Early Clinical and Inflammatory Predictors of Supplemental Oxygen Requirement in Children Hospitalized with Acute Lower Respiratory Infection: A Multivariable Analysis

Early identification of children at risk of severe acute respiratory infection remains a clinical challenge. Although C-reactive protein (CRP), procalcitonin (PCT), and white blood cell (WBC) count are commonly assessed at hospital admission, their predictive value for severity assessment remains uncertain. We aimed to evaluate the utility of clinical features, a respiratory multiplex PCR test detecting 23 pathogens, imaging findings, and early inflammatory biomarkers in predicting subsequent supplemental oxygen requirement, used as a marker of severe acute lower respiratory infection (ALRI), among hospitalized children. The primary outcome was the need for supplemental oxygen during hospitalization. We conducted a single-center retrospective cohort study of children admitted to the pediatric department between May 2022 and June 2023 with ALRI or other conditions for which respiratory infection was considered but subsequently excluded. All included patients underwent multiplex respiratory PCR testing. We included 439 children in the analysis [median age, 20 mo (IQR: 5–54); 52.7% males]. Out of them, 178 (40.5%) required supplemental oxygen. The strongest positive association with subsequent supplemental oxygen requirement was observed for low oxygen saturation on admission [OR 8.83 (95% CI, 3.00–26.02; p < 0.001)], although this variable was documented in only about half of the cohort (52.4%). The strongest predictors of subsequent supplemental oxygen requirement in multivariable analysis were age, abnormal auscultatory findings, and the cumulative number of danger signs, including hypoxemia, tachypnea, dyspnea, and increased work of breathing. Among the 23 pathogens tested, respiratory syncytial virus (RSV) was the only pathogen consistently associated with a more severe course: 60.3% of RSV-positive children required supplemental oxygen, compared with 36.9% of RSV-negative children (OR 2.59, 95% CI 1.53–4.40; p < 0.001). For predicting subsequent supplemental oxygen requirement, no single marker had strong utility. These findings support the importance of structured bedside clinical assessment for identifying hospitalized children who may subsequently require supplemental oxygen. Given incomplete documentation of some vital signs, associations involving these variables should be interpreted with caution.