Abstract
Background: Severe acute respiratory failure in adults causes high mortality despite improvements in ventilation techniques and other treatments (eg, steroids, prone positioning, bronchoscopy, and inhaled nitric oxide). We aimed to delineate the safety, clinical efficacy, and cost-effectiveness of extracorporeal membrane oxygenation (ECMO) compared with conventional ventilation support. Methods: In this UK-based multicentre trial, we used an independent central randomisation service to randomly assign 180 adults in a 1:1 ratio to receive continued conventional management or referral to consideration for treatment by ECMO. Eligible patients were aged 18-65 years and had severe (Murray score >3·0 or pH <7·20) but potentially reversible respiratory failure. Exclusion criteria were: high pressure (>30 cm H2O of peak inspiratory pressure) or high FiO2 (>0·8) ventilation for more than 7 days; intracranial bleeding; any other contraindication to limited heparinisation; or any contraindication to continuation of active treatment. The primary outcome was death or severe disability at 6 months after randomisation or before discharge from hospital. Primary analysis was by intention to treat. Only researchers who did the 6-month follow-up were masked to treatment assignment. Data about resource use and economic outcomes (quality-adjusted life-years) were collected. Studies of the key cost generating events were undertaken, and we did analyses of cost-utility at 6 months after randomisation and modelled lifetime cost-utility. This study is registered, number ISRCTN47279827. Findings: 766 patients were screened; 180 were enrolled and randomly allocated to consideration for treatment by ECMO (n=90 patients) or to receive conventional management (n=90). 68 (75%) patients actually received ECMO; 63% (57/90) of patients allocated to consideration for treatment by ECMO survived to 6 months without disability compared with 47% (41/87) of those allocated to conventional management (relative risk 0·69; 95% CI 0·05-0·97, p=0·03). Referral to consideration for treatment by ECMO treatment led to a gain of 0·03 quality-adjusted life-years (QALYs) at 6-month follow-up. A lifetime model predicted the cost per QALY of ECMO to be £19 252 (95% CI 7622-59 200) at a discount rate of 3·5%. Interpretation: We recommend transferring of adult patients with severe but potentially reversible respiratory failure, whose Murray score exceeds 3·0 or who have a pH of less than 7·20 on optimum conventional management, to a centre with an ECMO-based management protocol to significantly improve survival without severe disability. This strategy is also likely to be cost effective in settings with similar services to those in the UK. Funding: UK NHS Health Technology Assessment, English National Specialist Commissioning Advisory Group, Scottish Department of Health, and Welsh Department of Health.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 1351-1363 |
| Number of pages | 13 |
| Journal | The Lancet |
| Volume | 374 |
| Issue number | 9698 |
| DOIs | |
| State | Published - 2009 |
ASJC Scopus subject areas
- Medicine(all)
Access to Document
Other files and links
Fingerprint
Dive into the research topics of 'Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): a multicentre randomised controlled trial'. Together they form a unique fingerprint.Cite this
- APA
- Standard
- Harvard
- Vancouver
- Author
- BIBTEX
- RIS
Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR) : a multicentre randomised controlled trial. / Peek, Giles J.; Mugford, Miranda; Tiruvoipati, Ravindranath et al.
In: The Lancet, Vol. 374, No. 9698, 2009, p. 1351-1363.Research output: Contribution to journal › Article › peer-review
}
TY - JOUR
T1 - Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR)
T2 - a multicentre randomised controlled trial
AU - Peek, Giles J.
AU - Mugford, Miranda
AU - Tiruvoipati, Ravindranath
AU - Wilson, Andrew
AU - Allen, Elizabeth
AU - Thalanany, Mariamma M.
AU - Hibbert, Clare L.
AU - Truesdale, Ann
AU - Clemens, Felicity
AU - Cooper, Nicola
AU - Firmin, Richard K.
