Hemorrhagic Shock and Compensatory Circulation

Hemorrhage removes circulating volume, oxygen-carrying red cells, and components needed for hemostasis. These losses develop alongside compensatory changes in vascular capacity, resistance, cardiac performance, and fluid distribution. Pressure and hemoglobin concentration describe only parts of that evolving state. Understanding hemorrhagic shock requires following the source and rate of loss, the circulation's remaining reserve, and the relationship between oxygen delivery and tissue demand over time.

74.1 The Consequences of Blood Loss Depend on Rate, Amount, and Starting Conditions

Blood loss is a change in an existing circulation. Its consequences depend on the amount removed, how rapidly it leaves, whether loss continues, and the blood volume and cardiovascular reserve present beforehand. The same absolute loss represents different fractions of circulating volume in different people. Body size alone does not establish a precise starting blood volume, particularly during pregnancy, edema, chronic anemia, or prior resuscitation.

Rapid loss can reduce cardiac filling before redistribution and fluid conservation provide much support. Slower loss permits more compensation but progressively removes red-cell mass and iron. A person with chronic anemia may maintain resting flow through adaptation yet have little reserve for an additional acute fall in volume or increase in demand. An abrupt hemorrhage and a slowly developing anemia therefore cannot be interpreted from the final hemoglobin concentration alone.

The bleeding source also matters. A patent vascular injury continues to connect the circulation to a lower-pressure space, with flow influenced by vessel size, pressure, surrounding tissues, clot formation, and mechanical disruption. Blood accumulating internally remains outside the effective circulation even when it has not left the body. Local pressure can sometimes restrict bleeding, while simultaneously compromising nearby organs or venous return.

Associated injury adds pain, inflammation, endothelial disturbance, and sometimes direct cardiac or neurological damage. Hemorrhage from a gastrointestinal lesion, ruptured vessel, obstetric emergency, or traumatic wound shares important circulatory mechanisms but does not reproduce one identical clinical population.

In a large trauma-registry analysis, Guly and colleagues found that admission heart rate and systolic pressure changed across estimated blood-loss groups less than conventional shock classifications suggested. Blood loss was estimated rather than directly measured, and the observational findings do not provide a replacement bedside volume calculation. They reinforce the limited precision of assigning a fixed loss from vital signs. 1

74.2 Loss of Stressed Volume Reduces the Support for Venous Return

The first major circulatory effect of hemorrhage is often reduced effective filling. Systemic veins contain much of the blood volume and accommodate it according to their tone, compliance, and surrounding pressure. Removing blood changes the pressure–volume relationship supporting return to the heart. It does more than subtract an equal fraction from every vascular compartment.

Stressed volume describes the portion contributing to distending pressure in a vascular model; unstressed volume describes the amount accommodated before that pressure develops. They are functional descriptions rather than permanently separate reservoirs. Hemorrhage tends to reduce the pressure-generating volume, while sympathetic venoconstriction can recruit part of the remaining volume by reducing vascular capacity.

Let Qvr be venous return in L/min, Pms an effective mean systemic filling pressure and RAP intracavitary right atrial pressure in mmHg with the same reference, and Rvr resistance to venous return in mmHg·min/L. A simplified steady-state relationship is:

Loss of volume tends to lower Pms. RAP may also fall as the heart becomes less filled, partly preserving the pressure difference. Neither change guarantees preserved flow: venous resistance, vascular collapse, surrounding pressure, and the heart's operating function also matter. At severe central underfilling, further reduction in atrial pressure may fail to increase return because collapsible upstream veins limit flow.

Venoconstriction cannot manufacture red cells or replace all lost volume. It can redistribute the remainder, but its effects on resistance and regional drainage limit the resulting flow. Likewise, a low measured central venous pressure is not a quantitative measure of the missing blood volume.

Ventilation changes this relationship. Positive intrathoracic pressure can increase the downstream pressure encountered by systemic venous return and alter cardiac distension. A circulation relying on compensatory filling mechanisms may consequently respond strongly to a change in airway pressure, posture, or abdominal pressure. The venous-return equation is a framework for these interactions, not a complete model of all collapsible vessels.

74.3 Neural and Endocrine Responses Preserve Circulation at a Cost

Reduced stretch of arterial and cardiopulmonary receptors changes central autonomic output. Sympathetic activation increases heart rate and contractile support, constricts selected resistance vessels, and reduces venous capacity. Withdrawal of vagal cardiac influence can contribute to tachycardia. The relative contribution of each response varies with the starting state and the speed of hemorrhage.

