Evidence in Intensive Care Medicine - Respiratory

Systematic review and meta-analyses in blue.
Other high impact trials in red.


Lung protective ventilation strategy for the acute respiratory distress syndrome.
Petrucci et al. Cochrane database 2013, Issue 2
Update of previous review in 2007. No new studies eligible.
28 day and hospital mortality significantly reduced by lung protective ventilation. Mortality reduced by 26%.
Overall mortality not different if plateau pressure <31 in control group.

Lung protective mechanical ventilation and two year survival in patients with acute lung injury: prospective cohort study.
Needham et al. BMJ 2012; 344:
Studied 2 year survival rates in 485 consecutively ventilated patients with ALI/ARDS 2004-2007.
600 ventilator settings data measured twice daily median 8 ventilator settings per patient.
Only 41% of settings adhered to recommended settings (VT<6.5mls/kg, Pplat ,30.
37% never received lung protective vent.
Average VT showed a linear correlation with 2 year survival.
Even a small increase in average value associated with decreased mortality.
For each 1 ml/kg predicted body weight increase in VT there was an 18% increased risk of death during the subsequent 2 years.


Randomized, placebo- controlled clinical trial of an aerosolized beta-2 agonist for treatment of acute lung injury.
Matthay et al. for The National Heart, Lung, and Blood Institute ARDS Clinical Trials Network. Am J Respir Crit Care Med 2011
Nebulised albuterol given to ALI patients.
Trial stopped early due to futility. Signal towards increase LOV and mortality with albuterol.

Intravenous B-2 agonist in ARDS (BALTI-2)
Lancet 2012;379:229-235
IV salbutamol increases mortality in ARDS patients - trial stopped early.
BALTI-1 showed reduced EVLW and Pplat with IV salbutamol in ARDS

Note - Death from respiratory failure in ARDS is uncommon - it is usually MOF that kills.

AJRCCM 2011;183:59-66
Cohort study showing ARDS significantly declined over 8 years.
TVs were reduced from 10 to 6.8mls/kg over this time.
May have been some or many other factors (but clear from other evidence that this is the main reason).
In 1963 low TVs were noted to cause atelectasis and shunt (no PEEP was used) prompting recommendations of TVs of 10-15mls/kg which became common practice which persisted into the 21st century. In 2009 >¼ of H1N1 patients referred for ECHMO in Australia were receiving TVs >10mls/kg IBW. You could infer from the misguided application of high tidal volumes to improve oxygenation that measures like recruitment manoeuvres and going up to very high mean airway pressures in HFOV (>30) in order to improve oxygenation may actually be harmful overall - I am convinced they are for the latter which may stop any benefit of HFOV being shown in the OSCAR trial which allows MAPs of up to 50.

Higher vs lower PEEP in ALI and ARDS
JAMA 2010;303(9):865-873
All studies (ALVEOLI, LOVS, EXPRESS) used TV of approx 6mls/kg.
V. good data.
In the 1
st 2 trials PEEP titrated to oxygen and in EXPRESS titrated to plateau pressure.
Overall no mortality difference (ALI and ARDS)
Mortality reduction of 10% for ARDS in higher PEEP group.
Non significant increase in mortality for non ARDS group with higher PEEP.

Results support the use of higher PEEP in severe cases (ARDS) and not in less severe (ALI).

Recruitment manoeuvres
Cochrane review 2009
Non significant reduction in mortality.
No difference in barotrauma.
Transient improvement in oxygenation.
All studies used ventilatory strategies that included RM only as part of the package ie not studied alone.
CT evidence that individual response to RMs is varied – some benefit and others get over distended.
Need to do a study where RMs compared to no RMs with the same protocol for ot
her ventilatory parameters.

Optimal duration of a sustained inflation recruitment maneuver in ARDS patients.
Arnal JM et al (2011) Intensive Care Med 37:1588–1594
Most lung recruitment occurs in the 1st 10s of a recruitment manouever (40 cm PEEP)

Inhaled nitric oxide for ARDS and ALI - Systematic review and meta-analysis
Afshari et al Anaesth Analg 2011
No benefit except for transient improved oxygenation.
Harmful renal effects.

Inhaled nitric oxide
Cochrane review 2010
Improves V/Q matching and reduces PA pressure off-loading the RV.
Inhaled NO is a cytotoxic free radical and its metabolites can accumulate causing tissue damage.
Transient improvement in oxygenation.
Increased risk renal injury.
No outcome benefit.

