Covariates
Demographic and patient characteristics that might influence respiratory culture results or interpretation were collected from each encounter, including age and insurance status, location of culture collection (e.g., intensive care unit, acute care floor, office visit, emergency department, laboratory collection, bronchoscopy suite), and hospital department sending culture (e.g., Critical Care, Pulmonary Medicine). At this institution, children with baseline ventilator use do not require ICU admission unless they are clinically unstable.
To characterize clinical complexity and comorbid lung conditions, encounter-level diagnoses were pooled across each child’s repeated encounters. Diagnoses were coded to identify complex chronic conditions (CCC) using previously defined codes,13,14 and bronchopulmonary dysplasia (BPD, ICD -9 770.7, ICD -10 P27.1) or chronic lung disease of infancy (CLDI, ICD -10 P27.8).
Children with chronic P. aeruginosa isolation were additionally identified, defined using an adaption to the previously-described criteria in the tracheostomy population15 - P. aeruginosa isolation on ≥50% of cultures obtained in a 12 month period. Gram stain WBC semi-quantification, as categorized by the microbiology laboratory, was additionally examined.