Efficacy and Safety of Combination Therapy of Shenfu Injection and Postresuscitation Bundle in Patients With Return of Spontaneous Circulation After In-Hospital Cardiac Arrest: A Randomized, Assessor-Blinded, Controlled Trial

Zhang. Critical Care Medicine 2017; DOI:10.1097/CCM.0000000000002570

Clinical Question

  • In patients with in-hospital cardiac arrest, does Shenfu (a traditional Chinese
    medicine), improve 28 day mortality?


  • Traditional Chinese Medicine has been used for over 1000 years. Shenfu is a type of Chinese Medicine that contains ginseng and aconite. In patients with septic shock Shenfu has been reported to increase mean arterial pressure, reduce lactate and mortality. Studies in animals have reported positive effects of Shenfu on scavenging free radicals, inhibiting inflammatory mediators and suppressing cell apoptosis. With these benefits reported, the authors investigated whether Shenfu could have a potential benefit following in-hospital cardiac arrest.


  • Randomised controlled trial
  • Computer generated block randomisation
  • Stratified by centre
  • Treating physicians non-blinded
  • Blinding of assessors
  • Consecutive patients
  • Modified intention to treat analysis: excluded patients whose consent was
    absent/withdrawn, and patients who were placed under legal guardianship
  • Sample size calculation: 290 patients would provide 80% power to detect a reduction in mortality from 70% to 50% with a false positive rate of 5%. Considering a 20% drop-out rate and high mortality following return of spontaneous circulation (ROSC), planned to recruit 500 patients


  • 50 hospitals in China
  • Data collected: 2012 – 2015


  • Inclusion criteria: Adult patients with sustained return of spontaneous circulation following an in-hospital cardiac arrest
  • Exclusion criteria: Consent absent/withdrawn, placed under legal guardianship
  • 1022 patients assessed for eligibility of whom 978 included in modified intention to treat analysis
  • Comparing baseline characteristics of Shenfu vs. control group – no significant
    differences in all 37 baseline characteristics that were assessed

    • Age: 65.1 vs. 64.6
    • History of coronary artery disease: 32.1% vs. 31.3%
    • Cause of cardiac arrest
      • Cardiogenic shock: 27.8% vs. 27.9%
      • Acute coronary syndrome: 20.9% vs. 21.4%
      • Life threatening lethal arrhythmia: 17.5% vs. 17.1%
    • Cause of hospital admission
      • Acute cardiovascular disease: 29.9% vs. 29.8%
      • Acute respiratory disease: 24.2% vs. 24.1%
    • Location of cardiac arrest
      • Ward: 47% vs. 46%
    • Initial cardiac rhythm
      • VF: 10.4% vs. 10.1%
      • PEA 7.9% vs. 7.6%
      • Asytole: 81.7% vs. 82.3%
    • Duration of advanced life support, median (IQR): 13min (6-20) vs. 19 (9-30), p=0.27
    • Comatose at ICU admission: 96.3% vs. 96.7%
    • Ventilated: 99.2% vs. 99%
    • Therapeutic hypothermia: 18.9% vs. 17.9%
    • Angiography or percutaneous coronary intervention: 31.1% vs. 30.5%


  • Shenfu injection
    • 100ml twice daily for 14 days or until hospital discharge


  • Standard care

Management common to both groups

  • Post-resuscitation care bundle that included targeted temperature management (TTM), airway and ventilation management, haemodynamic management, early coronary angiography, and comprehensive critical care
  • Prognostication after 72 hours: Clinical +/- SSEP or EEG +/- biomarkers


  • Primary outcome: 28 days survival – significantly higher in Shenfu group
    • 42.7% vs. 30.1%, p=0.02
    • Absolute risk reduction = 12.6% (95% CI 6.67% to 18.62%)
    • NNT = 8
    • Fragility index = 33
  • Secondary outcomes – comparing Shenfu vs. control group
    • The Shenfu group had significantly greater rates of
      • Survival at 90 days
        • 39.6% vs. 25.9%, p=0.001
      • Alive at discharge with favourable neurological outcome
        • 29.1% vs. 17.1%, p=0.03
      • Cerebral Performance Category 1 or 2 in survivors to ICU discharge
        • 70% vs. 59.3%, p=0.03
    • Multiple organ failure – significantly higher in control group
      • 44.3% vs. 53.%, p=0.01
    • No significant differences in:
      • Renal failure
        • 22.1% vs. 20.2%, p=0.21
      • Cardiogenic shock
        • 10.2% vs. 11.1%, p=0.91
      • Cause of death
        • Irreversible brain death: 64.8% vs. 68%
        • Cardiac: 19.1% vs. 22.1%
        • Multiple organ failure: 10.9% vs. 7%
    • For longterm survivors, duration of mechanical ventilation – significantly shorter in Shenfu group
      • 8.6 vs. 12.7 days, p<0.001
    • Adverse events:
      • 2 patients in Shenfu group had pruritus that was not felt due to study drug

Authors’ Conclusions

  • Shenfu improves clinical outcomes including mortality following an in-hospital
    cardiac arrest.


  • Randomised controlled trial
  • Blinding of outcome assessors
  • Multi-centre


  • Physicians non-blinded, the authors state that this is because Shenfu is yellow. However it should have still been possible to make a yellow placebo.
  • The study was only carried out in China where traditional Chinese Medicine is more commonly accepted.
  • The sample size calculation planned recruitment 500 patients. It is unclear why so many more patients were recruited and why study recruitment was stopped when it was. It was not reported what the study outcomes were after the planned 500 patients had been recruited
  • Post cardiac arrest bundle included TTM for all patients. However, this treatment was only performed on a small proportion of patients. It is unclear why some patients received this treatments where as others did not.
  • We are not told how many patients had their prognostication based on clinical/SSEP/EEG/biomarkers
  • There was a longer duration of advanced life support in the control vs. the Shenfu group 19min (9-30) vs. 13 min (6-20). The author’s report that this difference was not statistically significant. The raw data is not provided that would allow us to check the accuracy of this calculation.

The Bottom Line

  • This large multi-centre randomised controlled trial demonstrated a significant reduction in mortality with the use of Shenfu following in-hospital cardiac arrest, with a number needed to treat of 8
  • This compares to a NNT of 6 for the original HACA therapeutic hypothermia trial
  • The sceptic in me tells me that if something appears too good to be true then it probably is
  • My concerns regarding the lack of blinding, the differences in baseline characteristics, the variation in compliance with the post-resuscitation bundle and the fact that the study was performed in a single country means that I want to see more evidence before I start using Shenfu in my practice.

External Links


Summary author: David Slessor
Summary date: 21 August 2017
Peer-review editor: Duncan Chambler

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.