Loading...
In two recent large randomized, controlled trials, researchers evaluated corticosteroids for managing patients with septic shock. The results were seemingly disparate — one trial showed lower sepsis-related mortality (APROCCHSS; NEJM JW Gen Med Apr 15 2018 and N Engl J Med 2018; 378:809), and one did not (ADRENAL; NEJM JW Gen Med Mar 1 2018 and N Engl J Med 2018; 378:797). With that background, participants in the BMJ Rapid Recommendations project — who respond to new evidence by providing timely practice guidelines in many fields — used an updated meta-analysis (with 42 randomized trials and >10,000 patients) to construct a clinical practice guideline on use of systemic corticosteroids for patients with septic shock.
Although a two–percentage point relative reduction in 30-day mortality with steroids compared with no steroids was not significant (27% and 29%; P=0.15), a similar difference in mortality at 60 days to 1 year barely reached significance because of appropriate statistical considerations (35% vs. 37%, P=0.05; number needed to treat [NNT], ≈50). Rapid reversal of shock (within 7 days) occurred significantly more often in the steroid group (73% vs. 63%; P<0.001; NNT, 10). However, mild-to-moderate adverse events also occurred more frequently: Hyperglycemia (33% vs. 29%; number needed to harm [NNH], 25), hypernatremia (7.7% vs. 4.0%; NNH, 27), and neuromuscular weakness (6.8% vs. 5.7%; NNH, 91) were all significantly more common with steroid treatment.
Rochwerg B et al. Corticosteroids in sepsis: An updated systematic review and meta-analysis. Crit Care Med 2018 Sep 1; 46:1411. (https://doi.org/10.1097/CCM.0000000000003262)
Lamontagne F et al. Corticosteroid therapy for sepsis: A clinical practice guideline. BMJ 2018 Aug 10; 362:k3284. (https://doi.org/10.1136/bmj.k3284)
Comment
This meta-analysis and guideline present “low certainty” evidence for steroid benefits on mortality in the setting of septic shock; however, because no large trials of steroids for treating sepsis are ongoing, this might be the best evidence we will get. Higher-risk patients — e.g., those with refractory shock who are unresponsive to fluids and vasopressors — were the most likely to benefit from steroid therapy. Although the incremental benefit of mineralocorticoid administration (added to glucocorticoid) could not be discerned in this analysis, their use in the positive APROCCHSS study reasonably could lead clinicians to add them when prescribing glucocorticoids. The downsides of steroid therapy seem to be relatively low, and the potential survival benefit would be relevant for a large population of patients with this clinical syndrome (NEJM JW Gen Med May 15 2018).