Arteriolar constriction helps maintain arterial pressure despite falling output and redistributes flow. Cutaneous and splanchnic flow may fall substantially while pressure remains deceptively familiar. This preserves a pressure-generating circulation through selective sacrifice of flow rather than maintaining every tissue's normal perfusion. The regional pattern is modified by local metabolism, autoregulation, injury, and administered drugs.

Cardiac stimulation also has costs. Faster rate and greater contractile work increase myocardial oxygen demand, while tachycardia shortens diastolic filling and coronary perfusion time. Reduced coronary pressure, anemia, or pre-existing coronary disease can make the same compensatory response less sustainable. Maintaining pressure does not imply an unlimited myocardial reserve.

Renin release, angiotensin signaling, aldosterone, vasopressin, thirst, and sympathetic renal effects support vascular regulation and fluid conservation through different time courses. Reduced renal filtration and altered tubular handling can limit water and sodium loss. These responses conserve available fluid; they do not promptly restore red-cell mass or stop a disrupted vessel from bleeding.

Stress hormones also mobilize substrates. Increased glucose availability and altered lactate production accompany the circulatory response, while pain, agitation, breathing effort, or shivering can raise demand. The organism is simultaneously attempting to support supply and using energy to sustain compensation.

A measured hormone concentration reflects production, clearance, binding, and sampling time rather than a direct reading of successful compensation. The autonomic and endocrine mechanisms developed in Chapters 64 and 65 therefore need interpretation within the current circulation, rather than as separate assurances that adequate perfusion has been preserved.

74.4 Compensatory Reserve Varies and Cannot Be Read from Heart Rate Alone

Compensatory reserve is the remaining ability to sustain useful circulation as central filling declines. It includes vascular responsiveness, cardiac function, autonomic regulation, circulating volume, and tissue extraction capacity. Two people with similar measured pressures may be operating at very different distances from circulatory failure.

Lower-body negative pressure provides a controlled human model of central hypovolemia by drawing blood toward dependent vascular compartments. It permits serial study of responses that cannot ethically be produced by uncontrolled hemorrhage. However, it redistributes existing blood rather than removing red cells and plasma, and it does not reproduce a wound, traumatic inflammation, or evolving coagulopathy.

In a randomized crossover experiment in healthy adults, Watso and colleagues found that morphine reduced tolerance to progressive lower-body negative pressure without a corresponding difference in muscle sympathetic burst-frequency responses. The result illustrates that one neural measurement does not fully describe circulatory reserve. It does not establish an analgesic strategy for injured patients. 2

Age, chronic hypertension, myocardial disease, autonomic dysfunction, pregnancy, conditioning, and medications can change the response to underfilling. Sedation or neurological injury can modify the usual clinical expression. Pain and anxiety can produce tachycardia even without major hemorrhage; beta-blockade or conduction disease can limit tachycardia despite substantial loss.

Clinical observations also resist a single sequence. Ley and colleagues described frequent absence of tachycardia among hypotensive trauma patients in a retrospective multicenter study. This heterogeneous injury cohort does not isolate a single vagal mechanism or establish a universal late-bradycardia stage, but it shows why a slow or ordinary pulse cannot independently exclude serious circulatory compromise. 3

The interpretation is strongest when trends in pulse, pressure, mental state, skin perfusion, output, and the suspected source are connected. A stable measurement can mean successful compensation, little loss, a drug effect, or a limited ability to respond. Context separates these possibilities.

74.5 Hematocrit Measures a Fraction Rather Than the Red-Cell Amount

Whole-blood loss initially removes red cells and plasma together. If their proportions in the shed blood resemble those in the circulation, the remaining blood can have a nearly unchanged hematocrit despite a major reduction in red-cell amount. An early concentration therefore cannot reconstruct the amount already lost.

Let Vred be circulating red-cell volume and Vblood total blood volume, both in L. Hematocrit, Hct, is a dimensionless volume fraction; multiplying it by 100 expresses a percentage. In a simple well-mixed accounting model:

Consider an invented starting blood volume of 5 L with Hct = 0.45. Red-cell volume is 5 × 0.45 = 2.25 L, leaving 2.75 L of plasma. Loss of 1 L of blood at the same hematocrit removes 0.45 L of red cells and 0.55 L of plasma.

