Icahn School of Medicine at Mount Sinai, New York, NY, USA
Definition of disease
A coagulopathy is a state of abnormal hemostasis in which an individual’s ability to form a clot to reduce bleeding is impaired, thereby increasing the bleeding risk; however coagulopathy may also describe pro‐thrombotic states.
Coagulopathy is characterized by abnormalities on tests assessing hemostasis: most commonly manifesting in the prothrombin time (PT), which monitors the tissue factor pathway, the activated partial thromboplastin time (aPTT), which monitors the contact activation and common pathways, and platelet count. The term coagulopathy is often meant to describe abnormalities in the PT and aPTT, while thrombocytopenia is used separately to distinguish a low platelet count. The PT, aPTT, and platelet count do not define the ICU patient’s entire hemostatic state, however, as abnormalities in other coagulation assay abnormalities and disorders of platelet function are common.
Coagulopathy in critically ill patients is common but the incidence and prevalence vary widely depending on the patient population studied and the cutoffs used to define a coagulopathic state.
An international normalized ratio (INR) >1.5 occurs in nearly 60% of critically ill patients.
Thrombocytopenia as defined by a platelet count of <150 × 109/L occurs in up to 60% of patients, although severe thrombocytopenia (<50 × 109/L), the point at which bleeding risk increases exponentially, is much less common.
In the majority of critically ill patients, deficiencies of coagulation factors and thrombocytopenia are acquired.
Disseminated intravascular coagulation (DIC), a condition of systemic intravascular activation of coagulation marked by consumption of coagulation factors, is the most common cause and can be secondary to many conditions, including sepsis, malignancy, and trauma.
Because the liver produces all clotting proteins except for factor VIII, liver disease is a common cause of dual pathway abnormalities.
The synthesis of factors II, VII, IX, and X are all dependent on vitamin K, which is frequently deficient in ICU patients due to inadequate intake and alterations in microbial synthesis due to antibiotic exposure. Vitamin K deficiency also causes dual pathway abnormalities, and can be distinguished from liver disease by measuring factor V levels, as factor V synthesis is not vitamin K dependent.
Losses of coagulation factors in the setting of hemorrhage without replacement (e.g. crystalloid/colloid resuscitation) will lead to dilutional coagulation defects.
Anticoagulants are common in the ICU, and depending on the underlying anticoagulant used, will variably affect tests of hemostasis.
Inherited factor deficiencies or the presence of inhibiting antibodies are seen less commonly in ICU patients.
DIC scoring system
Platelet count (× 109/L)
>3 and <6 seconds
Fibrin degradation products
≥5 points is compatible with DIC
Thrombocytopenia in the critically ill is caused by sepsis and DIC in 75% of cases, massive blood loss in 10%, and drug‐related thrombocytopenia (including heparin‐induced thrombocytopenia (HIT)) in 10%.
Drug‐induced thrombocytopenia can be due to immune‐mediated mechanisms or direct myelosuppression. Many drugs have been implicated and lists are available online (www.ouhsc.edu/platelets/ditp.html).
HIT is a life‐threatening condition characterized by platelet‐activating antibodies recognizing complexes bound to heparin–platelet factor 4 complexes. Classically, the fall in platelets is greater than 50% from baseline but not below 20 × 109/L and starts 5–10 days after heparin exposure. Heparin–platelet factor 4 antibodies are highly sensitive screening tests but require confirmation.
Table 56.1Pathogenesis of common causes of acquired coagulopathies in critically ill patients.
Liver insufficiency Vitamin K deficiency
Consumption of coagulation factors
DIC, sepsis, trauma
Loss of coagulation factors
Thrombotic microangiopathies such as thrombotic thrombocytopenic purpura (TTP), hemolytic‐uremic syndrome (HUS), malignant hypertension, and in obstetric patients the HELLP (hemolysis, elevated liver enzymes, low platelets) syndrome are responsible for about 1% of cases. These conditions have a similar pathogenic pathway of endothelial injury, platelet adhesion, and thrombin generation, with subsequent mechanical fragmentation of red blood cells, although treatment is vastly different among them.
Clotting disorders occur by various mechanisms in the ICU, but generally fall into three main categories: impaired synthesis, consumption, and loss/dilution (Table 56.1).
Anticoagulation agents variably alter coagulation pathways.
Since most clotting factors are produced in the liver, hepatic disease will lead to underproduction of many clotting factors.
Vitamin K deficiency leads to underproduction of the vitamin K‐dependent factors.
Consumption leads to progressive depletion of clotting factors and is the predominant mechanisms of coagulopathy in DIC, sepsis, and trauma.
Dilution usually occurs in the setting of the massively hemorrhaging patient who is resuscitated without replacement of clotting factors.
