ObjectivesTo address the following issues regarding the use of prone position ventilation in patients with severe acute respiratory distress syndrome (ARDS): a) response rate; b) magnitude and duration of improved oxygenation in responders during a 12-hr trial and the consequences of returning to the supine position; c) effects of the prone position on gas exchange and hemodynamics; d) consequences of oxygenation in nonresponders; and e) effects of repeated prone position trials.DesignProspective, nonrandomized interventional study.SettingMedical intensive care unit, university tertiary care center.PatientsNineteen consecutive, mechanically ventilated patients (age 45 +/- 20 yrs, mean +/- SD) with ARDS and severe hypoxemia, defined as PaO2/FIO2of <or=to150 with FIO2or=to24 hrs, and a pulmonary artery occlusion pressure of <18 mm Hg.InterventionsPatients were turned prone for 2 hrs. Nonresponders were returned supine, but responders were maintained prone for 12 hrs before being returned to the supine position. The procedure was repeated on a daily basis in all patients, until inclusion criteria were no longer met or the patients died.Measurements and Main ResultsHemodynamic, blood gas, and gas exchange measurements were performed at the following time points: a) baseline supine; b) after 30 mins prone; and c) after 120 mins prone. Additional measurements for nonresponders were taken after 30 mins supine. For responders, additional measurements were taken after 12 hrs prone and 30 mins supine. Patients were considered responders if an increase in PaO2or=to1.3 kPa), or increase in the PaO2/FIO2or=to20 occurred within 120 mins. Eleven (57%) patients responded to the prone position. There was no difference in initial baseline parameters between responders and nonresponders. After 30 mins, the prone position in responders increased PaO2and decreased calculated venous admixture (Qva/Qt). This improvement was the maximal obtained, and was maintained throughout the 12-hr prone period. After 12 hrs prone, mean FIO2had been lowered from 0.85 +/- 0.16 to 0.66 +/- 0.18 (p < .05). Thirty minutes after the patients were returned supine, PaO2, PaO (2/FIO)2, and Qva/Qt were not different from 12-hr prone values, and were improved in comparison with baseline supine values. There was no worsening of gas exchange or hemodynamics in nonresponders. After the initial trial, a total of 28 additional episodes of prone position ventilation were performed in nine of the 19 patients. Of the 24 additional episodes in the responders, there was a response in 17 (71%) of 24 episodes. In the four additional episodes in nonresponders, there was a response in only one (25%) of four episodes. Response was accompanied by the same beneficial effects on gas exchange and Qva/Qt and absence of effect on hemodynamics as in the initial trial. There was no worsening in gas exchange or hemodynamics in nonresponder trials.ConclusionsBased on the data from this study, the prone position can improve oxygenation in severely hypoxemic ARDS patients without deleterious effects on hemodynamics. This beneficial effect does not immediately disappear on return to the supine position. In our patients, an absence of response to this technique was not accompanied by worsening hypoxemia or hemodynamic instability. Repeated daily trials in the prone position should be considered in the management of ARDS patients with severe hypoxemia. (Crit Care Med 1998; 26:1977-1985)