The remaining red-cell volume is 2.25 − 0.45 = 1.80 L, and total blood volume is 5 − 1 = 4 L. Hematocrit is therefore 1.80 ÷ 4 = 0.45, still 45%. Red-cell volume has fallen by 20%, although its concentration has not changed. With unchanged average cellular hemoglobin, total hemoglobin mass has also fallen by 20%.

This is an accounting example, not an assertion that all hemorrhage remains undiluted until one fixed time. Transcapillary fluid movement, administered fluid, splenic effects, sampling, and ongoing loss can change concentrations during the episode. A large-vessel hematocrit also does not exactly equal the average red-cell fraction throughout every microvascular compartment.

Pottgiesser and colleagues measured total hemoglobin mass serially after controlled blood donation in male volunteers. Their work emphasizes that restoration of hemoglobin mass is a distinct process from changes in measured concentration; recovery varied between individuals. A donation model cannot supply a universal recovery timetable for injured or critically ill patients. 4

74.6 Plasma Refill Changes Concentration Before Red Cells Are Replaced

Falling microvascular hydrostatic pressure can reduce filtration and, under appropriate local conditions, permit transient absorption of interstitial fluid. Lymphatic return and redistribution between vascular regions also affect circulating volume. The magnitude and persistence depend on barrier properties, interstitial pressure, plasma proteins, tissue fluid availability, and the evolving circulatory state.

It is therefore misleading to imagine an unlimited interstitial reservoir that automatically restores a fixed fraction of blood volume. Injury-related permeability, edema, prior dehydration, and low protein concentrations alter the exchanges. Fluid entering the circulation may support filling while simultaneously changing protein, electrolyte, and hemoglobin concentrations.

Continue the constructed example with 1.80 L of red cells remaining in 4 L of blood. Suppose, purely for accounting, that 1 L of cell-free fluid subsequently enters and remains in the vascular compartment, with no further bleeding or red-cell addition. Total blood volume becomes 5 L. Hematocrit becomes 1.80 ÷ 5 = 0.36, or 36%.

The falling hematocrit now reveals dilution of a reduced red-cell amount. It does not mean that the later sample marks the moment those cells were lost. Conversely, restoration of total vascular volume in this model does not restore oxygen-carrying mass. If the starting hemoglobin concentration was 15 g/dL and cellular composition remains constant, the new concentration is 15 × 0.80 = 12 g/dL.

A small study by Wachsmuth and colleagues followed hemoglobin mass and derived blood compartments after donation in premenopausal women. Plasma-volume and red-cell-volume trajectories differed, reinforcing this separation of quantities. The pilot population and calculated compartment estimates do not establish a refill clock for trauma. 5

Renal sodium and water conservation can support later fluid balance, whereas erythropoiesis requires marrow activity, substrates, and time. Inflammation, kidney dysfunction, iron restriction, and repeated sampling may limit red-cell recovery in critical illness. Neither a lower concentration nor a later rising concentration identifies the underlying mass change without considering fluid movement.

74.7 Oxygen Delivery Depends on Both Flow and Red-Cell Concentration

Hemorrhage threatens oxygen delivery through reduced cardiac output and loss of oxygen-carrying cells. The relative contribution changes as plasma refills or fluid is administered. Arterial oxygen tension and saturation can remain satisfactory while total oxygen delivery falls substantially.

Let CaO₂ be arterial oxygen content in mL O₂/dL, Hb hemoglobin concentration in g/dL, and SaO₂ fractional arterial saturation. For this illustration, use an oxygen-binding coefficient of 1.34 mL O₂/g hemoglobin and neglect the relatively small dissolved component. Let CO be cardiac output in L/min and DO₂ systemic oxygen delivery in mL O₂/min; 10 converts L to dL.

At an invented baseline Hb of 15 g/dL and SaO₂ of 1.00, content is 1.34 × 15 = 20.1 mL/dL. With CO = 5 L/min, delivery is 5 × 20.1 × 10 = 1005 mL/min.

After the model's immediate whole-blood loss, suppose measured CO is 3 L/min while Hb remains 15 g/dL. Delivery becomes 3 × 20.1 × 10 = 603 mL/min. After cell-free refill, suppose Hb is 12 g/dL and measured CO rises to 3.75 L/min. Content is now 16.08 mL/dL, so delivery is 3.75 × 16.08 × 10 = 603 mL/min again. Better flow has been offset by lower concentration.

