Common haemostasis issues in major bleeding and critical illness

Diagnostic limitations in haemostasis

The BSH guidelines suggest that after the initial transfusions of FFP (and red cells), further transfusion management should be guided by results of conventional laboratory-based tests (PT, APTT, fibrinogen).

There are many general limitations when documenting abnormalities of PT and APTT tests and the PT and APTT provide only a very limited assessment of haemostatic status. Firstly, they measure the time taken for a clot to form, which occurs when only 5% of total thrombin generation has occurred. Secondly, they reflect only the procoagulant factors and are unaffected by the levels of natural anticoagulants. Thirdly, they are insensitive to modest but potentially clinically relevant levels of other factors, although the most frequently used test, the PT, is sensitive to functionally unimportant reductions in factor VII. Finally, they are poor at identifying a prothrombotic state.¹²

The BSH guidelines recommend clotting tests being taken every 30–60 minutes during a major haemorrhage episode, although this can be difficult to manage in clinical practice, and the time taken to receive the results of standard lab tests (often over 45 min) also limits their utility in emergency situations.¹ The potential value of conventional laboratory-based testing in individual patients is not as a single isolated measure of coagulation but for serial monitoring of a dynamic process.^13,14

Near patient tests (eg viscoelastic haemostatic assays [VHAs] such as TEG/ROTEM) are often considered promising alternative diagnostic tests but their exact role remains to be defined in clinical trials.¹ VHAs may potentially provide a more global assessment of coagulation status with faster turnaround of results (approximately 10 min); however, the main evidence for the value of VHAs has been in cardiac surgery. In trauma, the iTACTIC randomised trial is ongoing; this multinational European study aims to assess their use in the management of traumatic haemorrhage.¹⁵

Disseminated intravascular coagulation

A number of disorders are frequently encountered in the critical care environment that can result in disseminated intravascular coagulation (DIC). DIC remains a poorly understood and recognised entity.¹⁹ International Society for Thrombosis and Haemostasis (ISTH) guidelines exist to help standardised diagnosis of DIC, but their value in clinical practice remains less clear. Precipitating factors include sepsis, malignancies, trauma, obstetric complications and autoimmune disorders, with sepsis being by the far the most common.¹⁶ DIC remains a challenge to identify clinically, as there are no clinical findings sufficiently sensitive or specific to accurately diagnose the condition, and its overall course can vary considerably with different manifestations of the disease. ISTH has proposed a scoring method for overt and non-overt DIC based on standard laboratory assays (see Table 1).¹⁷ While these still remain the main method of establishing a diagnosis of DIC, they are often not sensitive enough to detect its presence – 50% of cases of DIC have a normal PT for example. Low or rapidly decreasing platelet levels are the most common abnormal haematological feature, though these too may be normal in some cases.¹⁸

Initial treatment of DIC is essentially based on supportive care and treating the underlying cause (eg initiating antibiotic administration and source control in sepsis-related DIC). In early cases this may be sufficient to halt the process and restore normal haemostasis; however, in more severe cases haemostatic derangements may persist.¹⁹

Thrombocytopenia in critical care

A number of conditions commonly encountered in critical care can result in a drop in the circulating platelet count, including sepsis, DIC, thrombotic microangiopathy, and heparin-induced thrombocytopenia (HIT) – the incidence of thrombocytopenia in intensive care units (ICUs) has been shown to range from 13–50%.²⁰ Typical thresholds for replacement therapy are at <30–50 × 10⁹/L in a patient suspected of bleeding, or <10 whether bleeding suspected or not. In patients with platelet dysfunction, eg due to antiplatelet medication, uraemia, or severe liver failure, the picture may be more complex, as platelet levels may not correspond to activity/effectiveness. However, the recent PATCH trial reported evidence of harm associated with use of platelet transfusions in the setting of intracranial bleeding in patients on antiplatelet drugs who were treated with platelet infusions.²¹ The reasons for this outcome are unclear, but serve to emphasise the point that platelet transfusions are biological products with a range of potential immunomodulatory and proinflammatory consequences.

Prolonged coagulation times in non-bleeding patients

Prolonged coagulation times are common in the ICU, with an international normalised ratio (INR) ≥1.5 (INR – a ratio to standardise the PT test that allows for differences between laboratories so that the test results can be compared) detected in almost two-thirds of patients; however, providing clear evidence-based recommendations for prophylactic use of plasma in non-bleeding critically ill patients remains challenging.²² FFP is widely used in the treatment of coagulopathy in the bleeding patient, but it is likely that much of this is inappropriate, and it is estimated that over one-third of its use in current UK ICU practice is outside standard recommendation.¹¹ Recently, BSH published guidelines on the use of FFP and cryoprecipitate in absence of bleeding.²³ The Intensive Care Study of Coagulopathy–1 (ISOC1) study described significant prophylactic FFP use in 1,923 critical care admissions with coagulopathies defined by standard coagulation assays.²⁴ The follow-up ISOC-2 study evaluated the use of an alternative global assay termed ‘thrombin generation’ (TG), that might provide more information on the summation of procoagulant activities or better address the net balance between procoagulant and anticoagulant forces.²⁵ Plasma samples were collected and analysed from 306 patients admitted to intensive care with prolonged PT. Overall, despite prolonged PT, 47.8% patients had endogenous thrombin potential (ETP) within normal limits. Another recent randomised trial by De Pietri et al evaluated the use of TEG in 60 patients with cirrhosis and significant coagulopathy who were undergoing an invasive procedure. The authors reported that TEG-guided transfusion strategy led to a significantly lower use of blood products without an increase in bleeding complications.²⁶ These publications support the need for better tests to determine whether interventions can be withheld safely in groups of nonbleeding patients with prolonged PTR to provide a rationale that would avoid complications of unnecessary transfusions of plasma to non-bleeding patients.