Clinical problem

Triple reassorted A[H1N1] influenza1 resulted in >300 deaths in the UK in 2010 (Health Protection Agency), with ∼70% of patients dying belonging to the age group of 16–64 years. The demographics of the patients suggest that those admitted to an intensive care unit (ICU) during the current outbreak are younger than patients with seasonal flu.2 Severity of illness and ICU mortality in A[H1N1] is associated with an increasing age, higher organ dysfunction (Sequential Organ Failure Assessment (SOFA) score), chronic organ dysfunction and immunocompromised states. Even though pregnancy was associated with higher critical care admission rates, as a cohort it was associated with a lower risk of death.1 2

Ventilatory management of refractory hypoxia

Histopathological data have shown similarities between acute respiratory distress syndrome (ARDS) and A[H1N1]-associated lung injury.3 Both cause diffuse alveolar damage, with varying degrees of inflammatory infiltrates and pulmonary haemorrhage. Severe acute lung injury causes a reduction in the static compliance and ventilation–perfusion mismatch. Subjecting the lung to excessive volume or pressure during mechanical ventilation (MV) results in an additional alveolar epithelial cell injury-induced inflammatory process that may worsen both lung compliance and gas exchange. Hence, prevention of ventilator-induced lung injury (VILI) due to volutrauma and barotrauma is at the heart of LPV strategy.

The basis of LPV strategy is to use a low tidal volume (V_T; 4–8 ml/kg of ideal body weight), plateau pressure ≤30 cm H₂O, inverse ventilation ratio (I:E) and reduction of oxygen toxicity, and to titrate PEEP to Fio₂ to maintain a Pao₂ >8 kPa. In the case of A[H1N1] this strategy would be tolerated by this young cohort of patients. However, resultant hypercapnoea could lead to increased intracranial pressure, myocardial depression, pulmonary hypertension and reduction in renal perfusion.4 An evidence-based LPV strategy has been shown to reduce in-hospital and 28 day mortality; currently we allow permissive hypercapnoea to a pH >7.2 and permissive hypoxaemia to Pao₂ >8 kPa. Our management algorithm is summarised in figure 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Our algorithm for management of severe refractory respiratory failure. FiO₂, fraction of inspired oxygen; V_T, Tidal volume; I:E, Inspiratory:Expiratory ratio; PEEP, Positive end-expiratory pressure; CVVHDF, Continuous Veno Venous Hemodiafiltration; P_Plat, Plateau pressure; ECMO, extracorporeal membrane oxygenation; CXR, chest x-ray; CT, computed tomography.

Adjunctive ventilator therapies

Prone ventilation (PV) can be used in ARDS with refractory hypoxaemia to improve oxygenation. However, improvement in mortality with PV is seen only in a subset of patients with severe refractory hypoxaemia. This would be difficult in your patient because of high BMI. High frequency oscillatory ventilation (HFOV) is an alternative technique to conventional ventilation in severe hypoxaemic respiratory failure, with little benefit in maintenance of Paco₂. HFOV uses small V_Ts (usually less than the anatomical dead space) and is delivered at high frequencies (3–15 Hz). A recent meta-analysis5 has shown a reduction in in-hospital and 30-day mortality along with a reduced risk of ‘treatment failure’ when compared with conventional ventilation in ARDS. A UK multicentre study into HFOV in ARDS is currently underway (OSCAR Trials No. ISRCTN10416500).

ECLS: extracorporeal membrane oxygenation (ECMO) and extracorporeal CO₂ removal (ECCO₂R)

ECLS was first used in 1972, and in an ICU setting it encompasses both an ECCO₂R device (iLA by Novalung, Talheim, Germany) and veno-arterial (VA) or veno-venous (VV) ECMO. ECLS enables the use of a lower V_T, at aptly named ‘rest settings’ to reduce VILI and any ventilator-associated haemodynamic compromise.6 The Glenfield Hospital, Leicester criteria for ECMO are: reversible cause of respiratory failure, <7 days high pressure or high Fio₂ ventilation, Murray score >3 (weighted score of: CXR quadrants, Pao₂/FiO₂ ratio on Fio₂ 1.0, PEEP and lung compliance), severe hypercarbic pH <7.20 with no contraindication(s) for anticoagulation. ECCO₂R requires arterial and venous cannulation of femoral vessels, and an adequate cardiovascular reserve to drive the pumpless arterio-venous circuit.

While the use of ECMO in neonates is an established practice,6 it is not so in adults. Previous trials in adults with ECMO have failed to show any benefit, but were done during the era prior to ‘rest setting’ LPV. A trial of conventional ventilatory support versus ECMO for severe adult respiratory failure7 (CESAR trial, single-centre study) suggested an improvement in mortality (one extra survivor for every six patients treated) without disability at 6 months. This trial has been criticised7 for heterogeneity of ventilatory management in the conventional ventilation arm, and was a pragmatic trial concluded without adequate pretrial efficacy data.

Out of the 14 patients with A[H1N1] who required MV at our hospital this winter, only one patient was referred for ECMO, but did not receive it, similar to a published series from North America.2 We had no mortality from A[H1N1] in patients treated on our ICUs. However, the UK experience with A[H1N1] over the last two winters suggests that ECLS could have a role in managing severe refractory respiratory failure. We feel that as yet there is no convincing evidence to routinely advocate ECLS in failed conventional ventilation While there are calls for its routine implementation for patients with ‘failing’ conventional ventilation, strict adherence to LPV strategies, together with meticulous control of fluid balance (if required early renal replacement therapy) and inotropes to support the circulation and minimise VILI, may avoid the need for ECLS. Further randomised trials are still needed to identify the optimal place of ECLS within the current best practice management of refractory hypoxaemia.2