These are constructed observed states, not predicted responses to a specified intervention. Dissolved oxygen, saturation changes, continued bleeding, and measurement error would modify the calculation. None of the numbers is a treatment target or critical delivery threshold.

Weiskopf and colleagues studied acute isovolemic anemia in selected conscious resting humans. Cardiovascular responses supported resting oxygen use despite reduced delivery. Volume was maintained and participants were not experiencing uncontrolled injury or hemorrhage; the findings demonstrate conditional reserve rather than a universally tolerable hemoglobin concentration. 6

74.8 Redistribution and Extraction Protect Some Functions While Leaving Regional Risk

When delivery falls, tissues may extract a greater fraction of the oxygen arriving. The available reserve depends on baseline extraction, diffusion, capillary transit, mitochondrial use, and metabolic demand. Whole-body extraction is an average that can conceal regions already unable to sustain their requirements.

The myocardium normally extracts a large fraction of delivered oxygen and relies strongly on increasing coronary flow when demand rises. During hemorrhage, lower perfusion pressure and anemia can oppose the increased flow required by sympathetic stimulation. A faster, more forceful heartbeat can therefore become vulnerable even while it helps preserve systemic pressure.

Cerebral autoregulation and redistribution can support brain flow within conditional limits. They do not guarantee protection during severe hypotension, associated brain injury, hypoxemia, or impaired autoregulation. An apparently preserved level of consciousness is useful information but is not a measurement of myocardial, renal, or intestinal oxygen balance.

The kidney and gut illustrate regional vulnerability. Total organ flow does not fully describe medullary oxygen availability or mucosal microvascular perfusion. Vasoconstriction can reduce delivery in vulnerable regions while systemic pressure is maintained. Renal filtration and urine production also depend on neurohumoral regulation and pressure history, so a change in urine output is neither instantaneous nor a direct measure of lost blood.

Venous oxygen measurements integrate extraction and flow from the territories they sample. A central venous sample is not identical to mixed venous blood, and neither describes every local region. Changes in demand, shunting, and impaired cellular extraction further complicate interpretation.

Pain, fever, respiratory effort, agitation, and shivering can narrow the delivery–demand margin. Sedation or cooling can alter measured demand but also change vascular tone, ventilation, and examination. The relevant question is whether supply supports current tissue function throughout the changing physiological state, rather than whether one global delivery or venous-saturation value appears reassuring.

74.9 Decompensation and Reperfusion Extend Beyond the Pressure Trace

Compensation can become insufficient when falling filling and oxygen content exceed cardiac, vascular, or extraction reserve. Output falls, regional oxygen availability deteriorates, and cellular ATP-dependent functions become harder to sustain. Reduced myocardial supply can impair the pump that is needed to maintain its own perfusion, creating a self-amplifying decline.

There is no universal transition at one blood-loss percentage or elapsed time. Continued bleeding, anesthesia, ventilation, hypothermia, acidosis, and associated injury can move the operating state rapidly. A previously stable pressure may have depended on substantial vasoconstriction and therefore provided little warning of the remaining reserve.

Lactate can rise when oxygen availability limits oxidative metabolism, but its concentration also reflects adrenergic glycolysis, tissue release, distribution, and clearance. Hepatic hypoperfusion can reduce clearance. A falling concentration is encouraging in context yet does not prove that bleeding has stopped or that every organ has recovered. Base deficit similarly combines several metabolic and fluid-related influences rather than measuring blood volume directly.

Restoring flow makes oxygen and substrates available again, but recovery can be uneven. Endothelial disturbance, altered microvascular distribution, edema, cellular calcium handling, and oxidant generation can persist or evolve during reperfusion. Tissue injury may reflect both the preceding supply limitation and events during restoration. These mechanisms do not imply that restoring perfusion should be withheld.

Blood leaving previously underperfused regions can also change systemic acid–base and electrolyte measurements. Renal and hepatic dysfunction modify subsequent handling. A satisfactory arterial pressure after source control therefore describes an important circulatory achievement without establishing immediate reversal of all cellular injury.

The useful distinction is between renewed macroscopic support and biological recovery. Serial organ function, gas exchange, acid–base status, temperature, and perfusion findings help describe their relationship. Persistent dysfunction may reflect incomplete perfusion, established injury, another shock mechanism, or several processes together; the pressure trace alone cannot separate them.

