Effect of Postextubation High-Flow Nasal Cannula vs Conventional Oxygen Therapy on Reintubation in Low-Risk Patients: A Randomized Clinical Trial.

IMPORTANCE: Studies of mechanically ventilated critically ill patients that combine populations that are at high and low risk for reintubation suggest that conditioned high-flow nasal cannula oxygen therapy after extubation improves oxygenation compared with conventional oxygen therapy. However, conclusive data about reintubation are lacking. OBJECTIVE: To determine whether high-flow nasal cannula oxygen therapy is superior to conventional oxygen therapy for preventing reintubation in mechanically ventilated patients at low risk for reintubation. DESIGN, SETTING, AND PARTICIPANTS: Multicenter randomized clinical trial conducted between September 2012 and October 2014 in 7 intensive care units (ICUs) in Spain. Participants were 527 adult critical patients at low risk for reintubation who fulfilled criteria for planned extubation. Low risk for reintubation was defined as younger than 65 years; Acute Physiology and Chronic Health Evaluation II score less than 12 on day of extubation; body mass index less than 30; adequate secretions management; simple weaning; 0 or 1 comorbidity; and absence of heart failure, moderate-to-severe chronic obstructive pulmonary disease, airway patency problems, and prolonged mechanical ventilation. INTERVENTIONS: Patients were randomized to undergo either high-flow or conventional oxygen therapy for 24 hours after extubation. MAIN OUTCOMES AND MEASURES: The primary outcome was reintubation within 72 hours, compared with the Cochran-Mantel-Haenszel χ2 test. Secondary outcomes included postextubation respiratory failure, respiratory infection, sepsis and multiorgan failure, ICU and hospital length of stay and mortality, adverse events, and time to reintubation. RESULTS: Of 527 patients (mean age, 51 years [range, 18-64]; 62% men), 264 received high-flow therapy and 263 conventional oxygen therapy. Reintubation within 72 hours was less common in the high-flow group (13 patients [4.9%] vs 32 [12.2%] in the conventional group; absolute difference, 7.2% [95% CI, 2.5% to 12.2%]; P = .004). Postextubation respiratory failure was less common in the high-flow group (22/264 patients [8.3%] vs 38/263 [14.4%] in the conventional group; absolute difference, 6.1% [95% CI, 0.7% to 11.6%]; P = .03). Time to reintubation was not significantly different between groups (19 hours [interquartile range, 12-28] in the high-flow group vs 15 hours [interquartile range, 9-31] in the conventional group; absolute difference, -4 [95% CI, -54 to 46]; P = .66]. No adverse effects were reported. CONCLUSIONS AND RELEVANCE: Among extubated patients at low risk for reintubation, the use of high-flow nasal cannula oxygen compared with conventional oxygen therapy reduced the risk of reintubation within 72 hours. TRIAL REGISTRATION: clinicaltrials.gov Identifier: NCT01191489.

Effect of aggressive vs conservative screening and confirmatory test on time to extubation among patients at low or intermediate risk: a randomized clinical trial.

PURPOSE: This study aimed to determine the best strategy to achieve fast and safe extubation. METHODS: This multicenter trial randomized patients with primary respiratory failure and low-to-intermediate risk for extubation failure with planned high-flow nasal cannula (HFNC) preventive therapy. It included four groups: (1) conservative screening with ratio of partial pressure of arterial oxygen (PaO2) to fraction of inspired oxygen (FiO2) ≥ 150 and positive end-expiratory pressure (PEEP) ≤ 8 cmH2O plus conservative spontaneous breathing trial (SBT) with pressure support 5 cmH2O + PEEP 0 cmH2O); (2) screening with ratio of partial pressure of arterial oxygen (PaO2) to fraction of inspired oxygen (FiO2) ≥ 150 and PEEP ≤ 8 plus aggressive SBT with pressure support 8 + PEEP 5; (3) aggressive screening with PaO2/FiO2 > 180 and PEEP 10 maintained until the SBT with pressure support 8 + PEEP 5; (4) screening with PaO2/FiO2 > 180 and PEEP 10 maintained until the SBT with pressure support 5 + PEEP 0. Primary outcomes were time-to-extubation and simple weaning rate. Secondary outcomes included reintubation within 7 days after extubation. RESULTS: Randomization to the aggressive-aggressive group was discontinued at the interim analysis for safety reasons. Thus, 884 patients who underwent at least 1 SBT were analyzed (conservative-conservative group, n = 256; conservative-aggressive group, n = 267; aggressive-conservative group, n = 261; aggressive-aggressive, n = 100). Median time to extubation was lower in the groups with aggressive screening (p < 0.001). Simple weaning rates were 45.7%, 76.78% (205 patients), 71.65%, and 91% (p < 0.001), respectively. Reintubation rates did not differ significantly (p = 0.431). CONCLUSION: Among patients at low or intermediate risk for extubation failure with planned HFNC, combining aggressive screening with preventive PEEP and a conservative SBT reduced the time to extubation without increasing the reintubation rate.

