How I manage major hemorrhage Free

Hematology consultation during life-threatening hemorrhage is challenging because it is infrequent, often occurs outside of daytime hours, and requires immediate response, often without the opportunity for direct assessment of an unfamiliar patient. This article provides a framework for the consultative discussion, outlines pathways for laboratory and transfusion decision-making for varied patient populations, and highlights strategies to optimize care. The focus of the article is on life-threatening hemorrhage requiring both red blood cell (RBC) and hemostatic transfusion support (plasma, platelets, cryoprecipitate, and coagulation factor concentrates). Management of less severe bleeding, particularly gastrointestinal tract hemorrhage, is usually with restrictive RBC transfusions, with exceptional use of plasma, and thus rarely requires hematology consultation.^1-3 Recent US and international guidelines for gastrointestinal tract bleeding recommend the use of a restrictive hemoglobin threshold (7-8 g/dL) for both variceal and nonvariceal bleeding.^2,4,5 In terms of recommendations for the use of hemostatic transfusion support for patients with variceal bleeding, the guidelines highlight that bleeding is from portal hypertension (and not a bleeding diathesis) and that no thresholds for use of hemostatic products are recommended.⁵ In a series of 53 836 massive transfusions (>10 units in 24 hours), the most common indications for massive transfusion were cardiac surgery (35%), noncardiac surgery (33%), traumatic injury (16%), and obstetrics (3%).⁶ Major hemorrhages account for 5% of transfusions.⁶ This article focuses on the management of hemorrhage in trauma, cardiac surgery, and obstetrics. The key contributors to the hemostatic derangement in these conditions are varied (Table 1). Given the diverse pathophysiological pathways that can lead to major hemorrhage, goal-directed, personalized care based on repeated laboratory assessments should be the primary strategy, and reliance on formula-based resuscitation (ratios of RBC to components) should be restricted to the early phase of hemorrhage while obtaining source control and awaiting results of laboratory tests of coagulation.²⁵ 

Quality improvement audits demonstrate it is a challenge to achieve high levels of compliance with massive hemorrhage protocols,^26,27 which is not surprising given the number and complexity of tasks that must be completed in the care of these patients. As such, the consulting hematologist should be astutely looking for strategies to optimize care to assist the bedside team to rapidly achieve hemostasis. Presented in Table 2 is a comprehensive list of key management strategies and suggested recovery tactics. Maintaining a collegial relationship with the interdisciplinary clinical team is essential for optimal patient outcomes. Thus, during these invariably brief interactions that take place over the telephone into an often tense operating room, the consulting hematologist must be professional and respectful but use an assertive communication style to ensure that the patient is provided an evidence-based hemostatic response.⁴⁴ 

Over the past 15 years, several randomized controlled trials (RCTs) in the setting of major hemorrhage across different patient types have been published. These trials have provided clarity on what we should (or perhaps should not) be doing to improve the care of bleeding patients. RCTs have found tranexamic acid (TXA), especially within 60 minutes of injury or hemorrhage, to be of value in traumatic injury³⁴ and postpartum hemorrhage (PPH),³⁵ and prophylactically in cardiac surgery⁴⁵ and noncardiac surgery³⁶; however, it is of little value (and potentially harmful) in patients with gastrointestinal tract bleeding.³⁷ 

Several RCTs have explored the role of coagulation products as part of early resuscitation in trauma settings. In a prehospital cluster trial involving 27 US air medical bases, there was a 10% improvement in survival when they compared 230 trauma patients administered a median of 2 units of plasma with 271 patients who received only crystalloid and red cells based on local protocols.⁴⁶ Subsequently, 3 RCTs comparing prehospital transfusion (including plasma) with crystalloid in traumatic injury were conducted and failed to observe improvements in outcomes.^47-49 It has been hypothesized that prehospital plasma may have clinically important benefits in a subset of trauma patients, such as those with severe blunt injury, traumatic brain injury,⁵⁰ or longer transport times.⁵¹ For trauma patients with severe bleeding in the emergency department, an RCT of 680 patients found no difference in mortality at 24 hours or 30 days when a 1:1:1 ratio of red cells to plasma to platelets was compared with a lower ratio of 2:1:1, but left residual uncertainty for other patient outcomes.^52,53 

