Steroids in Septic Shock- Debate continues

Fang Fang, Yu Zhang, Jingjing Tang et el published a meta-analysis of use of corticosteroids in septic shock in Dec, 2018 in JAMA internal medicine.

This ignites the age old debate of use of steroids in septic shock.  You may recall that B. Venkatesh, S. Finfer, J. Cohen, D. Rajbhandari et el published adrenal trial which showed that mortality at 90 days was similar in patients treated with steroids compared to placebo. However, there was quick resolution of shock in steroid group.  This was discussed here.  Another trial,  Activated Protein C and Corticosteroids for Human Septic Shock (APROCCHSS) , hydrocortisone plus fludrocortisone given at low doses reduced 90-day mortality among patients with septic shock.


MEDLINE, Embase,the Cochrane Central Register of Controlled Trials were searched electronically from inception until August 10, 2018. Total of 37 studies were included after eliminating studies which were biased, low quality and not having full results. Total of 9564 patients.


Steroid treatment was significantly associated with reduced 28-day mortality, ICU mortality, and in-hospital mortality among patients with sepsis. However, this survival benefit was not replicated with 90-day mortality.

Use of corticosteroids in sepsis was associated with a significant increase in shock reversal and vasopressor-free days to day 28 and with a marked decrease in ICU length of stay, SOFA score at 7 days, and time to resolution of shock.However, corticosteroid treatment was not associated with shorter hospital length of stay or fewer ventilation-free days to day 28.

No association between significant adverse effects and corticosteroid treatment when comparing rates of gastroduodenal bleeding, superinfection, or any severe adverse event. Corticosteroid administrationwas
associated with an increased risk of hypernatremia and hyperglycemia.


Rygård SL, Butler E, Granholm A et el published another meta-analysis off about 22 randomized controlled trials, involving 7297 patients.  This meta-analysis showed that short- and longer-term mortality are unaffected, adverse events increase, but duration of shock, mechanical ventilation and ICU stay were reduced.


Potential reasons for different outcomes in these two meta–analysis can be following

  1. This meta-analysis included All patients, low versus high-dose steroids.
  2. Few studies which did not report 28 day mortality or not included in this meta-analysis.
  3. The two meta-analyses used different statistical analysis, fixed-effects model, vs random-effects model


So far all  studies and meta-analyses have shown different outcomes in terms of mortality whether 28 days or 90 days.  None of the studies showed worse mortality with use of steroids.  Almost all studies showed quick resolution of shock, less vasopressor use, less ICU stay.  Furthermore, there were no significant adverse outcomes associated with steroid use.

Mortality reduction is a hard point, and very difficult to achieve.  We need further studies to ascertain that fact.  Until then, can use steroids in septic shock, especially when vasopressors dose requirement is going up .


Pantoprazole reduces GI bleeding in intubated patients without improving mortality.Read more