C HAPTER 139 
DISSEMINATED INTRAVASCULAR COAGULATION
Marcel Levi
A variety of disorders, including infectious or inflammatory condi-
tions and malignant disease, will lead to activation of coagulation. In
many cases, this activation of coagulation will not lead to clinical
complications and will not even be detected by routine laboratory
tests, but can only be measured with sensitive molecular markers for
activation of coagulation factors and pathways.1,2 However, if the
stimulus for activation of coagulation is sufficiently strong, the
platelet count may decrease and global clotting times may become
prolonged. In its most extreme form, systemic activation of coagula-
tion is known as disseminated intravascular coagulation (DIC). DIC
is characterized by the simultaneous occurrence of widespread (micro)
vascular thrombosis, thereby compromising blood supply to various
organs, which may contribute to organ failure.3,4 Because of ongoing
activation of the coagulation system and other factors, such as
impaired synthesis and increased degradation of coagulation proteins
and protease inhibitors, consumption of clotting factors and platelets
may occur, resulting in bleeding from various sites.
In view of the multiple, often contrasting mechanisms that
occur in patients with DIC, a consensual definition of DIC had
been a matter of debate. In 2001 the subcommittee on DIC of the
International Society on Thrombosis and Hemostasis proposed a
definition that reflects the central role of the microvascular milieu,
i.e., endothelial cells, blood cells, and the plasma protease system, in
the pathogenesis of DIC. This definition of DIC reads as follows:
“DIC is an acquired syndrome characterized by the intravascular
activation of coagulation without a specific localization and arising
from different causes. It can originate from and cause damage to
the microvasculature, which if sufficiently severe, can produce organ
dysfunction.”5
The diagnosis of DIC may be hampered by the nonspecific nature
of many indicators of coagulation activation, although newly devel-
oped scoring algorithms based on readily available routine laboratory
parameters show promising diagnostic accuracy.5 Owing to the
complexity of the clinical presentation, the variable and unpredictable
course, and the multitude of therapies given to patients with DIC,
properly conducted clinical trials are difficult to perform and even to
devise. Management relies on limited evidence from clinical trials in
combination with small studies employing surrogate outcome end-
points and experience from case series, as well as from an understand-
ing of the underlying pathophysiologic mechanisms.6
EPIDEMIOLOGY
Activation of coagulation in concert with inflammatory activation
can result in microvascular thrombosis, which contributes to multiple
organ failure in patients with severe sepsis.7 In support of this concept,
postmortem findings in patients with coagulation abnormalities and
DIC on the background of severe sepsis include diffuse bleeding,
hemorrhagic necrosis of tissues, microthrombi in small blood vessels,
and thrombi in midsize and larger arteries and veins. Ischemia and
necrosis were invariably the result of fibrin deposition in small and
midsize vessels. Importantly, intravascular thrombi appear to be the
driver of the organ dysfunction. Fibrin deposition in various organs
also is a characteristic finding in animal models of DIC. Thus experi-
mental bacteremia or endotoxemia causes intra- and extravascular
fibrin deposition in the kidneys, lungs, liver, brain, and other organs.
Amelioration of the hemostatic defect with various interventions
reduces fibrin deposition, improves organ function in these models,
2064
and, in some cases, reduces mortality. Finally, results of clinical studies
also support the concept that activation of coagulation is an impor-
tant determinant of clinical outcome. DIC has been shown to be an
independent predictor of organ failure and mortality.8 In a consecu-
tive series of patients with severe sepsis, 43% of the patients with
DIC were compared with the 27% without DIC. In that study, the
severity of the coagulopathy was directly related to mortality.9
In addition to microvascular thrombosis and organ dysfunction,
coagulation abnormalities may have other harmful consequences.
Thrombocytopenia in patients with sepsis places them at risk of
bleeding. For example, critically ill patients with a platelet count of
<50 × 109/L have a four- to fivefold higher risk for bleeding than
those with higher platelet counts.10 Although the overall risk of
intracerebral bleeding in intensive care unit (ICU) patients is less than
0.5%, up to 88% of patients with this complication have platelet
counts below 100 × 109/L. The use of anticoagulants in patients with
thrombocytopenia further increases the risk of bleeding. Regardless
of the cause, multivariate analyses indicate that thrombocytopenia is
an independent predictor of ICU mortality and increases the risk of
death by 1.9- to 4.2-fold. In particular, thrombocytopenia that per-
sists for more than 4 days after ICU admission, or a 50% or greater
decrease in platelet count during the ICU stay, is associated with a
four- to sixfold increase in mortality. In fact the platelet count appears
to be a stronger predictor of ICU mortality than composite scoring
systems, such as the Acute Physiology and Chronic Evaluation
(APACHE) II or Multiple Organ Dysfunction Score (MODS).
Decreased levels of coagulation factors, as reflected by prolonged
global coagulation times, also increase the risk of bleeding. Prolonga-
tion of the prothrombin time (PT) or activated partial thromboplastin
time (aPTT) to over 1.5 times the control is associated with an
increased risk of bleeding and mortality in critically ill patients.
PATHOBIOLOGY
Traditionally, DIC was thought to be the result of activation of both
the extrinsic and intrinsic pathways of coagulation. The classical
concept was that the extrinsic pathway was initiated by a tissue-
derived component, which activated factor VII leading to the direct
conversion of prothrombin to thrombin. This process would proceed
as long as there was tissue damage from systemic infection, trauma,
circulating placental components, or malignancy. In contrast, the
intrinsic or contact pathway of coagulation was initiated by contact
activation of factor XII which, together with its cofactors, kallikrein
and high molecular weight kininogen, then activated factor XI
leading to subsequent activation of factor IX. Until recently, the
initiators of contact activation were thought to include collagen and
artificial surfaces. In recent years the molecular mechanisms of coagu-
lation pathway activation have been better defined (Fig. 139.1),
thereby providing new insight into the pathogenesis of DIC. In
general, current thinking is that thrombin and fibrin generation in
patients with DIC is largely driven via the extrinsic pathway; the role
of the contact system is uncertain.
Tissue Factor-Factor VII(a) Pathway
The extrinsic pathway is initiated by the tissue factor (TF)–factor
VIIa complex. TF is a membrane-bound 4.5-kDa protein that is

Endothelial
cells
Cytokines Inflammatory cells
Tissue factor
expression
Impairment of
physiologic
anticoagulant
mechanisms
Inhibition of
fibrinolysis
due to high
levels of PAI-1
MICROVASCULAR THROMBOSIS AND
MODULATION OF INFLAMMATION
Fig. 139.1  PATHOGENESIS OF DISSEMINATED INTRAVASCULAR
COAGULATION (DIC). Pathways involved in the activation of coagulation
in DIC. Both perturbed endothelial cells and activated mononuclear cells
may produce proinflammatory cytokines that induce tissue factor expression,
thereby initiating coagulation. In addition, downregulation of physiologic
anticoagulant mechanisms and inhibition of fibrinolysis promotes intravas-
cular fibrin deposition. PAI-1, plasminogen activator inhibitor, type 1.
constitutively expressed on cells that are mostly in tissues and not in
direct contact with blood, such as the adventitial layer of larger blood
vessels.12 Subcutaneous tissue also contains substantial amounts of
TF. When expressed on the cell surface, TF interacts with factor VII,
either in its zymogen or activated form. The TF–factor VIIa complex
catalyzes the activation of both factor IX and factor X. Factors IXa
and Xa enhance the activation of factors X and prothrombin, respec-
tively. In cells in contact with the blood, TF is induced by the action
of mediators such as cytokines, C-reactive protein and advanced
glycosylation end products. Inducible TF is predominantly expressed
by monocytes and macrophages. Monocyte TF expression is enhanced
in the presence of platelets and granulocytes in a P-selectin dependent
fashion. This may reflect nuclear factor kappa-B (NFκB) activation
that occurs when activated platelets bind to neutrophils or mono-
nuclear cells. These cell-cell interactions also stimulate the production
of interleukin (IL)-1b, IL-8, MCP-1, and tumor necrosis factor
(TNF)-α. Under cell culture conditions, cytokines such as TNF-α,
and IL-1 can induce TF expression by vascular endothelial cells, but
the in vivo relevance of this finding is uncertain. Studies in vivo
suggest that IL-6 is the dominant mediator of TF expression by
mononuclear cells.
Increased monocyte TF expression and procoagulant activity has
been demonstrated in DIC associated with sepsis, cancer, or coronary
disease. Tissue expression of TF appears to be localized to certain
organs and vascular beds, but it is uncertain whether its expression is
under genetic control in an organ-specific fashion. With trauma, such
as extensive surgery, brain injury, or burns, it is likely that constitu-
tively expressed TF at the site of injury is the primary source of
procoagulant material, but direct support for this concept is lacking.
The Intrinsic Pathway
The role of the intrinsic pathway in the pathogenesis of DIC is
uncertain. Negatively charged substances, such as phospholipids,
Chapter 139  Disseminated Intravascular Coagulation 2065
polyphosphates, and glycosaminoglycans, are potential activators of
the contact pathway. Studies in patients with suspected DIC have
identified elevated levels of markers of activation of the contact
system. In patients with meningococcal septicemia, there was a nega-
tive correlation between plasma factor XII levels and factor XIIa–C1
inhibitor complexes. Although this finding implies consumption of
factor XII, and subsequent downstream activation of factor XI, an
alternative explanation is that there is a negative acute phase effect
with reduced synthesis of factor XII, in conjunction with thrombin-
mediated activation of factor XI.
