Clinical characteristics, treatment modalities, and mortality in pediatric acute respiratory distress syndrome: a single-center cohort study

Abstrakts

Background Pediatric acute respiratory distress syndrome (PARDS) remains an important cause of morbidity and mortality in pediatric intensive care units (PICUs), but data from Turkey and comparable middle-income settings are limited. We aimed to describe the etiology, management and outcomes of children with PARDS in a tertiary PICU and to explore factors associated with PICU mortality, with particular attention to in-hospital complications. Methods In this single-center retrospective cohort study, all 1,350 children admitted to a 20-bed tertiary PICU in Ankara, Turkey, between August 2019 and September 2021 were screened, and 38 children with PARDS defined by the PALICC-2015 criteria were included. The primary outcome was PICU mortality. Survivors and non-survivors were compared using non-parametric tests, and an exploratory multivariable logistic regression model, limited to three variables because of the small number of deaths, was fitted. Results The median age was 30.5 months (IQR 6.8–79.8), and infectious pneumonia was the leading cause (29/38, 76.3%). Thirty-one children (81.6%) required invasive mechanical ventilation and were classified as having mild ( n = 5), moderate ( n = 9) or severe ( n = 17) PARDS; seven were managed with non-invasive ventilation alone and could not be classified by severity. PICU mortality was 39.5% (15/38): 0/5 in mild, 2/9 in moderate and 13/17 in severe PARDS ( p = 0.001). At least one selected in-hospital complication (pulmonary hemorrhage, air leak or hospital-acquired infection) occurred in 20 children, 13 of whom died. Pulmonary hemorrhage (11/11 died; p < 0.001) and air leak syndromes (9/12 died; p = 0.004) were associated with mortality, whereas hospital-acquired infections were not ( p = 0.544). In the exploratory multivariable model, the composite of selected in-hospital complications was associated with mortality after adjustment for sex and prone positioning (adjusted OR 8.62, 95% CI 1.29–57.68; p = 0.026). Conclusion In this single-center cohort, PICU mortality was high and concentrated among children with severe PARDS. Pulmonary hemorrhage and air leak syndromes were frequent among non-survivors. Because complication timing was not available and the sample was small, these findings should be regarded as associations reflecting disease severity in our setting rather than as independent predictors of mortality.