Early Tracheostomy for Managing ICU Capacity During the COVID-19 Outbreak: A Propensity-Matched Cohort Study.

BACKGROUND: During the first wave of the COVID-19 pandemic, shortages of ventilators and ICU beds overwhelmed health care systems. Whether early tracheostomy reduces the duration of mechanical ventilation and ICU stay is controversial. RESEARCH QUESTION: Can failure-free day outcomes focused on ICU resources help to decide the optimal timing of tracheostomy in overburdened health care systems during viral epidemics? STUDY DESIGN AND METHODS: This retrospective cohort study included consecutive patients with COVID-19 pneumonia who had undergone tracheostomy in 15 Spanish ICUs during the surge, when ICU occupancy modified clinician criteria to perform tracheostomy in Patients with COVID-19. We compared ventilator-free days at 28 and 60 days and ICU- and hospital bed-free days at 28 and 60 days in propensity score-matched cohorts who underwent tracheostomy at different timings (≤ 7 days, 8-10 days, and 11-14 days after intubation). RESULTS: Of 1,939 patients admitted with COVID-19 pneumonia, 682 (35.2%) underwent tracheostomy, 382 (56%) within 14 days. Earlier tracheostomy was associated with more ventilator-free days at 28 days (≤ 7 days vs > 7 days [116 patients included in the analysis]: median, 9 days [interquartile range (IQR), 0-15 days] vs 3 days [IQR, 0-7 days]; difference between groups, 4.5 days; 95% CI, 2.3-6.7 days; 8-10 days vs > 10 days [222 patients analyzed]: 6 days [IQR, 0-10 days] vs 0 days [IQR, 0-6 days]; difference, 3.1 days; 95% CI, 1.7-4.5 days; 11-14 days vs > 14 days [318 patients analyzed]: 4 days [IQR, 0-9 days] vs 0 days [IQR, 0-2 days]; difference, 3 days; 95% CI, 2.1-3.9 days). Except hospital bed-free days at 28 days, all other end points were better with early tracheostomy. INTERPRETATION: Optimal timing of tracheostomy may improve patient outcomes and may alleviate ICU capacity strain during the COVID-19 pandemic without increasing mortality. Tracheostomy within the first work on a ventilator in particular may improve ICU availability.

CO2 response and duration of weaning from mechanical ventilation.

BACKGROUND: The CO2 response test measures the hypercapnic drive response (which is defined as the ratio of the change in airway-occlusion pressure 0.1 s after the start of inspiratory flow [ΔP(0.1)] to the change in P(aCO2) [ΔP(aCO2)]), and the hypercapnic ventilatory response (which is defined as the ratio of the change in minute volume to ΔP(aCO2)). OBJECTIVE: In mechanically ventilated patients ready for a spontaneous breathing trial, to investigate the relationship between CO2 response and the duration of weaning. METHODS: We conducted the CO2 response test and measured maximum inspiratory pressure (P(Imax)) and maximum expiratory pressure (P(Emax)) in 102 non-consecutive ventilated patients. We categorized the patients as either prolonged weaning (weaning duration > 7 d) or non-prolonged weaning (≤ 7 d). RESULTS: Twenty-seven patients had prolonged weaning. Between the prolonged and non-prolonged weaning groups we found differences in hypercapnic drive response (0.22 ± 0.16 cm H2O/mm Hg vs 0.47 ± 0.22 cm H2O/mm Hg, respectively, P < .001) and hypercapnic ventilatory response (0.25 ± 0.23 L/min/mm Hg vs 0.53 ± 0.33 L/min/mm Hg, respectively, P < .001). The optimal cutoff points to differentiate between prolonged and non-prolonged weaning were 0.19 cm H2O/mm Hg for hypercapnic drive response, and 0.15 L/min/mm Hg for hypercapnic ventilatory response. Assessed with the Cox proportional hazards model, both hypercapnic drive response and hypercapnic ventilatory response were independent variables associated with the duration of weaning. The hazard ratio of weaning success was 16.7 times higher if hypercapnic drive response was > 0.19 cm H2O/mm Hg, and 6.3 times higher if hypercapnic ventilatory response was > 0.15 L/min/mm Hg. Other variables (P(0.1), P(Imax), and P(Emax)) were not associated with the duration of the weaning. CONCLUSIONS: Decreased CO2 response, as measured by hypercapnic drive response and hypercapnic ventilatory response, are associated with prolonged weaning.

