Adapting prognostic respiratory variables of ARDS in children to small-scale community needs.

PURPOSE: The clinical literature on the incidence and subsequent mortality of adult respiratory distress syndrome (ARDS) has come primarily from the experiences of large tertiary referral centers, particularly in Western Europe and North America. Consequently, very little has been published on the incidence, management, and outcome of ARDS in smaller community-based intensive care units. We aimed to delineate early clinical respiratory predictors of death in children with ARDS on the modest scale of a community hospital.

MATERIALS AND METHODS: A retrospective chart review of children with ARDS needing conventional mechanical ventilation admitted to our pediatric intensive care unit from 1984 to 1997. The diagnosis of ARDS was based on acute onset of diffuse, bilateral pulmonary infiltrates of noncardiac origin and severe hypoxemia defined by partial pressure of oxygen <200 mm Hg during positive end-expiratory pressure (PEEP) of 6 cm H2O or greater for a minimum of 24 hours. Demographic, clinical, and physiological data including PaO2/ FIO2, A-aDo2, and ventilation index were retrieved.

RESULTS: Fifty-six children with ARDS aged 8 +/- 5.5 years (range, 50 days to 21 years) were identified. The mortality rate was 50%. Early predictors of death included the peak inspiratory pressure (PIP), ventilation index, and PEEP on the third day after diagnosis: Nonsurvivors had significantly higher PIP (35.3 +/- 10.5 cm H2O vs 44.4 +/- 10.7 cm H2O, P < .001), PEEP (8 +/- 2.8 cm H2O vs 10.7.0 +/- 3.5 cm H2O, P < .01), and ventilation index (49.14 +/- 20.4 mm Hg x cm H2O/minute vs 61.6 +/- 51.1 mm Hg cm H2O/minute) than survivors. In contrast, PAO2/FIO2 and A-a DO2 were capable of predicting outcome by day 5 and thereafter.

CONCLUSIONS: A small-scale mortality outcome for ARDS is comparable to large tertiary referral institutions. The PIP, PEEP, and ventilation index are valuable for predicting outcome in ARDS by the third day of conventional therapy. The development of a local risk profile may assist in decision-making of early application of supportive therapies in this population.