We read with interest Meyer and Hill's excellent review on noninvasive positive pressure ventilation [1]. The authors provide some selection guidelines for its use that are based on reports in the literature. According to those guidelines, noninvasive positive pressure ventilation should be avoided in patients with acute respiratory failure associated with hemodynamic instability.

We have noted that not all the cited reports included information on the hemodynamic status among their selection criteria in patients with acute disease. Further, when excluding hemodynamically unstable patients, the authors failed to define hemodynamic stability.

It is well known that conventional mechanical ventilation by endotracheal intubation may seriously alter hemodynamic status (particularly when external positive end-expiratory pressure, invasive positive pressure ventilation, or intermittent mechanical ventilation is used) [2]. Hemodynamic status is less affected in patients treated with pressure support ventilation [3]. We know of no reports of similar adverse effects of noninvasive positive pressure ventilation when nasal or facial masks are used. Therefore, could this type of ventilation represent a possible alternative ventilatory treatment in patients with hemodynamic instability?

We used bilevel positive airway pressure ventilation (BiPAP; Respironics, Inc.) through a nasal mask to ventilate 57 consecutive, unselected patients with acute respiratory failure who were unresponsive to oxygen and medical therapy alone. We applied noninvasive positive pressure ventilation for almost 22 hours daily, with a median inspiratory positive airway pressure of 15 cm H₂O (maximum, 20 cm H_2O; minimum, 8 cm H_2O) and a median expiratory positive airway pressure of 4 cm H₂O (range, 8 cm H₂O to 3 cm H₂O). Seven patients had hypotension (systolic blood pressure <100 mm Hg; mean, 88 ±6 mm Hg), 11 had radiologic and clinical findings of pulmonary edema, 20 had edema of the legs; 38 had tachycardia; 21 had hypertension; and 20 had arrhythmias. After 24 hours of BiPAP ventilation, only three patients with initial normal blood pressure and none of those with hypotension showed a decrease of 15 mm Hg or more in their mean blood pressure. None failed to respond to BiPAP. Of the 57 patients, 39 were successfully ventilated with BiPAP, and 18 patients required intubation or died. At the time of hospitalization, we found no significant variations in blood pressure, heart rate, presence of edema of the legs, arrhythmias, respiratory rate, or blood gas values between patients successfully ventilated with BiPAP and those who failed to respond to treatment. Only the prognostic score of the Acute Physiology and Chronic Health Evaluation III (APACHE III) was significantly higher in patients who failed to respond to noninvasive positive pressure ventilation (P < 0.01). Even in our previous report of 30 patients with severe exacerbations of chronic obstructive pulmonary disease, we found the APACHE III score to be the only determinant of success or failure of bilevel noninvasive positive pressure ventilation [4].

Clearly, our experience does not have the power of a perfectly designed study. It could, however, be of interest because it presents a series of unselected patients, including those who were hemodynamically unstable. Hemodynamic instability is often associated with exacerbation of obstructive airway disease and other causes of respiratory distress in patients with auto-positive end-expiratory pressure, even in the absence of pulmonary edema [5]. Clinical experience suggests that in acutely ill patients, there is often not enough time to evaluate the stability of the hemodynamic status before a decision to ventilate must be made. We therefore believe that when intensive care is available and promptly administered, noninvasive positive pressure ventilation may be attempted as a first-step treatment, even in hemodynamically unstable patients.