Aerosolised prostacyclin
Cochrane review 2010
Anti-inflamatory properties (inhibit PLT and neutrophil aggregation). Micro circulation improved.
Pulmonary vasodilation with lowered PA pressures and offloading of RV.
Only 1 trial of 14 patients
Transient improvement in oxygenation.
No outcome benefit.

Pharmalogical therapies in ARDS
Cochrane review 2004
No mortality difference in PGE2, NAC, high dose steroids, surfactant or pentoxifyline.
Increased adverse events in PGE2, high dose steroids and surfactant.
Mortality reduction in low dose steroids (see later)

Low dose steroids in ARDS
Crit Care Med 2009 37 no. 5
All studies showed trend towards benefit which became significant when results pooled.
Mortality benefit, reduced LOV, LOS, MODS, LIS, improved oxygenation.
No increase in adverse events
No difference between timing, formulation or tapering of treatment.

Crit Care Med 2010; 38(6):1398-1404
No benefit
2 trial were old and so not applicable.
Best evidence is just to look at CESAR trial.


CESAR - Lancet 2009
On an intention to treat basis no significant difference in mortality (only 75% referred for ECMO actually received it).
Significantly less death or disability at 6 months (primary endpoint) in ECMO arm.
LOS x2 in ECMO group.
Costs were x2 in ECMO group.
ECMO group received more steroids and 17% had MARS.
High rates of liver dysfunction and intracranial haemorrhage in ECMO group.
Conventional arm did not have a standardised protocol.
93% of those in ECMO centre treated with lung protective ventilation compared to 70% elsewhere.
ECMO survival rate (63%) compares with ARDSnet lung protective trial of 60%.
So, safety demonstrated but at significant cost.

Mechanical ventilation guided by oesophageal pressure in ALI
NEJM 2008;359:2095-2104
Distending pressure of the lung (transpulmonary pressure) is the alveolar pressure minus pleural pressure.
Pleural pressure can be estimated from oesophageal pressure measurement. Alveolar pressure from plateau pressure.
Trial maintained transpulmonary pressure at 0-10cmH2O.
High pleural pressures would mean the potential for underinflation so keeping transpulmonary pressures positive (but not too much) can prevent collapse but limit overdistension.
It effectively identifies patients who will benefit from higher PEEP.
Results were improved oxygenation and compliance and a trend towards improved 28 day mortality.


High-frequency oscillation in early acute respiratory distress syndrome.

Ferguson ND, Cook DJ, Guyatt GH et al. N Engl J Med 2013; 368:795-805.
Moderate to severe ARDS. Early application HFOV. Control group lung protective vent (20 PEEP).
Trial stopped early due to increased death in HFOV group RR 1.09 - 1.64 p = 0.005.

High-frequency oscillation for acute respiratory distress syndrome.
Young D, Lamb S, Shah S et al. N Engl J Med 2013; 368: 806-813
No mortality difference.

>1300 patients in the 2 trials.
More haemodynamic instability in HFOV arms.
HFOV does not improve and may increase mortality in moderate to severe ARDS.


Prone positioning in ARDS
Taccone. JAMA 2009;302:1977-1984

No mortality benefit in moderate or severe ARDS
More complications in prone group
Improved oxygenation in prone

Prior to this study 3 systematic reviews and a meta-analysis showed no mortality benefit but with improved oxygenation and less VAP


One thousand bedside percutaneous tracheostomies in the surgical intensive care unit: time to change the gold standard.
et al. J Am Coll Surg 2011;212:163–170.

Bedside percutaneous tracheostomy is safe even in high risk patients (PEEP >10, FiO2 >50%, cervical spine injury, cervical collar).
They used minimally invasive technique (dissected down until tracheal rings felt) and used rhino dilator and confirmed guide wire placement and trachy placement with bronch.
Of the 14 complications, half required urgent surgical intervention.

Early trachy (4 days) vs prolonged intubation after cardiac surgery
Ann Intern Med 2011;154:373-383
No difference in duration of MV, LOS, mortality or long term outcomes.
Less sedation and agitation and earlier mobilisation.
Note that data showing harm of long term intubation from 70s and 80s is not relevant now we have better tubes, cuffs and nursing care. No clear data to show prolonged intubation beyond 14 days is harmful.
Need to go on individual basis.