Thrombocytopenia in the ICU is often multifactorial, but one of four mechanisms usually predominates: increased consumption, underproduction, dilution, and sequestration (Table 56.2).
Consumption/destruction: this is the most common cause of thrombocytopenia in the ICU. It occurs via two mechanisms: immune or non‐immune mediated:
Immune‐mediated thrombocytopenia occurs via the production of platelet antibodies and their subsequent destruction. In the ICU, it is commonly secondary to drugs, especially heparin, but can also be due to viral infections.
Non‐immune mechanisms occur in patients with DIC, patients on cardiopulmonary bypass, and those with microangiopathic hemolytic anemias (MAHAs). TTP‐HUS is rare in the ICU, but important to recognize as, left untreated, it can be fatal in >90% of patients. TTP is due to an acquired or hereditary insufficiency of von Willebrand factor cleaving protease (ADAMTS‐13), leading to large von Willebrand factor multimers and platelet aggregation. HUS is caused by a cytotoxin released from a specific Escherichia coli serotype.
Table 56.2Pathogenesis of common causes of thrombocytopenia in critically ill patients.
Decreased production: this is usually due to bone marrow suppression. Medications are the most common culprit and should be reviewed. Toxins, viral infections, and nutritional deficiencies are also common causes.
Dilution: occurs in the setting of hemorrhage without adequate platelet replacement.
Sequestration: occurs in the setting of hypersplenism. It is common in cirrhosis.
Platelet count, PT, and aPTT are screening tests that detect many coagulopathies and are part of routine testing in critically ill patients. Further diagnostic testing will be based on the pattern of lab abnormalities and clinical context.
Patients on anticoagulation should have coagulation studies monitored per the specific drug protocol to avoid both under‐ and overtreatment.
For patients receiving heparin in whom clinicians consider the risk of HIT to be >1%, platelet count should be performed every 2–3 days from day 4 to 14 after heparin is started.
Administration of FFP and platelets to prevent coagulopathies and thrombocytopenia from developing is not recommended.
Routine transfusion of FFP to prevent bleeding in patients with acquired coagulopathies is not recommended.
Trauma patients should receive early and fixed repletion of clotting factors with blood products to prevent the dilutional coagulopathy that can develop (e.g. a 1:1:1 ratio of PRBCs : platelets : FFP). Despite a lack of evidence, these findings have been extended to resuscitation efforts in post‐surgical bleeding and gastrointestinal and obstetric hemorrhage. It is being evaluated in the North American Pragmatic, Randomized Optimal Platelets and Plasma Ratios study.
Dietary intake of vitamin K may be inadequate in the critical care setting, but there is no high quality evidence for routine supplementation for critical care patients at risk for deficiency.
Avoidance of drugs known to commonly cause thrombocytopenia is one method which may reduce the risk of thrombocytopenia, although the practice must be weighed against avoiding potentially necessary therapies. Use of low molecular weight heparins (LMWHs) in lieu of unfractionated heparin for routine venous thromboembolism prophylaxis leads to a lower odds ratio of thrombocytopenia, and may be cost effective.
Most critically ill patients will have minor coagulopathies that are asymptomatic and only noticed on routine blood tests. Those that manifest symptoms typically do so by bleeding, frequently with ecchymoses at puncture sites or frank hemorrhage.
Certain conditions, such as antiphospholipid syndrome and HIT, demonstrate prolonged clotting times but actually predispose to thrombosis and may require anticoagulation, which highlights the importance of elucidating the underlying etiology.
An important initial step in the critically ill patient with a coagulopathy is to assess if the patient has been receiving any anticoagulation.
A past medical or family history may elucidate both inherited and acquired causes of abnormal bleeding, such as von Willebrand disease (vWD), hemophilia, cirrhosis, and vitamin K deficiency.
It may be the critical condition itself or its treatment causing the coagulopathy. Sepsis, DIC, trauma, for example, all commonly cause coagulopathies and thrombocytopenia in the ICU, as do antibiotics, heparin, and hemorrhage.
The physical exam should assess for any signs of bleeding. Clotting factor deficiencies cause joint and soft tissue bleeding, while thrombocytopenia usually causes mucocutaneous bleeding.
Gastrointestinal bleeding can be confirmed by a rectal examination or a lavage via a nasogastric tube. Bleeding into a thigh or abdomen can be significant but may be clinically occult.
The clinician should also look for underlying diseases that may predispose to coagulopathy, such as the physical exam findings that accompany cirrhosis, sepsis, and trauma. Skin necrosis in the setting of heparin exposure and thrombocytopenia should raise the suspicion of HIT.
