Determining risk factors for candidemia among newborn infants from population-based surveillance - Baltimore, Maryland, 1998-2000

Background: Our objective was to determine risks factors for late onset candidemia, independent of birth weight, in newborn infants. Methods: We performed a matched case-control study. Cases were identified through active, population-based surveillance for candidemia, conducted in Baltimore City and County during 1998-2000, and were defined as the incident isolation of any Candida species from the bloodstream of an infant 3 months old or younger. Four controls, matched by age, hospital, birth weight category, hospital stay and admission date, were selected for each case. Potential risk factors included clinical, demographic and maternal prenatal data. Results: Of the 35 cases, 19 (54%) infections were with Candida albicans, 9 (26%) were with Candidaparapsilosis and 5 (14%) were with Candida glabrata. Cases had a median birth weight of 680 g (range, 430-3200 g); median gestational ages of cases and controls were 25 and 27 weeks, respectively. Compared with controls, cases had significant higher mortality (20% versus 4%; P = 0.004). No maternal factors were associated with increased risk of disease; cases were as likely as controls to be of black race. Multivariable conditional logistic regression analysis revealed that gestational age younger than 26 weeks [adjusted odds ratio, 6.5; 95% confidence interval (95% CI), 1.3-32], vaginal delivery (adjusted odds ratio, 4.3; 95% Cl 1.3-14.2) and abdominal surgery (adjusted odds ratio, 10.9; 95% C1 1.9-62) were independently associated with increased risk of candidemia. Conclusions: Independent of birth weight, infants born at < 26 weeks or those who have had abdominal surgery are at a significantly increased risk of candidemia. This study helps define a subgroup of preterm infants at high risk of developing bloodstream infections with Candida species.