The role of viscoelastic testing in bleeding management has also been studied in several settings. A cluster trial of viscoelastic testing and a transfusion algorithm in cardiac surgery showed a reduction in the rate of major bleeding⁵⁴; however, a smaller RCT in traumatic injury, called Implementing Treatment Algorithms for the Correction of Trauma-Induced Coagulopathy (ITACTIC), found no improvements in outcomes for patients managed with viscoelastic testing, compared with standard laboratory testing, despite receiving hemostatic products earlier and in greater number.⁵⁵ 

A large RCT of preemptive fibrinogen replacement in trauma, compared with fibrinogen replacement based on laboratory test results, found no improvement in patient outcomes.⁵⁶ Similarly, RCTs find preemptive fibrinogen replacement to be of no benefit in PPH.^57,58 For the type of fibrinogen replacement product, cryoprecipitate and fibrinogen concentrates have similar efficacy in cardiac surgery and trauma^33,59; and, fibrinogen concentrate is cost-effective, compared with cryoprecipitate.⁶⁰ Investigators in the area of trauma are continuing to study the differential effects on hemostatic laboratory tests after the administration of cryoprecipitate, compared with fibrinogen concentrates.^59,61 Prophylactic administration of 4-factor prothrombin complex concentrates (PCCs) as an adjunct to formula-based resuscitation failed to improve outcomes in trauma patients and increased the rate of thrombosis.⁶² Early-stage RCTs comparing PCCs with plasma for hemorrhage after cardiac surgery suggest similar efficacy to plasma, and a large definitive trial has completed enrollment (NCT05523297).^63-65 

Given the urgency of the situation, a targeted conversation with the bedside team can help to rapidly understand the current (general and hemostatic) status of the patient, hemorrhage source control, and the hemostatic treatments to date (Table 3). This conversation will direct your suggestions for additional laboratory testing, transfusions (types and dosages), anticoagulant reversal, and hemostatic treatments. When certain hemostatic tests of coagulation have not been done, facilitation of expedited testing with the hematology and coagulation laboratories is critical. It should be a core competency for hematologists in training to have a thorough understanding of hemostatic laboratory tests, including interpretation of viscoelastic tests, and the evidence base guiding the care of the patient with major hemorrhage. The inclusion of the management of major hemorrhage in postgraduate training is important in improving the competency in this area.⁶⁶ The subsequent 3 cases highlight the important contributions of the consulting hematologist in supporting investigations and suggesting additional management strategies.

It is also important that hematology is involved with development and dissemination of institutional policies or guidelines for patient bleeding management, including for anemia and anticoagulation management before elective surgery or obstetrical delivery (anemia is a contributing cause for impaired hemostasis and transfusion),⁶⁷ PPH prevention,⁶⁸ transfusion algorithms for high blood loss surgeries (eg, cardiac),⁵⁴ and massive hemorrhage.⁴¹ The hematology service needs to be involved in their creation, dissemination (education and simulation), and efforts to ensure high rates of adherence (quality metric data collection⁴¹). The promotion of a collaborative team approach for planning and reviewing the management of complex patients at risk for or with major hemorrhage can be invaluable for quality assurance and improvement purposes.

