Are Steroids Useful During Septic Shock?

This question was recently discussed during rounds regarding when we were faced with an ICU patient who had refractory hypoglycemia and hypotension.

A brief literature review revealed:

-          

-          Bone, et al 1987: N= 382. Methylprednisolone 30 mg/kg vs placebo. No mortality difference.

-          Jurney, et al 1987:  Non-responders (adrenal insufficiency reflected on cortisol/stim tests) benefit, otherwise no benefit.

-          Cronin, et al 1995: Review of literature. Suggestive of harm.

-          Annane, et al 2002: N = 300. Hydrocortisone (50 mg IV q6) + fludrocortisones (50 mcg/day) x 7 days vs placebo. Non-responders benefit (shorter length of pressors, better resusc).

-          CORTICUS, 2008: Hydrocortisone 50 mg IV q6 vs placebo. No benefit other than more rapid reversal of shock (though no mortality diff); significant adverse events.

-          COITSS, 2010: hydrocortisone + high dose insulin vs hydrocortisone + conventional insulin + fludrocortisone.  No benefit with either regimen.

-          Beale, 2010: Significant regional differences in use of steroids. Highest in Europe and South America. Lowest in Asia. ~15% of all patients in study received steroids and showed higher mortality rates.

-          Casserly, et al 2012: Patients requiring pressors randomized to steroids or none. Mortality higher in steroid group.

-          Surviving Sepsis (Feb 2013): If adequate response to fluid and vasopressors, no steroids. If shock refractory to these interventions, hydrocortisone 50 mg IV q6. No ACTH stim test to help identify responders/non-responsders. No use in absence of shock (eg severe sepsis).

-         In conclusion:  Conflicting findings, but studies variable in design. Most likely, steroids are most appropriate for high-risk patients with hemodynamic resuscitation refractory to aggressive fluid and vasopressor intervention. No need for predictive stim tests; monitor for adverse reactions.

from Andrew Pirotte, MD