Procalcitonin in pediatric emergency departments for the early diagnosis of invasive bacterial infections in febrile infants: results of a multicenter study and utility of a rapid qualitative test for this marker

Pediatr Infect Dis J. 2003 Oct;22(10):895-903. doi: 10.1097/01.inf.0000091360.11784.21.


Background: Procalcitonin (PCT) is a potentially useful marker in pediatric Emergency Departments (ED). The basic objectives of this study were to assess the diagnostic performance of PCT for distinguishing between viral and bacterial infections and for the early detection of invasive bacterial infections in febrile children between 1 and 36 months old comparing it with C-reactive protein (CRP) and to evaluate the utility of a qualitative rapid test for PCT in ED.

Methods: Prospective, observational and multicenter study that included 445 children who were treated for fever in pediatric ED. Quantitative and qualitative plasma values of PCT and CRP were correlated with the final diagnosis. To obtain the qualitative level of PCT the BRAHMS PCT-Q rapid test was used.

Results: Mean PCT and CRP values in viral infections were 0.26 ng/ml and 15.5 mg/l, respectively. The area under the curve obtained for PCT in distinguishing between viral and bacterial infections was 0.82 (sensitivity, 65.5%; specificity, 94.3%; optimum cutoff, 0.53 ng/ml), whereas for CRP it was 0.78 (sensitivity, 63.5%; specificity, 84.2%; optimum cutoff, 27.5 mg/l). PCT and CRP values in invasive infections (PCT, 24.3 ng/ml; CRP 96.5 mg/l) were significantly higher than those for noninvasive infections (PCT, 0.32 ng/ml; CRP, 23.4 mg/l). The area under the curve for PCT was 0.95 (sensitivity, 91.3%; specificity, 93.5%; optimum cutoff, 0.59 ng/ml), significantly higher (P < 0.001) than that obtained for CRP (0.81). The optimum cutoff value for CRP was >27.5 mg/l with sensitivity and specificity of 78 and 75%, respectively. In infants in whom the evolution of fever was <12 h (n = 104), the diagnostic performance of PCT was also greater than that of CRP (area under the curve, 0.93 for PCT and 0.69 for CRP; P < 0.001). A good correlation between the quantitative values for PCT and the PCT-Q test was obtained in 87% of cases (kappa index, 0.8). The sensitivity of the PCT-Q test (cutoff >0.5 ng/ml) for detecting invasive infections and differentiating them from noninvasive infections was 90.6%, with a specificity of 83.6%.

Conclusions: PCT offers better specificity than CRP for differentiating between the viral and bacterial etiology of the fever with similar sensitivity. PCT offers better sensibility and specificity than CRP to differentiate between invasive and noninvasive infection. PCT is confirmed as an excellent marker in detecting invasive infections in ED and can even make early detection possible of invasive infections if the evolution of the fever is <12 h. The PCT-Q test has a good correlation with the quantitative values of the marker.

Publication types

  • Comparative Study
  • Multicenter Study

MeSH terms

  • Bacteremia / complications
  • Bacteremia / diagnosis
  • Bacteremia / microbiology*
  • Biomarkers / analysis
  • C-Reactive Protein / analysis*
  • Calcitonin / analysis
  • Calcitonin / metabolism*
  • Calcitonin Gene-Related Peptide
  • Child, Preschool
  • Critical Illness
  • Diagnosis, Differential
  • Emergency Service, Hospital
  • Female
  • Fever of Unknown Origin / etiology*
  • Hospitals, Pediatric
  • Humans
  • Infant
  • Infant, Newborn
  • Male
  • Predictive Value of Tests
  • Probability
  • Protein Precursors / analysis
  • Protein Precursors / metabolism*
  • ROC Curve
  • Sensitivity and Specificity
  • Severity of Illness Index
  • Statistics, Nonparametric
  • Viremia / complications
  • Viremia / diagnosis
  • Viremia / virology*


  • Biomarkers
  • CALCA protein, human
  • Protein Precursors
  • Calcitonin
  • C-Reactive Protein
  • Calcitonin Gene-Related Peptide