AU - Elbourne, Diana
N1 - Funding Information: This study shows a significant improvement in survival without severe disability at 6 months in patients transferred to a specialist centre for consideration for ECMO treatment compared with continued conventional ventilation. For patients allocated to receive conventional management, outcome at 6-month follow-up was better than predicted when planning the study. However, outcome was even better for patients allocated for consideration to receive ECMO than for those allocated to receive conventional management. Mortality alone was also lower in the consideration for ECMO group than in the conventional management group, but the study was not powered to detect this outcome and the difference did not reach statistical significance. The quality of life and spirometry results at 6 months were better than that reported in previous studies for both treatment groups. 48,49 Although this effect could have been caused by participation in a trial, the definitive cause is unknown. Several factors could account for the improved outcome for patients allocated to consideration for treatment by ECMO. First, ECMO sustains life in acute lung failure long enough for diagnosis, treatment, and recovery. ECMO rests the lungs from high pressure and FiO 2 ventilation, thereby keeping to a minimum the iatrogenic contribution to lung injury. Correspondingly, we recorded a significant increase in the proportion of lung-protective ventilation in the consideration for ECMO group compared with the conventional management group. We believe that this effect indicates the advanced nature of lung injury in patients recruited into the CESAR trial, which means that these patients could not be ventilated gently without extracorporeal gas exchange. Risks associated with ECMO were small, but the procedure is complex and labour intensive, even in a highly experienced centre such as Glenfield Hospital, and at present very sick patients must undergo transfer to the centre. Three patients died before they could be transferred and two died in transit. If cannulation at the referring hospital and mobile ECMO support could be used for such patients, survival rates might be further improved. 50,51 Second, the study used a standardised protocol for disease management in a highly experienced centre. A fifth of patients treated at Glenfield Hospital improved on that protocol without the need for ECMO, of whom 82% survived to discharge. We believe that the lung disease in these patients was slightly less severe than in the four-fifths of patients who did not respond to conventional management and received ECMO, of whom 63% survived. We used careful minimisation to ensure that control and treatment groups were identical, therefore we expect that if the same protocol was uniformly applied in patients allocated to conventional management in conventional treatment centres, similar proportions of patients would have had similar proportions of patient survival. The possibility of bias caused by treatment preference of the clinical teams was eliminated by the design of the study because the ECMO team did not treat patients in the conventional management group, except for transferring 11 patients from referral hospitals to conventional treatment centres, none of whom died during transfer. A team treating a high volume of patients with a particular illness is expected to achieve good results compared with units that might treat such severe respiratory failure only once or twice per year. However, we emphasise that only a fifth of patients treated responded to expert conventional respiratory intensive care, and the remainder needed ECMO to achieve lung rest. Since very few patients with acute respiratory distress syndrome are transferred to other units for expert conventional care in the UK, the pragmatic design of our study examines the realistic situation for such patients. With the exception of use of the molecular albumin recirculating system and steroids, use of ancillary treatments was the same in both groups, and no conclusions can be drawn regarding the effectiveness of these treatments. Steroids are used in late fibrotic acute respiratory distress syndrome; more patients in the ECMO group survived long enough to enter this late stage of disease, which could account for the increased steroid use. The molecular albumin recirculating system was used for liver dysfunction from multiorgan failure in 15 patients allocated to ECMO, and