Hypercapnic respiratory failure in obesity-hypoventilation syndrome: CO2 response and acetazolamide treatment effects.

OBJECTIVE: In obesity-hypoventilation-syndrome patients mechanically ventilated for hypercapnic respiratory failure we investigated the relationship between CO2 response, body mass index, and plasma bicarbonate concentration, and the effect of acetazolamide on bicarbonate concentration and CO2 response. METHODS: CO2 response tests and arterial blood gas analysis were performed in 25 patients ready for a spontaneous breathing test, and repeated in a subgroup of 8 patients after acetazolamide treatment. CO2 response test was measured as (1) hypercapnic drive response (the ratio of the change in airway occlusion pressure 0.1 s after the start of inspiratory flow to the change in P(aCO2)), and (2) hypercapnic ventilatory response (the ratio of the change in minute volume to the change in P(aCO2)). RESULTS: We did not find a significant relationship between CO2 response and body mass index. Patients with higher bicarbonate concentration had a more blunted CO2 response. Grouping the patients according to the first, second, and third tertiles of the bicarbonate concentration, the hypercapnic drive response was 0.32 ± 0.17 cm H2O/mm Hg, 0.22 ± 0.15 cm H2O/mm Hg, and 0.10 ± 0.06 cm H2O/mm Hg, respectively (P = .01), and hypercapnic ventilatory response was 0.46 ± 0.23 L/min/mm Hg, 0.48 ± 0.36 L/min/mm Hg, and 0.22 ± 0.16 L/min/mm Hg, respectively (P = .04). After acetazolamide treatment, bicarbonate concentration was reduced by 8.4 ± 3.0 mmol/L (P = .01), and CO2 response was shifted to the left, with an increase in hypercapnic drive response, by 0.14 ± 0.16 cm H2O/mm Hg (P = .02), and hypercapnic ventilatory response, by 0.11 ± 0.22 L/min/mm Hg (P = .33). CONCLUSIONS: Patients with obesity-hypoventilation syndrome and higher bicarbonate concentrations had a more blunted CO2 response. Body mass index was not related to CO2 response. Acetazolamide decreased bicarbonate concentration and increased CO2 response.

Low Albumin Levels Are Associated with Poorer Outcomes in a Case Series of COVID-19 Patients in Spain: A Retrospective Cohort Study.

There is limited information available describing the clinical and epidemiological features of Spanish patients requiring hospitalization for coronavirus disease 2019 (COVID-19). In this observational study, we aimed to describe the clinical characteristics and epidemiological features of severe (non-ICU) and critically patients (ICU) with COVID-19 at triage, prior to hospitalization. Forty-eight patients (27 non-ICU and 21 ICU) with positive severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection were analyzed (mean age, 66 years, [range, 33-88 years]; 67% males). There were no differences in age or sex among groups. Initial symptoms included fever (100%), coughing (85%), dyspnea (76%), diarrhea (42%) and asthenia (21%). ICU patients had a higher prevalence of dyspnea compared to non-ICU patients (95% vs. 61%, p = 0.022). ICU-patients had lymphopenia as well as hypoalbuminemia. Lactate dehydrogenase (LDH), C-reactive protein (CRP), and procalcitonin were significantly higher in ICU patients compared to non-ICU (p < 0.001). Lower albumin levels were associated with poor prognosis measured as longer hospital length (r = -0.472, p < 0.001) and mortality (r = -0.424, p = 0.003). As of 28 April 2020, 10 patients (8 ICU and 2 non-ICU) have died (21% mortality), and while 100% of the non-ICU patients have been discharged, 33% of the ICU patients still remained hospitalized (5 in ICU and 2 had been transferred to ward). Critically ill patients with COVID-19 present lymphopenia, hypoalbuminemia and high levels of inflammation.

