An excellent summary:
Should we still be trying to modify the dysregulated immune response in sepsis? An interesting evolutionary perspective:
Although not improving (or worsening) mortality, if steroids reduce duration of shock and length of ICU stay then they are probably worth a go, given that the absolute possible increase in risk of noteworthy adverse outcomes is very low, although this must be borne in mind. Could still argue it either way though. When we can genotype patients we may be able to identify those specific individuals in whom steroid may do significant benefit or harm. Or maybe not!