Timing of tracheostomy in critically ill patients - systematic review and meta-analysis
Wang et al Chest 2011
No benefit in early trachy.

Pleural effusion

Pleural effusion drainage
Critical Care 2011, 15:R46
Common – up to 60% in critical care (probably more than this if use US).
Reduce compliance and increase V/Q mismatch.
In SV drainage only leads to minor improvements in lung mechanics and oxygenation but significantly improves SOB.
Presence of effusions associated with longer duration MV.
No difference in duration of MV, LOS or mortality.
Improves oxygenation.
Low risk of complications.
Alters diagnosis and management.
Volume requiring drainage remains unclear.
Use of US did not reduce risk PTX
– it is operator dependent.

Intraplural use of t-PA and DNase in pleural infection
Rahman et al NEJM 2011:365
Combination therapy (but not alone) reduced need for surgery and LOS.
Should be noted that pleuroscopy is minimally invasive and safe.
Note that streptokinase has been shown not to work.


Steroids for pneumonia
Cochrane review 2011
No mortality difference.
No significant evidence of harm.
Some evidence of improved oxygenation, resolution and less relapse.
Evidence not good quality. Need decent RCT.
Should not be part of routine practice.

Steroids to prevent post extubation stridor
Cochrane review 2009
Of benefit in high risk adults (identified by cuff leak test) if given in multiple doses 12-24h prior to extubation. Dex or hydrocort used.

Steroids to prevent extubation failure
ICM 2009;35:977-986
Reduce rate of reintubation.
Need to be given 12-24h prior to extubation.
20mg methylpred used in largest RCT.
Similar meta-analysis in 2009 Crit Care suggested the benefit only for high risk patients.
Relies on correctly identifying those patients who will benefit 12-24 h prior to extubation.

Weaning – protocol or not.
Cochrane review 2010 and 2011
75% wean easily; 25% are slow.
Increased morbidity in slow group.
Marked atrophy of diaphragm occurs within 3 days (Levine 2008) suggesting early SV is beneficial.
Physicians tend to underestimate probability of successful weaning.
Weaning protocols use collective knowledge, maximise consistency and objectivity.
Based on 3 components – readiness to wean, reducing support and extubation criteria.
Reduces MV by 25% (20h), weaning duration by 78% and ICU LOS by 10%. Greatest benefit in surgical ICUs.
No difference in mortality.
Heterogenicity in trials but supports notion that proactively and consistently addressing weaning reduces length of MV.

NIV for weaning
Cochrane review 2010
Significant mortality reduction
Reduced VAP (but easier diagnosis in ETT due to access for secretions).
Reduced LOS
Reduced LOV
No difference in re-intubation.
Reduced tracheostomy.

Benefit greatest if COPD.
Weaning methods and the timing of changing to NIV yet to be determined due to heterogenicity between trials.
Studies were small so these results better for hypothesis generation than changing practice.
But makes sense to consider NIV for weaning especially for COPD.

Heated humidification vs HME
Cochrane review 2010
Humidification reduces iatrogenic injury – inflammation, necrosis, keratinisation, airway collapse, deterioration of lung function and mucous plugging.
HH cause better humidification but can cause bacterial contamination.
HMEs increase dead space and WOB especially in paeds.
No clear difference in adults between the 2.
HMEs should be the standard of care based on their non-inferiority and reduced cost.
Use of HH for individual patients should have a clear rationale eg tenacious secretions.

NIV in acute cardiogenic pulmonary oedema
NEJM 2008; 359: 142-151 Gray et al
No mortality or intubation difference between standard therapy and NIV (CPAP and BiPAP).
Improved dyspnoea, acidosis and reduced heart rate.
Meta-analysis from 2006 Crit Care;10:R69 showed NIV reduced mortality and need for intubation.

Steroids for preventing reintubation
BMJ 2008;337:1088-1091
Reduce laryngeal oedema and need for reintubation.
Need multiple dosing regimen (for 12h pre extubation). 1 dose doesn’t work.
Only use for patients at high risk.

Airway complications
4th NAP part 2. BJA 2011;106:632-642
Airway complications more common out of hours with less experienced doctors.
Patient factors contributory in most suggesting can be anticipated.
Most common factor obesity - ½ of all complications.
nd most common factor was less experienced staff.
Routine use capnography recommended.