Objectives: Viral infections are common in children, but there is a lack of data on severe viral infections in critically ill children. We investigated testing for viral infections in children requiring PICU admission and describe the epidemiology and outcomes.
Design: Multicenter retrospective study. Results of viral testing for nine respiratory viruses using polymerase chain reaction were collected.
Participants: Children less than 16 years old nonelectively admitted to PICU over a 6-year period.
Setting: Two tertiary PICUs in Queensland, Australia.
Interventions: None.
Main outcome measures: Primary outcome was PICU length of stay. Secondary outcomes included need for and duration of intubation and mortality in PICU. Univariate and multivariate regression analyses were performed, adjusting for age, indigenous status, comorbidities, and severity of illness.
Results: Of 6,426 nonelective admissions, 2,956 (46%) were polymerase chain reaction tested for a virus of which 1,353 (46%) were virus positive. Respiratory syncytial virus was the most common pathogen identified (n = 518, 33%), followed by rhinovirus/enterovirus and adenovirus. Across all patients who underwent polymerase chain reaction testing, identification of a respiratory virus was not significantly associated with longer overall length of stay (multivariate odds ratio, 1.08; 95% CI, 0.99-1.17; p = 0.068) or longer intubation (p = 0.181), whereas the adjusted odds for intubation and mortality were significantly lower (p < 0.01). Subgroup analyses restricted to patients with acute respiratory infections (n = 1,241), bronchiolitis (n = 761), pneumonia (n = 311), confirmed bacterial infection (n = 345), and malignancy (n = 95) showed that patients positive for a virus on testing had significantly longer PICU length of stay (multivariate p < 0.05). In children with pneumonia, identification of a respiratory virus was associated with significantly increased duration of ventilation (p = 0.003). No association between positive test results for multiple viruses and outcomes was observed.
Conclusion: Viral infections are common in critically ill children. Viral infections were associated with lower intubation and mortality rates compared with all children testing negative for viral infections. In several subgroups studied, identification of viral pathogens was associated with longer PICU length of stay while mortality was comparable. Prospective studies are required to determine the benefit of routine testing for respiratory viruses at the time of PICU admission.