74.10 Hemostasis, Temperature, and Calcium Form Interacting Feedbacks

Hemorrhage removes platelets and coagulation proteins as well as red cells and plasma. Subsequent dilution can further alter their concentrations. However, traumatic coagulopathy is not simply the late result of replacing blood with cell-free fluid. Brohi and colleagues documented coagulation abnormalities early after injury, supporting a contribution from injury-associated biology before extensive hospital resuscitation. Their observational study does not establish one exclusive molecular pathway. 7

Endothelial signaling, fibrin formation, platelet function, fibrinolysis, and the local wound environment interact. The direction and degree of fibrinolytic disturbance vary. A single screening test cannot fully describe clot strength, platelet behavior, local bleeding, or the subsequent trajectory. The detailed hemostatic mechanisms are developed in Chapter 67.

Cooling can reduce coagulation-enzyme activity and impair cellular hemostasis. Exposure, reduced heat production, and cooler infusions can contribute to heat loss or falling temperature during hemorrhage. Rohrer and Natale demonstrated temperature-dependent changes in coagulation tests using normal plasma in vitro. This model supports a temperature effect but cannot specify an individual patient's bleeding risk. Tests performed after warming a specimen to the laboratory's standard temperature may not reproduce the patient's in-vivo thermal state. 8

Acidemia, hypoperfusion, and low temperature can coexist and affect cardiovascular and hemostatic function through different mechanisms. Their combination should not be reduced to a fixed sequence or one numerical boundary. Warming changes demand, vascular distribution, and electrolyte handling as well as enzyme kinetics.

Ionized calcium supports excitation–contraction coupling and coagulation-complex assembly. Citrate delivered with blood products can bind calcium, with the circulating effect influenced by delivery and metabolism. Total calcium is not interchangeable with the ionized fraction, which also varies with protein binding and pH.

Magnotti and colleagues prospectively associated admission ionized calcium with transfusion requirements and outcomes in critically ill trauma patients. That association identifies a physiological signal, not proof that a particular replacement strategy improves survival. 9 Loss, dilution, temperature, acid–base state, and treatment exposure therefore need to be interpreted together.

74.11 Controlling the Source and Supporting Circulation Address Different Problems

Source control limits further escape of circulating blood. Circulatory support attempts to maintain useful flow and oxygen transport while that problem is addressed. A rise in measured pressure can occur without stopping hemorrhage, and an initially modest pressure does not quantify whether the source is mechanically controlled.

Pressure across an open vascular injury contributes to bleeding, but the relationship also depends on wound geometry, local resistance, surrounding pressure, and clot stability. At the same time, organ perfusion depends on adequate driving pressure. These competing mechanisms cannot be converted into one universal pressure prescription, particularly when brain injury, coronary disease, or another perfusion-sensitive condition coexists.

Cell-free volume, red-cell-containing products, and hemostatic components address overlapping but different deficits. Increasing filling can improve output, replacing red cells can increase content, and hemostatic support can affect clot formation. Their physiological roles do not establish an optimal product strategy or demonstrate that a higher pressure or concentration has improved survival.

In TOWAR, 44 air-medical bases used cluster-randomized monthly whole-blood or component strategies. Among 1020 eligible assigned patients, 993 entered the primary analysis. Thirty-day mortality was 25.9% versus 20.5%; the adjusted odds ratio was 1.24, with a 95% confidence interval of 0.87–1.76. Whole blood did not lower the primary mortality outcome in this comparison. 10

SWiFT individually randomized 942 patients across English air-ambulance services. After exclusion of nontraumatic hemorrhage and traumatic cardiac arrest, 616 entered the analysis. Death or massive transfusion within 24 hours occurred in 48.7% with whole blood and 47.7% with components; the relative risk was 1.02, with a 95% confidence interval of 0.80–1.31. Superiority was not demonstrated. 11

The trials used different designs and primary endpoints. Both compared active prehospital blood-product strategies; neither compared blood with no blood or established universal product equivalence. Their results constrain strategy claims without changing the distinct physiological needs for source control, circulation, oxygen carriage, and hemostasis.

74.12 Serial Interpretation Connects Volume, Composition, Flow, and Recovery

Return to the constructed circulation. Immediately after loss, hematocrit remains 45% despite a 20% reduction in red-cell volume. Cardiac output has fallen from 5 to 3 L/min, and calculated oxygen delivery has fallen from 1005 to 603 mL/min. The early concentration looks unchanged because both red cells and plasma left together.