might have also contributed to their lengthened survival. Our study was limited by the absence of a standardised treatment protocol in the conventional management group, which was largely caused by the inability of participating units to reach a consensus on the constituents of best treatment. These units were not willing to participate in the study if a protocol was imposed. We considered transferring patients from both treatment groups to the ECMO unit, but participating units did not judge the ECMO unit to be competent providers of conventional management or intensive care, and were concerned about the ECMO unit's possible bias in favour of ECMO. We also rejected the possibility of transferring all control patients to one expert centre for conventional treatment since no centre had the necessary capacity. No NHS funding was available for this referral practice in the context of the trial, and participating units were unwilling to send patients to another unit when they did not perceive a treatment advantage. To secure the collaboration of participating units, we had to allow them freedom to choose a protocol for conventional management. The low-volume low-pressure ventilation strategy of the ARDSNet study 19 was recommended, but no specific treatment protocol was imposed. A randomised trial of a life-support technique in acute fatal illness is associated with unique ethical and logistical problems, especially when the endpoint includes death. ECMO, for example, is potentially expensive, and physicians and patients cannot be masked at treatment delivery. Families are told that the patient has a high probability of death from acute lung failure, and then asked to consent to a life-saving technique that is only available, but not guaranteed, if they agree to enrolment in the study. In a study in one centre, a patient on the life-support technique could possibly be next to a patient on conventional treatment. Crossover can dilute these effects but such a design can make the results of the study difficult to interpret because of the composite endpoint. For example, treatment failure in one group of a crossover study leads to change to the other treatment, and the endpoint then becomes death or ECMO, which in uninterpretable. Because of these problems, ECMO has been investigated with other designs such as matched-pair studies, adaptive design randomised controlled trials, and conventional randomised controlled trials. 7,8,16,52–54 The influential UK neonatal ECMO trial 16 compared the best available standard treatment in several experienced centres with the ECMO treatment algorithm in five specialist centres. We built on this study design in the CESAR trial, in which the families of only 33 eligible patients refused consent and treating intensivists refused to enter 28 patients. Our experience is that our design comes closest to a solution for randomised trials in patients with acute illness and high risk of death, while keeping ethical and logistical barriers to a minimum. We have shown that the additional average cost per patient of referral for ECMO is more than double the average cost of treatment with conventional management. The lifetime predicted cost-utility of about £19 000 ($31 000) per QALY is, however, well within the range regarded as cost effective by health technology assessment organisations. Furthermore, the number of patients with severe respiratory failure is small compared with other diseases, and so the effect on the health-care budget would also be small. The uncertainty around cost-effectiveness estimates underscores the wide range of predicted effects of ECMO treatment; such uncertainty is frequently the case in health-care planning, and an insurance-risk model of financing is needed. However, we have shown that referral for ECMO is likely to prove more efficient than conventional management. Our findings are relevant to other countries where ECMO is provided or being considered, although local costs, health services, practice, and distances from treatment centres might vary. The uncertainty around cost-effectiveness estimates indicates the combination of uncertainty in the trial data about patients' severity of illness, treatment outcomes, health systems, and costs. Further uncertainty is introduced in prediction of costs for the future and other settings. We found that our hospital cost estimates were sensitive to methods used to estimate costs in critical care units. National data about costs