Journal Club - Recent Additions

February - 2013


Showing Journal 1 of 15

High-Frequency Oscillation in Early Acute Respiratory Distress Syndrome

the OSCILLATE Trial Investigators and the Canadian Critical Care Trials Group N Engl J Med, 2013, 368:795-805


This multicenter RCT conducted in 39 ICU's in 5 countries planned to randomise 1200 patients with ARDS to early HFOV (3-12Hz, press amp 90cm H2O) vs control ventilation (Vt 6ml/kg, high titrated PEEP, Pplat

The study was stopped at 548 patients due to lack of efficacy/possible harm. The...


Previous Comments

Its fantastic to see these so called "novel ventilation strategies" finally being put to the test in RCTs. There are so many modes being introduced with little but bench data to support them, being trumpeted by ventilator manufacturers as the next big thing, with little sign of clinical outcome data to confirm it. If you assume for a moment that there is some physiological benefit associated with the ventilation strategy, this presumably means that the increased sedation and paralysis required has negated this benefit, and in fact worsened their outcomes. It would now be nice to see equally speculative modes such as NAVA and APRV put to the test
Gordon West-27 Jan, 2013 10:12:11 AM

most of the time it has been used in the ICU as rescue therapy not as a primary therapy. it is same as somebody is using vasopressor for severe sepsis patient.
muhammad faisal khan-29 Jan, 2013 09:14:09 PM

I would like to give analogy in our routine clinical practice. We never thrombolyse Non ST elevation Myocardial Infarction, nor we thrombolyse all CECT proven Pulmonary embolisms. Thrombolysis is reserved for transmural or ST elevation MI and in case of PE, its reserved for patients with shock and hypoxia or Echo evidence of RV dysfunction. Similarly HFOV should be reservred for Refractory Hypoxemia as Rescue Therapy and should not be MISUSED in early ARDS or Moderate ARDS when P/F ratio is < 200 as it was tested in the OSCILLATE trial Dr Sameer Jog
sameer-05 Mar, 2013 10:52:41 PM

They are fair comments, and it is interesting to delve into this. Is it reasonable to say that based on the evidence to date HFOV is not beneficial and may be harmful in moderate ARDS, but its role as a rescue therapy remains undefined?
Neilo-07 Mar, 2013 05:39:43 PM

Well, the trial seems well conducted, albeit with difficulties blinding. That aside, the results surely stand up to scruitiny, and that leaves us in the situation where there is little evidence to support its use, at least not widespread. I've seen plenty of commentary about the trials not including enough "sick" patients - so, yes, as you say, perhaps we still don't know in that group
Gemma-08 Mar, 2013 07:27:49 PM

Not sick enough? In the two trials, mortality in the four groups ranged between 35 and 47%. I think thats a pretty good definition of sick...
Gordon West-15 Mar, 2013 08:25:02 AM

I, like many others, was brought up believing that the gold standard for the evaluation of interventions was the randomised controlled trial, preferably double-blinded. When the target population is easily identified (eg acute myocardial infarction with ST elevation), lead-time bias minimal, and the intervention is 'clean' (eg single dose of a physician-administered intravenous drug) the noise-to-signal ratio is very low, and very small treatment effects can be detected by recruiting large populations of patients. Does this apply to trials like OSCILLATE and OSCAR? Is high frequency oscillation a treatment and, if so, how does it interrupt the pathophysiology of the underlying condition? Would you expect renal failure secondary to glomerulonephritis to respond to renal replacement therapy? First, improvements in ventilatory support can only be expected to reduce the mortality that is attributable to mechanical ventilation. With this in mind, in what proportion of the patients recruited to these studies was the attributable mortality due to mechanical ventilation alone greater than 10%? Now look at the size of the treatment effect that was sought. Do you think that makes sense? Second, HFOV is a very 'complex intervention'; a bit like a long and complicated recipe. We all think we know what a 'Sunday roast' looks like, but is your mum's 'Sunday roast' anything like my mum's? I think not. Sure, the studies had a recipe (protocol), but (a) the 'recipes' were hugely different, and (b) in which patients was the 'recipe' followed, and how well was it followed? I no longer believe that large RCTs with modest data collection is the appropriate format to investigate complex interventions in complex patients entangled in a web of confounding effects.
Iain Mac-03 Jun, 2013 12:37:08 AM