Airway Pressure Release Ventilation (APRV) and Permissive Hypercapnia in the Acute Respiratory Distress Syndrome (ARDS).


This case was chosen to address the difficulties of ventilation in patients with ARDS and strategies to improve outcome.

Clinical Problem

A 49 year old man presented to hospital with severe hypoxaemic respiratory failure. He had a 10 day history of flu-like symptoms (sore throat, dry cough, myalgia). His past medical history included Ca stomach in 2001 treated with a partial gastrectomy. He had had 2 previous admissions to hospital (not to critical care) with pneumonia.
On admission to Accident and Emergency he was in respiratory distress had oxygen saturations of 66%. His blood pressure was 110/64 and heart rate 140 (sinus rhythm) and his GCS was 15. The intensive care team were promptly called.


The patient was intubated and ventilated. 100% oxygen with a PEEP of 15cm water were required to maintain adequate oxygenation. BIPAP ASB inverse ratio ventilation was instituted at a ratio of 2:1 with lung protective tidal volumes of 6-8mls/kg. Plateau pressure was kept at 30 or below. The patient was paralysed in an effort to increase chest wall compliance and reduce oxygen consumption and CO2 production. Hypercapnia was permitted. Chest X-ray revealed bilateral pulmonary infiltrates which together with his PaO2 ratio of <26.6 and absence of clinical suspicion of raised left atrial pressure indicated ARDS which was presumed to be secondary to lung infection.
H1N1 swabs were taken which later were reported as negative. Tamiflu was given in the meantime along with coamoxyclav, clarythromycin and linezolid (previous MRSA pneumonia). Blood cultures were negative.
Septic shock rapidly developed with initial high doses of noradrenaline required to maintain BP. Activated protein C was instituted 18h post admission.
Oxygen requirements initially came down to 50% after 24h but then fluctuated between 60-90% over the next 7 days. A frusemide infusion was started on day 6 to achieve a neutral cumulative fluid balance and caspofungin started due to his poor response to antibiotics and positive sputum sample for candida.
On day 8 FiO2 was 85% and paralysis had been reinstituted. BIPAP pressures were 30/12 with tidal volumes of 9-10mls/kg. Lung ultrasound demonstrated no significant effusions, no pneumothorax and consolidated lungs. A recruitment manouvre (40cm CPAP for 40s) failed to improve oxygenation or lung compliance. At this point ventilation was changed to APRV with pressures of 30/5, Time high 5s and Time low 0.35s, which resulted in an I:E ratio of 14:1 a RR of 11 and tidal volumes of 6mls/kg. Paralysis was stopped and sedation lightened.

Over the next 28 hours there was a marked improvement in oxygenation. Initially there was a worsening of hypercapnia and acidosis but both slowly improved.
Over the next few days the ventilator settings were cautiously weaned until BIPAPASB was reinstituted once the FiO2 was down to 0.3. A tracheostomy was then performed and weaning continued.


High tidal volumes, pressures and FiO2, along with cyclical opening and closing of alveoli cause lung injury (1). This is avoided by controlling volumes and pressures, using strategies to maximise oxygenation and maintaining open alveoli. Oxygenation is proportional to mean airway pressure (MAP). Using higher mean pressures and ventilating a patient at the top of the volume pressure curve will recruit alveoli (improving compliance and VQ matching) and keep them open (minimising injury). Increasing MAP without increasing peak pressure means prolonging inspiratory time. This can be achieved using APRV or high frequency oscillation. In conventional Bilevel ventilation the ventilator cycles from a low level of CPAP to a higher one resulting in flow and thus CO2 removal. In APRV this is reversed with the ventilator maintaining a high airway pressure and cycling to a lower level for a very short time (‘release’) to allow CO2 removal. The release time is so short that there is incomplete lung emptying making use of intrinsic PEEP to maintain high MAP - thus setting the lower CPAP level above zero is often unnecessary. The 2 modes are essentially the same but with very different I:E ratios and without intrinsic PEEP generation in conventional Bilevel. Conventional ventilation ventilates lungs at the inspiratory bottom of the volume pressure curve whereas APRV ventilates at the top, expiratory portion of the curve. Mean airway pressure is increased, alveoli recruited, compliance improved and oxygenation increased all while maintaining safe pressures and volumes. Because with Bilevel modes (Bilevel, BIPAP, APRV) the patient can spontaneously breathe at any point in the ventilatory cycle (with or without pressure support), comfort is maintained and sedation can be minimised. Spontaneous ventilation is desirable as haemodynamic performance is enhanced (2), juxta-diaphragmatic alveoli are recruited (3), V/Q matching is improved (2, 4) and respiratory muscle function is preserved. Reduced length of ventilatory support, length of intubation and length of ICU stay have also been demonstrated when APRV with SV was compared to PCV (3).
To set up APRV the high CPAP should be set to the current level and the low CPAP to 0-5. The T high should be set to 5s and the T low to 0.5s. The T low should then be adjusted primarily to achieve a TV of 6mls/kg and the T high primarily to achieve the desired RR (8-12). Pressure support is optional, as long as pressures do not exceed 30, but is not necessary.

Permissive Hypercapnia
Using lung protective strategies for ventilation often means accepting high levels of CO2 for minimisation of lung injury from positive pressure ventilation. Hypercapnia is generally well tolerated (in the absence of contraindications such as raised ICP and PVR) with its detrimental effects often overstated. There is good evidence that significant haemodynamic effects are minimal with pH >7.15 although more severe derangements occur below this. There are however also theoretical reasons why hypercapnia may protect against lung and systemic organ injury independently of ventilatory strategies. Animal experiments demonstrate that both lung and systemic organ injury are reduced by hypercapnic acidosis via anti-inflammatory effects, effects on free radical generation and activity and regulation of gene expression. These beneficial effects seem to be related to the acidosis itself rather than hypercapnia on its own although it may be that hypercapnic acidosis is more beneficial than metabolic acidosis. It should be noted that in isolated lung, hypercapnia with a normal pH may cause lung injury. (5)
It is interesting to note in this case with a patient with normal renal function the speed with which renal compensation for the acidosis occured. The BE doubled in 4h and tripled in 8 with normalisation of pH within a little over 24h. This of course occurs by renal regulation of the strong anion chloride, to increase strong ion difference, although the ABG machine in this unit did not measure chloride to allow illustration of this.

Lessons learnt

I now know how to set up a ventilator with APRV settings.
This case has demonstrated to me the improvement in oxygenation that results from APRV even when ARDS has been prolonged and APRV is initiated at a late stage.
The results of the OSCAR (6) trial are eagerly awaited. If favourable then APRV will likely be considered redundant - a ‘poor man’s oscillator’. If negative then there will be little justification for the expense of buying a machine. In this case APRV will have a role in refractory hypoxaemia.
The case demonstrated the speed of metabolic correction of a respiratory acidosis.


Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med 2000; 342(18):1301-1308.

Putensen AJRCCM 164:2001

Spont breathing during APRV improves lung aeration in OA induced ALI. Wrigge et al. Anaesthesiology 2003; 99:376-84

APRV and spont vs PSV. Putensen C, AJCCM 159:1999

Bench-to-bedside review: Permissive hypercapnia. Donall O’ Croinin et al. Critical Care 2005, 9:51-59 (DOI 10.1186/cc2918) OSCAR Protocol Version 7 – 19 Oct 2009