Impact of the lab-score on antibiotic prescription rate in children with fever without source: a randomized controlled trial

PLoS One. 2014 Dec 11;9(12):e115061. doi: 10.1371/journal.pone.0115061. eCollection 2014.


Background: The Lab-score, based on the combined determination of procalcitonin, C-reactive protein and urinary dipstick results, has been shown accurate in detecting serious bacterial infections (SBI) in children with fever without source (FWS) on retrospective cohorts. We aimed to prospectively assess the utility of the Lab-score in safely decreasing antibiotic prescriptions in children with FWS and to determine its diagnostic characteristics compared to common SBI biomarkers.

Methods: Randomized controlled trial in children 7 days to 36 months old with FWS, allocated either to the Lab-score group (Lab-score reported, blinded WBC count) or to the control group (WBC, bands and C-reactive protein determined, blinded procalcitonin and Lab-score), followed up until recovery. Demographic data, antibiotic prescription rate, admission rate and diagnostic properties of the Lab-score were analyzed.

Results: 271 children were analyzed. No statistically significant difference concerning antibiotic prescription rate was observed: 41.2% (54 of 131) in the Lab-score group and 42.1% (59 of 140) in the control group (p = 1.000). If recommendations based on the Lab-score had been strictly applied, a hypothetical 30.6% treatment rate would have been encountered, compared to the overall 41.7% observed rate (p = 0.0095). A Lab-score ≥3 showed the following characteristics: sensitivity 85.1% (95% CI: 76.5-93.6%), specificity 87.3% (95% CI: 82.7-91.8%), positive predictive value 68.7% (95% CI: 58.7-78.7%), negative predictive value 94.1% (95% CI: 91.5-97.9%), positive and negative likelihood ratios: 6.68 and 0.17 respectively. Area under the receiver operating characteristic curve was best for the Lab-score (0.911, 95% CI: 0.871-0.950).

Discussion: No difference regarding antibiotic treatment rate was observed when using the Lab-score, due to lack of adherence to the related recommendations. However, if strictly followed, a significant 26.5% reduction of antibiotic prescriptions would have been encountered. Medical education needs to be reinforced in order to observe rather than treat low-risk well-appearing children with FWS.

Trial registration: NCT02179398.

Publication types

  • Randomized Controlled Trial
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Anti-Bacterial Agents / administration & dosage*
  • Bacterial Infections / drug therapy*
  • Bacterial Infections / pathology
  • Biomarkers / metabolism
  • C-Reactive Protein / metabolism
  • Calcitonin / blood
  • Calcitonin Gene-Related Peptide
  • Female
  • Fever / blood
  • Fever / drug therapy*
  • Fever / pathology
  • Humans
  • Infant
  • Infant, Newborn
  • Leukocyte Count
  • Male
  • Protein Precursors / blood


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

Associated data


Grants and funding

This work was financially supported by bioMérieux ( for data management and statistical analysis. The funder also provided technical support through the loan of the procalcitonin assay. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.