Corticosteroids for treating sepsis: A systematic review and network meta-analysis

Corticosteroids have been applied in sepsis treatment for decades, yet their clinical efficacy-particularly regarding optimal dosage, administration regimens, and impact on long-term prognosis-remains controversial and incompletely defined. This systematic review and network meta-analysis aimed to comprehensively evaluate corticosteroids' effect on improving sepsis patients' outcomes and clarify the comparative effectiveness of different corticosteroid dosages and regimens.

Four electronic databases (PubMed, Embase, Web of Science, Cochrane Library) were searched from inception to March 2023 to identify randomized controlled trials (RCTs) of corticosteroids in adult sepsis patients. R software, Stata 15.0, and RevMan 5.4 analyzed data. Primary outcomes (≤30-day short-term mortality, ≥90-day long-term mortality) and secondary outcomes (ICU length of stay [LOS], mechanical ventilator-free days, vasopressor-free days, renal replacement therapy-free days) were synthesized. Surface under the cumulative ranking curve (SUCRA) ranked efficacy; funnel plots assessed publication bias. The study was prospectively registered on PROSPERO (CRD 42023454288).

A total of 18 eligible RCTs involving 7591 patients were included. Corticosteroids reduced ≤30-day mortality (hydrocortisone [H] 100 mg/d: odds ratio [RR] = 0.22, 95% CI = 0.072-0.62), this finding is based on one small-scale RCT and requires confirmatory studies) and ≥90-day mortality (H 100 mg/d: RR = 0.33; H 200 mg/d + fludrocortisone 50 μg/d: RR = 0.79). Specific regimens improved secondary outcomes: H 200 mg/d shortened ICU LOS (mean difference [MD] = -2.6), H 200 mg/d + fludrocortisone increased ventilator/vasopressor-free days, and H 300 mg/d prolonged renal replacement therapy-free days.

Notably, the mortality benefit of H 100 mg/d is preliminary due to limited evidence from a single small RCT. Corticosteroids reduce sepsis patients' short/long-term mortality and organ support duration, but no "optimal" dosage consensus exists. Further research is needed to refine dosage and personalize therapy.