However, blockade of the contact system with a factor XIIa–
directed antibody failed to prevent DIC in a balloon model of
Escherichia coli sepsis, but diminished development of lethal hypoten-
sion. These findings provide reasonable support for the current view
that the contact pathway does not contribute to DIC, but may play
important roles in proinflammatory mechanisms related to vascular
permeability, vascular proliferation (kininogen induces smooth
muscle cell proliferation), and enhancement of fibrinolysis.11
Cytokines and Other Amplification Pathways
Activation of blood coagulation requires several cofactors. For devel-
opment of DIC, the surfaces of cell remnants or intact cells, inflam-
matory mediators, and coagulation proteins are required. The
stimulus for activation depends on the underlying disease and may
encompass bacterial cell compounds, such as endotoxin, TF on host
or cancer cells, other cancer cell procoagulants, fat, or amniotic fluid
emboli by unknown pathways. Each of these triggers interacts with
other mediators: TF assembles on anionic phospholipid surfaces,
which can be provided by activated platelets, leukocytes, or cancer
cells; cytokines interact with receptors and induce signaling pathways
that induce TF expression and other proinflammatory components
via the NFκB complex.
Endotoxin is a lipopolysaccharide compound of gram-negative
bacteria that induces the sepsis syndrome and DIC. Gram-negative
bacteria liberate endotoxins from their membrane, which interact
with cell surfaces via various pathways. In blood, endotoxin directly
binds to CD14 on monocytes, and binds to endothelial cells after
complexing with lipopolysaccharide binding protein (LBP) and the
Toll-like receptor 4 (TLR 4) complex. Through these interactions,
endotoxin induces signaling pathways that culminate in NFκB
activation, and initiates the expression of proinflammatory cytokines
and TF. Likewise, exotoxins, such as lipoteichoic acid (LTA) from
gram-positive bacteria can also induce proinflammatory cytokine
expression.
The molecular mechanisms underlying endotoxin-induced activa-
tion of coagulation have been studied in nonhuman primates. In
endotoxin or E. coli models of sepsis, inhibition of the TF pathway
abolishes the activation of coagulation, highlighting the importance
of TF. IL-6 is an important mediator of procoagulant effects, whereas
TNF-α is involved in the fibrinolytic response to endotoxin. Inhibi-
tion of TF with tissue factor pathway inhibitor (TFPI) reduces IL-6
levels in the baboon model, suggesting that there is extensive crosstalk
between coagulation and inflammatory mediators (see later). Mono-
cytes that express TF bind factor VII(a), shed TF, or bind to the
damaged vessel wall. After interacting with platelets, circulating
monocytes can trigger DIC. Microvesicles may accelerate this process,
and the complex interaction between cells, membrane fragments,
soluble mediators, and proteins may trigger the DIC syndrome. The
severity and duration of the consumptive process are mainly deter-
mined by the potency of the triggers and the capacity of inhibitory
mechanisms.
Cross Talk Among Coagulation Proteases Results in
Proinflammatory Effects
In addition to activating coagulation protein zymogens, coagulation
proteases also interact with specific cell receptors and trigger signaling
 2066 Part XII  Hemostasis and Thrombosis
Thrombomodulin +
thrombin
Endothelial
protein C
receptor
Endothelial
cell apoptosis
–
TFPI
–
Glycosaminoglycans
Antithrombin
Fig. 139.2  PHYSIOLOGIC ANTICOAGULANT MECHANISMS IN DISSEMINATED INTRAVASCU-
LAR COAGULATION. Physiologic anticoagulant mechanisms (activated protein C system, tissue factor
pathway inhibitor (TFPI), and antithrombin) are not only involved in blocking thrombin generation and
thrombin activity but also affect inflammatory pathways.
pathways that elicit proinflammatory mediators.7 Factor Xa, throm-
bin, and the factor VIIa–TF complex have such effects. Factor Xa
injection into rats induces localized inflammation, probably as a
result of its interaction with EPR-1 and not because of thrombin
generation. Exposure of cultured endothelial cells to factor Xa stimu-
lates the production of monocyte chemotactic protein 1 (MCP-1),
IL-6 and IL-8, and upregulates the expression of adhesion proteins
that tether neutrophils to the cell surface. Further evidence for the
crosstalk between inflammation and coagulation comes from the
observations that IL-6 and IL-8 elicit TF-dependent procoagulant
activity in monocytes and the fact that IL-6 has been identified as
the critical mediator of procoagulant activity either on its own or after
endotoxin challenge in vivo. Therefore cytokine production induced
by factor Xa may be an important driver of coagulation in DIC.
In addition to its procoagulant functions, thrombin has a variety
of noncoagulant effects. Thrombin induces the release of MCP-1 and
IL-6 from fibroblasts, epithelial cells, and mononuclear cells in vitro.
Thrombin also induces IL-6 and IL-8 production in endothelial cells.
When generated in whole blood, IL-8 production has a procoagulant
effect that is TF dependent. Cell activation by thrombin is likely
mediated by protease activated receptors (PARs). The factor VIIa–TF
complex also activates cells by binding and activating PAR 2.
Direct evidence of the in vivo relevance of these phenomena
comes from a study showing that recombinant factor VIIa infusion
in volunteers induces an increase in plasma levels of IL-6 and IL-8.
Although the concentrations of factor VIIa infused far exceed those
found in patients with sepsis, it is possible that factor VIIa-induced
cytokine production is of physiologic importance. Therefore this
information adds to the concept that several coagulation proteases
induce proinflammatory mediators that augment procoagulant
Protein C
Activated
protein C
–
Factor Va
Factor VIIIa
Blocking
NFκB
translocation
Tissue factor –
Decreased
– TNF-α and
IL production
in monocytes
Thrombin
Factor Xa
Antiinflammatory
– cytokines (IL-10)
Increased TNF-α
activity and amplify the consumptive process. Endogenous antico-
agulant pathways are essential to regulate these proteases and to
prevent uncontrolled DIC.
Endogenous Anticoagulant Pathways in Disseminated
Intravascular Coagulation
The development of DIC is counteracted by several mechanisms.
First, coagulation inhibitors regulate the coagulation mechanism.
Those inhibitors include antithrombin (AT), the protein C system,
and TFPI (Fig. 139.2).7 AT, which complexes and inhibits thrombin
and factor Xa, is one of the most important inhibitors and reduced
AT levels are a characteristic of DIC. Reduction in AT levels reflects
a combination of reduced protein synthesis as well as increased clear-
ance through the formation of protease–AT complexes, and by deg-
radation by neutrophil elastase. In addition, cytokines may impair
proteoglycan synthesis in the vessel wall thereby reducing the avail-
ability of heparan sulfate for potentiation of AT activity.
In animal models of experimental bacteremia, AT concentrate
infusion increases survival, reduces the severity of DIC, and lowers
the levels of IL-6 and IL-8. Therefore in addition to its anticoagulant
function, AT may also have an antiinflammatory effect.
Activated protein C and its cofactor protein S provide a second
line of defense. Thrombin binds to thrombomodulin on the endo-
thelial cell membrane and the thrombin–thrombomodulin complex
converts protein C to its active form, activated protein C (APC).12
In addition the thrombin–thrombomodulin complex converts the
latent carboxypeptidase B–like enzyme thrombin activatable fibrino-
lytic inhibitor (TAFI) to its activated form. APC inactivates factors

Va and VIIIa by proteolytic cleavage, thereby downregulating the
coagulation cascade. Endothelial cells, primarily of large blood vessels,
express endothelial protein C receptor (EPCR) on their surface.
EPCR augments protein C activation by binding protein C and
presenting it to the thrombin–thrombomodulin complex on the cell
surface. APC has antiinflammatory effects on mononuclear cells and
granulocytes, which may be distinct from its anticoagulant activity.
Administration of APC prevented thrombin-induced thromboembo-
lism in mice, mainly through its antithrombotic effect.
Defects in the protein C mechanism enhance the vulnerability to
inflammatory reactions and DIC. In patients, reduced levels of
protein C and protein S are associated with increased mortality. Mice
with a one-allele targeted disruption of the protein C gene, causing
heterozygous protein C deficiency, developed a more severe form of
DIC and associated inflammatory response compared with their
wild-type counterparts. Blockade of protein C activity by infusion of
C4 binding protein converted a sublethal model of E. coli in baboons
into a lethal model. In addition, blockade of EPCR with a neutral-
izing monoclonal antibody also increased mortality in the E. coli
baboon model, whereas infusion of PC protected against DIC and
lethality. Therefore the protein C pathway is important in the host
defense against sepsis and DIC. In situations associated with DIC
and systemic inflammation, cell culture experiments suggest that
TNF-α and IL-1 may downregulate thrombomodulin expression.
However, in vivo studies suggest that EPCR is upregulated in sepsis;
an effect mediated by thrombin. The generation of thrombin may
also trigger EPCR shedding due to the activation of metalloprotein-
ases by thrombin. It is presently unknown whether thrombomodulin
is also cleaved by similar mechanisms.
TFPI provides a third inhibitory mechanism. TFPI exists in
several pools including endothelial cell associated and lipoprotein
bound in plasma. It inhibits the TF–factor VIIa complex by forming
a quaternary complex in which factor Xa is the fourth component.
Clinical studies in patients with sepsis have not provided clues as to
its importance because in the majority of patients the TFPI levels are
normal. This may be explained by the lack of downregulatory effects
of inflammatory mediators on cultured endothelial cells. The relevance
of TFPI in DIC is illustrated by two lines of experimentation. First,
depletion of TFPI sensitizes rabbits to DIC induced with tissue factor
infusion. Second, TFPI infusion protects against the harmful effects
of E. coli in primates. TFPI not only blocks DIC in baboons given
lethal doses of E. coli, but improves vital functions and survival. A
study in human volunteers confirmed the potential of TFPI to block
the procoagulant effects of endotoxin.
