skip to Main Content

Plasma Use in the United States


Triulzi D. et al. A multicenter study of plasma use in the United States. Transfusion 2015; 55: 1313-1319

Plasma transfusion, historically, has been steeped in much mystery and myth, with wide variability and inappropriate practice.1,2 The literature and guidelines for plasma transfusion have lagged somewhat behind the more robust studies for red cell transfusions. The NBCUS report noted 3.9 million plasma transfusions occurred in 2011 which clearly reflects the broad scope of use that persists in the United States.3

The recent publication by Triulzi et al. in Transfusion, cited above, represents an epidemiological study to more optimally define plasma use in ten U.S. hospitals over a one-year period. Over 9,200 patients were identified and over 72,000 plasma units transfused. Several characteristics were noted which were likely not unanticipated, however, there were other findings that might prove to be less expected.

IV bag flat icon with long shadow,eps10Plasma transfusions were typically not given in isolation. Approximately 77% of the patients received other blood components. Intuitively this is not a surprise, particularly in the arena of massive hemorrhage protocols secondary to trauma or non-trauma situations. Cardiovascular surgery had the highest use. The primary reason documented for plasma transfusion, across all facilities, was “treatment of coagulopathy” i.e., not for diagnosis of TTP or other illnesses that might necessitate plasma exchange. Again, not unanticipated. Over 22,000 invasive procedures were performed on the same day as transfusion with central venous catheter placement “leading the pack”, followed closely behind by UGI endoscopy, paracentesis, thoracentesis and bronchoscopy. All of these minimally invasive procedures have been identified in previous meta-analyses as not benefiting from prophylactic plasma transfusion, in terms of bleeding outcomes.4,5

The study found pre-transfusion INR to be available in 71% of patients and the majority of patients had a post-transfusion INR as well, to identify treatment effect. These are much higher than in my anecdotal experience, so perhaps a bit comforting? The median INR for which plasma transfusion was given was 1.9. Thirty-three percent of plasma was given for INR between 1.6 and 2.0 and 22.5% for INR < 1.6. No change in INR or a minimal decrease (median change = 0.2) was noted in 25% of patients. This data is in synch with prior studies that found similar minimal changes particularly when INR is ≤ 1.7.6

The use of standard FFP, plasma 24, or thawed plasma made no differences in pre- or post-INR values.

Unexpected and interesting findings, in my view, included: Over one-third of units were given on general Med/Surg wards and not within operating suites or emergency departments. Only 10.7% of plasma transfusions occurred within the O.R. and only 3.7% in the E.D. Single unit and double unit transfusions were reported at 15% and 46% respectively.

Thus, this publication provides further documentation of the broad and often unnecessary use of plasma, particularly outside of surgical suites and emergency departments for minimally invasive procedures. The use of prophylactic plasma transfusion for a perceived coagulopathy (INR <2.0) obviously persists in spite of more recent literature that speaks to the contrary. Inadequate dosing is also pervasive as >60% of transfusions were single or double-unit which is not considered an adequate adult dose.

The authors state that “…the opportunities for practice improvement are clear and substantial.” Based on their results I would strongly agree. As advocates for patient blood management, we must leverage these types of studies to help push forward the message of avoiding unnecessary, inadequate, and inappropriate transfusions.


  1. Tinmouth A et al. Transfusion 2013; 53: 2222-2229
  2. Stanworth S et al. Transfusion 2011; 51: 62-70
  4. Segal J and Dzik W. Transfusion 2005; 45: 1413-1425
  5. Yang L et al. Transfusion 2012; 52: 1673-1686
  6. Holland L and Brooks J. Amer J Clin Pathol 2006; 126: 133-139
Back to top