Retrospective Application of New Pediatric Ventilator-Associated Pneumonia Criteria Identifies a High-Risk Population.

OBJECTIVES

To promote standardization, the Centers for Disease Control and Prevention introduced a new ventilator-associated pneumonia classification, which was modified for pediatrics (pediatric ventilator-associated pneumonia according to proposed criteria [PVAP]). We evaluated the frequency of PVAP in a cohort of children diagnosed with ventilator-associated pneumonia according to traditional criteria and compared their strength of association with clinically relevant outcomes.

DESIGN

Retrospective cohort study.

SETTING

Tertiary care pediatric hospital.

PATIENTS

Critically ill children (0-18 yr) diagnosed with ventilator-associated pneumonia between January 2006 and December 2015 were identified from an infection control database. Patients were excluded if on high frequency ventilation, extracorporeal membrane oxygenation, or reintubated 24 hours following extubation.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Patients were assessed for PVAP diagnosis. Primary outcome was the proportion of subjects diagnosed with PVAP. Secondary outcomes included association with intervals of care. Two hundred seventy-seven children who had been diagnosed with ventilator-associated pneumonia were eligible for review; 46 were excluded for being ventilated under 48 hours (n = 16), on high frequency ventilation (n = 12), on extracorporeal membrane oxygenation (n = 8), ineligible bacteria isolated from culture (n = 8), and other causes (n = 4). ICU admission diagnoses included congenital heart disease (47%), neurological (16%), trauma (7%), respiratory (7%), posttransplant (4%), neuromuscular (3%), and cardiomyopathy (3%). Only 16% of subjects (n = 45) met the new PVAP definition, with 18% (n = 49) having any ventilator-associated condition. Failure to fulfill new definitions was based on inadequate increase in mean airway pressure in 90% or FIO2 in 92%. PVAP was associated with prolonged ventilation (median [interquartile range], 29 d [13-51 d] vs 16 d [8-34.5 d]; p = 0.002), ICU (median [interquartile range], 40 d [20-100 d] vs 25 d [14-61 d]; p = 0.004) and hospital length of stay (median [interquartile range], 81 d [40-182 d] vs 54 d [31-108 d]; p = 0.04), and death (33% vs 16%; p = 0.008).

CONCLUSIONS

Few children with ventilator-associated pneumonia diagnosis met the proposed PVAP criteria. PVAP was associated with increased morbidity and mortality. This work suggests that additional study is required before new definitions for ventilator-associated pneumonia are introduced for children.