The objective of the present research is a combination of a one-sided test of non-inferiority and a one-sided test of superiority. A stepped approach will be used to evaluate these hypotheses:
1. a less invasive intervention (i.e., no indwelling arterial catheter insertion until felt absolutely needed, according to consensual and predefined safety criteria) is non inferior to usual care (i.e., systematic indwelling arterial catheter insertion in the early hours of shock) in terms of mortality at day 28 (non-inferiority margin of 5%).
2. a less invasive intervention is not only non-inferior but also superior to usual care in terms of mortality.
Multi-centre, pragmatic, randomised, controlled, open, two-parallel group, non-inferiority clinical trial.