Efficacy of VV ECMO in Severe ARDS: A Meta-Analysis of the CESAR and EOLIA Trials

Introduction: Venovenous ECMO is a potentially life-saving therapy in individuals suffer-
ing from severe ARDS. However, clinical trial data has been conflicting. This
paper synthesizes data from two landmark trials to see if there is a significant
improvement in mortality among ARDS patients who receive VV-ECMO.


Methods: Data from the CESAR and EOLIA trials were extracted with the assistance of
the Cochrane data extraction and assessment form. Mathematical calculations
performed in the generation of this meta-analysis were done with the assistance
of R. A random effects meta-analysis was performed using the risk ratio as a summary measure. A random-effects meta-analysis was preferentially selected over fixed-effects due to the small differences in trial design, such as ECMO initiation criteria and crossover that occurred. Risk ratio was selected over odds ratio due to the relative ease of clinical interpretation, higher event rates and the fact that both trials reported risk-based outcomes. The statistical heterogeneity was calculated using T², I² and the Q test.
ChatGPT-4o was used in the initial generation of the R code used in the formulation of the forest plot and in the calculation of the pooled effect estimates and helped troubleshoot coding issues. However, final adjustments to the code were performed by the authors.


Results: Between the two studies, a total of 429 patients were evaluated. 214 of these
patients were on ECMO and 215 patients were controls who received conventional management. The pooled risk ratio for mortality was 0.763 (95 percent CI, 0.594
to 0.981; p=0.0345). This value indicates that there was a 24 percent reduction
in all-cause mortality in patients who received ECMO treatment. Using the 95
percent confidence interval, we can determine a true mortality benefit which can
range between 1.9 percent and 40.6 percent. Finally, after calculating T², I² and
Q, it was determined that there was no significant heterogeneity between the two
studies, (T² = 0, I² = 0 percent, Q = 0.12, p = 0.73).


Conclusions: Using the data from the EOLIA and CESAR trials, we can see that there is a
significant reduction in all-cause mortality in patients with severe ARDS who
receive venovenous ECMO. This supports the use of this potentially life-saving
therapy in the appropriate patient population.