of NHS critical care were not available at the outset of our study, but are now published as tariffs for providers (NHS hospital trusts) 8 to use in contracts with third party payers. Although these costs are likely to be reliable estimates of true resource costs, the NHS financial system uses different values (not case-mix-adjusted) that predict reduced costs per outcome gained. The cost-effectiveness of ECMO would be improved if costs of both transport and provision of the technique could be reduced. These two factors might be inversely related. Provision of ECMO will probably be most clinically and economically efficient (reduced cost per successful case treated) in large critical care units, and the clinical effectiveness of small units would be lower than that of busy units. However, long-distance air travel could be kept to a minimum with a large number of well placed critical care units, which would inevitably be small and less economically efficient than large units. In our trial, almost all air transport was provided by the Royal Air Force (RAF), which was quite expensive, and unrealistic since the RAF is not a routine service provider for the NHS. Air transport costs could be reduced by use of a dedicated air ambulance for patient retrieval. We recommend further careful modelling of the most cost-effective solution for different settings. We are confident that ECMO is a clinically effective treatment for acute respiratory distress syndrome, which also promises to be cost effective in comparison with other techniques competing for health resources. Contributors GJP was lead clinical investigator. MM was lead investigator for economics input, and MMT and CLH were investigators for economics input. AW was lead investigator for patient follow-up. DE was lead investigator for statistics, and study design and management. All authors were members of the project management team. AT and GJP obtained ethical approval for the study. RT, AT, GJP, and RKF participated in recruitment of centres or patients, or both. GJP, AW, EA, AT, FC, NC, RKF, and DE participated in study design and data collection; MM, MMT, CLH participated in the study design and data collection for economics research; and RT participated in data collection for clinical research. DE, EA, MM, and CLH participated in data analysis. All authors participated in data interpretation and reporting of results. All authors have seen and approved the final version. The CESAR trial collaboration Steering Committee : R Adfield, E Allen, F Clemens, E Coates, N Cooper, K Diallo, D Edbrooke, D Elbourne, G Faulkner, J Fawcett, D Field, R Firmin, D Goldhill, B Gutteride, P Hardy, S Harris, C Hibbert, S Holden, N Jones, H Killer, M Mugford, W Nganasurian, G Peek, M Pepperman, D Piercy, S Robertson, J Scott, A Tattersfield, M Thalanany, R Tiruvoipati, K Tomlin, A Truesdale, N Webster, A Wilson. Project Management Group: E Allen, F Clemens, N Cooper, K Diallo, Y Doyle, D Edbrooke, D Elbourne, G Faulkner, R Firmin, D Francis, P Hardy, S Harris, C Harvey, C Hibbert, N Jones, H Killer, M Mugford, G Peek, D Piercy, S Robertson, M Thalanany, R Tiruvoipati, K Tomlin, A Truesdale, A Wilson. Data Coordination, London : E Allen, F Clemens, K Diallo, D Elbourne, P Hardy, D Piercy, S Robertson, K Tomlin, A Truesdale. Clinical Coordination, Glenfield : M Aslam, G Faulkner, R Firmin, S Harris, C Harvey, H Killer, N Jones, C McCulloch, G Peek, J Redfern, R Reeves, N Roberts, L Russell, A Sheward, L Smith, A Sosnowski, A Tebbat, R Tiruvoipati. Data Monitoring Committee : D Altman, R Doll (died in July, 2005), T Evans, D Macrae Follow-up Group : J Sanderson-Mann, P Sinfield, C Tarrant, H Watkinson, A Wilson. It support: A King, M Bennett. Randomisation service : G McPherson, A Walker. Independent categorization of causes of deaths: C Waldmann, D Goldhill. Participating centres For all centres that recruited patients, we have listed each hospital with the names of collaborating medical staff. The number in brackets represents the number of patients recruited by that centre. Airedale General Hospital, Airedale (2) J Scriven, K Price; Alexandra Hospital, Redditch (1) T Leach, D Bagnall, L Clements; Arrowe Park Hospital, Liverpool (1) JP Gannon, J Chambers, P Grice, C Taylor; Ayr Hospital, Ayr (6) I Taylor, M Dunlop, D Kerr; Bassetlaw District General Hospital, Worksop (1) R Harris, W Lee, P Wootton; Bedford Hospital, Bedford (10) D Niblett, F Barchard, F Bertasius; Castle Hill Hospital, Cottingham (8) S Gower, J Dickson, K Roberts; Cheltenham General Hospital, Cheltenham (2) W Doherty, A Culpepper, S Maisey; Chesterfield & North Derbyshire Royal Hospital, Chesterfield (2), RP Wroth, L Barton, D Handley; Chorley & South Ribble District General Hospital, Chorley (1) M Calleja, J Baldwin; Derby Hospitals NHS Foundation Trust, Derby (2) P Harris, K Greatorex, J Herring, L Thomas; East Surrey Hospital, Redhill (1) B Bray, B Keeling; Frimley Park Hospital, Camberley (3) L Shaikh, J Thomas; Glan Clwyd District General Hospital, Rhyl (1) B Tehan, L Burgoyne, K Owen; Glenfield Hospital, Leicester (8) R Firmin, G Peek, D Turner, L Marriot, J Morton, L Randall; Gloucestershire Royal Hospital, Gloucester (8) C Roberts, A Bailey, E Maggs; Hereford County Hospital, Hereford (1) J Hutchinson, L Davies, L Kehoe; Huddersfield Royal Infirmary, Huddersfield (2) J O'Riordan, S Ainley, S Maguire; Hull Royal Infirmary, Hull (3) I Smith, D Muir, N Smith; Kent & Sussex Hospital, Royal Tunbridge Wells (1) P Sigston, A Collins; Kettering General Hospital, Kettering (3) L Twohey, C Harland, J Thomas; King's Mill Hospital, Sutton in Ashfield (1) M Ross, M Platt, A Tinsley; Leicester General Hospital, Leicester (2) P Spiers, J Cadwallader; Leicester Royal Infirmary, Leicester (6) D Turner, K Coulson; Leighton Hospital, Leighton (3) A Martin, T Schiavone, M Smith; Llandough Hospital, Llandough (1) A Turley, C Taylor, S Bennett, R Kyte; Luton & Dunstable Hospital, Luton (10), M Patten, M Kermack; Macclesfield District General Hospital, Macclesfield (4), J Hunter, H Cooper, J Rhodes; Manchester Royal Infirmary, Manchester (1) R Slater, W Cook; Milton Keynes General Hospital, Milton Keynes (1) P Chambers, J McHugh; Newham University Hospital, London (1) S Holbeck, C McMullen, L Woodbridge; Ninewells Hospital and Medical School, Dundee (1) JR Colvin, B Soutar; North Manchester General Hospital, Manchester (1) M Longshaw, E Jones; Northern General Hospital, Sheffield (3) S Michael, J Sutherland, L Wadsworth; Nottingham City Hospital, Nottingham (2) M Levitt, C Crocker, M Hope; Pilgrim Hospital, Boston (3) M Spittal, D Connolly, I Hamilton; Prince Charles Hospital, Merthyr Tydfil (1) B Jenkins, J Davies; Prince Philip Hospital, Llanelli (4) M Esmail, L Evans; Queen Elizabeth Hospital, Gateshead (6) F McAuley, E Britton-Smith, A Jackson, V McLean; Raigmore Hospital, Aberdeen (1) CA Lee, G Calder; Rotherham District General Hospital, Rotherham (2) D Harling, D O'Malley, H Proctor; Royal Albert Edward Infirmary, Wigan (1) R Saad, J Hilton, M Taylor; Royal Bolton Hospital, Bolton (4) W Price, S Westwell; Royal Hallamshire Hospital, Sheffield (7) D Edbrooke, K Bailey, S Smith; Royal Liverpool University Hospital, Liverpool (1) G Masterson, T Rowan; Royal Preston Hospital (1) P Duncan, C Richarson; Sandwell General Hospital, Birmingham (1) J M Bellin, A Markham, M Willis; Scunthorpe General Hospital, Scunthorpe (1) T Samuel, R Sharawi, A Holmes, S Snelson; Southend Hospital, Southend (1) D Higgins, J Lee, P Tyler; Southport & Ormskirk Hospital NHS Trust, Southport (1) D Jayson, G Levens, H Rymell, M Smith, M Vangikar, J Webb; St Mary's Hospital, Portsmouth (1) C Wareham, J Bean, A Read; Staffordshire General Hospital, Stafford (1) J Hawkins, J Lewis, N Worral; Stepping Hill Hospital, Stockport (3) J Rigg, K Berry, S Swire; Horton Hospital, Banbury (2) J Everatt, G Walker, K Marchant; Ipswich Hospital, Ipswich (1) M Garfield, C Calder, M Parfitt; Royal London Hospital, London (4) D Kennedy, S Nourse, I O'Connor; University Hospital Aintree, Liverpool (1) E Shearer, P Hale, S Tabener; University Hospital of Hartlepool, Hartlepool (3) V Gupta, L Morgan; University Hospital of North Staffordshire NHS Trust, Stoke on Trent (1) B Carr, T Proctor, A Normington; University Hospital of Wales, Cardiff (1) G Findlay, M Smithie, E Hutcheon; Victoria Hospital, Blackpool (1) D Kelly, M Drummond; Walsgrave Hospital, Coventry (2) J Little, D Watson, T Mason, G McMillan; Warrington Hospital, Warrington (1) J Little, T Mason, G McMillan; Warwick Hospital, Warwick (5) J Aulakh, H Reading; Watford General Hospital, Watford (1) V Page, T Stambach, C Armstrong, W Dore; West Suffolk Hospital, Bury St Edmunds (4) J Cardy, P Oats, S Humphreys; Worcestershire Royal Hospital, Worcester (4) N Volpe; Wrexham Maelor Hospital, Wrexham (3) WC Edmondson, K Miller; Wycombe Hospital, High Wycombe (2) T Dexter, R Bryson, G Toovey. For all other hospitals supplying data, we have listed