What INR thresholds are being used for prophylactic and therapeutic use of FFP?

A physician reports that his hospital has seen a progressive increase in the prophylactic use of frozen plasma to "normalize" the PT/INR prior to invasive procedures (e.g. percutaneous liver and kidney biopsies, fine needle aspiration of intra-abdominal masses for cytopathology, central line placement, etc). The commonly accepted clinical "threshold" INR of his medical staff for use of FFP pre-procedure is 1.5. Based on their 'review' of the literature, the transfusion utilization committee would like to promulgate a threshold INR of 1.9 for most procedures, and 1.7 for lumbar puncture or other CNS procedures to minimize unnecessary use of frozen plasma. He wants to know what INR thresholds are being used elsewhere for both prophylactic and therapeutic use of FFP.

The following comments have been received.

ADDENDA May 7, 2006

1. Dr. Sunny Dzik of Massachusetts General Hospital (attribution used with permission) comments that the inquiring physician's local practice reflects a wider reality. Dr. Dzik noted that: "According to a recent publication, FFP usage in the United States is rising each year and rising at a rate faster than RBCs. In the USA, usage relative to RBCs is higher than in other nations with advanced economies. The reason for this increasing usage is not completely known, but one hospital-based study found that 30% of the requests for FFP outside the operating room was for attempted 'correction' of an elevated PT (INR) prior to an invasive procedure. If this finding applies more generally, then the issue raised by the inquiring physician, namely the proper indication for FFP prior to invasive procedure, is a very important topic. The importance is underscored by the finding that the rising use of FFP has occured in parallel with an increasing recognition of the morbidity and mortality associated with TRALI. Pre-procedure FFP is based on three assumptions-- none of which are validated with good clinical evidence. These assumptions are: 1) that an elevated PT (INR) identified patients at increased risk of bleeding; 2) that pre-procedure FFP will 'correct' the abnormality; and 3) that the risk-benefit or cost-benefit of FFP given before the procedure is higher than FFP given after the procedure to those individuals who actually bleed". Dr. Dzik points out that several reviews have underscored that the pre-procedure PT (INR) does not predict the risk of bleeding at the time of invasive bedside procedures and that a study soon to be published in TRANSFUSION shows that 1-4 units of FFP fails to correct mild to moderate prolongation of the INR in nearly all cases. "There are no clinical studies to show that FFP administered to all patients with an elevated test before a procedure is a preferred strategy to FFP given to those patients who demonstrate bleeding after a procedure regardless of their laboratory test results. At the Massachusetts General Hospital, the Transfusion Committee recommends FFP prophylaxis prior to an invasive bedside procedure for patients with an INR > 2.0. However, some physicians retain a strong bias that milder elevations of the INR are an indication for FFP and so not all physicians follow the above recommendation. Currently, the NIH Clinical Trials Network in Transfusion Medicine and Hemostasis is conducting a multicenter randomized prospective trial of FFP or no therapy for patients with an INR in the range of 1.3 to 1.9 who are undergoing an invasive hepatobiliary procedure. The results of this study are expected to provide higher quality evidence of the appropriate indication for FFP."

ADDENDA May 8, 2006

2. Dr. Mick Scanlan, Medical Director Transfusion Service at the Oregon Health & Sciences University (attribution used with permission) reports that when they last reviewed this topic, they found that good scientific studies were not especially helpful, and that they had to fall back on the 'time honored practice of medicine' standard. Like the inquiring physician, they found that there was something of a consensus that coagulopathic bleeding was more of an issue when the INR was at or above 1.8. Their medical staff did not want to wait until the patient was coagulopathic before they could give FFP, so they stayed with an INR of 1.5 as their recommended 'threshold' for FFP infusions. Despite building this buffer into their recommendation, 20-30% of their transfusions are reportedly given to patients whose INRs are under 1.5. Many of these transfusions are ordered by neurosurgeons who argue that the dangers of closed space bleeding warrants a more aggressive approach to INR normalization. After extended negotiations, their transfusion committee agreed to adopt a "threshold" of INR >1.3 for neurosurgical cases.

ADDENDA May 15, 2006

3. A physician in Barcelona, Spain comments that in his opinion there is limited published evidence on which to base a decision to transfuse FFP prior to an invasive procedure for a patient with prolonged PT (or INR). He adds that in his experience the PT is a crude measure of the state of plasma coagulation, as the test may be prolonged due to reduced production and/or accelerated consumption of coagulation proteins. He has observed that this test's ability to predict hemorrhage is not uniform across patients with different pathologies and underlying mechanisms for a prolonged PT. For instance, he cites an example of a PT ratio of 1.6 which may be "safe" in a patient with recent-onset hepatic failure (e.g. severe acute hepatitis), in whom both the procoagulants and the natural anticoagulants are decreased alike, but may prove to be "unsafe" in a patient who has low-grade DIC and increased fibrinolysis, as it often occurs in patients with advanced cirrhosis. He acknowledges that in patients on oral anticoagulant therapy the INR has proven to be a helpful predictor of hemorrhagic risk, in part because these patients are homogeneous with regard to the mechanism of the prolonged PT. As mentioned above, this homogeneity does not hold in many other clinical situations.
   