Hypercapnia test as a predictor of success in spontaneous breathing trials and extubation.

BACKGROUND: The ventilatory capacity of the respiratory neuromuscular system can be studied with the hypercapnia test. OBJECTIVE: To determine whether decreased response to the hypercapnia test is associated with failure to pass a spontaneous breathing trial (SBT) or extubation failure. METHODS: We studied 103 intubated patients ready for SBT. We used a hypercapnia test in which we approximately doubled the dead space and thus caused re-inhalation of expired air. We calculated 3 ratios: the ratio of P(0.1) (airway occlusion pressure 0.1 s after the onset of inspiratory effort) during hypercapnia test to baseline P(0.1); the ratio of the change in minute volume [DeltaV(E)] to the change in P(aCO(2)) (we call this ratio the hypercapnic ventilatory response); and the ratio of the change in P(0.1) [DeltaP(0.1)] to the change P(aCO(2)) (we call this ratio the hypercapnic-respiratory-drive response). RESULTS: Thirty-six patients failed the SBT, and 11 patients failed extubation. The mean values for the SBT/extubation-success group, the extubation-failure group, and the SBT-failure group, respectively, were: ratio of hypercapnia-test P(0.1) to baseline P(0.1): 4.3 +/- 2.7, 3.7 +/- 1.3, and 3.0 +/- 1.8 (P = .03); hypercapnic ventilatory response: 0.60 +/- 0.35 L/min/mm Hg, 0.50 +/- 0.26 L/min/mm Hg, and 0.31 +/- 0.21 L/min/mm Hg (P < .001); hypercapnic respiratory-drive response: 0.48 +/- 0.24 cm H(2)O/mm Hg, 0.42 +/- 0.19 cm H(2)O/mm Hg, and 0.27 +/- 0.15 cm H(2)O/mm Hg (P < .001). For predicting SBT/extubation success, the sensitivities and specificities, respectively, were: ratio of hypercapnia-test P(0.1) to baseline P(0.1) 0.80 and 0.47; hypercapnic ventilatory response 0.86 and 0.53; hypercapnic respiratory-drive response 0.82 and 0.55. CONCLUSIONS: The SBT/extubation-failure patients had less response to the hypercapnia test than did the SBT/extubation-success patients, and the hypercapnia test was not useful in predicting SBT or extubation success.

Role of First-Line Noninvasive Ventilation in Non-COPD Subjects With Pneumonia. / El papel de la ventilación no invasiva como tratamiento ventilatorio inicial en pacientes con neumonía y sin enfermedad pulmonar obstructiva crónica.

INTRODUCTION: The use of noninvasive ventilation (NIV) in non-COPD patients with pneumonia is controversial due to its high rate of failure and the potentially harmful effects when NIV fails. The purpose of the study was to evaluate outcomes of the first ventilatory treatment applied, NIV or invasive mechanical ventilation (MV), and to identify predictors of NIV failure. METHODS: Historical cohort study of 159 non-COPD patients with pneumonia admitted to the ICU with ventilatory support. Subjects were divided into 2 groups: invasive MV or NIV. Univariate and multivariate analyses with demographic and clinical data were performed. Analysis of mortality was adjusted for the propensity of receiving first-line invasive MV. RESULTS: One hundred and thirteen subjects received first-line invasive MV and 46 received first-line NIV, of which 27 needed intubation. Hospital mortality was 35, 37 and 56%, respectively, with no significant differences among groups. In the propensity-adjusted analysis (expressed as OR [95% CI]), hospital mortality was associated with age (1.05 [1.02-1.08]), SAPS3 (1.03 [1.00-1.07]), immunosuppression (2.52 [1.02-6.27]) and NIV failure compared to first-line invasive MV (4.3 [1.33-13.94]). Compared with invasive MV, NIV failure delayed intubation (p=.004), and prolonged the length of invasive MV (p=.007) and ICU stay (p=.001). NIV failure was associated with need for vasoactive drugs (OR 7.8 [95% CI, 1.8-33.2], p=.006). CONCLUSIONS: In non-COPD subjects with pneumonia, first-line NIV was not associated with better outcome compared with first-line invasive MV. NIV failure was associated with longer duration of MV and hospital stay, and with increased hospital mortality. The use of vasoactive drugs predicted NIV failure.

