[Nosocomial bacteremias in pediatrics]. / Bactériémies nosocomiales en pédiatrie.

OBJECTIVES: To identify pathogenic microorganisms responsible for hospital-acquired bloodstream infections and to evaluate the associated risk factors in pediatric units, in a case-control study over 30 months from January 1st 1997 to June 30th 1999. RESULTS: Forty-six of 855 (5.4%) positive blood cultures were attributed to nosocomial infections. They were related to 32 infectious episodes in 28 patients hospitalized for more than 48 hours. The incidence rate was 0.11 per 100 admissions. Gram-positive cocci (n = 14; 38.8%) were the most frequently isolated pathogens (7 cases of Staphylococcus aureus, 5 of coagulase-negative staphylococci), followed by enterobacteria (n = 9; 25%), Pseudomonas aeruginosa (n = 5; 13.8%) and yeasts (n = 5; 13.8%). The major risk factors for hospital-acquired bloodstream infections were: length of stay before positive blood culture (32 +/- 51 days in cases vs 15 +/- 43 days in controls, p < 0.01), presence of central venous catheter [odds ratio (OR): 6.05, 95% confidence interval (CI): 1.87-20.42], number of days with central venous catheter (p < 0.001) and parenteral nutrition (OR: 9.44, 95% CI: 2.03-50.05). CONCLUSION: Central venous catheter use, length of stay, parenteral nutrition and particularly intravenous lipids are major risk factors for the acquisition of bloodstream infection in hospitalized children.

[Bacteriologic surveillance of nosocomial septicemia and bacteremia in a pediatric hospital]. / Surveillance bactériologique des septicémies et bactériémies nosocomiales dans un hôpital pédiatrique.

BACKGROUND: Nosocomial bloodstream infections in pediatrics are an important cause of morbidity and mortality. To identify pathogens causing nosocomial bloodstream infections, evaluate associated risk factors and take preventive measures, we conducted a prospective study from January 1995 to December 1995 at Saint-Vincent-de-Paul Hospital (Paris). PATIENTS AND RESULTS: All patients hospitalized more than 48 hours were included in the study. During this period, we recorded 21 bloodstream infections in 20 children. The incidence rate of nosocomial bloodstream infection was 1/1,000 admissions. Sixteen children were hospitalized in surgery, three in medical intensive care unit; the median day onset of infection was approximately 20 days. Recorded risk factors were: surgery, invasive procedures, central catheterization, bladder catheters, parenteral nutrition, device, endotracheal tube, antibiotic therapy before infection. The number of risk factors ranged from zero to six per patient. The most common isolated pathogens were in ten cases Gram positive cocci: five methicillin-sensible Staphylococcus aureus, four methicillin-resistant coagulase-negative staphylococci and one Streptococcus milleri. Other bacteria were seven enterobacteria, three Pseudomonas sp and three Candida sp. In 11 cases, the same bacteria as in bloodstream infection could be found: in three urine samples, in two tracheal samples, in two gastro-intestinal samples, two puncture sites, one device, and one umbilical catheter. CONCLUSION: In our study, 6.2% of positive blood culture were due to a nosocomial infection. We confirm the importance of Gram positive cocci, and particularly of methicillin-resistant coagulase negative staphylococci.