ICU patients with thoracic trauma who had no other indication for intubation than marked hypoxaemia (pO2/FiO2 < 200 mmHg) were randomised to intubation vs non-invasive ventilation (NIV). Analgesia was via epidural bupivacaine / fentanyl or iv remifentanil. Numbers are small (total 50 patients) - partly because the trial was stopped early due to large difference in the outcome of tracheal intubation between the two groups favouring NIV. Length of hospital stay was significantly shorter in the NIV group but there was no survival difference. Noninvasive ventilation reduces intubation in chest trauma-related hypoxemia: a randomized clinical trial
Chest. 2010 Jan;137(1):74-80