each hospital with the names of collaborating medical staff. Addenbrookes Hospital, Cambridge; Amersham Hospital, Amersham; Biggleswade Hospital, Biggleswade; Cannock Chase Hospital, Cannock; Chapel Allerton Hospital, Leeds; Coventry & Warwickshire Hospital, Coventry; Freeman Hospital, Newcastle upon Tyne; Goodmayes Hospital, Illford; Hammersmith Hospital, London; Harefield Hospital, Harefield; Hawthornes Care Centre, Peterlee; Hope Hospital, Salford; Leigh Infirmary, Wigan; Lister Hospital, Stevenage; Mile End Hospital, London; North Middlesex Hospital, London (A Chan, R Lo, GL Dabuco, N Mathew); Northwick Park Hospital, London; Papworth Hospital, Cambridge; St James's University Hospital, Leeds; St Thomas' Hospital, London; Southern General Hospital, Glasgow (M Garrioch); University Hospital, North Tees, Hartlepool (P Ritchie, F Bage, L Williams, J Tint); Wythenshawe Hospital, Manchester . Conflicts of interest GJP, RKF, and RT are clinicians who provide extracorporeal membrane oxygenation services. GJP received a travel grant to present results of the CESAR trial at the Children's National Medical Centre conference (February, 2008) from Chalice Medical. RKF received a travel grant to attend the Extracorporeal Life Support Organization meeting (September, 2008) from Chalice Medical. All other authors declare that they have no conflicts of interest. Acknowledgments We thank all the patients and their families who participated in the trial, and Elizabeth Coates for help with research for the economic analysis at different stages during the study.
PY - 2009
Y1 - 2009
N2 - Background: Severe acute respiratory failure in adults causes high mortality despite improvements in ventilation techniques and other treatments (eg, steroids, prone positioning, bronchoscopy, and inhaled nitric oxide). We aimed to delineate the safety, clinical efficacy, and cost-effectiveness of extracorporeal membrane oxygenation (ECMO) compared with conventional ventilation support. Methods: In this UK-based multicentre trial, we used an independent central randomisation service to randomly assign 180 adults in a 1:1 ratio to receive continued conventional management or referral to consideration for treatment by ECMO. Eligible patients were aged 18-65 years and had severe (Murray score >3·0 or pH <7·20) but potentially reversible respiratory failure. Exclusion criteria were: high pressure (>30 cm H2O of peak inspiratory pressure) or high FiO2 (>0·8) ventilation for more than 7 days; intracranial bleeding; any other contraindication to limited heparinisation; or any contraindication to continuation of active treatment. The primary outcome was death or severe disability at 6 months after randomisation or before discharge from hospital. Primary analysis was by intention to treat. Only researchers who did the 6-month follow-up were masked to treatment assignment. Data about resource use and economic outcomes (quality-adjusted life-years) were collected. Studies of the key cost generating events were undertaken, and we did analyses of cost-utility at 6 months after randomisation and modelled lifetime cost-utility. This study is registered, number ISRCTN47279827. Findings: 766 patients were screened; 180 were enrolled and randomly allocated to consideration for treatment by ECMO (n=90 patients) or to receive conventional management (n=90). 68 (75%) patients actually received ECMO; 63% (57/90) of patients allocated to consideration for treatment by ECMO survived to 6 months without disability compared with 47% (41/87) of those allocated to conventional management (relative risk 0·69; 95% CI 0·05-0·97, p=0·03). Referral to consideration for treatment by ECMO treatment led to a gain of 0·03 quality-adjusted life-years (QALYs) at 6-month follow-up. A lifetime model predicted the cost per QALY of ECMO to be £19 252 (95% CI 7622-59 200) at a discount rate of 3·5%. Interpretation: We recommend transferring of adult patients with severe but potentially reversible respiratory failure, whose Murray score exceeds 3·0 or who have a pH of less than 7·20 on optimum conventional management, to a centre with an ECMO-based management protocol to significantly improve survival without severe disability. This strategy is also likely to be cost effective in settings with similar services to those in the UK. Funding: UK NHS Health Technology Assessment, English National Specialist Commissioning Advisory Group, Scottish Department of Health, and Welsh Department of Health.