After the stipulated cell-free entry, total blood volume returns to its starting value, hematocrit falls to 36%, and hemoglobin to 12 g/dL. Output rises to 3.75 L/min, but calculated delivery remains 603 mL/min. Filling, concentration, and flow have changed in different directions. None of these observations demonstrates that the source has stopped bleeding.

Suppose arterial pressure also rises during this interval. Improved filling, greater vascular resistance, altered cardiac function, or several changes could explain it. Pressure alone cannot identify the contribution of each. A later fall in lactate would add information about production and clearance over time, while temperature and demand could have changed simultaneously.

Conversely, a subsequent lower hemoglobin concentration could reflect further hemorrhage, additional cell-free entry, sampling differences, or combinations. A rising concentration could reflect plasma loss rather than restored red-cell mass. The account becomes clearer when the source, timing, volume exposure, blood products, and laboratory sampling are placed on the same timeline.

A transient improvement after a circulatory intervention describes responsiveness under those conditions. It does not prove durable recovery, identify the amount still missing, or exclude an obstructive, distributive, or cardiac contribution. Repeated deterioration can reflect ongoing loss or changing physiology even when an earlier response was substantial.

The integrated assessment therefore follows several linked quantities: whether loss continues, what supports venous return, what forward flow is achieved, what oxygen each unit of blood carries, and how tissues are functioning at their current demand. Trends help connect these questions while retaining the limits of each measurement. Hemorrhagic shock is an evolving system of losses and compensations rather than a single pressure or laboratory value.

References

1. Guly HR, Bouamra O, Spiers M, et al. Vital signs and estimated blood loss in patients with major trauma: testing the validity of the ATLS classification of hypovolaemic shock. Resuscitation. 2011;82:556–559. doi:10.1016/j.resuscitation.2011.01.013.

2. Watso JC, Belval LN, Cimino FA III, et al. Low-dose morphine reduces tolerance to central hypovolemia in healthy adults without affecting muscle sympathetic outflow. Am J Physiol Heart Circ Physiol. 2022;323:H89–H99. doi:10.1152/ajpheart.00091.2022.

3. Ley EJ, Salim A, Kohanzadeh S, Mirocha J, Margulies DR. Relative bradycardia in hypotensive trauma patients: a reappraisal. J Trauma. 2009;67:1051–1054. doi:10.1097/TA.0b013e3181bba222.

4. Pottgiesser T, Specker W, Umhau M, et al. Recovery of hemoglobin mass after blood donation. Transfusion. 2008;48:1390–1397. doi:10.1111/j.1537-2995.2008.01719.x.

5. Wachsmuth NB, Haupt S, Bauer P, et al. Impact of a single blood donation on hemoglobin mass, iron stores, and maximum oxygen uptake in premenopausal women—a pilot study. Transfusion. 2024;64:1481–1491. doi:10.1111/trf.17917.

6. Weiskopf RB, Viele MK, Feiner J, et al. Human cardiovascular and metabolic response to acute, severe isovolemic anemia. JAMA. 1998;279:217–221. doi:10.1001/jama.279.3.217. Table correction, JAMA. 1998;280:1404.

7. Brohi K, Singh J, Heron M, Coats T. Acute traumatic coagulopathy. J Trauma. 2003;54:1127–1130. doi:10.1097/01.TA.0000069184.82147.06.

8. Rohrer MJ, Natale AM. Effect of hypothermia on the coagulation cascade. Crit Care Med. 1992;20:1402–1405. doi:10.1097/00003246-199210000-00007.

9. Magnotti LJ, Bradburn EH, Webb DL, et al. Admission ionized calcium levels predict the need for multiple transfusions: a prospective study of 591 critically ill trauma patients. J Trauma. 2011;70:391–395. doi:10.1097/TA.0b013e31820b5d98.

10. Sperry JL, Guyette FX, Cotton BA, et al. Prehospital resuscitation with type O whole blood for trauma and hemorrhage. N Engl J Med. 2026;394:2317–2328. doi:10.1056/NEJMoa2602167.

11. Smith JE, Cardigan R, Sanderson E, et al. Prehospital whole blood in traumatic hemorrhage—a randomized controlled trial. N Engl J Med. 2026;394:2305–2316. doi:10.1056/NEJMoa2516043.