRS, MODS, positivity of bacteriology, hemoculture and apparent infection were also assessed. Statistical methods comprised the chi-square test, Wilcoxon unpaired test, nonparametric Kruskal-Wallis test, ROC analysis and comparison of area under the curve (AUC).

Results

AUC is given for the following variables (only AUC > 0.6 is mentioned). MODS: temperature 0.700, PCT 0.683, platelets 0.657. SIRS: temperature 0.852, leukocytes 0.761, PCT 0.694, fibrinogen 0.612. SOFA: PCT 0.756, platelets 0.711, temperature 0.654, leukocytes 0.646. Hemoculture: albumin 0.703, fibrinogen 0.686, orosomucoid 0.666, temperature 0.649. APACHE II score: platelets 0.685, PCT 0.605, leukocytes 0.602. Bacteriology: temperature 0.653, PCT 0.627. Apparent infection: temperature 0.694, fibrinogen 0.680. There was a significant difference between survivors and nonsurvivors with respect to ICU stay for PCT (P = 0.00), platelets (P = 0.00), leukocytes (P < 0.03) and temperature (P < 0.05).

Conclusion

PCT was a more effective marker of sepsis-related complications in ICU patients than CRP or other acute phase reactants. ROC analysis was a suitable tool for confirmation of these relations.