Written by Seth Walsh-Blackmore
Spoon Feed
This Bayesian network meta-analysis of RCTs indirectly compares hydrocortisone + fludrocortisone to hydrocortisone alone or placebo in septic shock and finds superiority with the combination regimen.
Forget vitamins and remember mineralocorticoids
The 2021 Surviving Sepsis Guidelines expanded the recommendation for corticosteroids in septic shock. RCTs have examined hydrocortisone or hydrocortisone+fludrocortisone regimens but rarely head-to-head.
A Bayesian network meta-analysis can compare two interventions indirectly when sufficient data meet appropriate assumptions. Bayesian statistics report a credible interval (CrI) instead of the usual confidence interval, which functions similarly to determine significance at a given level. A prediction interval is reported as well, which represents the probable range of future observation.
A search for RCTs comparing hydrocortisone or hydrocortisone-fludrocortisone vs control in patients 16 years or older with septic shock (sepsis diagnosis + vasopressor) yielded 17 RCTs, totaling 7,688 randomized patients for the analysis. Only one trial (509 patients) compared the steroid regimens directly.
The headline result is a 15% reduction in relative risk (RR) of all-cause mortality at the last follow-up (median 30 days) when fludrocortisone+hydrocortisone was compared to placebo. Hydrocortisone alone did not demonstrate a significant benefit. Hydrocortisone vs hydrocortisone+fludrocortisone favors a combination regimen. Though the 95% CrI crossed 1.0 (table below), the probability of superiority for the combination was 94.2%. Safety outcomes were not significantly different between regimens.
The authors were diligent in testing and adjusting for necessary statistical assumptions. The overall bias assessment was low; however, heterogeneity among trial protocols represents a potential source. Without individual patient data, no subgroup analysis could be performed to identify a cohort driving the effect.
How will this change my practice?
A mortality reduction in a highly prevalent disease is clinically relevant even when the effect size is small. A mortality benefit with fludrocortisone was also observed in a large target trial emulation last year. Given low risk and mounting evidence of benefit, adding fludrocortisone in refractory septic shock seems more than reasonable.
Another Spoonful
In writing this article I had to learn some new statistical terms and found posts from towards data science contrasting prediction and credible intervals to confidence intervals a helpful primer.
Editor’s note: If you are interested in using a combination in practice, the APROCCHSS trial used 50mcg fludrocortisone via NG tube daily + hydrocortisone 50mg IV Q6h. ~Clay Smith
Source
Effectiveness of Fludrocortisone Plus Hydrocortisone Versus Hydrocortisone Alone in Septic Shock: A Systematic Review and Network Meta-Analysis of Randomized Controlled Trials. Am J Respir Crit Care Med. Published online January 25, 2024. doi:10.1164/rccm.202310-1785OC
