Pre-ECMO congestion as an independent predictor of mortality in cardiogenic shock supported with venoarterial extracorporeal membrane oxygenation

Venoarterial extracorporeal membrane oxygenation (VA-ECMO) is used in patients with severe cardiogenic shock (CS) to reverse peripheral hypoperfusion and associated multiorgan failure. Congestive status among candidate patients to VA-ECMO support and its clinical impact during and after the therapy is not well known.

To assess pre-cannulation congestion prevalence and its impact on in-hospital mortality in patients undergoing VA-ECMO for refractory CS.

This was a retrospective study, between January 2021 and August 2025, in which we consecutively enrolled 65 patients requiring VA-ECMO for medical CS and refractory cardiac arrest (ECPR). Pre-ECMO congestion was diagnosed based on radiographic evidence of pulmonary edema, elevated post capillary wedge pressure (PCWP) >18 mmHg or clinical signs of volume overload.

The mean age was 54.7 ± 14.4 years, with 73.8 % (n=48) male participants. VA-ECMO was initiated due to cardiogenic shock complicating acute myocardial infarction in 61.5% (n=40). Left ventricle unloading was used in 47.7% (n=31), 74.2% (n=23) being IABP and 25.8% (n=8) being microaxial flow pump device. Overall mortality was 52.3% (n=34). Pre-ECMO congestion diagnosis was found in 36.9% (n=24) patients. Main demographic and clinical characteristics are listed in Table 1. Patients with Pre-ECMO congestion required higher circulatory support to reach adequate organic perfusion, with increased needs of noradrenaline (mcg/kg/min) during the first 24 hours (0.6 ± 0.5 vs 0.3 ± 0.3; 0.014) and VA-ECMO support flow (l/minute) (3.5 ± 0.7 vs 3.0 ± 0.8; p = 0.025). In the univariate analysis, Pre-ECMO congestion was associated with acute pulmonary edema at any moment of the support (45.8% vs 14.6 %; p = 0.006), need for continuous renal replacement therapy (29.2% vs 7.3%; p = 0.019) and overall mortality (70.1% vs 41.6%; p = 0.048). In the multivariate analysis, after adjustment for age, sex, etiology of shock, and ECPR, Pre-ECMO congestion was independently associated with mortality (OR 6.7; [95% CI 1.52–29.24]; p = 0.012), together with ECPR (OR 11.23; [95% CI 2.12–59.35]; p = 0.004). (Figure 1)

Pre-ECMO congestion not only was found in a significant number of patients but also identified a high-risk phenotype of cardiogenic shock when requiring VA-ECMO, as it showed an independent association to overall in-hospital mortality. Early recognition and active decongestive strategies prior to cannulation may improve outcomes in this critically ill population.Table 1  Figure 1