Why do physicians request fresh frozen plasma?

Each year approximately 3 million units of fresh frozen plasma (FFP) are transfused in the US. FFP transfusions risk several complications including volume overload, allergic reactions, acute lung injury, and transmission of infectious pathogens. Most consensus guidelines list only a few clear indications for transfusion of FFP including treatment of bleeding due to multiple factor deficiencies, emergency reversal of vitamin K antagonists, and treatment of thrombotic thrombocytopenic purpura (TTP). 1 These limited indications would not appear to account for the widespread use of FFP and published audits of FFP transfusion have frequently suggested inappropriate usage. 2 Nevertheless, most published audits report indications as interpreted by chart review of transfusions given, rather than indications as stated by the prescribing physician before the FFP is administered. Massachusetts General Hospital (MGH) uses a computerized system for physician order entry (POE) for all transfusions given outside of the operating rooms. This software requires that the individual ordering a transfusion select from a “pick list” to declare the reason for the transfusion. The software is used for all requests on both adult and pediatric services except those originating in the operating rooms and those originating within the transfusion service for the treatment of TTP by plasma exchange. Requesting physicians can select “other” for cases not meeting one of the pick list entries. We report in Table 1 the reasons stated by prescribing physicians for all FFP transfusions requested at the hospital using POE during the period from February through April 2003. Of note, the most common reason for prescribing FFP was to “prepare” a patient with an elevated international normalized ratio (INR) for an invasive procedure. This single “indication” represents one-third of all plasma requested by the physicians with POE. MGH performs blood utilization review of all transfusions given outside the operating room. Electronic mail feedback is given to the ordering physician within 24 hour following the transfusion. Because review before blood release is not done, POE requests are generally filled as requested. Although our data are intended to reflect the reasons why physicians request FFP (rather than report on FFP transfused), the POE data presented in Table 1 represent a substantial proportion of FFP transactions. For example, during the period from February to April 2003 the total number of transfused products at MGH (both operating room and non-operating room) was 3636 units of FFP, 8988 units of red blood cells, and 3219 units other components. There is little published evidence to support the use of FFP transfusion as an appropriate prophylaxis before invasive procedures. In fact, published retrospective studies examining the ability of the prothrombin time or INR to predict bleeding at the time of invasive procedures have failed to demonstrate a correlation between the INR and hemorrhagic outcomes. 3-5 Likewise we could find no published evidence that preprocedure FFP transfusion reduces bleeding at the time of a wide range of invasive procedures including central venous catheter placement, thoracentesis, paracentesis, liver biopsy, or angiography. Nevertheless, the use of FFP before procedures among patients with an elevated INR appears to be quite common. 6