In general, normal levels and function of inhibitors is important
in the defense against DIC. It should be noted, however, that there
are no strong indications that patients with congenital deficiencies of
inhibitors are at increased risk of DIC, but this issue requires greater
exploration. In addition, the capacity of inhibitors to modify the
interaction between coagulation and inflammation deserves further
attention.
Fibrinolysis
In experimental models of DIC, fibrinolysis is activated, demon-
strated by an initial activation of plasminogen, followed by impair-
ment caused by the release of type 1 plasminogen activator inhibitor
(PAI-1). This results in a net procoagulant state. The molecular basis
for this procoagulant state is cytokine-mediated activation of vascular
endothelial cells, thus TNF-α and IL-1 decrease tPA production and
increase PAI-1 production. Although TNF-α increases urokinase
type plasminogen activator (uPA) production by endothelial cells,
endotoxin and TNF-α stimulate PAI-1 production in the liver,
kidney, lung, and adrenals of mice.
The net procoagulant state is manifested by a late rise in fibrin
breakdown fragments after E. coli challenge in baboons. Experimental
data also suggest that the fibrinolytic mechanism is active in clearing
fibrin from organs and the circulation. Endotoxin-induced fibrin
formation in the kidneys and adrenals is most dependent on decreased
Chapter 139  Disseminated Intravascular Coagulation 2067
levels of uPA. In contrast to wild-type mice challenged with endo-
toxin, PAI-1 knockout mice do not develop thrombi in the kidney.
Endotoxin administration to mice with a functionally inactive
thrombomodulin gene (TMProArg mutation) and defective protein
C activator cofactor function caused fibrin plugs in the pulmonary
circulation, whereas wild-type mice did not develop macroscopic
fibrin. This phenomenon in thrombomodulin deficient mice proved
to be temporary; thrombi developed 4 hours after endotoxin admin-
istration and disappeared at 24 hours in animals killed at that time
point. These experiments demonstrate that the fibrinolytic system is
essential for clearance of intravascular fibrin.
Fibrinolytic activity is regulated by PAI-1, the principal inhibitor
of this system. Previous studies have shown that a functional muta-
tion in the PAI-1 gene, the 4G/5G polymorphism, not only influ-
ences plasma levels of PAI-1, but also affects the clinical outcome of
meningococcal septicemia. Patients with the 4G/4G genotype had
significantly higher PAI-1 concentrations in plasma and an increased
risk of death. Further investigations demonstrated that the PAI-1
polymorphism did not influence the risk of contracting meningitis,
but probably increased the likelihood of developing septic shock from
meningococcal infection. These studies provide the first evidence that
genetically determined differences in the level of fibrinolysis influence
the risk of developing complications of gram-negative infection. In
other clinical studies in cohorts of patients with DIC, high plasma
levels of PAI-1 were one of the best predictors of mortality. These
data suggest that DIC contributes to mortality in this situation, but
as indicated earlier, because PAI-1 is an acute-phase protein, higher
plasma levels may also be a marker of disease rather than a causal
factor.
CLINICAL MANIFESTATIONS
The clinical manifestations of DIC vary depending on the underlying
disorder. At the extreme end of the spectrum is acute, severe DIC,
which often occurs in the setting of sepsis, major trauma, obstetric
calamities, and severe immunologic responses. Diffuse multiorgan
bleeding, hemorrhagic necrosis, microthrombi in small blood vessels,
and thrombi in medium and large blood vessels are common find-
ings at autopsy, although patients with unequivocal clinical and
laboratory signs of DIC may not have confirmatory postmortem
findings. Conversely, some patients in whom clinical and laboratory
signs were not consistent with DIC have typical autopsy findings.
Although this occasional lack of correlation among clinical, labora-
tory, and pathologic findings can partly be explained by excessive
fibrinolysis post mortem, this does not account for all cases. The
organs most frequently involved by diffuse microthrombi are the
lungs and kidneys, followed by the brain, heart, liver, spleen, adrenal
glands, pancreas, and gut. Specific immunohistological techniques
and ultrastructural analysis have revealed that most thrombi consist
of fibrin monomers or polymers in combination with platelets. In
addition, involvement of activated mononuclear cells and other
signs of inflammation are frequently present. In cases of long-lasting
DIC, organization and endothelialization of the microthrombi are
often observed. Acute tubular necrosis is more frequent than renal
cortical necrosis. Clinically, thrombotic occlusive events occur first
as a consequence of microvascular obstruction by microthrombi
consisting of fibrin or platelets. These thrombi result from clots that
form either in the circulation or in situ in arterioles, capillaries, or
venules. Circulatory obstruction reduces organ perfusion and may
lead to ischemia, infarction, and necrosis. The process is disseminated
throughout the microcirculation, therefore all organs are potentially
vulnerable.
In contrast to acutely ill patients with severe DIC, others may
have mild or protracted clinical manifestations of consumption or
even subclinical disease manifested only by laboratory abnormalities.1
The clinical picture of subacute to chronic DIC generally occurs in
patients with malignancy, in particular those with mucin-producing
adenocarcinomas or acute promyelocytic leukemia (APL). The latter
usually is dominated by a hemorrhagic presentation, whereas venous
 2068 Part XII  Hemostasis and Thrombosis
Clinical Conditions Most Frequently Complicated by
BOX 139.1
Disseminated Intravascular Coagulation
• Sepsis/severe infection
• Trauma/burn/heatstroke
• Malignancy
Solid tumors
Acute leukemia
• Obstetric conditions
Amniotic fluid embolism
Abruptio placentae
HELLP syndrome
• Vascular abnormalities
Kasabach-Merrit Syndrome
Other vascular malformations
Aortic aneurysms
• Severe allergic/toxic reactions
• Severe immunologic reactions (e.g., transfusion reaction)
thrombotic manifestations are more common in the former. In addi-
tion, patients with solid tumors may develop nonbacterial thrombotic
endocarditis with systemic arterial embolization and infarction.
Another cause of subacute to chronic DIC is the retained dead fetus
syndrome. These patients have an extremely variable presentation
ranging from asymptomatic to mild or moderate skin and mucous
membrane bleeding.
It is important to stress that DIC is not a disease in itself but is
always secondary to an underlying disorder, which causes the activa-
tion of coagulation. The underlying disorders most commonly known
to be associated with DIC are listed in Box 139.1 and are described
in detail below.
Disseminated Intravascular Coagulation in
Infectious Disease
Systemic infections are among the most common causes of DIC.
Immunocompromised patients, asplenic patients whose ability to
clear bacteria (particularly pneumococci) is impaired, and newborns
whose anticoagulant systems are immature are particularly prone to
infection-induced DIC. Infections may be superimposed on trauma
or malignancies, which themselves are potential triggers of DIC. In
addition, infections can aggravate bleeding and thrombosis by directly
inducing thrombocytopenia, hepatic dysfunction, and shock, which
can lead to diminished blood flow in the microcirculation.
Clinically overt DIC occurs in 30%–50% of patients with gram-
negative or gram-positive sepsis. Extreme examples of sepsis-related
DIC are streptococcus A toxic shock syndrome, which is character-
ized by deep tissue infection, vascular collapse, vascular leakage, and
multiple organ dysfunction. M protein released from streptococci
forms complexes with fibrinogen that bind to β2 integrins on neu-
trophils leading to their activation, and meningococcemia, a fulmi-
nant gram-negative infection characterized by extensive hemorrhagic
necrosis, DIC, and shock. More frequent gram-negative infections
associated with DIC are caused by Pseudomonas aeruginosa, E. coli,
and Proteus vulgaris. Patients with these infections may only have
laboratory evidence of activated coagulation or they may present with
severe DIC.
Activation of the coagulation system has also been documented
with nonbacterial pathogens, such as viruses, protozoa (malaria), and
fungi.13 Common viral infections, such as influenza, varicella, rubella,
and rubeola, are rarely associated with DIC. Some viral infections
can cause hemorrhagic fever characterized by fever, hypotension,
bleeding, and renal failure. Laboratory evidence of DIC can accom-
pany Korean, rift valley, and dengue-related hemorrhagic fevers.
Protozoan infections, such as cerebral malaria, may be associated with
overt DIC. In these cases, secondary deficiency of a disintegrin-like
metalloprotease with thrombospondin type 1 repeats (ADAMTS13),
the von Willebrand cleaving protease, may occur. Such deficiency
may also complicate other types of sepsis-induced DIC and may
enhance platelet–vessel wall interactions, thereby contributing to
microvascular thrombosis.14
Purpura fulminans is an extreme form of DIC, which is often
lethal. This disorder is characterized by extensive hemorrhagic necro-
sis of the skin over the extremities and buttocks. The disease predomi-
nantly affects infants and children and is rare in adults. Diffuse
microthrombi in small blood vessels leads to necrosis and vasculitis
may also be found in biopsies of skin lesions. The disorder can occur
2 to 4 weeks after mild infection, such as scarlet fever, varicella, or
rubella, or can occur during an acute viral or bacterial infection in
patients with acquired or hereditary deficiencies of protein C or
protein S. The syndrome mimics neonatal homozygous protein C or
protein S deficiency where purpura fulminans, with or without
extensive thrombosis, develops soon after birth.
