The Association Between Alveolar Dead Space Fraction and Mortality in Pediatric Acute Respiratory Distress Syndrome: A Prospective Cohort Study

Abstract

BackgroundAlveolar dead-space fraction (AVDSF), the volume of alveolar gas that does not participate in gas exchange, has been reported to predict mortality and morbidity in adults with acute respiratory distress syndrome (ARDS). This study aims to characterize AVDSF in patients with pediatric ARDS (PARDS), to determine its association with clinical outcomes and examine the validity of a previously studied cutoff (AVDSF > 0.25).MethodsThis was a prospective cohort study performed in the setting of a lung protective mechanical ventilation protocol. AVDSF was calculated by the equation: AVDSF = [partial pressure of arterial carbon dioxide (PaCO2) – end tidal carbon dioxide (etCO2)]/PaCO2. Receiver operating curve and Youden index were used to identify an AVDSF cutoff associated with mortality, which characterized “high or low AVDSF” groups. Correlation between AVDSF and clinical indices of severity were determined [including daily oxygenation index (OI), admission Pediatric Index of Mortality 2 (PIM 2) and Pediatric Logistic Organ Dysfunction (PELOD) scores]. The primary outcome, mortality in PARDS patients, was compared between the high and low AVDSF groups and analyzed in a multivariable logistic regression adjusting for inotrope use and PIM 2 score. Secondary outcomes included 28-day ventilator-free (VFD) and intensive care unit-free (IFD) days.ResultsSixty-nine PARDS patients with a median (interquartile range) age of 4.5 (0.8, 10.6) years were included in this analysis. Daily AVDSF correlated with daily OI (R2 = 0.10; p < 0.001). Mean AVDSF over the first 7 days of PARDS correlated with PIM 2 (R2 = 0.10; p = 0.010) and PELOD (R2 = 0.12; p = 0.004) scores. The greatest area under the curve identified an AVDSF cutoff of 0.22, which was close to the previously suggested 0.25. The high AVDSF group had higher mortality [7/19 (36.8%) vs. 5/50 (10.0%); p = 0.009] and lower VFD [2 (0, 18) vs. 21 (15, 24); p = 0.007] and IFD [0 (0, 16) vs. 16 (5, 21); p = 0.013]. In the multivariable model, being in the high AVDSF group [adjusted odds ratio 4.67 (95% CI: 1.12, 19.39)] was significantly associated with mortality.ConclusionsHigh AVDSF was independently associated with increased mortality and decreased VFD and IFD. AVDSF may be complementary to oxygenation indices in risk stratifying PARDS and warrant further study.