A 68-year-old pedestrian struck by a vehicle at high speed is transported by helicopter to a trauma center for definitive surgical control of hemorrhage. She has a head injury with a bleeding scalp laceration and decreased Glasgow Coma Score of 9, with high suspicion for traumatic brain injury. During transport, she is hypotensive (systolic blood pressure, 60-65 mm Hg) and tachycardic (heart rate, 120-130 beats per minute). She is administered 500 mL of crystalloid and 2 units of RBCs during transport, without hemodynamic improvement. On arrival to the hospital (43 minutes after injury), the massive hemorrhage protocol is activated as her Assessment of Blood Consumption score is 3 (systolic blood pressure < 90 mm Hg, heart rate > 120 beats per minute, and +focused assessment with sonography in trauma [FAST]).⁶⁹ She is noted to have multiple other injuries in addition to the traumatic brain injury, including a hemothorax and a pelvic fracture with a retroperitoneal hematoma. Hemostatic laboratory tests, including viscoelastic testing, are drawn at 72 minutes after injury. Initial resuscitation is commenced with a formula-based resuscitation (2:1:1 units of RBC to plasma to platelets). TXA, 1 g, is administered at 95 minutes. Despite the formula-based transfusion strategy, the patient continues to have profuse bleeding from her stapled and bandaged scalp laceration and her central line insertion site. She is being transported to interventional radiology for embolization of bleeding pelvic vessels. You receive a page from the interventional radiology suite from the Trauma Team Leader requesting your help with navigating a coagulation testing problem. The coagulation laboratory has requested redraw of the coagulation tests as no results could be obtained (international normalized ratio [INR] > 10, partial thromboplastin time [PTT] > 150 seconds, and fibrinogen < 0.5 g/L). The baseline viscoelastic testing by rotational thromboelastometry (ROTEM) is shown in Figure 1. Her hemoglobin is 10.2 g/dL, platelet count is 106 000 × 10⁶/L, pH is 7.12, and lactate is 10.3 mmol/L. Her temperature is 34.7°C. You are called at 120 minutes from injury.

This patient has hemostatic failure attributable to acute traumatic coagulopathy (Table 1), with the presence of marked hypofibrinogenemia. The hemostatic derangement is consistent with the clinical scenario of major uncontrolled hemorrhage after trauma, severe shock, delayed TXA administration, and hypothermia. The clinical picture of persistent bleeding from her scalp wound and central line site is also indicative of a severe hemostatic derangement. Acute traumatic coagulopathy with hyperfibrinolysis is associated with a high mortality rate of ≈80%.⁷⁰ Every effort needs to be made to achieve an INR <1.8 and fibrinogen >1.5 g/L (or normalization of viscoelastic testing). It is communicated to the Trauma Team Leader that sample redraw is not required, and the hemostatic failure should be treated immediately. The recommendation is made to administer another 1 g of TXA and 3 pools of cryoprecipitate (15 units of cryoprecipitate or the equivalent amount of fibrinogen concentrate) to increase the fibrinogen level to >1.5 g/L, 4 units of plasma, and 1 dose of platelets (with the expectation that the platelet count would decrease below 100 000 × 10⁶/L because of ongoing bleeding). The bedside team should also address the acid-base abnormalities, hypocalcemia, and hypothermia that will further exacerbate her coagulopathy. The posttransfusion laboratory tests are improved, with an INR of 2.2, fibrinogen of 1.3 g/L, platelet count of 107 000 × 10⁶/L, and hemoglobin of 8.9 g/dL. The posttransfusion viscoelastic testing is shown in Figure 1. Embolization is complete, the patient’s hemodynamics have stabilized, and the patient is being transferred to the operating room. In anticipation of bleeding during surgery and to prevent postoperative rebleeding, you recommend transfusing an additional 4 units of plasma, to correct the prolonged EXTEM-clot time (CT) and INR, and 2 pools of cryoprecipitate, to achieve a fibrinogen level >1.5 g/L.

There were several issues with the patient’s hemostatic management. First, there was a delay to administering TXA, which should have been prioritized within the first 60 minutes of injury to mitigate hyperfibrinolysis that can result in severe hypofibrinogenemia.³⁸ TXA has the greatest survival benefit if administered within 60 minutes of injury.^71,72 TXA is not associated with a clinically significant increase in thromboembolic events in the setting of trauma.⁷³ Increasing patient age is also not associated with an increase in thromboembolic events in the trauma setting after TXA administration.⁷⁴ Last, administration of a single dose of 2 g of TXA is effective, is not associated with safety concerns, and reduces the logistical burden of 2 doses.^75,76 A single 2-g bolus infusion was used in a large prehospital traumatic brain injury study without safety concerns (345 patients randomized to the 2-g bolus arm of the study).⁷⁵ 

Second, hemostatic laboratory tests should also have been prioritized to more rapidly assess the degree of hemostatic derangement and expedite targeted hemostatic treatments. The coagulation laboratory had failed to follow the protocol for rapid release of results despite no clot found. This laboratory strategy has been validated and found to be safe.⁴² 

