Introduction: Apart from age and underlying illness, predictors of adverse outcome in children hospitalized with influenza are poorly understood. Our goal is to determine clinical and laboratory predictors that will help identify children at increased risk for an unfavorable outcome.
Methods: A retrospective, observational cohort study conducted at the Rambam Healthcare Campus in Haifa. We analyzed data from electronic records of children < 18 years with influenza A or B infection hospitalized between 2009 and 2020. Multivariate regression analyzes were used to identify predictors of unfavorable outcome, defined as mortality, ICU admission, intubation, prolonged length of stay, or bacterial co-infection.
Results: A total of 1077 children were included, of whom 54% were male. The median age was 2.5 (1-5.9) years. Influenza A was detected in 797 (74%) and influenza B in 286 (26%) of the cases. Underlying illness was reported in approximately 35% and admission to the PICU in 5% of cases. In multivariate analyses, desaturation < 90% on admission was associated with an 8-fold increase in risk of unfavorable disease [p < 0.001, CI (3.59-17.66))]. Also, C-reactive protein > 5 mg/dL at admission [p < 0.001, OR 3.25 (2.16-4.91)], viral coinfection [p 0.036, OR 1.71(1.03-2.83)], and presence of underlying disease [p < 0.001, OR 2.77 (1.96-3.9)].
Conclusions: Data available on admission can help identify children hospitalized with influenza who are at increased risk for complications and unfavorable outcome, promoting aggressive treatment and care. These factors may help establish a future score predicting unfavorable outcome in children hospitalized with influenza.