Objective: To identify risk factors for central line-associated bloodstream infection (BSI) in patients receiving care in a pediatric cardiac intensive care unit.
Design: Matched case-control study.
Setting: CICU at Children's Hospital Boston.
Patients: Central line-associated BSI cases were identified between April 2004 and December 2006. We identified two randomly selected control patients who had a central vascular catheter and were admitted within 7 days of each index case.
Measurements and main results: Univariate and multivariate conditional logistic regression analyses were used to identify risk factors for central line-associated BSI. In a secondary analysis, risk factors for central line-associated BSI in those cases who underwent cardiac surgery were sought. During the study period, 67 central line-associated BSIs occurred in 61 patients. Independent risk factors for central line-associated BSI were nonelective admission for medical management (odds ratio [OR] = 6.51 [1.58-26.78]), the presence of noncardiac comorbidities (OR = 4.95 [1.49-16.49]), initial absolute neutrophil count <5000 cells/uL (OR = 6.17 [1.39-27.48]), blood product exposure > or =3 units (OR = 5.56 [1.35-22.87]), central line days > or =7 (OR = 6.06 [1.65-21.83]), and use of hydrocortisone (OR = 28.94 [2.55-330.37]). In those patients who underwent cardiac surgery (n = 37 cases and 108 controls), independent risk factors for central line-associated BSI were admission weight < or =5 kg (OR = 3.13 [1.01-9.68]), Pediatric Risk of Mortality III score > or =15 (OR = 3.44 [1.19-9.92]), blood product exposure > or =3 units (OR = 3.38 [1.28-11.76]), and mechanical ventilation for > or =7 days (OR = 4.06 [1.33-12.40]).
Conclusions: Unscheduled medical admissions, presence of noncardiac comorbidities, extended device utilization, and specific medical therapies are independent risk factors for central line-associated BSI in patients receiving care in a pediatric cardiac intensive care unit.