Glucocorticoids in sepsis: dissecting facts from fiction

Dr Marik's recent commentary [1] states that "..as a result of an overwhelming selection bias, only approximately 5% of eligible patients were enrolled in the CORTICUS study [2]." As its authors, we are unaware as to what constitutes this overwhelming bias and from where he has plucked a figure of 5%. CORTICUS enrolled patients with clinical evidence of infection, systemic inflammatory response syndrome, shock within 72 hours (defined by a systolic blood pressure <90 mmHg despite adequate fluid replacement OR need for vasopressors for ≥1 hour), and hypoperfusion or organ dysfunction attributable to sepsis [2]. The placebo group received a maximum norepinephrine dose of 0.4 ± 0.5 mcg/kg/minute and their 28-day mortality was 31.5%. We openly acknowledged that slow recruitment was partly related to loss of equipoise; however, the above data not only appear representative of real life practice but are comparable to other contemporary septic shock studies, for example, VASST [3]. His contention that 7 to 10 days of low-dose hydrocortisone should be considered in patients receiving norepinephrine or equivalent at doses >0.1 mcg/kg/minute within 12 hours of shock onset is not supported by any evidence base, contradicts presently accepted international recommendations [4] and portrays a far more striking example of bias.