From Wiki Journal Club
Jump to navigation Jump to search
Guérin C, et al. "Prone positioning in severe acute respiratory distress syndrome". The New England Journal of Medicine. 2013. 368(23):2159-2168.
PubMedFull textPDF

Clinical Question

Among patients with severe ARDS, does prone positioning reduce all-cause mortality at 28 days when compared to supine-only positioning?

Bottom Line

Among patients with severe ARDS (P:F ratio <150 mm Hg), prone positioning reduces 28-day mortality.

Major Points

Acute respiratory distress syndrome (ARDS) is associated with a high mortality and prolonged periods of intubation. While some ground was gained through low tidal volume (Vt) ventilation (ARDSNet; 2000), low fluid volume (FACTT; 2006), and therapeutic paralysis (ACURASYS; 2010), the mortality approaches 25% even in the era of these and other lung-protective strategies.[1] Prone positioning, which seems to improve ventilation-perfusion matching through aeration of previously atelectatic dependent alveoli, has been shown in unblinded studies to improve survival particularly when used as rescue therapy in severe ARDS (ie, P:F <100 mmHg). While this finding was corroborated in a 2010 meta-analysis of seven well-designed randomized trials in severe ARDS,[2] a large randomized trial was lacking.

The 2013 Proning Severe ARDS Patients (PROSEVA) trial randomized 466 patients with severe ARDS in European ICUs to early (<36h after intubation), lengthy (goal 16 hours daily), intermittent prone positioning or to a standard supine position. Severe ARDS was defined as a P:F ratio of <150 mmHg with FiO2 >60% and PEEP ≥5 cm H2O. At 28 days, the prone group had a 51% relative and 17% absolute reduction in all-cause mortality when compared to the supine group (16.0% vs. 32.8%; NNT 6). Importantly, PROSEVA incorporated low-tidal ventilation and therapeutic paralysis, suggesting that prone positioning confers a survival benefit beyond that seen with the standard of care. Adverse events more common with prone positioning included unscheduled extubation (13.3% vs. 10.9%) and ET tube obstruction (4.9% vs. 2.2%). PROSEVA did not report incidence of pressure ulcers, which have been shown to occur more frequently in prone-positioned patients.[2]

The large survival benefit in PROSEVA was markedly higher than the 16% relative risk reduction in the 2010 meta-analysis referenced earlier.[2] Some of this discrepancy may reflect prior trials' inadequate duration of prone positioning and their inclusion of less critically ill patients who are unlikely to benefit from prone positioning.[3] However, an alternative explanation stems from differences in baseline characteristics between study groups at the time of inclusion. In particular, although the prone group had a similar SAPS II score, their SOFA scores were lower (mean 9.6 vs. 10.4) and fewer prone patients received ECMO (0.8% vs. 2.6%), inhaled nitric oxide (9.7% vs. 15.7%), vasopressors (73% vs. 83%), renal replacement therapy (11% vs. 17%), and glucocorticoids (40% vs. 45%), all of which suggest that the prone group was relatively healthier than the supine group at baseline.

While PROSEVA is an imperfect study whose results are confounded by baseline between-group differences, it adds to the growing body of evidence that prone positioning improves survival in severe ARDS. This intervention should be restricted to select patients meeting this study's inclusion criteria or perhaps the more stringent criterion of ARDS with P:F ratio <100 mmHg.


Surviving Sepsis Campaign severe sepsis and septic shock (2016, adapted)[4]

  • Recommend prone over supine positioning in adults with ARDS from sepsis and a PaO2/FIO2 ratio <150 (strong recommendation, moderate quality of evidence)


  • Multicenter, randomized, open-label trial
  • N=466
    • Supine (n=229)
    • Prone (n=237)
  • Setting: 27 European ICUs
  • Enrollment: 2008-2011
  • Follow-up: 90 days
  • Analysis: Intention-to-treat
  • Primary outcome: All-cause mortality at 28 days


Inclusion Criteria

  • Severe ARDS
    • P:F ratio <150 mmHg
    • FiO2 >60%
    • PEEP ≥5 cm H2O
    • Vt 6 mL/kg
  • ETT in place and mechanical ventilation for <36 hours

Exclusion Criteria

  • ICP >30 mmHg, CPP <60 mmHg, or MAP <65 mmHg
  • Massive hemoptysis
  • Recent tracheal surgery, sternotomy, facial trauma, DVT, PPM placement
  • Unstable fractures of spine, femur, or pelvis
  • Anterior chest tube with air leak
  • Nitric oxide, almitrine bismesylate, or ECMO therapy
  • Lung transplant, pregnancy, or >20% BSA burn
  • Chronic O2 or NIV therapy
  • Other disease resulting in life expectancy <1 year
  • End-of-life decision
  • Inclusion in other research protocol in prior 30 days with mortality as endpoint
  • Prone positioning prior to inclusion

Baseline Characteristics

From the supine group. Comparisons are supine vs. prone when given.