AB - Background: Severe acute respiratory failure in adults causes high mortality despite improvements in ventilation techniques and other treatments (eg, steroids, prone positioning, bronchoscopy, and inhaled nitric oxide). We aimed to delineate the safety, clinical efficacy, and cost-effectiveness of extracorporeal membrane oxygenation (ECMO) compared with conventional ventilation support. Methods: In this UK-based multicentre trial, we used an independent central randomisation service to randomly assign 180 adults in a 1:1 ratio to receive continued conventional management or referral to consideration for treatment by ECMO. Eligible patients were aged 18-65 years and had severe (Murray score >3·0 or pH <7·20) but potentially reversible respiratory failure. Exclusion criteria were: high pressure (>30 cm H2O of peak inspiratory pressure) or high FiO2 (>0·8) ventilation for more than 7 days; intracranial bleeding; any other contraindication to limited heparinisation; or any contraindication to continuation of active treatment. The primary outcome was death or severe disability at 6 months after randomisation or before discharge from hospital. Primary analysis was by intention to treat. Only researchers who did the 6-month follow-up were masked to treatment assignment. Data about resource use and economic outcomes (quality-adjusted life-years) were collected. Studies of the key cost generating events were undertaken, and we did analyses of cost-utility at 6 months after randomisation and modelled lifetime cost-utility. This study is registered, number ISRCTN47279827. Findings: 766 patients were screened; 180 were enrolled and randomly allocated to consideration for treatment by ECMO (n=90 patients) or to receive conventional management (n=90). 68 (75%) patients actually received ECMO; 63% (57/90) of patients allocated to consideration for treatment by ECMO survived to 6 months without disability compared with 47% (41/87) of those allocated to conventional management (relative risk 0·69; 95% CI 0·05-0·97, p=0·03). Referral to consideration for treatment by ECMO treatment led to a gain of 0·03 quality-adjusted life-years (QALYs) at 6-month follow-up. A lifetime model predicted the cost per QALY of ECMO to be £19 252 (95% CI 7622-59 200) at a discount rate of 3·5%. Interpretation: We recommend transferring of adult patients with severe but potentially reversible respiratory failure, whose Murray score exceeds 3·0 or who have a pH of less than 7·20 on optimum conventional management, to a centre with an ECMO-based management protocol to significantly improve survival without severe disability. This strategy is also likely to be cost effective in settings with similar services to those in the UK. Funding: UK NHS Health Technology Assessment, English National Specialist Commissioning Advisory Group, Scottish Department of Health, and Welsh Department of Health.
UR - http://www.scopus.com/inward/record.url?scp=70350018319&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=70350018319&partnerID=8YFLogxK
U2 - 10.1016/S0140-6736(09)61069-2
DO - 10.1016/S0140-6736(09)61069-2
M3 - Article
C2 - 19762075
AN - SCOPUS:70350018319
SN - 0140-6736
VL - 374
SP - 1351
EP - 1363
JO - The Lancet
JF - The Lancet
IS - 9698
ER -