Background: Acute respiratory distress syndrome is a well-recognized condit
ion resulting in high permeability pulmonary edema associated with a high m
orbidity.
Objectives: To examine a 10 year experience of pre disposing factors, descr
ibe the clinical course, and assess predictors of mortality in children wit
h this syndrome.
Methods: The medical records of all admissions to the pediatric intensive c
are unit over a 10 year period were evaluated to identify children with ARD
S'. Patients were considered to have ARDS if they met all of the following
criteria: acute onset of diffuse bilateral pulmonary infiltrates of non-car
diac origin and severe hypoxemia defined by <200 partial pressure of oxygen
during greater than or equal to 6 cm H2O positive end-expiratory pressure
for a minimum of 24 hours. The medical records were reviewed for demographi
c, clinical, and physiologic information including PaO22/forced expiratory
O-2, alveolar-arterial O-2 difference, and ventilation index.
Results: We identified 39 children with the adult respiratory distress synd
rome. Mean age was 7.4 years (range 50 days to 16 years) and the male:femal
e ratio was 24:15. Predisposing insults included sepsis, pneumonias, malign
ancy, major trauma, shock, aspiration, near drowning, burns, and envenomati
on. The mortality rate was 61.5%. Predictors of death included the PaO2/FIO
2, ventilation index and A-aDO(2)(3) on the second day after diagnosis. Non
survivors had significantly lower PaO2/FIO2 (116+/-12 vs. 175+/-8.3, P<0.00
1), and higher A-aDO(2) (368+/-28.9 vs. 228.0+/-15.5, P<0.001) and ventilat
ion index (43.3+/-2.9 vs. 53.1+/-18.0, P<0.001) than survivors.
Conclusions: Local mortality outcome for ARDS is comparable to those in ter
tiary referral institutions in the United States and Western Europe. The Pa
O2/FIO2, A-aDO(2) and ventilation index are valuable for predicting outcome
in ARDS by the second day of conventional therapy. The development of a lo
cal risk profile may allow early application of innovative therapies in thi
s population.