[Thrombopenia and the nature of the microorganism in infected critically ill children]

Arch Pediatr. 2014 Oct;21(10):1073-8. doi: 10.1016/j.arcped.2014.06.027. Epub 2014 Aug 11.
[Article in French]

Abstract

Background: Thrombopenia is correlated with sepsis and mortality in pediatric intensive care units. The purpose of this study was to find an association between thrombopenia and the microorganism type to guide the choice of empiric antibiotic therapy in infected critically ill children.

Materials and methods: We conducted a prospective descriptive study, including all newborns, infants, and children admitted to a pediatric surgical intensive care unit from 1st January to 31st December 2009. We identified patients who developed an infection and/or thrombopenia (platelet count less than 100,000/mm(3)) during hospitalization.

Results: One hundred ninety-seven patients were included (57 newborns, 41 infants, 99 children). Ninety patients developed 100 infectious episodes during the study period. Of the 57 newborns enrolled in the study, 31 (54%) developed 37 infections. Seventy-six microorganisms (55 Gram-negative bacilli [GNB], 17 Gram-positive cocci, two Gram-negative cocci, two fungal pathogens) were identified during 65 infectious episodes in 55 patients. Thirty-four episodes of thrombopenia were observed in 30 patients. Thrombopenia was observed only in infected patients (P<0.001). Thrombopenia was associated with infections caused by GNB (26/28 vs 20/37, P=0.001) and by Klebsiella (16/28 vs 6/37, P=0.001) and may be associated with infections caused by GNB producing extended-spectrum beta-lactamases (P=0.07). Gram-positive cocci infections were correlated to the non-occurrence of thrombopenia (P=0.02). Postoperative peritonitis was also significantly associated with thrombopenia (P=0.03). The mortality rate in our patients was 12.7% (22.8% in neonates). There was an association between thrombopenia and death in univariate analysis (11/25 vs 19/172, P<0.001). Multivariate logistic regression analysis did not confirm thrombopenia as an independent predictive factor of mortality in children.

Conclusion: Because of the relatively high proportion of resistant GNB, an empiric antibiotic therapy combining a carbapenem and an aminoglycoside may be indicated in infected critically ill children developing thrombopenia.

Publication types

  • English Abstract

MeSH terms

  • Child, Preschool
  • Critical Illness
  • Female
  • Gram-Negative Bacteria / isolation & purification
  • Gram-Negative Bacterial Infections / epidemiology*
  • Hospital Mortality
  • Humans
  • Infant
  • Infant, Newborn
  • Intensive Care Units, Pediatric
  • Male
  • Peritonitis / epidemiology
  • Postoperative Complications
  • Prospective Studies
  • Sepsis / epidemiology
  • Sepsis / microbiology*
  • Thrombocytopenia / epidemiology*
  • Tunisia / epidemiology