Monday, July 22, 2013

A Break in Massive Transfusion Evidence

The standard of care in trauma centers for massive transfusion in the setting of trauma has rapidly evolved to a fixed-ratio protocol, attempting to provide a physiologically balanced 1:1:1 mixture of PRBCs, FFP, and platelets.  The evidence upon which this is based stems from observational battlefield data, as well as retrospective trauma service registries.  However, as I've noted before (parroted, really, from folks smarter than me), these retrospective reviews are prone to survivorship bias – folks too sick to thaw FFP in time will die, and appear to reflect increased mortality association with not receiving FFP.

There is a large, multi-center prospective trial underway attempting to determine the optimal ratio of blood products – testing PRBC:FFP:platelets in 1:1:1 vs. 2:1:1 – because there are concerns especially with complications & costs associated with increasing FFP and platelet transfusions.  This article describes a single-center, prospective study of the feasibility of even implementing a 1:1:1 ratio, given the difficulty of having plasma products on hand – but has the interesting side effect of providing some rather interesting and unexpected comparative outcomes data.

These authors enrolled, over a two year period, 78 patients from a pool of 203 screened for eligibility, and randomized them in unblinded fashion to 1:1:1 fixed ratio transfusion or their "usual care" control.  "Usual care" for this institution consists of transfusion product balance guided by laboratory results (Hgb, INR, PTT, and fibrinogen).  They found, as the primary outcome of their study, that the 1:1:1 ratio was feasible – but resulted in over twice as many wasted units of FFP (22% vs. 10% of thawed units).

The secondary outcomes reported include coagulation monitoring targets and mortality data.  There was, for the most part, no statistically significant difference in any reported outcome.  The coagulation monitoring targets all had p-values ranging from 0.4 to 0.8 and, truly, are not different.  The mortality data, on the other hand, showed 29.7% mortality in the 1:1:1 group and 9.4% mortality in the usual care group – 20.3% difference (95% CI 2.5 to 38.2).

This is not practice-changing evidence.  It's a small sample size data coming from secondary outcomes in a feasibility study.  But, regardless, it is very interesting to see.

"Effect of a fixed-ratio (1:1:1) transfusion protocol versus laboratory-results–guided transfusion in patients with severe trauma: a randomized feasibility trial"


  1. Interesting. I wonder if the added fluid volume that comes with more aggressive FFP-resus is to blame?

  2. I don't see how to read anything other than clinical equipoise for further study into their secondary outcomes. Could be the magic of small samples, unbalanced samples, or any number of unmeasured or controlled confounders. I agree with "interesting".

  3. FFP will soon be outdated in 'real' trauma centers...authors, especially military authors, publishing studies using 1:1:1 are just padding resumes (hint)

    (comments are rare here, but I suspect there are many! docs who are reading your blog and considering your opinions deeply -- please stay motivated: my residents have twice referenced your blog when discussing EBM with me)


    1. Well, hopefully your residents are learning the value of close reading of the literature for themselves, rather than just echoing what the author/the guidelines/some yahoo on the Internet has to say!


Comments on posts > 10 days old will be moderated; blame spammers.