Relationship of Viral Detection with Duration of Ventilation in Critically Ill Infants with Lower Respiratory Tract Infection

Ann Am Thorac Soc. 2021 Oct;18(10):1677-1684. doi: 10.1513/AnnalsATS.202008-996OC.

Abstract

Rationale: Although respiratory virus testing is frequently done for critically ill infants with bronchiolitis, the prognostic value of this testing is unknown for those requiring positive-pressure ventilation (PPV). Objectives: To determine the differences in PPV use according to viral detection and to explore the association between viral detection and duration of PPV in critically ill children with presumed respiratory infection. Methods: This is a retrospective cohort study in a quaternary pediatric intensive care unit from February 2014 until February 2017. We evaluated 984 children less than 1 year of age who received PPV for presumed respiratory infection without significant congenital heart disease, care limitations, baseline PPV usage, or tracheostomy. Respiratory viruses were identified using a PCR panel. Analyses of duration of PPV according to viral etiology were performed using univariate and multivariable logistic regression and truncated negative binomial regression with calculated mean marginal effects (MME). Results: Overall, 85 (9%) infants had no viruses identified, 629 (64%) had a single virus detected, most commonly respiratory syncytial virus (417, 42%) followed by rhinovirus/enterovirus (145, 15%), 230 (23%) had two viruses detected, and 40 (4%) had three viruses detected. Compared with those with one or no virus detected, infants with ⩾2 viruses received longer total PPV duration in adjusted analysis (relative risk [RR], 1.4; 95% confidence interval [CI], 1.2-1.6; P < 0.001; MME = 29 h). Detection of rhinovirus/enterovirus alone, compared with respiratory syncytial virus alone, was associated with significantly shorter duration of total PPV (RR, 0.7; 95% CI, 0.62-0.87; P = <0.001; MME = -23 h), noninvasive PPV (RR, 0.7; 95% CI, 0.60-0.85; P < 0.001; MME = -15 h), and invasive PPV (RR, 0.7; 95% CI, 0.54-0.83; P < 0.001; MME = -54 h) when adjusted for weight, prematurity, and administration of early antibiotic therapy. Conclusions: Identification of viral type and number in severe bronchiolitis is an important predictor of duration of PPV.

Keywords: artificial respiration; bronchiolitis; pediatric intensive care unit.

MeSH terms

  • Bronchiolitis*
  • Child
  • Critical Illness
  • Humans
  • Infant
  • Respiratory Syncytial Virus, Human*
  • Respiratory Tract Infections* / diagnosis
  • Respiratory Tract Infections* / epidemiology
  • Retrospective Studies