Although there may be disputes over its efficacy, there are few people left who do not think that venovenous extracorporeal membrane oxygenation (VV-ECMO) is lifesaving to some extent in patients with acute respiratory distress syndrome (ARDS); this belief has been reflected by a sharp increase in its deployment over the last decade despite lack of clear pos. randomized controlled trials.One of the problems is that the benefits in terms of lifesaving are offset by complications of ECMO support itself.Analogously to, for example, the case of patients with hematol. malignancies, mortality is in part disease-related, but there is also significant treatment-related mortality.Exposure of blood to the nonbiol. surfaces of an extracorporeal circuit initiates a complex inflammatory response involving both the coagulation and the inflammatory response pathway.More importantly, there seems to be no relationship between the level of anticoagulation and the occurrence of a rare thromboembolic stroke; however, there is a strong relationship between the level of anticoagulation and the frequent occurrence of bleeding complications (55%) as well as the need for a blood transfusion, both of which are directly related to poor outcome.Taken together, one might postulate that anticoagulation with heparin during ECMO might lead to more problems than benefits. However, there is a paucity of studies evaluating different anticoagulation strategies inpatients supported with ECMO and no randomized trials comparing one strategy with another.The hypothesis, based on experiences with renal replacement therapy, was that the addition of PGE1couldextendthelifespanoftheECMOcircuit, but as this was a safety study, the primary outcome was the rate of transfused packed red cells per day of ECMO support.Fewer patients in the PGE1group had any membrane lung clotting (7 vs. 16;P= 0.020) and the time to first membrane change was longer in patients allocated to PGE1(hazard ratio 0.30;95% confidence interval 0.12-0.75).The secondary endpoints even suggest a reduction in thromboembolic and bleeding events with PGE1administration, but given the pharmacol. features of PGE1,inhibition of platelet aggregation and arterial vasodilatation, the mechanism by which that would work is obscure.As blood pressure was unaffected with this low dose of PGE1the mechanism of action is most probably restricted to inhibition of platelet aggregation, which also suggests that platelet aggregation plays a major role in circuit lifespan.Currently, the main purpose of anticoagulation during ECMO support seems to be the prevention of thromboembolic events other than stroke and preventing the ECMO circuit from clotting, which is also potentially fatal.Circuit exchange might be an objective measure for the study of the latter assuming that the exchange is protocolized; which is complicated, as the decision for exchange is guided by multiple parameters in clin. practice.Although examining the results of the primary endpoint, blood transfusion, yields no concern about safety, 90-day survival was lower, although statistically nonsignificant, in the treatment group; accordingly, mortality should be included in the (combined)endpoint of a definitive phase-III trial.In the end, the authors need to be complimented on having performed one of the first double-blind, placebo-controlled medication trials during ECMO support; we welcome their results.We hope this study paves the way for further medication trials to find, in addition to the optimal anticoagulant, drugs such as the optimal inotrope or vasopressor for administration during ECMO support.