Last, hypothermia, shock, and metabolic abnormalities were not adequately addressed, likely because of competing priorities during her initial resuscitation. These factors are known to be prognostically important, and their management needs to be prioritized.⁷⁷ 

You receive a page from the operating room to provide advice regarding the indication and dosing of recombinant activated factor VII (rFVIIa) for the management of refractory bleeding after cardiopulmonary bypass (CPB) in an 80-kg man with an acute type A aortic dissection, who is undergoing emergency aortic valve and aortic hemiarch replacement. His medications include antihypertensive drugs and acetylsalicylic acid. On presentation, he was stable and neurologically intact. Baseline laboratory results showed anemia (hemoglobin, 12.6 g/dL), thrombocytopenia (platelet count, 145 000 × 10⁶/L), elevated creatinine (125 μmol/L), and normal INR and PTT. Activated clotting time (ACT) before heparin administration was within the normal range at 102 seconds.

The surgery was complex and required 210 minutes on CPB, including 35 minutes of hypothermic circulatory arrest. He received 2 g of TXA before CPB, and a total of 60 000 IU of unfractionated heparin during surgery. Following termination of CPB and reversal of heparin with 400 mg of protamine, the ACT was 132 seconds, and massive bleeding was noted. Another 100 mg of protamine was administered, but the ACT did not improve, at which point a formula-based transfusion strategy was initiated. Over the next 45 minutes, the first set of blood products (4 units of RBCs, 2 units of plasma, and 1 dose of platelets) was administered. However, bleeding continued at a rate of 200 mL every 10 minutes. A second set of blood products was ordered, and laboratory assays were sent. At 120 minutes after CPB, when hematology was consulted, the patient was bleeding at a rate of 100 mL every 10 minutes, despite having received the second set of blood products (4 units of RBCs, 2 units of plasma, and 1 dose of platelets). Laboratory testing that was sent after the first set of blood products showed a hemoglobin of 7.2 g/dL, platelet count of 45 000 × 10⁶/L, fibrinogen level of 1.1 g/L, INR of 1.9, and PTT of 39 seconds. Repeat samples had just been sent to the laboratory when you were consulted, but the results were not yet available. A recent point-of-care arterial blood gas result showed a normal pH and calcium (after 2 g of calcium chloride), and a hemoglobin of 7.5 g/dL.

The patient has had major bleeding for 120 minutes after CPB, with only a modest response to a large-volume, formula-based transfusion strategy. Ongoing bleeding in this patient may prove catastrophic as there is a direct, nonlinear relationship between the amount of transfusions and risk of major morbidity and mortality after cardiac surgery.^78,79 Because this patient is continuing to experience major hemorrhage, rapid restoration of hemostasis is paramount.

Although rFVIIa can restore hemostasis in the setting of refractory bleeding by generating a thrombin burst at sites of injury,^7,79,80 you recommend that it not be used at this time because it carries significant thromboembolic risks, and the patient continues to have major coagulation defects that are correctable with lower-risk hemostatic agents, which may avoid the need for rFVIIa or improve its efficacy should it be needed as rescue therapy.⁷⁹ 

Your first recommendation is to administer a second 2-g dose of TXA as hyperfibrinolysis is a major contributor to post–cardiac surgery bleeding.^7,81 The initial 2 g (20 mg/kg) of TXA that was administered pre-CPB is consistent with the recommended low-dose regimens for prevention of hyperfibrinolysis.^7,81,82 However, because the initial dose was administered ≈5 hours (>2 half-lives) ago and the patient has lost the equivalent of ≈1 blood volume since then, redosing at this time is indicated to ensure that the concentration of TXA is at or above therapeutic levels.

Your next recommendation is to initiate an aggressive but targeted transfusion therapy based on the coagulation abnormalities that were present before the second set of blood products, as the second set is unlikely to have corrected the abnormalities given the continued major hemorrhage and the lack of hemoglobin increment. It is recommended to administer, in order of availability, 4 g of fibrinogen concentrate (or 10 units or 2 pools of cryoprecipitate) to increase the fibrinogen level to >2.0 g/L, 2 doses of platelets to increase the platelet count to >100 000 × 10⁶/L, 3 to 4 units of plasma to a target INR ≤1.5, and RBCs (or cell-saved blood) to maintain a hemoglobin of >9.0 g/dL.

