MMU banner 1

JC: AVP in Haemorrhagic Shock. St Emlyn’s

Blood product transfusion can be lifesaving for patients who have suffered major trauma, but the associated side-effects and risks, mean that most people would agree that less is more. This paper by Sim et al examines whether giving arginine vasopressin (AVP) can reduce blood product requirements in patients receiving massive transfusions following trauma. 

Effect of Low-Dose Supplementation of Arginine Vasopressin on Need for Blood Product Transfusions in Patients With Trauma and Hemorrhagic ShockA Randomized Clinical Trial

JAMA Surg. 2019;154(11):994-1003. doi:10.1001/jamasurg.2019.2884

Who was involved?

This was a single-centre study and involved adult trauma patients who had required more than six units of blood products (packed red blood cells, platelets or fresh frozen plasma). There were quite a number of exclusions including criteria such as age > 65 years, recent corticosteroid use, chronic renal disease, significant coronary artery disease and pregnancy.  

What was the intervention?

After receiving the sixth blood product, patients were randomised to receive either AVP (4u bolus followed by infusion at 0.04 U/min) or the equivalent volumes of saline administered from packs that appeared identical. The infusion was continued for 48 hours with the intention of maintaining the patient’s MAP at 65 mmHg or higher. Patients could receive other vasopressors if required, but open-label AVP was not permitted.  

What were the endpoints?

The primary outcome was the cumulative volume of all blood products used for that patient. Secondary outcomes included total volume of crystalloid and total vasopressor requirement within the first 48 hours. The authors also looked at 30-day outcomes including mortality, length of stay and complication rates. 

What were the results?

The authors had performed a power calculation assuming a  baseline mean volume of 5.4l of blood products per patient and a  50% reduction associated with AVP use. This suggested that they required 100 patients. The estimated blood product volume was based on results from a previous study by the authors. I did attempt the calculation on two on-line calculators and on each occasion, the result I obtained suggested that twice as many patients would be required. 

Given the inclusion and exclusion criteria for this study, it took the authors four years to reach their target number of patients. The patients were nearly all male, young and most had penetrating trauma. As might be expected from the inclusion criteria, they were all pretty sick with the study group having a median ISS of 19 (SD 14-26) and the placebo group having a median value of 26 (SD 17 – 34). 

In terms of the blood products used, AVP use appeared to be associated with reduced requirements of each of the components. 

AVP GroupPlaceboP value
PRBC, median, IQR, L0.9 (0.1-1.8)1.5 (0.6-3.0)0.8
FFP, median, IQR, L0.9 (0.8-1.3)1.0 (0.5-1.8)0.03
Platelets, median, IQR, L200 (0-300)300 (0-600)0.02
Cryoprecipitate, mean (SD) ml12.6 (75.4)34.7 (84.8)0.04
Total blood products, median (IQR), L1.7 (0.7 – 3.1)3.0 (1.4 – 5.2)0.03
Blood Products Used

The volume of crystalloids used in each group was not significantly different (9.7L (7.2 -13.0) in the AVP group vs. 10.7L (8.7 – 14.4), p = 0.24). The use of additional vasopressors, calculated in terms of norepinephrine equivalence, also failed to show a significant difference (581 µg (1.2 – 11255) in the AVP group vs. 1536 µg (227 – 8491), p=0.40).

AVP was not associated with a significant reduction in length of stay 16 days (IQR 10 – 32) vs. 22 days (IQR 14 – 44)  (p = 0.46) or with a significant difference in 30 day mortality with 6 deaths in each group. 

There was an unexpected reduction in the risk of DVT, with 5 out of 44 patients in the treatment group were affected versus 16 out of 47 in the placebo group. Patients were scanned weekly for three weeks as it was thought patients receiving AVP may have been at increased risk of DVT due to activation of platelets by AVP. The reason for the observed difference is not clear.

Bottom Line?

This paper suggests that the use of AVP did reduce the administration of blood products in major trauma patients requiring transfusion of more than six units; however, this failed to translate into any measurable clinical outcome, other than the unexpected reduction in DVT incidence, so I don’t think I’m ready to change my practice yet. 

The authors suggest a further, larger study is required to demonstrate the true effect but given the lack of demonstrated clinical benefit, and the small number of patients that this is applicable to, it may be some time before this is completed. 


Craig Ferguson

Cite this article as: Craig Ferguson, "JC: AVP in Haemorrhagic Shock. St Emlyn’s," in St.Emlyn's, March 2, 2020,

Posted by Craig Ferguson

Dr Craig Ferguson MB ChB, FRCEM, PhD is an Editorial Board Member of the St Emlyn’s blog and podcast. He is a Consultant in Emergency Medicine in Manchester where is operational lead for the emergency department. His research interest include diagnostics, heart failure, human factors and EBM. You will find him on twitter as @doccjf

  1. […] not been entirely Coronvirus (not quite). There are also posts about the Diploma in Immediate Care, the use of AVP in haemorrhagic shock, a great article on “Patient Flow” (yes, the four hour target will still be a thing […]


Thanks so much for following. Viva la #FOAMed