DOI: 10.1097/CCM.0b013e3181701513

,  

,  

PMID: 18448917

Issn Print: 0090-3493

Publication Date: 2008/05/01


Print

Trust one who has gone through it*

Jürgen Graf; Alexander Torossian; Uwe Janssens

    loading  Checking for direct PDF access through Ovid

Excerpt

Mechanical cardiopulmonary support and extracorporeal membrane oxygenation (ECMO) for cardiac or pulmonary assist are achievements of the 20th century. Stimulated by a patient with lethal pulmonary embolism, John Gibbon was among the first to propose the concept of mechanical cardiac assist for acute circulatory failure (1). In the 1950s the so called Gibbon-type pump oxygenator commenced the era of open heart surgery in humans (2). In the 1970s ECMO has been applied for respiratory support in both adults and infants with acute respiratory failure (3). Both randomized controlled clinical trials of ECMO for the treatment of acute respiratory distress syndrome (ARDS) completed to date failed to demonstrate a survival benefit in the intervention group (4, 5). Based on evidence ECMO for respiratory support is currently only recommended in infants (6) or in controlled clinical trials (7). Whether ECMO is effective in patients with primarily circulatory failure has never been thoroughly investigated. A systematic review of 84 articles of 1,494 patients with cardiogenic shock or cardiac arrest reported a median survival to hospital discharge of about 40% (8). This review mainly consisted of case series and the funnel plots for both survival of patients with cardiogenic shock and survival of patients with cardiac arrest, showed smaller effects (i.e., fewer patients who left the hospital alive) in larger trials, and vice versa. Whether this is indicative for publication bias, attributable to the heterogeneity of the included trials with regard to the applied systems for extracorporeal support, or simply a consequence of limited methodologic quality of the trials remains speculative. A recent retrospective single center analysis of 91 patients which was not included in the aforementioned systematic review revealed a hospital survival of 44% in patients who underwent ECMO support for circulatory failure (9).
In this issue of Critical Care Medicine Combes et al. present a retrospective, single center analysis of 81 patients supported by ECMO for refractory cardiogenic shock between 2003 and 2006 (10). Thirty-four patients survived to hospital discharge (42%) and 28 patients (35%) were still alive a median of 11 months thereafter. At that time quality-of-life assessment revealed significant physical limitations when compared to an age- and gender-matched French control, whereas mental health domains were judged satisfactorily. Nevertheless, patients’ quality of life was considered superior in most dimensions when compared to patients with congestive heart failure, renal failure, or 1-yr survivors of ARDS suffering similar morbidity and mortality. Notably, a regression analysis revealed a strong association of cardiopulmonary resuscitation, renal failure, and liver dysfunction before ECMO initiation with adverse outcome, i.e., patient death. In addition, female gender was associated with adverse outcome (10).
The effectiveness of ECMO to reduce mortality in patients with circulatory failure still awaits confirmation. Although Nichol et al. concluded their systematic review suggesting that ECMO may well be an efficacious intervention, an adequately powered experimental study was deemed necessary (8). This inevitably means a randomized controlled clinical trial would be required.
Based on recent survival data, the hospital mortality in patients with acute circulatory failure with and without ECMO is close to 60% (8–10). Anticipating a decrease in hospital mortality from 60% to 50%, i.e., an absolute risk reduction of 10%, with a statistical power of 80%, more than 300 patients per group would be required. A more conservative estimation may expect a risk reduction of only 5%, i.e., more than 1,000 patients would have to be enrolled per group, and a risk reduction of merely 2% would lead to more than 6,600 patients in each study arm.
    loading  Loading Related Articles