The recommended thresholds for administration of fibrinogen replacement and platelets are more liberal than those in current bleeding management guidelines.^83-85 In our view, these more liberal thresholds are warranted because this patient has excessive consumption and loss of fibrinogen and platelets attributable to the underlying condition (ie, acute aortic dissection),⁸⁶ prolonged CPB duration with circulatory arrest, presence of a new aortic graft (which absorbs fibrinogen and consumes platelets),⁸⁷ and ongoing major hemorrhage.

Similarly, the 9.0-g/dL hemoglobin target is also higher than commonly recommended restrictive thresholds of 7.0 to 9.0 g/dL^83,84 to not only allow for ongoing blood loss, but also to take advantage of the contributions of RBCs to hemostasis (which include enhanced platelet function and thrombin generation).^88,89 

Additional coagulation factor replacement is needed because the elevated INR indicates that, despite the plasma transfusions, the patient still has impaired thrombin generation because of coagulation factor deficiency, which is an important cause of bleeding in cardiac surgery.^7,10,90 Emerging evidence suggests that PCCs at a dose of 25 IU/kg may be an alternative to plasma in this setting as it may be able to better restore hemostasis without increasing the risk of thromboembolic complications,⁶³ in contrast to the elevated risk seen with rFVIIa.⁹¹ A large multicountry, multicenter RCT of plasma vs PCC in cardiac surgery has completed enrollment and may provide additional clarity on their use in cardiac surgery (NCT05523297).

After the recommendations were performed, bleeding slowed to a rate that allowed for chest closure and transfer to critical care. Repeat laboratory tests before transfer show a hemoglobin of 10.5 g/dL, platelet count of 140 000 × 10⁶/L, fibrinogen level of 2.1 g/L, and INR of 1.6.

There were several deficiencies in hemostatic management of this patient. First, the prophylactic TXA dose was likely too low given the presence of multiple factors that placed the patient at high risk for massive bleeding, including the aortic dissection, baseline renal dysfunction, thrombocytopenia, antiplatelet therapy, and anemia, as well as emergency, complex surgery that requires hypothermic circulatory arrest and prolonged CPB.^86,92,93 Second, the clinical team should have assessed the coagulation status before termination of CPB to allow for goal-directed rather than formula-based transfusion management. Ideally, coagulation assessment should include viscoelastic and platelet function assays that can be conducted on heparinized samples, providing timely and specific results that can be addressed immediately after CPB.^7,54 If viscoelastic testing is not available, then platelet count and fibrinogen level can, in some centers, be conducted on CPB. The Clauss fibrinogen uses both a high dilution (1:10) and a high concentration of thrombin, mitigating some of the heparin interference, but the result should be interpreted with caution with high levels of heparin.⁹⁴ Third, the post-CPB coagulation status should be measured frequently (every 30-60 minutes) to guide hemostatic therapy. Last, the focus should be on normalizing hemostasis as rapidly as possible rather than minimizing transfusions, which entails the use of liberal rather than conservative transfusion thresholds.

The anesthesiologist on labor and delivery calls you for advice with an obstetric hemorrhage in a patient who is in interventional radiology for uterine artery embolization. The patient is a 36-year-old (body mass index, 41 kg/m²) gravida 4, para 3 woman who had a vaginal birth of twins 2 hours ago. Her obstetric history includes a cesarean delivery for breech presentation, followed by 2 vaginal deliveries. Her third delivery had a mild PPH (1100 mL) attributable to uterine atony. During that delivery, she received additional uterotonics and 1 g of TXA but did not require any surgical intervention or transfusion. This twin pregnancy was complicated by preeclampsia, requiring admission to the hospital at 35 weeks’ gestation. On admission, her blood pressure was 162/90 mm Hg, alanine aminotransferase was 70 U/L, lactate dehydrogenase was 550 U/L, hemoglobin was 11.5 g/dL, and platelet count was 138 000 × 10⁶/L. She was started on labetalol and magnesium sulfate.