Disseminated Intravascular Coagulation in Trauma,
Brain Injury, Burns, and Heat Stroke
The time interval between trauma and medical intervention correlates
with the development and magnitude of DIC. Experience during
wars proved that fast evacuation and prompt medical care reduce the
risk of DIC. Extensive exposure of TF to the blood and hemorrhagic
shock are the most immediate triggers of DIC in such instances,
although direct proof of this mechanism is lacking. An alternative
hypothesis is that cytokines play a pivotal role in the occurrence of
DIC in trauma patients. In fact, the changes in cytokine levels in
trauma patients are virtually identical to those in patients with
sepsis.15 The levels of TNF-α, IL-1β, PAI-1, circulating TF, neutro-
phil elastase, and soluble thrombomodulin can be elevated in patients
with signs of DIC, predicting multiorgan dysfunction (adult respira-
tory distress syndrome [ARDS] included) and death. Careful moni-
toring for laboratory signs of DIC, reduced fibrinolytic activity, and
perhaps low AT levels may be useful to predict the outcome of such
patients.
In adults and children with head injuries, mortality is high
when DIC occurs. A laboratory DIC score has predictive value for
prognosis in patients with head injuries, thereby supplementing the
Glasgow coma score. Brain injury can be associated with DIC, most
likely because the injury exposes the abundant TF of the brain to
blood. Specimens of contused brain, obtained during surgery in
patients with head injury, and of liver, lungs, kidneys, and pancreas
obtained during autopsy, reveal microthrombi in arterioles and
venules.
Bleeding, laboratory tests indicative of DIC, and vascular micro-
thrombi in biopsies of undamaged skin have been described in
patients with extensive burns. Kinetic studies with labeled fibrinogen
and platelets suggest that in addition to systemic consumption of
hemostatic factors, there is significant local consumption in burned
areas. TF exposed at sites of burned tissue, the systemic inflammatory
response syndrome induced by the burn, and the presence of super-
imposed infections can trigger DIC. Local activation of coagulation
in the bronchoalveolar compartment may contribute to acute lung
injury in these patients.
A severe hemorrhagic diathesis and multiple organ failure indica-
tive of DIC can complicate heat stroke.16 Diffuse fibrin deposition
and hemorrhagic infarctions are found in fatal cases. DIC associated
with profound fibrin(ogen) degradation is evident in such patients.
Potential triggers of DIC in patients with heat stroke include endo-
thelial cell damage and TF released from heat-damaged tissues. In 18
critically ill patients with heat stroke during the 2003 heat wave in
Western Europe that caused numerous deaths in France,16 there were
high levels of IL-6 and IL-8. In addition, there was marked activation
of white blood cells, as evidenced by β2-integrin upregulation and
increased production of reactive oxygen species. All patients had
evidence of systemic activation of coagulation and DIC was present
in about 35%. There was good correlation between the extent of
activation of inflammation and coagulation and the clinical severity
of the heat stroke.

Disseminated Intravascular Coagulation in
Obstetric Complications
Placental abruption is a leading cause of perinatal death.17 Older
multiparous women or patients with one of the hypertensive disorders
of pregnancy are thought to be at highest risk. The severe hemostatic
failure accompanying abruptio placentae is the result of acute DIC
emanating from the introduction of large amounts of TF into the
circulation from the damaged placenta and uterus. Amniotic fluid
has been shown to activate coagulation in vitro, and the degree of
placental separation correlates with the extent of DIC, suggesting that
leakage of thromboplastin-like material from the placental system is
responsible for the DIC. Abruptio placentae occurs in 0.2% to 0.4%
of pregnancies, but only 10% of these cases are associated with DIC.
Different grades of severity are found among those who develop DIC,
with only the more severe forms resulting in shock and fetal death.
Amniotic fluid embolism is a rare but serious complication of
pregnancy and delivery. A maternal mortality rate of 86% was
reported in a 1979 review of 272 cases, but in a more recent
population-based study, the maternal mortality was 26%. Patients
predisposed to amniotic fluid embolism are multiparous women
whose pregnancies are postmature with large fetuses and women
undergoing a tumultuous labor after pharmacologic or surgical
induction. Amniotic fluid is introduced into the maternal circulation
through tears in the chorioamniotic membranes, rupture of the
uterus, and injury of uterine veins. TF in amniotic fluid is the likely
trigger for DIC. Mechanical obstruction of pulmonary blood vessels
by fetal debris, meconium, and other particulate matter in the
amniotic fluid enhances local fibrin–platelet thrombus formation and
fibrinolysis. The extensive occlusion of the pulmonary arteries and
an acute anaphylactoid response reminiscent of severe systemic
inflammatory response syndrome provoke sudden dyspnea, cyanosis,
acute cor pulmonale, left ventricular dysfunction, shock, and convul-
sions. These symptoms are followed within minutes to several hours
by severe bleeding in 37% of patients. Hemorrhage is particularly
severe from the atonic uterus, puncture sites, gastrointestinal tract,
and other organs.
Some studies provide evidence of activation of coagulation, in its
most extreme form DIC, in preeclampsia and eclampsia. In a large
series of patients, a good correlation was noted between the clinical
severity and abnormalities in platelet counts and fibrin(ogen) degra-
dation products. Also consistent with DIC were results of assays of
sensitive parameters of thrombin generation and activation of fibri-
nolysis, such as TAT complexes, D-dimer, and fibrinopeptide Bβ1–42.
Despite these observations, administration of heparin to patients with
preeclampsia and eclampsia has not resulted in convincing benefits.
The HELLP syndrome, which is comprised of hemolysis, elevated
liver enzymes, low platelet count, and severe epigastric pain, is
another complication of pregnancy-induced hypertension. Liver
biopsy findings of fibrin deposition in hepatic blood vessels and labo-
ratory tests consistent with DIC are found in a significant proportion
of patients suggesting that DIC plays a role in the pathogenesis of
the syndrome. However, according to current insights, thrombotic
microangiopathy rather than DIC causes the coagulation abnormali-
ties in patients with hypertensive disorders of pregnancy.
Disseminated Intravascular Coagulation in Malignancy
Patients with solid tumors are vulnerable to risk factors and additional
triggers for DIC that can aggravate thromboembolism and bleed-
ing.18 Risk factors include advanced age, stage of the disease, and use
of chemotherapy or antiestrogen therapy. Triggers include septicemia,
immobilization, and hepatic metastases, which can compromise the
capacity of the liver to control DIC. Microangiopathic hemolytic
anemia frequently is induced by DIC in patients with malignancies
and is particularly severe in patients with widespread intravascular
metastases of mucin-secreting adenocarcinomas. Solid tumor cells
can express different procoagulant molecules including tissue factor,
Chapter 139  Disseminated Intravascular Coagulation 2069
which forms a complex with factor VII(a) to activate factors IX and
X, and a cancer procoagulant (CP), a cysteine protease with factor
X–activating properties. In breast cancer, TF is expressed by vascular
endothelial cells as well as by the tumor cells. TF also appears to be
involved in tumor metastasis and angiogenesis. CP is an endopepti-
dase that can not only be found in extracts of neoplastic cells but also
in the plasma of patients with solid tumors. The exact role of CP in
the pathogenesis of cancer-related DIC is unclear. Interactions of
P- and L-selectins with mucin from mucinous adenocarcinoma can
induce the formation of platelet microthrombi, which probably
constitute a third mechanism of cancer-related thrombosis. Depend-
ing on the rate and quantity of exposure or influx of shed vesicles
from tumors containing TF, a covert or overt DIC develops.
There are numerous reports of DIC and excessive fibrinolysis
complicating the course of acute leukemia. In 161 consecutive
patients presenting with acute myeloid leukemia, DIC was diagnosed
in 52 (32%). In acute lymphoblastic leukemia, DIC was diagnosed
in 15% to 20%.19 Some reports suggest that the incidence of DIC
in acute leukemia patients may increase even more during remission
induction with chemotherapy. In patients with APL, DIC is present
in more than 90% of patients at the time of diagnosis or after initia-
tion of remission induction. The pathogenesis of hemostatic distur-
bance in APL is related to properties of the malignant cells and their
interaction with host endothelial cells. APL cells express TF and CP,
which can initiate coagulation, and they release IL-1β and TNF-α,
which downregulate endothelial thrombomodulin, thereby compro-
mising the protein C anticoagulant pathway. APL cells also express
increased amounts of annexin II, which binds plasminogen and tPA
and promotes plasmin generation. The net result of these processes
is DIC and hyperfibrinolysis, which can lead to bleeding that can be
fatal. All-trans-retinoic acid, which is used for induction and main-
tenance therapy of APL, inhibits the deleterious effect of APL cells
in vitro and in vivo and has reduced the frequency of early hemor-
rhagic death.
Disseminated Intravascular Coagulation With
Vascular Disorders
Rarely, vascular anomalies can trigger DIC. With some large aortic
aneurysms, localized consumption of platelets and fibrinogen can
produce coagulation abnormalities and bleeding.20 In a series of
patients with aortic aneurysms, 40% had elevated levels of fibrin(ogen)
degradation products, but only 4% had laboratory evidence of DIC
or bleeding.20 Factors that predispose such patients to the develop-
ment of DIC include large aneurysms, dissection, and expansion.
Coagulation in these cases likely is triggered by the abundant TF in
atherosclerotic plaques.
Kasabach and Merritt were the first to describe bleeding in associa-
tion with giant cavernous hemangiomas, benign tumors found in
newborns or children that can evolve into convoluted masses of
abnormal vascular channels that sequester and consume platelets and
fibrinogen. Localized pain may occur, likely from thrombosis of these
vascular channels and there may be bleeding after trauma or surgery
because of the consumptive process. Increased fibrinogen consump-
tion reflects hyperfibrinolysis, which may be driven by tPA released
from the abnormal endothelium lining the tumor walls. Patients with
this syndrome exhibit accelerated platelet turnover, and accumulation
of labeled platelets and fibrinogen in the hemangiomas.
