Subscribe to RSS
DOI: 10.3233/JPI-2010-0265
Risk factors for nosocomial infections in children who had open-heart surgery
Subject Editor:
Publication History
02 December 2009
22 April 2010
Publication Date:
28 July 2015 (online)

Abstract
Nosocomial infections at pediatric intensive care units (PICUs) often lead to substantial morbidity, mortality, and prolonged hospital stays in children who have open-heart surgery. Little is known about the risk factors in this population. Our aim was to identify the incidence and clinical factors associated with infections at the PICU in children with congenital heart disease after surgery. Clinical records of patients ≤ 3-year-old with congenital heart disease, admitted for surgery, were evaluated for clinical, nutritional, and other risk factors. Infection was determined during the patient's stay at the PICU. Two hundred eighty-nine patients were included in the study. Factors related to overall infection were: higher risk adjustment for congenital heart surgery, age > 1-year-old, cyanotic cardiac anomaly, parenteral or mixed nutrition support, more than 5 days of fasting, and mechanical ventilation ≥ 48 . Factors related to pneumonia were higher risk adjustment for congenital heart surgery, cyanotic cardiac anomaly more than three inotropics infused during surgery, parenteral or mixed nutritional support, more than 5 days of fasting, and mechanical ventilation ≥ 48 h. Factors related to mediastinitis and sepsis were: age > 1-year-old, aortic clamp ≥ 120 min, parenteral or mixed nutritional support, more than 5 days of fasting, and mechanical ventilation ≥ 48 h. In the multivariate survival analysis, parenteral or mixed nutrition and mechanical ventilation ≥ 48 h increased the risk for overall infection (relative risk 1.949, 95% confidence interval 1.108–3.43, P = 0.021 and relative risk 25, 95% confidence interval 2.53–246.19, P = 0.006 respectively). Early enteral nutrition after surgery and early weaning from mechanical ventilation (less than 48 h) will reduce infection incidence.