As both fetuses were cephalic presenting, an induction of labor was initiated, and she delivered 32 hours later. With delivery of the second infant, she received carbetocin, 100 μg intravenously, to prevent uterine atony, but 4 minutes after delivery, the vaginal bleeding was brisk and uterine tone unsatisfactory. During uterine massage, the anesthesiologist administered misoprostol, 200 μg sublingually, ergometrine, 250 μg intramuscularly, and 1 g of TXA. After the additional uterotonics and an ultrasound-guided curettage of the uterine cavity, tone had improved. The quantitative blood loss (QBL), the weight of blood loss in surgical materials and volume measured with a volumetric under-buttocks drape, was 2100 mL; and with ongoing bleeding, embolization of the uterine arteries was believed to be indicated. The obstetrical massive hemorrhage protocol was activated before transfer to interventional radiology, triggering the delivery of pack 1 (4 units of uncrossmatched O-negative, K-negative RBCs). RBC transfusion was commenced at 8 minutes after hemorrhage protocol activation. Before transfer to the interventional radiology, an arterial line was placed and blood samples were drawn to assess hemostasis (75 minutes from the onset of hemorrhage). A Bakri balloon was placed into the uterine cavity, and the patient was emergently transported to interventional radiology.

While attempting uterine artery embolization, the patient’s hemodynamics deteriorate, and the Bakri balloon blood reservoir measured an additional 900 mL. The laboratory testing sent before leaving the delivery suite reveal a hemoglobin of 8.1 g/dL, platelet count of 109 000 × 10⁶/L, fibrinogen level of 0.8 g/L, and INR of 1.5. The bedside team concludes that a hysterectomy is required immediately because of inability to control bleeding with embolization and hemodynamic instability. The request to the consulting hematologist is to provide guidance regarding the optimal hemostatic response given the recent laboratory testing results.

Hemostatic resuscitation is paramount until surgical control of hemorrhage can be obtained via hysterectomy and to prevent rebleeding. While coordinating an emergency hysterectomy, the hemoglobin should be maintained in a safe zone (7.0-9.0 g/dL), the platelet count maintained at >50 000 × 10⁶/L to 75 000 × 10⁶/L, fibrinogen at >2.0 g/L, and INR at <1.8.⁹⁵ The targets for resuscitation are all based on expert opinions, derived from evidence from observation data and trials outside of postpartum hemorrhage. Although it is likely critical that the fibrinogen level is maintained >2.0 g/L during hemorrhage,²⁹ preemptive use of fibrinogen replacement (before fibrinogen levels are known or decrease below 2.0 g/L) has not been shown in RCTs to improve outcomes.⁹⁶ The recommended hemoglobin target during resuscitation is based on a cluster-randomized trial involving 411 trauma patients at 22 centers, finding similar outcomes when a target of 7 to 9 g/dL was compared with 10 to 12 g/dL, as no RCTs have been performed in the PPH setting.⁹⁷ The recommended INR target of 1.8 is based on expert opinion and not clinical trials; however, given the difference in clotting factor concentrations between a target of 1.5 and 1.8 is <10%, the targets of 1.5 and 1.8 may be considered equivalent.⁹⁸ 

This patient opted for a vaginal delivery of twins after weighing the benefits and risks, particularly in light of her multiple risk factors for PPH, including a prior cesarean delivery, history of previous PPH, multiple gestation, elevated body mass index, hypertensive disorder of pregnancy, and prolonged labor of >24 hours.⁹⁹ Effective management of PPH requires a multicomponent bundled clinical intervention focused on risk assessment, immediate readiness, active third-stage management, and early intervention.¹⁰⁰ Early detection of PPH and use of bundled treatment reduce the incidence of severe PPH, need for surgical intervention, and death.¹⁰¹ 

Transition from routine postpartum bleeding to PPH (QBL, >1000 mL) can be abrupt, rapidly escalating to multiorgan dysfunction and death. Underestimating blood loss and the severity of PPH leads to treatment delays. Traditional nonquantitative estimates of blood loss underestimate blood loss. Accurate monitoring with QBL is essential for longitudinal risk assessment after delivery. Given this patient’s multiple risk factors for PPH, preparation with all potentially needed materials and medications should be made in advance of delivery; these include dual intravenous access, RBC availability, pharmaceuticals (uterotonics, TXA), and medical devices (Bakri balloon).