High-flow nasal cannula to prevent postextubation respiratory failure in high-risk non-hypercapnic patients: a randomized multicenter trial.

BACKGROUND: Extubation failure is associated with increased morbidity and mortality, but cannot be safely predicted or avoided. High-flow nasal cannula (HFNC) prevents postextubation respiratory failure in low-risk patients. OBJECTIVE: To demonstrate that HFNC reduces postextubation respiratory failure in high-risk non-hypercapnic patients compared with conventional oxygen. METHODS: Randomized, controlled multicenter trial in patients who passed a spontaneous breathing trial. We enrolled patients meeting criteria for high-risk of failure to randomly receive HFNC or conventional oxygen for 24 h after extubation. Primary outcome was respiratory failure within 72-h postextubation. Secondary outcomes were reintubation, intensive care unit (ICU) and hospital lengths of stay, and mortality. Statistical analysis included multiple logistic regression models. RESULTS: The study was stopped due to low recruitment after 155 patients were enrolled (78 received high-flow and 77 received conventional oxygen). Groups were similar at enrollment, and all patients tolerated 24-h HFNC. Postextubation respiratory failure developed in 16 (20%) HFNC patients and in 21 (27%) conventional patients [OR 0.69 (0.31-1.54), p = 0.2]. Reintubation was needed in 9 (11%) HFNC patients and in 12 (16%) conventional patients [OR 0.71 (0.25-1.95), p = 0.5]. No difference was found in ICU or hospital length of stay, or mortality. Logistic regression models suggested HFNC [OR 0.43 (0.18-0.99), p = 0.04] and cancer [OR 2.87 (1.04-7.91), p = 0.04] may be independently associated with postextubation respiratory failure. CONCLUSION: Our study is inconclusive as to a potential benefit of HFNC over conventional oxygen to prevent occurrence of respiratory failure in non-hypercapnic patients at high risk for extubation failure. Registered at Clinicaltrials.gov NCT01820507.

Role of Respiratory Drive in Hyperoxia-Induced Hypercapnia in Ready-to-Wean Subjects With COPD.

BACKGROUND: Hyperoxia-induced hypercapnia in subjects with COPD is mainly explained by alterations in the ventilation/perfusion ratio. However, it is unclear why respiratory drive does not prevent CO2 retention. Some authors have highlighted the importance of respiratory drive in CO2 increases during hyperoxia. The aim of the study was to examine the effects of hyperoxia on respiratory drive in subjects with COPD. METHODS: Fourteen intubated, ready-to-wean subjects with COPD were studied during normoxia and hyperoxia. A CO2 response test was then performed with the rebreathing method to measure the hypercapnic drive response, defined as the ratio of change in airway-occlusion pressure 0.1 s after the start of inspiratory flow (ΔP(0.1)) to change in P(aCO2) (ΔP(aCO2)), and the hypercapnic ventilatory response, defined as the ratio of change in minute volume (ΔV̇(E)) to ΔP(aCO2). RESULTS: Hyperoxia produced a significant increase in P(aCO2) (55 ± 9 vs 58 ± 10 mm Hg, P = .02) and a decrease in pH (7.41 ± 0.05 vs 7.38 ± 0.05, P = .01) compared with normoxia, with a non-significant decrease in V̇(E) (9.9 ± 2.9 vs 9.1 ± 2.3 L/min, P = .16) and no changes in P(0.1) (2.85 ± 1.40 vs 2.82 ± 1.16 cm H2O, P = .97) The correlation between hyperoxia-induced changes in V̇(E) and P(aCO2) was r(2) = 0.38 (P = .02). Median ΔP(0.1)/ΔP(aCO2) and ΔV̇(E)/ΔP(aCO2) did not show significant differences between normoxia and hyperoxia: 0.22 (0.12-0.49) cm H2O/mm Hg versus 0.25 (0.14-0.34) cm H2O/mm Hg (P = .30) and 0.37 (0.12-0.54) L/min/mm Hg versus 0.35 (0.12-0.96) L/min/mm Hg (P = .20), respectively. CONCLUSIONS: In ready-to-wean subjects with COPD exacerbations, hyperoxia is followed by an increase in P(aCO2), but it does not significantly modify the respiratory drive or the ventilatory response to hypercapnia.

Prone positioning in patients with acute respiratory distress syndrome.