Hemangiomas may regress spontaneously; some respond to radia-
tion or laser therapy.
Disseminated Intravascular Coagulation With
Liver Disease
The levels of most coagulation factors, endogenous anticoagulants,
and major components of the fibrinolytic system are reduced in
patients with severe liver disease reflecting reduced synthesis. In addi-
tion, the capacity of the liver to clear factors IXa, Xa, XIa, and tPA
 2070 Part XII  Hemostasis and Thrombosis
is decreased. Thrombocytopenia is common because of hypersplen-
ism and decreased hepatic production of thrombopoietin. The simi-
larities between the hemostatic defects of liver disease and those of
DIC have evoked controversy as to the contribution of DIC to the
coagulopathy of liver disease. Several laboratory and clinical observa-
tions support the concept that DIC accompanies hepatic disorders.
These include a reduced half-life of radiolabeled fibrinogen that is
reversed with heparin administration; failure of replacement therapy
to significantly increase the levels of hemostatic factors (suggesting
ongoing consumption); and increased levels of markers of activation
of coagulation. All of these findings are consistent with increased
thrombin generation. Against the DIC hypothesis are the observa-
tions that microthrombi are found in only 2% of the tissues from
patients who die of liver disease, and the fact that the increased
fibrinogen turnover can be explained by extravascular accumulation.
Current thinking is that DIC is rare in patients with liver disease, but
such patients are sensitive to the triggers of DIC because of the
decreased synthetic capacity of the diseased liver and its inability to
clear activated clotting factors. In patients who undergo peritoneove-
nous shunting for ascites, those with underlying liver disease are more
likely to develop DIC than those with normal liver function.
Disseminated Intravascular Coagulation With Toxic
Reactions or Snake Bites
The venom of certain snakes, particularly vipers and rattlesnakes, can
produce a coagulopathy similar to DIC.21 Prominent among these
species are the Vipera, Echis (E. carinatus or E. coloratus), Aspis,
Crotalus, Bothrops, and Agkistrodon. Venoms of these snakes contain
enzymes or peptides that (1) release fibrinopeptide A (Agkistrodon
rhodostoma); (2) activate prothrombin even in the absence of calcium
(E. carinatus); (3) activate factors X and V (Russell viper venom); (4)
degrade fibrinogen (Agkistrodon acutus); (5) induce platelet aggrega-
tion; (6) inhibit platelet aggregation because of the presence of
arginine-glycine-aspartic acid–containing peptides; (7) activate
protein C; and (8) damage endothelial cells which leads to bleeding,
tissue ischemia, and edema. Interestingly, victims of snake bites rarely
have excessive bleeding or thromboembolism despite the abnormal
coagulation tests and DIC-like picture.
LABORATORY MANIFESTATIONS
Thrombocytopenia or a rapidly declining platelet count is an impor-
tant diagnostic hallmark of DIC. However, only 35% to 44% of
critically ill patients develop thrombocytopenia (platelet count <150
× 109/L). Consequently, the specificity of thrombocytopenia for the
diagnosis of DIC is limited.10 A platelet count of <100 × 109/L is
seen in 50% to 60% of patients with DIC, whereas 10% to 15% of
patients have a platelet count <50 × 109/L. In surgical or trauma
patients with DIC, over 80% have platelet counts less than 100 ×
109/L.
Consumption of coagulation factors leads to low levels of coagula-
tion factors in patients with DIC. In addition, impaired hepatic
synthesis, for example due to impaired liver function or vitamin K
deficiency, and loss of coagulation proteins, due to massive bleeding,
may also play a role in DIC. Although the accuracy of the measure-
ment of one-stage clotting assays in DIC has been contested (because
of the presence of activated coagulation factors in plasma), the levels
of coagulation factors appear to correlate with the severity of the DIC.
The low levels of coagulation factors are reflected by prolonged global
tests of coagulation, such as the PT and the aPTT. A prolonged PT
or aPTT is found in 14% to 28% of intensive care patients but is
present in more than 95% of patients with DIC.
Plasma levels of factor VIII are paradoxically increased in most
patients with DIC, probably due to massive release of von Willebrand
factor from the endothelium in combination with the acute phase
behavior of factor VIII. Recent studies have pointed to a relative
deficiency of ADAMTS-13, the von Willebrand cleaving protease,
which results in high concentrations of ultralarge von Willebrand
multimers in plasma. These large multimers promote platelet-vessel
wall interaction, which can lead to thrombotic microangiopathy and
organ dysfunction.14
Although determination of the fibrinogen concentration has been
advocated as a useful tool for the diagnosis of DIC, this test is rarely
helpful.22 Fibrinogen is as an acute-phase protein whose levels increase
with inflammation. Therefore the fibrinogen level may remain within
the normal range for a long period of time despite ongoing consump-
tion. In a consecutive series of patients, the sensitivity of a low
fibrinogen level for the diagnosis of DIC was 28%, and hypofibrino-
genemia was only found in very severe cases of DIC.
Markers of Fibrin Generation and Degradation
Plasma levels of fibrin split products are frequently used for the
diagnosis of DIC.23 Fibrin split products were detectable in 42% of
consecutive patients in the intensive care unit, in 80% of trauma
patients, and in 99% of patients with sepsis and DIC. The levels of
fibrin degradation products (FDP) can be quantified by immunoas-
say. Latex agglutination assays can be used for rapid point-of-care
determination in emergency cases. None of the available FDP assays
discriminate between degradation products derived from cross-linked
fibrin or from fibrinogen, which may cause spuriously high levels.
Therefore the FDP assay lacks specificity because in addition to DIC,
high levels can be found with trauma, recent surgery, inflammation,
venous thromboembolism, and many other conditions. Because FDP
are metabolized in the liver and cleared by the kidneys, FDP levels
are influenced by liver and kidney function.
D-dimer is a degradation product of cross-linked fibrin. Therefore
this test is not influenced by fibrinogen degradation products.
D-dimer levels are high in patients with DIC, but high levels can also
be found in patients with venous thromboembolism, recent surgery,
or inflammatory conditions. Theoretically, soluble fibrin or fibrin
monomers would be useful markers of intravascular fibrin formation
in DIC. Indeed, initial clinical studies suggest that if the soluble fibrin
concentration exceeds a threshold level, a diagnosis of DIC can be
made. Unfortunately there are no reliable tests to quantify plasma
levels of soluble fibrin. Since plasma levels of soluble fibrin reflect
intravascular fibrin formation, the test is not influenced by extravas-
cular fibrin formation, which can occur with local inflammation or
trauma.
Endogenous Coagulation Inhibitors
Plasma levels of physiologic coagulation inhibitors, such as protein
C or antithrombin, may be useful indicators of ongoing activation of
coagulation.23 Reduced levels of these inhibitors are found in 40% to
60% of critically ill patients and in 90% of DIC patients.
Levels of protein C may correlate with the severity of the DIC.
In patients with meningococcal septicemia, plasma levels of protein
C are markedly reduced, which likely contributes to the purpura
fulminans that occurs in these patients. Downregulation of throm-
bomodulin and reduced levels of protein S further compromise the
capacity to generate APC. Therefore it is not surprising the low levels
of protein C are a strong predictor of poor outcome in DIC patients.
Plasma levels of antithrombin also are reduced in patients with
DIC. The low levels reflect consumption due to ongoing thrombin
generation, decreased synthesis, and degradation by neutrophil elas-
tase. Like protein C, low levels of antithrombin also a strong predictor
of mortality in patients with sepsis and DIC.
Fibrinolytic Markers
Increased fibrinolytic activity in DIC can be monitored by measur-
ing plasma levels of plasminogen and α2-antiplasmin. Low levels
may indicate consumption of these proteins. The concentration of

α2-antiplasmin is a helpful test for assessing the dynamics of fibrino-
lysis. Markers of plasmin generation include plasmin–α2-antiplasmin
(PAP) complexes, which are moderately elevated in patients with
DIC. However, because the normal plasma concentration of α2-
antiplasmin is about half that of plasminogen, α2-antiplasmin is sus-
ceptible to consumption when there is excessive plasmin generation.
Therefore PAP levels may underestimate total fibrinolytic activity.
With α2-antiplasmin consumption, other protease inhibitors, such
as antithrombin, α2-macroglobulin, α1-antitrypsin, and C1-inhibitor
may act as plasmin inhibitors as well. Fibrinolytic activity may be
insufficient to degrade intravascular fibrin in patients with DIC.
Fibrinolysis is suppressed by high levels of PAI-1, the major inhibitor
of tPA and uPA, PAI-1 levels are often elevated in patients with DIC,
and correlate with an unfavorable outcome. A functional mutation in
the PAI-1 gene, the 4G/5G polymorphism, is associated with high
levels of PAI-1, DIC and an increased risk of death in patients with
meningococcal septicemia.
Point of Care Tests
Although thrombelastography (TEG) was developed decades ago,
modern techniques, such as rotational thrombelastography (ROTEM),
enable bedside assessment of coagulation and fibrinolysis, which can
be helpful in acute care settings.24 A theoretical advantage of TEG
over conventional coagulation assays is that because TEG uses whole
blood, it provides a global assessment of platelet function, coagula-
tion, and fibrinolysis. Increased and decreased coagulation as dem-
onstrated with TEG was shown to correlate with clinically relevant
morbidity and mortality in several studies, although the superiority
of TEG over conventional tests has not been unequivocally proven.
