BACKGROUND:Approximately 0.1-0.2 of pregnancies are complicated by respiratory failure requiring mechanical ventilatory support, but few data exist to inform clinical management. This study aimed to characterize current practice and the effect of delivery on respiratory function.METHODS:A retrospective review was performed of pregnant women who received mechanical ventilation for more than 24h, from four intensive care units in institutions with large-volume obstetric units.RESULTS:Data were collected from 29 patients with a mean gestation at intensive care unit admission of 25.3 ± 6 weeks. Tidal volumes were 7.7 ± 1.7 mL/kg predicted body weight. Estimated respiratory system compliance was reduced, but was higher in four patients ventilated for neurological conditions without lung disease. Three maternal and three neonatal deaths occurred. Ten patients delivered while on ventilatory support: one spontaneous delivery, four for obstetric indications and five for worsening maternal condition. Following delivery of these 10 patients, three demonstrated a greater than 50 decrease in oxygenation index and five a greater than 50 increase in compliance. No characteristics identified which patients may benefit from delivery.CONCLUSIONS:Review of current practice in four centers suggests that mechanical ventilation in pregnant patients follows usual guidelines applicable to non-pregnant patients. Delivery was associated with modest improvement in maternal respiratory function in some patients. Any potential benefit of delivery in improving maternal physiology must be weighed against the stress of delivery. The risks of premature birth for the fetus must be weighed against continued exposure to maternal hypoxemia and hypotension.