Acute respiratory distress syndrome (ARDS) is a severe form of respiratory failure that is characterized by marked hypoxemia, bilateral infiltrates on chest radiograph, and no clinical evidence of left ventricular failure. Mechanical ventilation with positive end-expiratory pressure (PEEP) is a cornerstone therapy for ARDS patients. Because the fundamental aim of supportive treatment is to improve arterial oxygenation, several alternatives to mechanical ventilation with PEEP have been used. One of these alternative therapies is prone positioning, which has been used safely to improve oxygenation in many patients with ARDS. Despite encouraging results, however, the use of prone positioning is not widely accepted as an adjunct to therapy in hypoxemic patients because, aside from temporarily improving gas exchange, it does not seem to affect the outcome of these patients. This article reviews the rationale for using prone positioning in ARDS patients who require intubation and mechanical ventilation.

Central respiratory drive in patients with neuromuscular diseases.

BACKGROUND: The contribution of the central respiratory drive in the hypercapnic respiratory failure of neuromuscular diseases (NMD) is controversial. OBJECTIVE: To compare the CO2 response and the duration of weaning of mechanical ventilation between a group of NMD patients and a group of quadriplegic patients due to ICU-acquired weakness (ICU-AW). METHODS: We prospectively studied 16 subjects with NMD and 26 subjects with ICU-AW ready for weaning, using the method of the re-inhalation of expired air. We measured the hypercapnic drive response, defined as the ratio of change in airway occlusion pressure 0.1 second after the start of inspiration (ΔP0.1) to the change in Paco2 (ΔPaco2), and the hypercapnic ventilatory response, defined as the ratio of the change in minute ventilation (ΔVe) to ΔPaco2. We considered a value of ≤ 0.19 cm H2O/mm Hg as reduced hypercapnic drive response. RESULTS: Hypercapnic drive response (ΔP0.1/ΔPaco2 = 0.14 ± 0.08 cm H2O/mm Hg vs 0.37 ± 0.27 cm H2O/mm Hg, P = .002) and hypercapnic ventilatory response (ΔVe/ΔPaco2 = 0.21 ± 0.19 L/min/mm Hg vs 0.44 ± 0.40 L/min/mm Hg, P = .02) were lower in the NMD than in the ICU-AW subjects. Duration of weaning values, according to the Kaplan-Meier curves, were similar in both groups (Log-rank = 0.03, P = .96). Eleven NMD (69%) and 9 ICU-AW (35%) subjects had hypercapnic drive response ≤ 0.19 cm H2O/mm Hg. The duration of weaning was longer in subjects with hypercapnic drive response ≤ 0.19 cm H2O/mm Hg (log-rank = 15.4, P < .001). CONCLUSIONS: Subjects with acute hypercapnic respiratory failure due to NMD had reduced hypercapnic drive response, compared to ICU-AW subjects. The duration of weaning was longer in subjects with reduced hypercapnic drive response.

A multicenter trial of prolonged prone ventilation in severe acute respiratory distress syndrome.

RATIONALE: Ventilation in the prone position for about 7 h/d in patients with acute respiratory distress syndrome (ARDS), acute lung injury, or acute respiratory failure does not decrease mortality. Whether it is beneficial to administer prone ventilation early, and for longer periods of time, is unknown. METHODS: We enrolled 136 patients within 48 h of tracheal intubation for severe ARDS, 60 randomized to supine and 76 to prone ventilation. Guidelines were established for ventilator settings and weaning. The prone group was targeted to receive continuous prone ventilation treatment for 20 h/d. RESULTS: The intensive care unit mortality was 58% (35/60) in the patients ventilated supine and 43% (33/76) in the patients ventilated prone (p = 0.12). The latter had a higher simplified acute physiology score II at inclusion. Multivariate analysis showed that simplified acute physiology score II at inclusion (odds ratio [OR], 1.07; p < 0.001), number of days elapsed between ARDS diagnosis and inclusion (OR, 2.83; p < 0.001), and randomization to supine position (OR, 2.53; p = 0.03) were independent risk factors for mortality. A total of 718 turning procedures were done, and prone position was applied for a mean of 17 h/d for a mean of 10 d. A total of 28 complications were reported, and most were rapidly reversible. CONCLUSION: Prone ventilation is feasible and safe, and may reduce mortality in patients with severe ARDS when it is initiated early and applied for most of the day.