TEG may be overly sensitive to some interventions, such as fibrinogen
administration, which is of questionable therapeutic value. Although
there are no systematic studies on the diagnostic accuracy of TEG for
the diagnosis of DIC, the test may be useful for assessing the global
status of the coagulation and fibrinolytic systems in critically ill
patients.
The aPTT biphasic waveform analysis is a sensitive and specific
test for hypercoagulability in critically ill patients. This test, which
requires specific instrumentation, detects the presence of precipitates
of complexes of very low density lipoprotein and C-reactive protein
that form early in DIC. The appearance of such complexes in the
plasma of individuals with diseases known to predispose to hyperco-
agulability confer a greater than 90% sensitivity and specificity for
subsequent development of DIC and fatal outcome.
Diagnostic Algorithm for Disseminated
Intravascular Coagulation
A simple scoring system for the diagnosis of overt DIC was developed
by the subcommittee on DIC of ISTH (Box 139.2).5 The score can
be calculated using routinely available laboratory tests including the
platelet count, the PT, a fibrin-related marker (usually D-dimer), and
the fibrinogen concentration. Tentatively, a score of 5 or higher is
compatible with DIC, whereas a score below 5 may be indicative but
is not affirmative for nonovert DIC. More refined scoring systems
have been developed for the diagnosis of nonovert DIC. A recent
study showed that the international normalized ratio (INR) can be
used in place of the PT, further facilitating international exchange
and standardization. Using receiver-operating characteristic curves,
an optimal D-dimer cut-off was identified, thereby optimizing the
sensitivity and the negative predictive value of the test. Prospective
studies show that the sensitivity and specificity of the DIC score are
93% and 98%, respectively. Studies in series of patients with specific
underlying disorders causing DIC (e.g., cancer or obstetrical compli-
cations) show similar results. The severity of DIC according to this
scoring system is related to the mortality in patients with sepsis.9
Linking prognostic determinants from critical care measurement
scores such as APACHE-II to DIC scores is an important means to
Chapter 139  Disseminated Intravascular Coagulation 2071
Diagnostic Algorithm for the Diagnosis of Overt
BOX 139.2
Disseminated Intravascular Coagulationa
1. Presence of an underlying disorder known to be associated with
disseminated intravascular coagulation (DIC) (Table 139.2)
(no = 0, yes = 2)
2. Score global coagulation test results
• Platelet count (>100 = 0; <100 = 1; <50 = 2)
• Level of fibrin markers (e.g., D-dimer, fibrin degradation
products)
(no increase: 0; moderate increase: 2; strong increase: 3)b
• Prolonged prothrombin time
(<3 s = 0; >3 s but <6 s = 1; >6 s = 2)
• Fibrinogen level
(>1.0 g/L = 0; <1.0 g/L = 1)
3. Calculate score
4. If ≥5: compatible with overt DIC; repeat scoring daily
If < 5: suggestive (not affirmative) for nonovert DIC; repeat next
1–2 days
a
According to the Scientific Standardization Committee of the International
Society of Thrombosis and Haemostasis.5
b
Strong increase, greater than 5× upper limit of normal; moderate increase,
greater than upper limit of normal but less than 5× upper limit of normal.
assess prognosis in critically ill patients. Similar scoring systems have
been developed and extensively evaluated in Japan. The major differ-
ence between the international and Japanese scoring systems is a
slightly higher sensitivity of the Japanese algorithm, which may reflect
differences in the patient populations because the Japanese series
include relatively large numbers of patients with hematologic
malignancies.
DIFFERENTIAL DIAGNOSIS
There are several other causes of coagulation abnormalities in patients
with underlying disorders known to be associated with DIC. A
complicating factor is that patients may have multiple explanations
for their coagulopathy. For example, a patient with sepsis with DIC
may also have liver failure and vitamin K deficiency.
The differential diagnosis of thrombocytopenia in patients with
suspected DIC is shown in Table 139.1. Sepsis itself is a risk factor
for thrombocytopenia in critically ill patients and the severity of
sepsis correlates with the extent of thrombocytopenia. The princi-
pal factors that contribute to thrombocytopenia in patients with
sepsis are decreased platelet production, increased consumption
or destruction, or sequestration platelets in the spleen or on the
endothelial surface. The decreased production of platelets in the
bone marrow in patients with sepsis occurs despite high levels of
proinflammatory cytokines, such as TNF-α and IL-6, and increased
levels of thrombopoietin. Platelet production is impaired because of
hemophagocytosis, a pathologic process characterized by phagocyto-
sis of megakaryocytes and other hematopoietic cells by monocytes
and macrophages. Hemophagocytosis may be driven by the high
levels of macrophage colony-stimulating factor (M-CSF) in patients
with sepsis.
Heparin-induced thrombocytopenia (HIT) is caused by a heparin-
induced antibody that binds to the heparin-platelet factor 4 (PF4)
complex on the platelet surface (see Chapter 133). This may result
in platelet activation and consumption and arterial and venous
thrombosis. A consecutive series of critically ill patients who received
heparin revealed an incidence of HIT of 1%. The risk of HIT is
higher with unfractionated heparin than with low-molecular-weight
heparin (LMWH). Thrombosis may occur in 25% to 50% of patients
with HIT (with fatal thrombosis in 4%–5%). The diagnosis of HIT
is based on detection of HIT antibodies in combination with the
occurrence of thrombocytopenia in a patient receiving heparin, with
or without concomitant arterial or venous thrombosis. The immu-
noassay for heparin-PF4 antibodies has a high negative predictive
 2072 Part XII  Hemostasis and Thrombosis

TABLE Differential Diagnosis of Thrombocytopenia in TABLE Differential Diagnosis of Prolonged aPTT and/or PT in
139.1 Suspected Disseminated Intravascular Coagulation 139.2 Suspected Disseminated Intravascular Coagulation
Differential Diagnosis Additional Diagnostic Clues Test Result Cause
DIC Prolonged aPTT and PT, increased PT prolonged, aPTT Factor VII deficiency
FDP, low levels of antithrombin or normal Mild vitamin K deficiency
protein C Mild liver insufficiency
Sepsis without DIC Positive (blood) cultures, positive Low doses of vitamin K antagonists
sepsis criteria, hematophagocytosis PT normal, aPTT Factor VIII, IX, or XI deficiency
in bone marrow prolonged Unfractionated heparin
Massive blood loss Major bleeding, low hemoglobin, Inhibitory antibody and/or antiphospholipid
prolonged aPTT and PT antibody
Factor XII or prekallikrein deficiency
Thrombotic microangiopathy Schistocytes evident on blood smear,
Coombs-negative hemolysis, fever, Both PT and aPTT Factor X, V, II, or fibrinogen deficiency
neurologic symptoms, renal prolonged Severe vitamin K deficiency
insufficiency, coagulation tests Vitamin K antagonists
usually normal, ADAMTS13 levels Global clotting factor deficiency
decreased • Decreased synthesis: liver failure
• Increased loss: massive bleeding, DIC
Heparin-induced Use of heparin, venous or arterial
aPTT, Activated partial thromboplastin time; PT, prothrombin time.
thrombocytopenia thrombosis, positive HIT test
(usually immunoassay for
heparin-platelet factor 4
antibodies), increase in platelet
count after cessation of heparin;
by cytostatic agents), or by immune-mediated mechanisms. Drug-
coagulation tests usually normal
induced thrombocytopenia is a difficult diagnosis in patients suspected
of DIC because these patients are often receiving multiple drugs and
Immune thrombocytopenia Antiplatelet antibodies, normal or have several other potential reasons for the thrombocytopenia. Drug-
increased number of induced thrombocytopenia is often diagnosed based upon the timing
megakaryocytes in bone marrow of initiation of a new agent in relationship to the development of
aspirate, normal levels of TPO thrombocytopenia, after exclusion of other causes of thrombocyto-
(TPO levels are usually normal or penia. The observation of rapid restoration of the platelet count after
slightly increased in ITP); discontinuation of the suspected agent is highly suggestive of drug-
coagulation tests usually normal induced thrombocytopenia.
Drug-induced Decreased number of megakaryocytes A prolongation of global coagulation tests may be due to a defi-
thrombocytopenia in bone marrow aspirate or ciency of one or more coagulation factors (Table 139.2). In addition,
detection of drug-induced but more rarely, the prolonged tests may be due to an inhibitory
antiplatelet antibodies, increase in antibody. Some of these antibodies may be clinically important, such
platelet count after cessation of as antibodies to factor VIII that lead to acquired hemophilia (see
drug; coagulation tests usually Chapter 136), whereas others may be less important, such as
normal antiphospholipid antibodies (see Chapter 141). However, patients
ADAMTS13, A disintegrin and metalloproteinase with thrombospondin 13; with antiphospholipid antibodies may have thrombocytopenia and
aPTT, activated partial thromboplastin time; DIC, disseminated intravascular may be at increased risk of thrombosis particularly if they also have
coagulation; FDP, fibrin degradation products; HIT, heparin-induced a lupus anticoagulant. Inhibitory antibodies and lupus anticoagulants
thrombocytopenia; PT, prothrombin time; ITP, immune thrombocytopenia; can be identified and distinguished with mixing studies (see Chapters
TPO, thrombopoietin.
136 and 141).