   He offers a context for deciding to transfuse FFP based on a balance between risks and benefits. For instance, TRALI is estimated to complicate about 1:5,000 – 1:10,000 plasma unit transfusions, but the risk of severe bleeding after performing a liver biopsy in cirrhotics is reportedly about 1:200 - 1:250, with a 0.1% - 1% mortality rate (Reverter JC. Abnormal hemostasis tests and bleeding in chronic liver disease: are they related? Yes. J Thromb Hemost 2006; 4: 717-20). Given the aforementioned risk estimates, one would need to enroll several thousand patients in a clinical trial to reach a quantitative conclusion about the balance of risks. This is further compounded by the fact that other variables influence the risk of hemorrhage when doing an invasive procedure, including the skill of the person performing the procedure to mere chance (e.g., cutting a hepatic vessel in a percutaneous liver biopsy). Thus, he thinks that the indication for prophylactic FFP prior to invasive procedures in patients with prolonged PT (INR) will elude the precepts of "evidence-based medicine" and remain in the sole domain of good clinical judgement for a long while. He concludes saying... "Unfortunately, in the current climate of malpractice litigation, clinical judgement is probably biased in favor of minimizing the risk of bleeding, a complication that can be clearly imputed to the person who performed the procedure, instead of preventing TRALI or other transfusion reactions, complications whose imputability is quite less precise".

ADDENDA May 23, 2006

4. Frank Boulton, Chair of the Transfusion Task Force, British Committee for Standards in Haematology (BCSH) refers readers to the Guidelines from the British Committee for Standards in Haematology regarding the use of FFP. The guidelines were published in 2004 (British Journal of Haematology Volume 126 Page 11 - July 2004) and are available from the BCSH website. According to Section 10.8, on the use of FFP in Liver Disease (the situation which may be most pertinent to the question here), it states in full;
   
   "A variety of abnormalities of coagulation is seen in patients with liver disease. The magnitude of haemostatic abnormalities correlates with the degree of parenchymal damage. Reduced clotting factor synthesis, reflected in a prolonged PT, may predispose to bleeding, which may be exacerbated by dysfibrinogenaemia, thrombocytopenia and increased fibrinolysis. However, bleeding seldom occurs without a precipitating factor such as surgery, liver biopsy, or variceal rupture.
   
   "Fresh-frozen plasma is still advocated by some for the prevention of bleeding in patients with liver disease and a prolonged PT, although complete normalization of the haemostatic defect does not always occur (Williamson et al, 1999). Routine use of FFP in these circumstances is therefore questionable. Platelet count and function, as well as vascular integrity, may be more important in these circumstances. Although PCCs have been shown to sufficiently correct abnormal clotting in liver disease (Green et al, 1975; Mannucci et al, 1976), their use, even of the more recently available and less thrombogenic preparations, is not recommended in view of the high risk of DIC. For similar reasons, it may also be advisable to avoid giving SDFFP in this situation in view of the relative depletion of protein S.
   
   "Many liver units will only undertake liver biopsy if the PT is no more than 4 s longer than the upper limit of the normal range. There is no evidence to substantiate this. Other tests, such as the APTT and thrombin time, do not normally help the decision-making. The response to FFP in liver disease is unpredictable. If FFP is given, repeat coagulation tests should be conducted as soon as the infusion is completed if it is to inform future decision-making. The merits of different infusion regimens, such as 5 ml/kg/h versus intermittent boluses, have not been addressed. These are areas that need more research. More work needs to be conducted to establish the role, if any, of FFP in patients with liver disease to correct the bleeding tendency prior to biopsy.
   "Recommendation "Available evidence suggests that patients with liver disease and a PT more than 4 s longer than control are unlikely to benefit from FFP (grade C recommendation, level IV evidence)."

ADDENDA July 2, 2006

5. Editor's note: Colleagues might find the publication: Holland LL, Brooks JP. Toward Rational Fresh Frozen Plasma Transfusion, the Effect of Plasma Transfusion on Coagulation Test Results. Am J Clin Pathol. 2006 Jul;126(1):1-7, to be germane to the present discussion.

Submit comments to the e-Network Forum at enetworkforum@cbbsweb.org

Ira A. Shulman, MD
CBBS e-Network Forum Senior Editor & Moderator

W. Tait Stevens, MD
CBBS e-Network Forum Editor & Moderator

Elizabeth M. St. Lezin, MD
CBBS e-Network Forum Associate Editor & Moderator