Risk factors for candidemia in the neonatal intensive care unit of the Philippines General Hospital from October 2003 to August 2006: A case-control study

ABSTRACT

Candidemia is a major cause of nosocomial morbidity and mortality in neonates. Prompt diagnosis and treatment is crucial. Risk factor analyses have been conducted worldwide, but limited local data are available. This study was conducted to help pediatricians practicing locally decide when to suspect if a neonate has candidemia; therefore, helping them in the judicious use of empirical antifungal therapy. Objective: To determine if there was a difference in the risk factors among neonates with candidemia and those without it, who were admitted at the Neonatal Intensive Care Unit of the Philippine General Hospital from October 2003 to August 2006. Methods: Neonates admitted within the mentioned period, surviving at least on the third day of life, and had at least one blood culture on or after day 3 of life were included in the study. A retrospective review of records was performed to identify the presence or absence of known risk factors for candidemia. The outcome of interest was the presence of candidemia. Each variable was analyzed initially using the bivariate analysis chi-square. Cut-off value for inclusion into multivariate analysis was p<0.25. Multivariate analysis, through backward elimination, was done to narrow down independent variables (p value for retention <0.25). Results: One hundred thirty-eight neonates (69 cases and 69 controls) were included. Based on bivariate analysis, patients exhibiting the following characteristics showed increased risk for candidemia birth weights of 1250 to 1499g (OR 3.24; 95% CI 1.04-10.07) and 1500 to 2449g (OR 3.84; 95% CI 1.31-11.27); pediatric aging < 28 weeks (OR 1.42; 95% CI 1.07-8.5) and 28 to 32 weeks (OR 1.89; 95% CI 0.74-4.84); central vascular access (OR 0.52; 95% CI 0.26-1.03); prolonged broad-spectrum antibiotic use (OR 2.0; 95% CI 0.95-4.2); and increased hospital stay (OR 0.5; 95% CI 0.24-1.05). Intralipid use was also associated with candidemia, but was excluded due to insufficient data available. In the multivariate analysis, only patients with birth weights of 1500 to 2449g (OR 3.65; 95% CI 1.24-10.77) and 1250 to 1499g (OR 3.24; 95% CI 1.04-10.07) qualified. A clinical predictive model in diagnosing candidemia was not possible due to insufficient variables available. Conclusion: Based on the study, infants with lower birth weights (<2500 g) were at most risk for developing subsequent candida infection.