In general, acquired deficiencies of coagulation factors can be due
to impaired synthesis, massive loss, or increased turnover (consump-
tion). Impaired synthesis is often due to hepatic insufficiency or
value (100%) but a low positive predictive value (10%). Although vitamin K deficiency. Vitamin K deficiency may be caused by poor
the gold standard for the diagnosis of HIT is a sensitive platelet nutrition in combination with the use of antibiotics that impair
activation assay, this test is not routinely available in most hospitals. bacterial vitamin K production in the intestine. The PT is sensitive to
Normalization of the platelet count 1–3 days after discontinuation both conditions because this test is highly dependent on the plasma
of heparin may further support the diagnosis of HIT. levels of factor VII, the vitamin K–dependent coagulation factor with
The group of thrombotic microangiopathies includes thrombotic the shortest half-life. Liver failure may be differentiated from vitamin
thrombocytopenic purpura (see Chapter 134), hemolytic–uremic K deficiency by measuring the levels of factor V, which is not vitamin
syndrome (see Chapter 134), malignant hypertension, chemotherapy- K dependent. In fact, the factor V level is included in several scoring
induced microangiopathic hemolytic anemia, and the HELLP syn- systems for acute liver failure. Uncompensated loss of coagulation
drome. A common pathogenic feature of these disorders is endothelial factors may occur after massive bleeding, which can occur in trauma
damage, which triggers platelet adhesion and aggregation, thrombin patients or those undergoing major surgical procedures. This is
generation, and an impaired fibrinolysis. The clinical consequences common in patients with major bleeding who receive intravascular
of extensive endothelial dysfunction include thrombocytopenia, volume replacement with crystalloids, colloids, and red cells without
mechanical fragmentation of red cells with hemolytic anemia, and simultaneous administration of coagulation factors. This results in
microvasular occlusion, which leads to multiorgan dysfunction, a dilutional coagulopathy, which can exacerbate the bleeding. In
including renal insufficiency and neurologic symptoms. Despite this addition, transfusion in these patients may lead to systemic activation
common final pathway, the various thrombotic microangiopathies of inflammatory processes and may contribute to further coagula-
have different underlying etiologies (see Chapter 134). tion derangements. In hypothermic patients (e.g., trauma patients)
Drug-induced thrombocytopenia is another frequent cause of measurement of the global coagulation tests may underestimate the
thrombocytopenia in critically ill patients (see Chapter 131). Throm- extent of the coagulopathy because these assays are performed at 37°C
bocytopenia may be caused by drug-induced myelosuppression, (e.g., to mimic normal body temperature.
 Chapter 139  Disseminated Intravascular Coagulation 2073

BOX 139.3 Mainstays of Supportive Treatment of Disseminated Intravascular Coagulation

Modality Details Expectations/Rationale

Treating the underlying
disorder
Antithrombotic agents
Transfusion
Anticoagulant factor
concentrates
Fibrinolytic inhibitors
Dependent on the primary diagnosis
Prophylactic heparin to prevent venous
thromboembolic complications
(Low dose) therapeutic heparin in case of
confirmed thromboembolism or if clinical
picture is dominated by (micro) vascular
thrombosis and associated organ failure
Infuse platelets, plasma and fibrinogen
(cryoprecipitate) if there is overt bleeding or a
high risk of bleeding
Recombinant human activated protein C may be
effective in sepsis and DIC (24 µg/kg/h for 4
days); Currently withdrawn from the market
Tranexamic acid (e.g., 500–1000 mg q8–12 h or
ε-aminocaproic acid 1000–2000 mg q8–12 h)
Inhibit or block the complicating pathologic mechanism of
disseminated intravascular coagulation (DIC) in parallel
with the response (if any) of the disorder
Risk of thromboembolism is increased in critically ill
patients, trauma patients, or patients with cancer
Prevent fibrin formation; tip the balance within the
microcirculation toward anticoagulant mechanisms and
physiologic fibrinolysis; allow reperfusion of the skin,
kidneys, and brain
Bleeding should diminish and stop over the course of hours
Platelet count, coagulation tests and fibrinogen should
return toward normal
Restore anticoagulation in microvascular environment and
may have antiinflammatory activity
Latest trials were negative.
May be useful if there is (hyper) fibrinolysis
Bleeding ceases, but there is a risk of microvascular
thrombosis and renal failure

THERAPY
Management of patients with DIC depends on vigorous treatment
of the underlying disorder to alleviate or remove the inciting injurious
cause. For sepsis-induced DIC, treatment includes aggressive use of
intravenous organism-directed antibiotics and source control (e.g., by
surgery or drainage). Other examples of vigorous treatment of the
underlying condition include cancer surgery or chemotherapy, uterus
evacuation in patients with abruptio placentae, resection of aortic
aneurysm, and debridement of crushed tissue in the case of trauma.
In addition to intensive support of vital functions, supportive treat-
ment aimed at the coagulopathy may be helpful (Box 139.3),6 as
outlined later.
Platelet and Plasma Transfusion
Low levels of platelets and coagulation factors may increase the risk
of bleeding. However, plasma or platelet transfusion should not be
instituted on the basis of laboratory results alone; it is only indicated
in patients with active bleeding and in those requiring an invasive
procedure or at risk for bleeding complications.6 The presumed
efficacy of treatment with plasma, fibrinogen concentrate, cryopre-
cipitate, or platelets is not based on the results of randomized con-
trolled trials. Nonetheless, transfusion of these products is rational in
bleeding patients or in patients at risk of bleeding who have significant
depletion of these hemostatic factors. The suggestion that administra-
tion of blood components might “add fuel to the fire” has never been
proven in clinical or experimental studies.
Replacement therapy for thrombocytopenia consists of 5 to 10 U
platelet concentrate to raise the platelet count to 20–30 × 109/L and
to 50 × 109/L in patients requiring an invasive procedure.
One of the major challenges of infusion of fresh-frozen plasma in
patients with DIC who are bleeding is the propensity of the added
volume, which is necessary to correct the coagulation defect, to
exacerbate the capillary leak syndrome. This increases the risk of
causing or worsening pulmonary edema and, by extension, predis-
poses to ARDS and induces ascites. Coagulation factor concentrates,
such as prothrombin complex concentrate, may partially overcome
this obstacle, but these concentrates do not contain essential factors,
such as factor V. Moreover, caution is advocated with the use of
prothrombin complex concentrates in DIC because they can contain
trace amounts of activated coagulation factors, which may worsen the
coagulopathy. Specific deficiencies of coagulation factors, such as
fibrinogen, may be corrected by administration of purified coagula-
tion factor concentrates.
Cryoprecipitate (if available) can be used to rapidly raise the levels
of fibrinogen, von Willebrand factor, and factor VIII levels in patients
with DIC, particularly when there is bleeding and the fibrinogen level
is less than 1 g/L. Cryoprecipitate has at least four- to fivefold more
fibrinogen in each milliliter than fresh-frozen plasma. Nonetheless,
fresh-frozen plasma contains sufficient amounts of fibrinogen to treat
mild to moderate hypofibrinogenemia.
Anticoagulant Treatment
Heparin therapy in patients with DIC remains controversial. Experi-
mental studies have shown that heparin can at least partly inhibit
activation of coagulation in DIC. However, clinical trials have failed
to demonstrate a beneficial effect of heparin on clinically important
outcome events in patients with DIC and the safety of heparin is
debatable in those at risk for bleeding. A large trial in patients with
severe sepsis showed a small but nonsignificant benefit of low dose
heparin on 28-day mortality in patients and no major safety
concerns.25
There is general consensus that administration of heparin is benefi-
cial in some categories of DIC, such as metastatic cancer, purpura
fulminans, and aortic aneurysm (prior to resection). Heparin also is
indicated for treating thromboembolic complications in large vessels
and before surgery in patients with chronic DIC. Heparin administra-
tion may be helpful in patients with acute DIC when intensive blood
component replacement fails to improve excessive bleeding or when
thrombosis threatens to cause irreversible tissue injury (e.g., acute
cortical necrosis of the kidney or digital gangrene).
Heparin should be used cautiously in these conditions. In patients
with chronic DIC because of metastatic cancer or aortic aneurysm,
continuous infusion of unfractionated heparin (500 to 750 U/hour
without a bolus) has been advocated. If no response is obtained within
24 hours, higher dosages can be used. In hyperacute DIC cases, such
as amniotic fluid embolism, septic abortion, and purpura fulminans,
intravenous bolus injection of 5000 to 10,000 U heparin may be given
simultaneously with replacement therapy with blood products. Some
experts, however, recommend against administration of a bolus dose
of heparin under these circumstances. A heparin infusion of 500 to
1000 U/hour may be necessary to maintain the benefit until the
underlying disease responds to treatment. Aside from these consider-
ations, current guidelines recommend universal thromboprophylaxis
with low doses of heparin or LMWH in critically ill patients.6
Theoretically, the most logical anticoagulants to use in DIC would
be those directed against TF. Potential agents include recombinant
TFPI, inactivated factor VIIa, and recombinant NAPc2, a potent and
 2074 Part XII  Hemostasis and Thrombosis
specific inhibitor of the ternary complex of TF/factor VIIa and factor
Xa. Phase II trials of recombinant TFPI in patients with sepsis showed
promising results but phase III trials in patients with severe sepsis or
severe pneumonia and organ failure failed to show a survival advan-
tage with TFPI adminsitration.26
Recombinant human soluble thrombomodulin binds thrombin
to form a complex that blocks the coagulant activity of thrombin and
promotes its capacity to activate protein C. In a phase III randomized
double-blind clinical trial in patients with DIC, soluble thrombo-
modulin was superior to heparin at reducing bleeding and improving
the coagulation parameters. A systematic review and meta-analysis of
mostly retrospective studies on the effect of recombinant human
soluble thrombomodulin in severe sepsis demonstrated a pooled rela-
tive risk of 0.81 (95% confidence interval (CI) 0.62–1.06) in three
randomized trials encompassing 838 patients, and a relative risk of
0.59 (95% CI 0.45–0.77) in nine observational studies including 571
patients.27 Interestingly, meta-regression of the combined results
revealed a strong relationship between the effect size of recombinant
soluble thrombomodulin treatment and baseline mortality risk,
which is reminiscent of a similar relationship in patients who were
treated with activated protein C.9,28 Importantly, there was no evi-
dence of an increased risk of major bleeding or other safety concerns
with soluble thrombomodulin. Ongoing randomized trials with
soluble thrombomodulin are focusing on its effects on DIC, organ
failure, and mortality.
