Risk factors for Invasive Acinetobacter Infections in Children: A Case – Control Study

Background: Acinetobacter infections in adults increasingly are nosocomial and multidrug resistant (MDR). Data in children are limited but suggest that Acinetobacter is emerging as a clinical pathogen. Invasive infections are relatively infrequent in children. We aimed to identify risk factors for invasion of Acinetobacter spp. in children. 

Methods: Patients (pts) with invasive and noninvasive Acinetobacter infections were identified from Jan Ô08-Dec Õ15 by medical record review. Clinical & microbiologic data were analyzed by retrospective case-control design with multivariate analysis. Cases were patients with invasive infection (INV), defined by Acinetobacter isolation from sterile body sites, deep tissue, or in significant colony count in urine or bronchial lavage. Controls were pts with noninvasive infection (non-INV), defined by isolation of Acinetobacter from upper respiratory culture or superficial skin/wound culture taken for clinical suspicion of infection and gram stain performed. Isolates obtained as outpatient were excluded. In pts with >1 isolate of Acinetobacter, only the first clinical episode was used for analysis. MDR Acinetobacter was defined as resistance to at least ³1 agent in ³3 drug classes.

Results: 50 pts with INV (29 blood, 10 urine, 3 CSF, 6 deep soft tissue, 2 BAL) and 146 pts with non-INV (88 upper respiratory, 58 superficial skin/wound) were included. Only one infection was health care associated (CLABSI). Acinetobacter baumannii was most frequent isolate (Fig 1). Antimicrobial susceptibility profiles were similar between cases and control isolates (Fig 2). Two isolates (non-INV) were carbapenem resistant. 11 isolates were MDR (7 INV, 4 non-INV). Majority pts had predisposing factors (Fig 3); only 10% with INV had no identifiable risk factors. Multivariate analysis with logistic regression identified risk factors for INV: age ³10 years {p=0.004; OR 3.2 [CI 1.44-7.01]} and hospitalization within prior 60 days {p=0.024; OR 2.7 [CI 1.1 to 6.5]}. Presence of >1 pathogen in clinical specimen was associated with decreased risk of INV {p=0.002; OR 0.32 [CI 0.16 to 0.67]}.

Conclusion: Age ³10 years and hospitalizations in prior 60 days were independent predictors of invasive Acinetobacter infection.