Useful clinical decision rules and calculators
4T scoring system for HIT
Platelet ↓ >50% and nadir >20 × 109/L
Platelet ↓ 30–50% or nadir 10–20 × 109/L
Platelet ↓ <350% or nadir <10 × 109/L
Timing (from first heparin dose)
Day 5 to 10
Unknown exposure, > day 10, or < day 1 (with heparin exposure 30–100 days ago)
* A low score has a very high negative predictive value (97–99%) and is useful for ruling out HIT. The positive predictive value is 10–20% for intermediate and 40–80% for high scores, depending on the clinical setting.
Disease severity classification
Since laboratory abnormalities of the clotting factor pathways correlate poorly with bleeding risk, there is no formal severity classification based on PT and aPTT levels.
Thrombocytopenia is further subdivided into mild (<150 × 109/L), moderate (<100 × 109/L), and severe (<50 × 109/L). This is not specific to the ICU population and the correlation between platelet count and bleeding risk also varies according to the underlying etiology.
Differential diagnosis of laboratory abnormalities of hemostasis
MAHA, sepsis, malignancy, ↑LDH
MAHA, ↑LDH, diarrhea, AMS, AKI
MAHA, ↑LDH, ↑LFTs, pregnancy, HTN, proteinuria
Liver disease Early Late
Normal factor VIII levels, cirrhosis
Vitamin K deficiency
↔aPTT if mild, normal factor V levels
von Willebrand disease (vWD)
Platelet inhibitors, ↑BT
↑PT when supratherapeutic
Monitor anti‐Xa activity assay, ↑PT when supratherapeutic
Errors in sampling commonly cause abnormal clotting assays. This can be due to underfilling of the tube or using the wrong tube type. Repeating an unexpected test in accordance with proper collection technique will avoid costly and potentially invasive further testing.
Exclude exposure to anticoagulation.
If the aPTT is elevated and the PT is normal, an aPTT‐based mixing study should be done to delineate the presence of an inhibitor to the contact activation pathway. In the setting of an inhibitor, the mixing study will not correct the aPTT, whereas the aPTT will normalize if there is a simple factor deficiency. Further testing for specific factor levels (VIII, IX, XI), antiphospholipid antibody testing, fibrinogen disorders, or vWD can be considered depending on such results.
An isolated prolongation of the PT should prompt investigation of systemic disease, such as sepsis, DIC, early liver failure, or mild vitamin K deficiency. If there is no underlying etiology, a PT‐based mixing study and factor VII activity will delineate the presence of an inhibitor or factor deficiency, which is rare.
Likewise, prolongation of both the PT and aPTT warrants investigation of systemic disease, which is the most common cause. If none is found, a prolonged thrombin time will suggest fibrinogen disorders, while a normal thrombin time suggests factor inhibitors or deficiencies.
Pseudothrombocytopenia due to platelet clumping is common; consider redrawing with heparin‐ or citrate‐containing collection tubes.
The dynamics of the platelet count are particularly important in regards to HIT, which typically has its nadir between 5 and 10 days after heparin exposure.
Thrombocytopenia should be interpreted in the context of other tests of coagulation; such as the PT, aPTT, fibrinogen, and fibrin degradation products, as well as a complete blood count.
An examination of the peripheral smear should be an early step in the assessment of thrombocytopenia, with attention paid to all three cell lines.
White blood cells:
Leukemic cells diagnose hematologic malignancies.
Neutrophilia, lymphocytosis, and toxic granulation suggests infection.
Red blood cells:
Schistocytes indicate a MAHA. In the presence of other coagulopathies, consider DIC; otherwise, rule out TTP/HUS.
Macrocytosis suggests vitamin B12 or folate deficiency.
Dacrocytes (tear drop cells) suggest myelofibrosis.
Nucleated RBCs suggest hemolytic anemia, myelofibrosis, and infiltrative processes.
Large platelets indicate increased turnover or hereditary disorders.
Small platelets are usually seen in production disorders.
Viral‐associated thrombocytopenia: parvovirus, hepatitis C, human immunodeficiency virus.
Connective tissue diseases.
Creatinine elevation and uremia may identify renal failure, which is itself a risk factor for bleeding and may indicate diseases such as HUS.
Liver function tests:
Abnormalities may support a diagnosis of cirrhosis and or the HELLP syndrome.
Pregnancy test in pre‐menopausal women.
Blood cultures if sepsis and DIC are suspected.
A type and screen should be drawn in case the patient will require blood product support.
Abdominal ultrasound to assess for splenomegaly and cirrhosis.
Potential pitfalls/common errors made regarding diagnosis of disease
PT and aPTT are relatively insensitive; both remain normal until factor levels are significantly depleted (<50%).
The degree of PT and aPTT abnormalities does not necessarily correlate with the risk of bleeding, and may actually be indicative of a prothrombotic state.