Physiologic Anticoagulant Factor Concentrates
Restoration of the levels of physiologic anticoagulants in DIC may
be a rational approach.29,30 Based on successful preclinical studies, AT
concentrates without or with concomitant heparin have been evalu-
ated in patients with DIC. All trials have shown some beneficial effect
in terms of improvement of laboratory parameters, shortening of the
duration of DIC, or even improvement in organ function. In several
small clinical trials, use of very high doses of AT concentrate produced
a modest but not statistically significant reduction in mortality. A
large, multicenter, randomized controlled trial also showed no signifi-
cant reduction in mortality with AT concentrate in patients with
sepsis.31 Interestingly, post hoc subgroup analysis suggested that AT
was of benefit in patients who did not receive concomitant heparin,
but this observation needs validation. In a small randomized trial in
patients with burns and DIC, AT administration decreased mortality,
reduced multiple organ failure, and improved coagulation parameters
compared with placebo.
Because decreased function of the protein C system contributes
to the pathogenesis of DIC, recombinant APC was predicted to be
beneficial.32 Indeed, in a dose-ranging study, a continuous infusion
of recombinant human APC at a dose of 24 µg/kg per hour was
considered optimal based on reductions in D-dimer levels. A subse-
quent placebo-controlled phase III trial of APC concentrate in
patients with severe sepsis was stopped early because of a 19% reduc-
tion in 28-day all cause mortality from 30.8% to 24.7% with APC.
Interestingly, patients who manifested with “overt DIC” (see Diag-
nosis section, earlier) benefited more from APC than those without
overt DIC.9 However, meta-analyses of the APC trials concluded that
APC had little or no effect. A series of negative trials in patients with
severe sepsis has added to the skepticism regarding the use of APC.
In addition to questionable efficacy, APC increases the risk of bleed-
ing in patients with severe sepsis. A placebo-controlled trial in patients
with severe sepsis and septic shock was stopped early because there
was no benefit with APC. Subsequently, the manufacturer elected to
withdraw APC from the market, which has resulted in a revision of
current guidelines for treatment of DIC.
Fibrinolytic Inhibitors
Most guidelines recommend against the use of antifibrinolytic agents,
such as ε-aminocaproic acid or tranexamic acid, in patients with
DIC. This is because these drugs block already suppress endogenous
fibrinolysis, and may further compromise tissue perfusion. In support
of this concept there are reports of severe thrombosis in DIC patients
treated with these agents. However, in patients with DIC accompa-
nied by primary fibrino(geno)lysis, which occurs in some cases of
acute promyelocytic leukemia, giant cavernous hemangioma, heat
stroke, and metastatic carcinoma of the prostate, the use of fibrino-
lytic inhibitors can be considered if the patient has profuse bleeding
that does not respond to replacement therapy and there is evidence
of excessive fibrino(geno)lysis. In these situations, it is important to
replace depleted blood components before initiating treatment with
fibrinolytic inhibitors.
REFERENCES
1. Seligsohn U: Disseminated intravascular coagulation. In Handin RI,
Lux SE, Stossel TP, editors: Blood: Principles and practice of hematology,
Philadelphia, 2000, J.B. Lippincott.
2. Levi M, Seligsohn U: Disseminated intravascular coagulation. In
Kaushansky K, Lichtman M, Beutler E, et al, editors: Williams hematol-
ogy, Philadelphia, 2010, McGraw Hill.
3. Levi M, ten Cate H: Disseminated intravascular coagulation. N Engl J
Med 341(8):586–592, 1999.
4. Levi M: Disseminated intravascular coagulation. Crit Care Med
35:2191–2195, 2007.
5. Taylor FBJ, Toh CH, Hoots WK, et al: Towards definition, clinical and
laboratory criteria, and a scoring system for disseminated intravascular
coagulation. Thromb Haemost 86(5):1327–1330, 2001.
6. Levi M, Toh CH, Thachil J, et al: Guidelines for the diagnosis and
management of disseminated intravascular coagulation. Br J Haematol
145(1):24–33, 2009.
7. Levi M, van der Poll T: Inflammation and coagulation. Crit Care Med
38(2 Suppl):S26–S34, 2010.
8. Fourrier F, Chopin C, Goudemand J, et al: Septic shock, multiple
organ failure, and disseminated intravascular coagulation. Compared
patterns of antithrombin III, protein C, and protein S deficiencies. Chest
101(3):816–823, 1992.
9. Dhainaut JF, Yan SB, Joyce DE, et al: Treatment effects of drotrecogin
alfa (activated) in patients with severe sepsis with or without overt dis-
seminated intravascular coagulation. J Thromb Haemost 2:1924–1933,
2004.
10. Levi M, Opal SM: Coagulation abnormalities in critically ill patients.
Crit Care 10(4):222, 2006.
11. Colman RW, Schmaier AH: Contact system: a vascular biology modulator
with anticoagulant, profibrinolytic, antiadhesive, and proinflammatory
attributes. Blood 90(10):3819–3843, 1997.
12. Esmon CT: The regulation of natural anticoagulant pathways. Science
235(4794):1348–1352, 1987.
13. Keller TT, Mairuhu AT, de Kruif MD, et al: Infections and endothelial
cells. Cardiovasc Res 60(1):40–48, 2003.
14. Lowenberg EC, Meijers JC, Levi M: Platelet-vessel wall interaction in
health and disease. Neth J Med 68(6):242–251, 2010.
15. Gando S, Nakanishi Y, Tedo I: Cytokines and plasminogen activator
inhibitor-1 in posttrauma disseminated intravascular coagulation:
relationship to multiple organ dysfunction syndrome. Crit Care Med
23(11):1835–1842, 1995.
16. Huisse MG, Pease S, Hurtado-Nedelec M, et al: Leucocyte activation:
the link between inflammation and coagulation during heatstroke. A
study of patients during the 2003 heat wave in Paris. Crit Care Med
36:2288–2295, 2008.
17. Levi M: Disseminated intravascular coagulation (DIC) in pregnancy and
the peri-partum period. Thromb Res 123(Suppl 2):S63–S64, 2009.
18. Colman RW, Rubin RN: Disseminated intravascular coagulation due to
malignancy. Semin Oncol 17(2):172–186, 1990.
19. Barbui T, Falanga A: Disseminated intravascular coagulation in acute
leukemia. Semin Thromb Hemost 27(6):593–604, 2001.
20. Fisher DF, Jr, Yawn DH, Crawford ES: Preoperative disseminated intra-
vascular coagulation associated with aortic aneurysms. A prospective
study of 76 cases. Arch Surg 118(11):1252–1255, 1983.

21. Isbister GK: Snake bite doesn’t cause disseminated intravascular coagula-
tion: Coagulopathy and thrombotic microangiopathy in snake envenom-
ing. Semin Thromb Hemost 36:444–451, 2010.
22. Levi M, de Jonge E, Meijers J: The diagnosis of disseminated intravas-
cular coagulation. Blood Rev 16(4):217–223, 2002.
23. Levi M, Meijers JC: DIC: Which laboratory tests are most useful. Blood
Rev 2010.
24. Dempfle CE, Borggrefe M: Point of care coagulation tests in critically ill
patients. Semin Thromb Hemost 34(5):445–450, 2008.
25. Levi M, Levy M, Williams MD, et al: Prophylactic heparin in patients
with severe sepsis treated with drotrecogin alfa (activated). Am J Respir
Crit Care Med 176(5):483–490, 2007.
26. Abraham E, Reinhart K, Opal S, et al: Efficacy and safety of tifacogin
(recombinant tissue factor pathway inhibitor) in severe sepsis: a random-
ized controlled trial. JAMA 290(2):238–247, 2003.
27. Yamakawa K, Aihara M, Ogura H, et al: Recombinant human soluble
thrombomodulin in severe sepsis: a systematic review and meta-analysis.
J Thromband Haemost 13(4):508–519, 2015.
Chapter 139  Disseminated Intravascular Coagulation 2075
28. Bernard GR, Vincent JL, Laterre PF, et al: Efficacy and safety of
recombinant human activated protein C for severe sepsis. N Engl J Med
344:699–709, 2001.
29. Levi M, de Jonge E, van der Poll T: Rationale for restoration of physi-
ological anticoagulant pathways in patients with sepsis and disseminated
intravascular coagulation. Crit Care Med 29:S90–S94, 2001.
30. de Jonge E, van der Poll T, Kesecioglu J, et al: Anticoagulant factor
concentrates in disseminated intravascular coagulation: rationale for use
and clinical experience. Semin Thromb Hemost 27:667–674, 2001.
31. Warren BL, Eid A, Singer P, et al: Caring for the critically ill patient.
High-dose antithrombin III in severe sepsis: a randomized controlled
trial. JAMA 286(15):1869–1878, 2001.
32. Levi M: Activated protein C in sepsis: a critical review. Curr Opin
Hematol 15(5):481–486, 2008.