The trusted source for
healthcare information and
Synopsis: Despite early improvement in oxygenation, continuous positive airway pressure administered by face mask failed to reduce the need for intubation or survival in patients with hypoxic, nonhypercarbic acute lung injury.
Source: Delclaux C, et al. JAMA 2000;284:2352-2360.
Continuous positive airway pressure administered by face mask (Mask CPAP) is occasionally used to treat patients with acute hypoxic lung injury. The benefits of this approach are believed to be improved oxygenation and a reduced need for invasive ventilation. This intervention has been documented to improve gas exchange parameters, but the actual effect on patient outcome is not well documented. Delclaux and colleagues report the results of a multicenter prospective, randomized, controlled trial of CPAP in patients presenting with non-hypercarbic acute lung injury. They examined the effects of the therapy on oxygenation as well as the need for intubation and survival. Patients with a history of heart failure were included if they had a non-cardiogenic cause for their current illness; a stratified randomization procedure allowed equal distribution of these patients in the study.
One hundred twenty-three consecutive adult patients presenting to one of six hospitals with acute hypoxic respiratory failure were randomized. Entrance criteria included a PaO2/FiO2 ratio less than 300 mm Hg on 10 L/min of supplemental oxygen for 15 minutes, plus bilateral diffuse infiltrates on chest radiograph. Patients were accepted only if less than three hours had passed since meeting entrance criteria. Patients with coma, altered mental status, respiratory acidosis, ventricular arrhythmias, a requirement for vasopressor support, an oxyhemoglobin saturation by pulse oximetry (SpO2) of less than 80% were excluded. Patients received either standard treatment (mask oxygen) or Mask CPAP. Failure of either treatment arm included any of the following: deceased alertness, agitation requiring sedative medications, signs of exhaustion, hemodynamic instability or cardiac arrest, or refractory hypoxemia (SpO2 < 85% on 100% O2). Patients were considered treated successfully when their PaO2/FiO2 ratio was greater than 300 mm Hg.
The patients were matched in age (58 years), sex (64% male), and the presence of cardiac disease (18% each group), its cause and severity. One patient in each group had a contraindication to intubation but completed the trial. Within one hour of beginning treatment, the CPAP group had an improved average PaO2/FiO2 ratio, decreased respiratory rate and increased pH, as well as feeling subjectively better. During the remainder of the study, however, there were no differences in any of these assessments. There were no differences in rate of intubation (CPAP = 37.5% vs 44% in patients without cardiac disease; CPAP = 27% vs 30% in patients with cardiac disease), length of hospital stay (about 16 days), or hospital mortality (CPAP = 30% vs 27% in patients without cardiac disease; CPAP = 32:% vs 35% in patients with cardiac disease). Intubation occurred earlier in the oxygen alone group. The total number of complications was higher in the CPAP group, but no specific complication difference reached statistical significance.
On first blush, this paper seems to demonstrate no effect of CPAP on prevention of intubation or improvement in outcome in patients with hypoxic lung injury. Intubation was performed at a later point in the disease process in the CPAP group. As usual, though, the devil is in the details. The CPAP group had the highest number of complications, including facial skin necrosis, stress ulcer (no prophylaxis), and nosocomial pneumonia. There were four patients who sustained cardiac arrest in the CPAP group, and none in the O2 alone group. This is interesting because one of these was when the CPAP was accidentally removed during a nursing maneuver during which profound hypoxemia ensued. The other three cardiac arrests were during elective intubation of CPAP patients. It’s probable that the delayed intubation in these patients resulted in patients who were in extremis requiring intubation and thus, faring poorly. This suggests that the criteria for intubation were applied differently in the two patient groups, making the comparison unreliable. It also suggests that early rather than late intubation may have benefits in treating hypoxic lung injury. There were no differences in important outcomes, however, so this delay did not create a statistical difference in outcome between the two treatments.
A significant problem with this study is the actual use of CPAP. Patients were treated with face mask CPAP less than continually. In fact, if the patient received only six hours of CPAP in a 24-hour period, this was considered a success. This means that those patients who received CPAP only received it when their gas exchange variables deteriorated to the entrance criteria. It was withdrawn when they improved and was not used therapeutically. With this use pattern in mind, the study demonstrates the value of this therapy in improving oxygenation in the short term with no ultimate outcome worsening.
While this is an interesting study, it really does not answer many of the important questions about using mask CPAP. Does CPAP, continuously used, modify the course of hypoxic lung injury? Is there an increased risk of any particular complication with its use? Does it delay intubation to a point where intubation is more dangerous? And what is the cost (in terms of caregiver time) needed to safely provide this therapy? These and other important questions remain to be evaluated.