In the June 2017 issue of the ASAIO Journal, Truby et al. described a retrospective, monocenter cohort of 121 patients who underwent veno-arterial extracorporeal membrane oxygenation (VA-ECMO) implantation because of refractory cardiogenic shock (1). Authors stratified patients in three distinct groups according to a new proposed definition of left ventricular distention (LVD) severity based on clinical, hemodynamic and radiologic criteria: (I) clinical LVD (LVD++) if "mechanical intervention to decompress the LV was initiated immediately after VA-ECMO because of pulmonary edema, ventricular arrhythmia, or significant stagnation of blood within the LV"; (II) subclinical LVD (LVD+) if "evidence of pulmonary edema on chest radiograph and pulmonary artery diastolic pressure greater than 25 mmHg within the first 2 hours of VA-ECMO support"; and (III) no LVD (LVD−) in the absence of the above criteria.Significant baseline differences existed between groups: 33% of LVD++ patients had post myocardial infarction refractory cardiogenic shock while 55% of LVD+ had postcardiotomy shock. More importantly, cardiopulmonary resuscitation was performed in 67% of LVD++ compared to 11% and 19% of LVD+ and LVD−, respectively. Furthermore, all LVD++ patients had an LV ejection fraction lower than 30% compared to 50% in the LVD+ group and 40% in the LVD−. After a mean duration of VA-ECMO support of 4.11±2.98 days, the overall survival to discharge was 43% and did not differ between groups. However, only 20% of LVD++ survivors achieved myocardial recovery defined as absence of longterm mechanical support. In comparison, 60% of LVD+ and 20% of LVD− achieved myocardial recovery. One major finding was that survival to discharge was 44.4% (4/9) in those patients who benefited from early decompression, all of which were in the LVD++, versus only 10% (1/10) in those of the LVD+ group who underwent a late decompression. The only preoperative factor that was significantly predictive of the need of decompression in the multivariate model was extracorporeal cardiopulmonary resuscitation (odds ratio: 3.64, confidence interval: 1.21-10.98; P=0.022).Although the results are interesting, authors advocate significant limitations that may have impacted the interpretation of study findings. Only 105 out of 226 VA-ECMO runs were included in the analysis, the other ones being excluded because of insufficient hemodynamic data.