Compared with placebo, ergometrine, carboprost, or misoprostol, oxytocin is the drug of choice for the active management of the third stage of labor. The other agents are adjunctive for the treatment of uterine atony and PPH. Carbetocin is a synthetic, long-acting analogue of oxytocin with equivalent affinity for the oxytocin receptor. Prophylactic carbetocin, 100 μg intravenously given over 30 to 60 seconds, provides similar or better clinical effects compared with oxytocin.⁹⁵ 

An additional hemostatic adjunct to uterotonics is TXA. Despite mixed findings of RCTs of prophylactic or treatment TXA, it remains a guideline recommendation based on its benefit-to-risk considerations.^102-104 TXA should be considered whenever a second-line adjunctive uterotonic is administered prophylactically in patients at high risk of PPH, such as in this case with multiple PPH risk factors. A second dose of 1 g should be considered 30 minutes after the initial dose if bleeding is not controlled.

Emergent uterine artery embolization is an effective method in the management of PPH. If hemodynamically stable with persistent low-volume bleeding, uterine artery embolization is able to achieve hemostasis and reduce the incidence of hysterectomy.^105,106 Uterine artery embolization can also reduce blood loss if a hysterectomy is still necessary to treat the atony and ongoing PPH. It should also be considered after hysterectomy in rare cases with persistent bleeding.

Several issues with hemostasis management were noted with this case. First, the patient was at high risk for a PPH, and a predelivery multidisciplinary review with the team should have been conducted to ensure adequate preparation in the event of PPH (eg, prophylactic TXA, ensuring all medications and devices were on hand, and cell saver setup on hand¹⁰⁷). Second, the patient’s bleeding was likely too brisk for consideration of uterine artery embolization. In cases of massive PPH, where basic resuscitation and hemodynamic management cannot stabilize the patient, a hysterectomy should not be delayed. Uterine artery embolization requires transporting the patient to a less ideal environment for anesthesia and resuscitative efforts, so only reasonably stable patients should be considered for uterine artery embolization. Third, failure to monitor the fibrinogen level early and frequently (every 30-60 minutes)⁴⁰ may have delayed the achievement of hemostasis because of the decrease in the level to <2.0 g/L.

Managing the hemostatic response to acute, life-threatening hemorrhage is an important competency for all consulting hematologists. The clinical response to a bleeding patient is challenging, often with key treatments or tests inadvertently delayed or missed because of the sheer number of tasks required. This article provides a review of these pitfalls to assist the consulting hematologist. Through these varied case vignettes, we have highlighted the importance of considering the clinical context, the rate of hemorrhage, pharmaceutical adjuncts, and laboratory test results in decision-making. In all cases of major hemorrhage, obtaining surgical or radiologic control of bleeding is paramount. The hematologist plays an important communication role between the clinical team and both the coagulation and blood bank laboratories. In cases such as these, an aggressive liberal (and as much as possible, personalized) approach to hemostatic replacement is favored to assist with rapid bleeding control to mitigate the downstream morbidity of hemodynamic instability from ongoing hemorrhage.

The authors acknowledge the critical review by Sarah Ryan of Queen’s University.

Contribution: All authors were involved in the drafting of the manuscript and approved the final version.

Conflict-of-interest disclosure: J.L.C. has received research funding from Canadian Blood Services (produces and distributes blood components) and Octapharma (manufactures prothrombin complex concentrate and fibrinogen concentrates). K.K. has received research funding, consulting fees, and honoraria from Octapharma; and consulting fees from Werfen (provides viscoelastic point-of-care testing). R.B.G. declares no competing financial interests.

Correspondence: Jeannie L. Callum, Kingston Health Sciences Centre, 201b, 88 Stuart St, Kingston, ON K7L 2V7, Canada; email: jlc17@queensu.ca.