Effect of Low-Dose Supplementation of Arginine Vasopressin on Need for Blood Product Transfusions in Patients with Trauma and Hemorrhagic Shock: A Randomized Clinical Trial (AVERT Shock)

Sims et al, JAMA Surgery, Published Online August 28, 2019, doi:10.1001/jamasurg.2019.2884

Clinical Question

  • In trauma patients with hemorrhagic shock, does the use of low-dose arginine vasopressin compared to placebo reduce the volume of blood products transfused?


  • Traumatic hemorrhagic shock is a leading cause of death among adults under 45 years
  • Traumatic hemorrhagic shock is associated with neuroendocrine dysfunction
  • Arginine Vasopressin (AVP) deficiency underlies this neuroendocrine dysfunction. AVP deficiency in trauma occurs due to depletion of AVP stores, loss of AVP in haemorrhage, and resuscitation with AVP-deficient blood products
  • Replacement of AVP can help with:
    • Vasomotor tone
    • Haemostasis from improvement in platelet function
    • Vasoconstriction of injured vessels and shunting blood to vital organs
  • Retrospective human studies have found increased mortality with AVP use in this population


  • Randomised controlled trial
  • Computerized block randomization
  • Blinding of clinical team, research personnel, patients & families
  • Both intention-to-treat and per-protocol analyses performed
  • Per-protocol excluded patients dying early in operating room, deemed as “nonsurvivable” injuries
  • Registered on clinicaltrials.gov
  • 100 patients required to give 80% power and a 2 sided alpha of 0.05 to detect a 50% reduction in total volume of blood products from a baseline mean use of 5.4L of blood products


  • Single urban level 1 trauma centre, USA
  • Data collected 2013 – 2017


  • Inclusion: Adult trauma patients receiving more than 6 units of any blood product within 12 hours of admission
  • Exclusion: interhospital transfers, prehospital CPR, emergency department thoracotomy, recent corticosteroid use, chronic renal insufficiency, significant coronary artery disease, traumatic brain injury requiring neurosurgical intervention, pregnancy, incarcerated patients
  • 257 patients identified, with 157 excluded predominantly for insufficient blood transfusion (< 6units), or ED thoracotomy being performed
  • 101 randomized, with 1 post-randomization refusal
  • Comparing baseline characteristics of intervention group (n=49 ) vs. placebo (n=51) no significant differences for:
    • Patient demographics (age, sex, ethnicity) – only 7 women in total
    • Injury and mechanism characteristics including admission vital signs, site of bleeding, enrolment vital signs, ISS Score
    • Penetrating trauma: 80% vs. 78%
    • Injury Severity Score: 19 (14-26) vs. 26 (17-34)
    • Initial lab values (lactate, creatinine, hemoglobin, coagulation panel)
    • Pre-enrolment resuscitation (crystalloids, blood components, tranexamic acid, EBL)


  • Arginine Vasopressin (AVP)
    • 4U bolus of AVP following randomization
    • 0.04U/min infusion (0.4U/mL concentration) started after bolus
    • Titrated 0-0.04U/min following surgical control of bleeding to maintain MAP >= 65 for 48 hours


  • Placebo
    • Bolus of normal saline in equivalent volume as AVP
    • Infusion of saline placebo of equivalent volume

Management common to both groups

  • 1:1:1 transfusion strategy in both groups was suggested (though not protocolized
  • Crystalloid and blood products at discretion of treating physician
  • Additional vasopressors at discretion of physician (phenylephrine, norepinephrine, epinephrine)
  • These were titrated and stopped before tapering study infusion
  • Study coordinator monitored all resuscitations to ensure compliance in protocol


  • Primary outcome: AVP vs. placebo
    • Cumulative volume of blood products transfused within 48hrs (including red blood cells, fresh frozen plasma & platelets) – significantly reduced in AVP group
      • 1.7 (0.7-3.1) vs 3.0 (1.4-5.2)
      • Median Difference: -1.00L (-2.03 to 0.00L, P = 0.03)
    •  For each component of primary outcome
      • No significant difference in:
        • PRBC, median (IQR) in L
          • 0.9 (0.1-1.8) vs 1.0 (0.6-3.0) P = 0.08
      • In AVP group significantly reduced:
        • FFP, median (IQR) in L
          • 0.9 (0.8-1.3) vs 1.0 (0.5-1.8) P = 0.03
        • Platelets, median (IQR) in mL
          • 200 (0-300) vs 300 (0-600) P = 0.02
        • Cryoprecipitate, mean (SD) in mL
          • 12.6 (75.4) vs 34.7 (84.8) P = 0.04
  • Secondary outcome:
    • No difference in any of the secondary outcomes studied:
      • Volume of crystalloid transfused in 48hr
        • 9.6 vs. 10l
      • Fluid balance in 48hr
        • 6l vs. 7l
      • Total vasopressor requirements for 48hr
      • 30 day mortality
        • 12% vs. 12%
      • Hospital/ICU length of stay
      • Organ dysfunction (AKI, ARDS, open abdomen free days, ventilator free days)
    • No difference in adverse events except:
      • DVT – significantly reduced in AVP group
        • 20% vs. 39%, P=0.05
        • RR (95% CI): 0.33 (0.13-0.83) p = 0.02

Authors’ Conclusions

  • Early administration of AVP during the resuscitation of trauma patients with hemorrhagic shock leads to decreased use of blood products at 48hours without increased complications


  • Randomized control trial with appropriate blinding
  • Balanced pragmatic design vs. protocolized design allow modest generalizability
  • Intention-to-treat and per-protocol analysis allowing for standardization of patients with non-survivable injuries
  • Physiologic basis for the findings are well supported
  • Registered on clinicaltrials.gov


  • Single center trial limits generalizability, open to institutional practice bias
  • Study population predominantly young males with penetrating trauma, limiting generalizability
  • Not adequately powered for secondary endpoints, or safety markers
  • Sample size calculation based on baseline volume of blood products of 5.4l where as actual baseline was only 3l

The Bottom Line

  • Trauma patients presenting with hemorrhagic shock may benefit from low-dose vasopressin supplementation to prevent blood transfusion, however larger multicentre trials are needed to determine efficacy & safety of this practice. The findings of this study do not suggest changes in practice at this time

External Links


Summary author: Vatsal Trivedi

Summary date: 01/09/2019

Peer-review editor: David Slessor

One comment

  • George Theodoropoulos

    Does the result of this study sound meaningful? How can a vasoconstrictor agent reduce the volume of transfusion? Is it only the enhancement of exocytosis of vWf from vasopressin that made the difference?

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.