Résumé
Background: Patients in cardiogenic shock (CS) can benefit from temporary mechanical circulatory support (tMCS) but data in heterogeneous populations with diverse CS aetiologies are rare.Aims: To compare baseline characteristics, management and independent correlates of 30-day and 1-year mortalities between patients managed with and without tMCS for CS.Methods: The FRENSHOCK registry (NCT02703038) included 772 unselected patients with CS admitted in 49 French critical care units between April and October 2016.Results: Among 770 patients with CS and available data, 135 (17.5%) received tMCS (63 extracorporeal life support, 35 intra-aortic balloon pump, 13 micro-axial flow pump and 24 combination), of whom 69.0% were implanted during the first 24hours. Patients with tMCS were less likely to present with histories of cardiac or peripheral artery diseases or chronic renal failure. An ischaemic or mechanical complication trigger for CS tripled the probability of receiving tMCS, while increased lactate doubled the probability. Thirty-day mortality was higher among patients with versus without tMCS (34.1% vs. 24.3%; P<0.001), but after adjustment for age and CS severity (LVEF and arterial lactates at admission, noradrenaline and invasive mechanical ventilation use), the difference was no longer significant (hazard ratio: 1.37, 95% confidence interval: 0.93-2.01). Among 30-day survivors, mortality was similar regardless of initial receipt of tMCS (P=0.312).Conclusion: In real-life practice, tMCS may not be associated with improvements in 30-day or 1-year mortality in an all-comers cohort of patients with CS. However, it should be noted that there were substantial differences in patient characteristics and management between patients who received tMCS and those who did not. Additional targeted studies should help to determine more precise algorithms for granting assistance within the complex and heterogeneous population of patients with CS.