Efficacy and safety of Shenfu injection on sepsis-induced myocardial dysfunction: a meta-analysis and trial sequential analysis

The efficacy and safety of Shenfu Injection (SFI) as an adjunctive treatment for sepsis-induced myocardial dysfunction (SIMD) remain uncertain despite preliminary investigations. Eight databases were searched for qualified randomized controlled trials (RCTs) from inception to 10 July 2025, with no language restrictions. The primary outcomes included 28-day mortality, SIMD incidence and left ventricular ejection fraction (LVEF). The secondary outcomes included length of hospital stay (LOHS), ICU length of stay, duration of mechanical ventilation (DMV), N-terminal pro-B-type natriuretic peptide (NT-proBNP), cardiac troponin I (cTnI) and adverse events (AEs). The meta-analyses were conducted by R (version 4.5.0) and Stata software (version 17.0). Trial sequential analysis (TSA) was performed by TSA program. Finally, the Grading of Recommendation, Assessment, Development, and Evaluation system (GRADE) was applied for evaluating the quality of evidence. This meta-analysis of 10 RCTs ( n = 691) found that adjunctive SFI was associated with a lower incidence of SIMD (OR = 0.44, 95% CI: 0.26–0.74, p = 0.002), lower NT-proBNP levels (SMD = -0.91, 95% CI: -1.78 to -0.05, p = 0.039), lower cTnI levels (SMD = -0.87, 95% CI: -1.25 to -0.50, p < 0.001), and higher LVEF (SMD = 0.92, 95% CI: 0.17–1.67, p = 0.016) compared with CWT alone. However, no significant differences were observed in 28-day mortality (OR = 0.72, 95% CI: 0.42–1.24, p = 0.234), LOHS (SMD = -0.81, 95% CI: -2.36 to 0.74, p = 0.306), ICU length of stay (SMD = -0.13, 95% CI: -0.40 to 0.13, p = 0.323), DMV (SMD = 0.21, 95% CI: -0.09 to 0.52, p = 0.174), or AEs (OR = 1.18, 95% CI: 0.38–3.72, p = 0.773). Evidence certainty was moderate for ICU length of stay and DMV, low for 28-day mortality, SIMD incidence, and AEs, and very low for LVEF, LOHS, NT-proBNP, and cTnI. TSA showed that the required information size was achieved for SIMD incidence and cTnI. For all other outcomes, the required information size was not reached and/or the TSA monitoring boundaries were not crossed. Adjunctive SFI may improve cardiac function and reduce the incidence of SIMD; however, these physiological improvements did not translate into reduction in 28-day mortality or length of hospital stay. TSA showed that the required information size was not reached for several major outcomes, and the certainty of the evidence ranged from very low to moderate. Therefore, the current evidence is insufficient to support the routine use of SFI in clinical practice. Larger, adequately powered, multicenter, multinational RCTs are warranted to confirm these findings.