Morgenstern, J. Massive hemorrhage: The clinical approach, First10EM, October 23, 2023. Available at:
Your patient is bleeding. A lot. Their vital signs are abnormal. They look very sick. (I am being intentionally vague, because the post is going to look at massive hemorrhage in general, rather than getting into the specifics of any one condition.)
This post is a condensed clinical summary of appropriate actions in the setting of massive hemorrhage. If you are interested, there is an accompanying post that goes into the evidence in excruciating detail.
Step 1 is to stop any bleeding that you can. Put direct pressure on any visible source of bleeding. Place a tourniquet on extremity bleeding. Apply a pelvic binder if indicated. Call for help early if you are going to need a surgeon to stop the bleeding. It is much better to keep the patient’s blood in their body than to try to replace it later.
Get adequate vascular access. Personally, I stick to peripheral IV for the initial resuscitation of most bleeding patients. Largely, that is a pragmatic decision, working in community hospitals, but as long as I have multiple 18 gauge IVs, I am generally going to have adequate flow rates, and the logistics of placing a central line just complicate resuscitation. Unlike medical resuscitations, an intraosseous line will not save you in the context of bleeding, so if you don’t have adequate peripheral access, a central line will be necessary. The best option is probably a hemodialysis catheter. A standard triple lumen central line is not ideal, as flow rates will be lower than through peripheral IVs.
Start resuscitation with uncrossed packed red blood cells (PRBCs). Not every patient will need the full massive transfusion protocol. Unless the patient is critical, before activating your hospital’s code transfusion, which diverts resources away from other patients, start with 3-4 units of PRBCs. O positive is fine except for women of childbearing age or younger. If there is a delay to blood product arrival, it is fine to start with crystalloids, but avoid crystalloids once blood is available. (Practically speaking, if the only way to get STAT blood in your hospital is through a massive hemorrhage protocol, activate immediately, but good hospitals should have other systems.)
Unless there is a neurologic emergency, target a low-normal blood pressure. Also sometimes called “permissive hypotension”, the goal is to maintain perfusion, but not increase blood pressure to a level that you might increase ongoing bleeding. The target is a MAP of approximately 60-65. In traumatic brain injury, instead target a MAP of 90-100.
Activate your massive hemorrhage protocol. There is no single criteria on which to activate. I use the ‘critical administration threshold’. If you hit your third unit of PRBCs and the patient still looks sick, activate. (Note that hemoglobin is not mentioned at all in this decision.)
Follow your hospital’s policy with regards to balanced transfusion. (You will read in the main post that there is very little evidence for balanced transfusion, and higher ratios of FFP and platelets might actually be doing harm. I anticipate that future research will tell us to use less empiric FFP and platelets, but these are not issues you want to be debating with a critically ill patient in front of you. Just follow your hospital’s protocol.)
Don’t forget about calcium and fibrinogen. I aim to give a gram of calcium chloride (or equivalent) with every round of 1:1:2 transfusion that I give initially, and treat by lab values after that. I don’t give fibrinogen empirically, but replace it when the fibrinogen level falls below 1.5 grams/L. The usual dose is either 4 grams of fibrinogen concentrate or 10 units of cryoprecipitate.
Transition from empiric to lab guided therapy as soon as possible. Empiric 1:1:2 therapy only really makes sense in the absence of data. If you can get rapid and repeated bedside measurement of INR, hemoglobin, platelets, calcium, and fibrinogen, use that to guide therapy instead. (If you are relying on your lab to run tests, you need to have a protocol to ensure the results are communicated rapidly, without waiting for checks on abnormal results, and directly to the physician in charge.) A VBG and lactate might also help monitor perfusion state, and are often recommended, but I think perfusion is probably best monitored clinically. The early data on viscoelastic testing (TEG/ROTEM) is also reasonable, so use that if it is available to you. The sooner we can move from empiric to lab guided therapy the better, as it will limit the number of unnecessary products we give our patients.
Once in the laboratory guided phase, my general targets are:
- Hemoglobin > 80
- Platelets > 50 (>100 if there is cranial or spinal injury)
- INR 1.8
- Fibrinogen >1.5 grams /L
- Ionized calcium > 1.15 mmol/L
Try to prevent further bleeding. The easy step is to reverse any known coagulopathy. TXA probably only has a role in trauma.
- Warfarin: Prothrombin complex concentrate 2,000 units plus vitamin K 10 mg IV
- DOACs: PCC 2000 IU; consider a second dose if ongoing significant bleeding at 1 hour.
Supportive care matters. Don’t let supportive care interfere with other life saving procedures, like obtaining vascular access, or chest tubes, but try to prevent hypothermia and acidosis as much as possible. In the midst of a chaotic resuscitation, no one is going to remember to take repeat temperature measurements, so an esophageal, rectal, or bladder probe makes sense in these patients. Blood should be going through a rapid infuser, so should be warmed. The primary supportive measures are removing wet clothing, warm blankets, and a forced air heater.
Reconsider source control. We are generally very good at controlling external bleeding in the very first step, but don’t forget the importance of getting the patient to the OR, or endoscopy, or interventional radiology, especially if you are unable to stabilize them with initial resuscitation.
Remember to stop the protocol. Massive hemorrhage protocols are incredibly resource intensive for hospitals. Normally the entire blood bank and lab is shut down to focus soley on your patient. For the sake of all the other patients in the hospital, it is essential to let everyone know when your patient has stabilized, and you can cancel the ‘code transfusion’.
In order for a massive transfusion to run smoothly, you need to practice. Someone on the team needs to be an expert in managing a rapid infuser. You need to practice with your porters, to ensure blood gets from the blood bank to wherever it is needed rapidly. You need to practice communication with the lab.
Some hospitals attach a phone to the first cooler of blood, so the lab always knows how to get in contact with the resuscitation team. That sounds like a brilliant idea to me, but I have not seen it in action.