OSCILLATE: High-Frequency Oscillation in Early Acute Respiratory Distress Syndrome

Ferguson et al for the OSCILLATE Trial Investigators. N Engl J Med 2013; 368:795-805

Clinical Question

• Does High Frequency Oscillatory Ventilation (HFOV) improve mortality in patients with new onset moderate-severe ARDS compared with standard ventilation?

Design

• Randomised controlled trial
• Block randomisation, stratified by centre

Setting

• 38 centres in 5 countries (Canada, USA, Saudi Arabia, Chile and India)
• July 2009 – August 2012

Population

• Intubated patients who had an onset of pulmonary symptoms within the previous 2 weeks
• PaO2:FiO2 <200mmHg (26.6 kPa), FiO2 ≥0.5 and bilateral air-space opacities on Chest x-ray
• Exclusions: hypoxaemia primarily related to left atrial hypertension; suspected vasculitic pulmonary haemorrhage; neuromuscular disorders; severe chronic respiratory disease; pre-exisiting conditions with an expected 6 month mortality exceeding 50%; risk of intracranial hypertension; MV < 48 hours; < 16 or >85 years of age; weight <35kg; if met eligibility criteria > 72 hours or those whose physicians declined to enroll them
• 1149 eligible patients of whom 571 enrolled
  • the trial was terminated early based on the 12% higher mortality in the HFOV group and concerns about harm

After enrollment, standardised ventilator settingswere used for all patients; PC mode, tidal volume of 6mls/kg, FiO2 of 0.6 with PEEP 10cmH2O or higher if needed for oxygenation. After 30 minutes, if PaO2:FiO2 ratio remained at 200mmHg or less, patients underwent randomisation. Otherwise, the standardised ventilator setttings were maintained and the patients were reassessed at least once daily for up to 72 hours. If recruited, both intervention and control groups had a recruitment maneuever first.

Intervention

• HFOV ventilation (SensorMedics 3100B) until airway pressure <20cm H2O, or at discretion of physician if <24cm H2O for >12 hours

Control

• Conventional ventilation, pressure control mode, target 6ml/kg, plateau pressure <35cm H2O and high PEEP (started with 20cm H2O)
• FiO2 and PEEP adjusted according to protocol
• Volume-assist control modes and pressure support modes with same limits for tidal volumes and airway pressures permitted
• Once PEEP < 10cmH2O and FiO2 0.4 or less, there were no limits on tidal volume or airway pressures
• Weaning protocol included daily trials of spontaneous breathing

For both groups: if requiring FiO2 of 0.9 or greater, clinicians could use other therapies including pronin or inhaled NO. Prescriptions for fluids, sedatives and use of neuromuscular blockade was at the physicians discretion

Outcome

• Primary Outcome: In-Hospital Mortality
  • Terminated early as mortality 47% in HFOV vs. 35% in control group. RR 1.33 (95%C.I. 1.09-1.64, P=0.005)
  • Sub-group analysis showed no interaction of mortality with baseline severity or centre experience
• Secondary Outcome
  • Refractory hypoxaemia 7% in HFOV vs. 14% in conventional, P=0.007

• Other Outcomes
  • HFOV group received more neuromuscular blockers (83% vs. 68%, P<0.0001), vasoactive drugs (91% bs. 84% P=0.001) and for a longer period (5 days vs. 3 days P=0.01) and more sedation with midazoloam (199mg per day vs. 141mg per day P<0.001)

Authors’ Conclusions

• In adults with moderate-severe ARDS, the early application of HFOV as compared to standard ventilation does not reduce mortality and may be harmful.

Strengths

• Randomised
• Multi-centre
• Appropriate power calculation
• Conventional ventilation group had strict protocol and unlike OSCAR, adherence to this was mandated rather than suggested

Weaknesses

• Small number of patients
• Prior to this study HFOV was used predominately as a rescue treatment. This study had wide inclusion criteria and therefore cannot give a definitive answer for the best strategy to utilise in patients for whom we need to consider rescue therapy
• Interesting exclusion criteria ‘those whose physicians declined to enroll them’. This accounted for 23% of the patients that did not meet eligibility criteria