Frozen plasma (FP) is commonly transfused to patients with abnormal coagulation results prior to medical procedures associated with a risk of bleeding (e.g. involving a needle or incision). The most frequent group that requires FP transfusions are patients admitted to an intensive care unit. These patients frequently have abnormal coagulation tests and also frequently require invasive medical procedures that can be associated with bleeding. While FP transfusions can improve abnormal coagulation test results, there is no evidence to suggest that prophylactic FP transfusions will reduce bleeding associated with medical procedures in patients with mild coagulation abnormalities. Additionally, it is known that important adverse complications occur with FP transfusions.

The investigators will undertake a pilot randomized controlled trial in non-bleeding ICU patients who require an invasive medical procedure (central venous catheterization, chest tube, thoracocentesis, paracentesis, biopsy, fluid drainage) in 3 Canadian hospitals. Patients with an abnormal coagulation test (INR between 1.5 to 2.5) will receive a FP transfusion or no treatment prior to an invasive procedure.

The primary outcome for this pilot study will be feasibility (the number of patients enrolled per month). Other important outcomes will include those that will be evaluated in the definitive trial including bleeding post procedure, red cell transfusions, adverse transfusion reactions, mortality and hospital length of stay.

The objective of the study is to enrol 80 patients over 2 years in 3 Canadian hospitals.

There are 3 accepted indications for transfusing FP:

1. bleeding in patients with prolonged coagulation tests that cannot be corrected with other products or medications,
2. prior to surgery or invasive procedures in patients who have prolonged coagulation test that cannot be corrected with other products or medications and
3. plasmapheresis for Thrombotic Thrombocytopenia Purpura.

While the indications for FP seem relatively straightforward, there are still large variations in use and important inappropriate use. This is due to the lack of properly conducted studies evaluating the true risk of bleeding in patients with an abnormal coagulation tests and, as a consequence, a lack of knowledge about the true benefits or harms of transfusing FP to reduce bleeding.

The current paradigm for the use of FP to treat or prevent bleeding is based on the following assumptions:

1. elevated coagulation tests represent a decrease in coagulation factors that will contribute to bleeding,
2. FP transfusions will increase the levels of coagulation factors and correct the coagulation test abnormalities,
3. the correction of the coagulation test abnormality will decrease bleeding. However, there is a lack of evidence to support each of the assumptions in this paradigm.

First, while decreases in coagulation factor levels may increase the risk of bleeding, this only occurs when coagulation factor levels decrease below a minimum hemostatic threshold. The minimum hemostatic for most coagulation factors is 30%. In vitro data demonstrates that, in samples with a single factor coagulation deficiency, the commonly used coagulation tests, the prothrombin time (usually reported as an international normalized ratio (INR) and the activated partial thromboplastin time (aPTT) will become abnormal when the levels are 25-40% or below. However, in the setting of multiple factor deficiencies, which is the situation usually encountered in clinical practice, mild to moderate increases in INR are seen with coagulation factor levels that are considerably higher than the minimum hemostatic threshold of 30%. Work by the group and others confirm that in patients with critical illness and liver disease factor levels are consistently above the hemostatic threshold even when the INR is elevated. This lack of correlation between abnormal coagulation test results and increased bleeding risk is supported by a systematic review of observational studies that showed no increased bleeding risk in patients with abnormal coagulation tests as compared to patients with normal coagulation tests.

Second, the expected improvement in coagulation factor tests following a FP transfusion is often not seen especially in mild to moderate coagulation test abnormalities. Given the concentration of 1 unit/ml for individual coagulation factor levels in FP, transfusing four units of FP (approximately 800-1000 mls) to a 70 kg patient with a 5-litre blood volume, would be expected to increase levels by a maximum of 16%. However, the effect on the INR will also depend on the starting level of the coagulation factors. If the levels are very low (less than 10%), this will result in a significant improvement in the INR, but if the initial levels are only mildly decreased, the same increase in coagulation factor levels will make little change in the INR. Furthermore, the INR FP units can be as high as 1.3. The lack of efficacy of FP to correct the INR or aPTT in patients with mild increases in the INR has been demonstrated in a number of studies.

Finally, there is little data from clinical studies that demonstrates that attempting to correct abnormal coagulation tests with FP reduces bleeding. Based on clinical experience, in patients with a high INR and active bleeding, FP may be beneficial in controlling bleeding. However, the prophylactic use of FP has never been shown to reduce bleeding. The only prospective clinical trials showed no reduction in bleeding following FP transfusion in neonates, patients with pancreatitis or after cardiac surgery. A retrospective study of FP use in critical care patients showed no differences in bleeding after the administration of FP.

