Objective: Identify factors associated with the survival of pediatric patients who are submitted to mechanical ventilation (MV) for more than 12 hours.
Design: International prospective cohort study. It was performed between April 1 and May 31 1999. All patients were followed-up during 28 days or discharge to pediatric intensive care unit (PICU).
Setting: 36 PICUs from 7 countries.
Patients: A total of 659 ventilated patients were enrolled but 15 patients were excluded because their vital status was unknown on discharge.
Results: Overall in-UCIP mortality rate was 15,6%. Recursive partitioning and logistic regression were used and an outcome model was constructed. The variables significantly associated with mortality were: peak inspiratory pressure (PIP), acute renal failure (ARF), PRISM score and severe hypoxemia (PaO2/FiO2 < 100). The subgroup with best outcome (mortality 7%) included patients who were ventilated with a PIP < 35 cmH2O, without ARF, or PaO2/FiO2 > 100 and PRISM < 27. In patients with a mean PaO2/FiO2 < 100 during MV mortality increased to 26% (OR: 4.4; 95% CI 2.0 to 9.4). Patients with a PRISM score > 27 on admission to PICU had a mortality of 43% (OR: 9.6; 95% CI 4,2 to 25,8). Development of acute renal failure was associated with a mortality of 50% (OR: 12.7; 95% CI 6.3 to 25.7). Finally, the worst outcome (mortality 58%) was for patients with a mean PIP >/= 35 cmH2O (OR 17.3; 95% CI 8.5 to 36.3).
Conclusion: In a large cohort of mechanically ventilated pediatric patients we found that severity of illness at admission, high mean PIP, development of acute renal failure and severe hypoxemia over the course of MV were the factors associated with lower survival rate.