A positive HIT antibody should be interpreted only in the appropriate clinical context and confirmed with a serotonin‐release assay or a heparin‐induced platelet aggregation assay.
Disorders of platelet function are common but are not reflected in platelet counts. This can occur due to medications (aspirin, glycoprotein IIb/IIIa inhibitors), renal disease, and intrinsic defects (vWD).
Many conditions in the ICU are associated with coagulopathies and thrombocytopenia. It is the responsibility of the clinician to not overtreat (i.e. unnecessary transfusions) while the primary etiology is being elucidated. The most common causes (e.g. DIC, sepsis, trauma) have no specific therapy and will improve with supportive care aimed at the underlying etiology.
A platelet count threshold of 10 × 109/L is widely used as prophylaxis against spontaneous bleeding, but there is a lack of high quality evidence suggesting benefit. Avoiding prophylactic platelet transfusions at any level may be reasonable in patients with autologous stem cell transplantation.
As mentioned previously, trauma patients should receive early and fixed repletion of clotting factors and platelets while receiving blood products to avoid dilutional coagulopathies and thrombocytopenia. There is an unproven benefit in other hemorrhaging patients, but most providers adhere to this practice.
In non‐massive hemorrhage, FFP should be transfused as long as the INR or aPTT is >1.5× the upper limit of normal. PCC can be given for patients in whom excessive volume is a concern and in patients on vitamin K antagonist therapy with intracranial bleeding.
Platelets should be transfused to at least >50 × 109/L; higher values may be necessary for high risk bleeding (e.g. intracranial).
The management of a patient who presents with supratherapeutic clotting assays (e.g. a patient on warfarin who presents with an INR >3–3.5) but not actively bleeding may be difficult. The risk of bleeding with observation must be weighed against the risk of thrombosis if the patient is actively reversed. Those risks vary with the level of coagulopathy, patient comorbidities, and the condition being treated.
If the bleeding risk is low, coagulopathies related to excessive anticoagulation can usually be observed.
Procedures are frequently necessary in the critical care patient, but coagulopathy is considered a relative contraindication to most procedures. Most simple procedures (e.g. central venous catheter placement) however appear to be safe with modest derangements of the PT and aPTT. Surgical and higher risk procedures will require input from the individual operator, and there is significant discrepancy among clinicians as to acceptable laboratory abnormalities.
Minimum platelet counts for common ICU interventions are given in Table 56.3.
Table 56.3Minimum platelet counts for common interventions in the ICU.
Platelet count (× 10 9 /L)
Transjugular liver puncture, Central venous catheter placement
Bone marrow biopsy, Gastrointestinal endoscopy with biopsy, Lumbar puncture (emergent), Bronchoscopy
Bronchoscopy with biopsy, Lumbar puncture (elective), Spinal anesthesia
Coagulopathy and thrombocytopenia follow the clinical course of the disease and will resolve as the underlying condition improves There is no specific treatment. Therapy with intravenous heparin is controversial but not currently supported by clinical evidence
There is no specific treatment. Treatment is aimed at supportive care The Surviving Sepsis Campaign recommends that platelets be administered prophylactically when counts are <10 × 109/L in the absence of bleeding, and <20 × 109/L when at high risk of bleeding
Vitamin K deficiency
Coagulopathy will generally improve promptly with oral or intravenous administration of vitamin K
Managing a critically ill patient with hereditary factor deficiencies or inhibitors is complicated and may require hematology consultation
Care is supportive In the setting of fulminant hepatic failure, no FFP is given unless there is active bleeding or need for invasive procedures
Early plasma exchange therapy improves outcomes; immunosuppressives (glucocorticoids, rituximab) are adjunctive therapies Transfusion of platelets only when bleeding
There is an unclear benefit of plasma exchange Eculizumab for complement‐mediated HUS
Delivery will usually lead to reversal within 72 hours Consider platelet transfusion for delivery if platelet counts <20 × 109/L
Stop all heparin (including LMWH and heparin flushes) once HIT is suspected Initiate treatment dose with fundaparinux, argatroban, lepirudin, or danaparoid, as dictated by comorbid conditions Transition to vitamin K antagonists once the platelet count is >150 × 109/L with a 5 day overlap Duration of anticoagulation therapy is unclear; most experts recommend 2–3 months in the absence of thrombosis
Avoid future drug administration by documentation of allergy Intravenous immunoglobulin (in addition to platelet transfusions) can be considered if actively bleeding
Bone marrow suppression/failure
Treatment is supportive care and targeted to the underlying condition
Intravenous immunoglobulin and glucocorticoids if platelets <30 × 109/L and bleeding Glucocorticoids preferred over intravenous immunoglobulin in the non‐bleeding patient