Overall there have been few randomized controlled studies that have examined the efficacy or effectiveness of FP transfusions. Only 3 randomized controlled studies evaluating the effectiveness of FP (as compared to no treatment) prior to invasive procedures have been carried out. All 3 studies failed to recruit sufficient patients due to important issues in the design of these trials. The main failure in the first two studies was the narrow eligibility criteria. Two studies only included patients undergoing a single type of procedure and they included only patients with very mild coagulation abnormalities (i.e. an INR equivalent of less than 2.0). Thus, many patients with modest coagulation abnormalities were excluded thereby leaving only a small pool of potentially eligible patients. The third study failed to recruit patients for a number of reasons including restrictive inclusion criteria and inability to obtain consent for emergency procedures especially after hours. While this study recruited patients undergoing a variety of invasive procedures and used a wider INR threshold, they did reduce the number of potentially eligible patients by not including patients with thrombocytopenia or patients on anti-platelet agents (both common in ICU patients). Less than 30% of eligible patients were enrolled as informed consent from the patients or a surrogate was required prior to enrolling the patient. As ICU patients frequently have prolonged INR tests and require urgent procedures at the time of admission, a deferred consent process and the ability to enroll patients after hours is absolutely essential to ensure that eligible ICU patient who are usually not able to give consent are not systematically excluded from the trial due to the need for informed consent.

In summary, FP is commonly transfused prior to invasive procedures in patients with an elevated INR. Remarkably, there is no evidence that transfusing FP reduces bleeding complications when transfused prophylactically. Observational data suggests that an elevated INR does not increase risk of bleeding from invasive procedures. However, there is no randomized clinical trial data to support not transfusing FP. This lack of evidence combined with fear of causing bleeding during an invasive procedure and an under-appreciation of the potential harms associated with blood transfusions contributes to continued practice of transfusing FP prior to invasive procedures. Given that mild to moderately increased coagulation tests have not been associated with increased bleeding following invasive procedures and the FP transfusion are known to only partially correct abnormal coagulation tests, the prophylactic use of FP may be exposing patients to potentially life-threatening adverse reactions without providing any benefit. Previous randomized controlled trials of FP transfusions prior to invasive procedures have failed to enroll sufficient patients, but they had important limitations in their design that affected their ability to recruit patients.

A pilot study is required to determine if, by using specific strategies to increase enrollment, a large pragmatic clinical trial can successfully be implemented to determine the benefit and safety of transfusing FP prior to invasive procedures. The planned study will be pragmatic and open to the most common procedures in ICU patients who are known to represent the largest group of patients receiving FP.

This study is a multi-center, open-label, randomized controlled pilot trial with blinded outcome assessment. In this trial evaluating feasibility, the investigators will randomly allocate ICU patients who have an elevated INR and require an invasive procedure to receive either a FP transfusion or no transfusion. Patients will be followed for bleeding complications and adverse transfusion reactions for 48 hours following the invasive procedure, and for mortality for the duration of their hospital stay. All bleeding complications will be recorded using a validated bleeding score and adjudicated by blinded assessors.

Inclusion Criteria:

• Age 18 or greater
• Admission or planned admission (e.g. patients in emergency department who are being seen by the ICU team) to an intensive care unit.
• An elevated INR between 1.5 and 2.5.
• Requiring an invasive procedure in the next 24 hours including central venous line, arterial line, paracentesis, thoracocentesis, bronchoscopy, endoscopy, and ultrasound guided biopsy (mass or organ) or fluid drainage.

Exclusion Criteria:

• Active bleeding, defined as visible or suspected blood loss in last 48 hours, resulting in a fall in hemoglobin greater than or equal to 20 g/L, requiring a red cell transfusion or an intervention to control bleeding.
• Full dose therapeutic anticoagulation with warfarin, heparin, low molecular weight heparin, or other novel oral anticoagulants.
• Congenital bleeding disorders including hemophilia, von Willebrand Disease or platelet function disorders.
• Acquired coagulation factor deficiencies.
• Frozen plasma transfusion during this ICU admission.
• Use of other hemostatic blood products (recombinant factor VIIa, prothrombin complex concentrate, cryoprecipitate, fibrinogen concentrate) during the ICU admission
• Previously enrolment in the study.

Patients will not be excluded for thrombocytopenia or anti-platelet drugs. As a pilot trial for a pragmatic large randomized controlled trial, both thrombocytopenic patients and patients on anti- platelet agents will be enrolled as they are routinely encountered in clinical practice. Specific therapy (i.e. platelet transfusions) will not be mandated, but left to local routine practice. Information regarding platelet counts, anti-platelet medications, platelet transfusions and other hemostatic therapies will be collected.