  • Demographics: Mean age 60 years, male 66.4%
  • ARDS from pneumonia: 58.1%
  • Sepsis: 85.2% vs. 82.2% (P<0.05)
  • Source of admission to ICU: ED 42.8%, acute care facility 38.0%, home 11.4%, ICU 3.9%, other 3.9%
  • McCabe score: A 79.9%, B 19.7%, C 0.4%
  • SAPS II: 47
  • Mean SOFA score 10.4
  • PMH: DM 17.0%, ESRD 5.2%, liver disease 7.0%, CAD 10.5%, cancer 13.1%, COPD 12.7%, immunodeficiency 16.6%
  • BMI: 29
  • Interventions:
    • Vasopressors: 83.0% vs. 72.6% (P<0.05)
    • NM blockade 82.3% vs. 91.0% (P<0.05)
    • RRT 17.1%
    • glucocorticoids 44.9%
  • Ventilator settings: Vt 381 mL (6.1 mL/kg), RR 27/min, PEEP 10 cm H2O, FiO2 79%
  • Respiratory measurements: Plateau pressure 23 cm H2O, compliance 35 mL/cm H2O
  • PaO2:FiO2 100
  • Bloodwork: pH 7.30, PaCO2 52 mmHg, PaO2 80 mmHg, HCO3 25


  • Stabilization for 12-24 hours prior to enrollment
  • Randomized to supine or prone position
    • Supine position: Semirecumbent position with physiologic measurements every 6 hours
    • Prone position: Turning to prone for ≥16h, repeated daily for ≥28 days; cessation of prone positioning if:
      • Improved oxygenation, defined as P:F ratio ≥150 mmHg with PEEP ≤10 cm H2O and FiO2 ≤60% ≥4 hours after last prone session
      • Decreased P:F ratio before 2 consecutive prone sessions of ≥20% relative to the supine ratio
      • Complications during a prone session leading to its interruption
  • Both groups were kept in standard ICU beds (ie, did not use beds specific for prone-positioning)
  • Mechanical ventilation was modeled after the low Vt ARDSNet protocol, with end-inspiratory plateau pressure goal ≤30 and goal pH 7.20-7.45


Comparisons are as supine vs. prone.

Primary Outcome

28-day mortality
32.8% vs. 16.0% (HR 0.39; 95% CI 0.25-0.63; P<0.001; NNT=6)
SOFA score-adjusted: HR 0.42; 95% CI 0.26-0.66; P<0.001

Secondary Outcomes

All-cause mortality at 90 days
41.0% vs. 23.6% (HR 0.44; 95% CI 0.29-0.67; P<0.001; NNT=6)
SOFA score-adjusted: HR 0.48; 95% CI 0.32-0.72; P<0.001
Successful extubation at 90 days
65.0% vs. 80.5% (HR 0.45; 95% CI 0.29-0.70; P<0.001)
Time to successful extubation
Survivors: 19 vs. 17 days (P=0.87)
Non-survivors: 16 vs. 18 days
ICU stay duration
Survivors: 26 vs. 24 days (P=0.05)
Non-survivors: 18 vs. 21 days
Ventilator-free days
At 28 days: 10 vs. 14 (P<0.001)
At 90 days: 43 vs. 57 (P<0.001)
5.7% vs. 6.3% (HR 0.89; 95% CI 0.39-2.02; P=0.85)
Noninvasive ventilation
At 28 days: 4.7% vs. 1.8% (HR 0.36; 95% CI 0.7-3.50; P=0.11)
At 90 days: 1.5% vs. 1.8% (HR 1.22; 95% CI 0.23-6.97; P=1.00)
At 28 days: 5.2% vs. 3.8% (HR 0.71; 95% CI 0.27-1.86; P=0.37)
At 90 days: 8.1% vs. 6.4% (HR 0.78; 95% CI 0.36-1.67; P=0.59)

Additional Analyses

2.6% vs. 0.8% (P=0.14)
Inhaled NO
15.7% vs. 9.7% (P=0.05)
Almitrine bismesylate
6.6% vs. 2.5% (P=0.04)
NM blockade
5.6 days vss. 5.7 days (P=0.74)
9.5 vs. 10.1 days (P=0.35)

Adverse Events

Cardiac arrest
31 vs. 16 events (P=0.02)


  • Unclear if centers with less experience in prone positioning would have higher rates of complications
  • Unclear if an inclusion bias present as little data from excluded patients was collected
  • Fluid balance and catacholamine dosing were not measured or reported
  • Groups imbalanced based on their SOFA score, NM blockade, and vasopressor use
  • Highly selected patients[3]
  • Significant mortality difference may have been due to a decrease in deaths due to refractory shock. There was minimal difference in deaths due to hypoxemia.


Both public and private funding.

Further Reading