Neonatal late-onset bloodstream infection: attributable mortality, excess of length of stay and risk factors

Eur J Epidemiol. 2001;17(8):715-20. doi: 10.1023/a:1015665810739.

Abstract

Background: Bloodstream infection represents a major threat among neonates under intensive care with considerable impact on morbidity and mortality. This study evaluated extra stay, attributable mortality and the risk factors associated with late-onset bloodstream infection (LO-BSI) among neonates admitted to a neonatal intensive care unit during a 4-year period.

Methods: A retrospective matched cohort study was conducted. For each case, there was one control patient without LO-BSI matched for sex, birth weight, gestational age, duration of hospitalization prior to the date of LO-BSI in the respective cases, underlying illness and birth date. A novel test, sequential plan, was employed for attributable mortality analysis in addition to standard tests. Multiple logistic regression was employed for risk factor analysis.

Results: Fifty pairs of cases and controls were compared. LO-BSI prolonged hospital stay of 25.1 days in pairs where both subjects survived. Overall attributable mortality was 24% (95% CI: 9-39% p < 0.01) and specific attributable mortality due to Staphylococcus epidermidis was 26.7% (95% CI: 23-30.4%; p = 0.01). Blood and/or blood components transfusion was independently associated with neonatal LO-BSI (OR: 21.2; 95% CI: 1.1-423).

Conclusions: LO-BSI infection prolongs hospital stay and is associated with increased mortality among neonates. In the present series, blood transfusion was a significant risk factor for LO-BSI.

MeSH terms

  • Brazil / epidemiology
  • Case-Control Studies
  • Female
  • Humans
  • Infant, Newborn
  • Intensive Care Units, Neonatal
  • Length of Stay / statistics & numerical data*
  • Logistic Models
  • Male
  • Retrospective Studies
  • Risk Factors
  • Sepsis / microbiology
  • Sepsis / mortality*
  • Staphylococcus epidermidis / isolation & purification
  • Statistics, Nonparametric