Practice recommendations for transfusion support are offered in the Discussion. See Figure Figure11.Figure 1Three-strategy approach to transfusion support in trauma patients at risk for massive hemorrhage. DAPT secretase Gamma-secretase FFP, fresh frozen plasma; RBC, red blood cell; TEG?/ROTEM?, thromboelastography/rotational thromboelastometry.Panel consensus: unanimous agreement.Question 1a. To what extent is the evidence on 1:1:1 formula-driven resuscitation affected by survivorship bias? Question 1a.i. What is the magnitude of the problem of survivorship bias? Question 1a.ii. What are the options to correct the bias?Each study of formula-driven resuscitation reviewed by the panel was found to be susceptible to survivorship bias [10]. Two reports that attempted to correct for survivorship bias by treating the blood component ratio as a time-dependent covariate [11,12] found no benefit on mortality.

The relationship between blood ratios and survival rates is not linear [13,14], although reported comparisons may have assumed linearity. Moreover, current studies of ratio-driven blood support are further complicated by other sources of bias, including those commonly found in retrospective studies, registry studies, and studies without random allocation. These methodologic concerns include secular trends, poor generalizability of single-site studies, selection bias, and imbalance of measured and unmeasured confounders [9].Failure to adequately address survivorship bias is a serious impediment to the interpretation of retrospective studies of blood ratios and has probably contributed substantially to the observed interpretation and uptake of results.

It is unlikely that further retrospective studies will overcome survivorship bias or resolve questions regarding the value of ratio-driven resuscitation.While treating blood ratios as a time-dependent covariate may improve analysis, this analytic approach assumes that the risk of mortality is constant over the period of observation. Exclusion of early deaths from analysis, while partially accounting for survivorship bias, excludes the key subpopulation with the highest, and perhaps modifiable, mortality. Randomized controlled trials will require a very large sample size in order to demonstrate any statistically significant effect.

In the absence of randomized controlled trials, better organized observational studies in which the exact timing of blood infusions is captured may allow analysis that partially corrects for survivorship bias. Studies based on a cluster randomized trial design or a before-after design would have the advantage that each participating site need only follow one protocol.Panel Carfilzomib consensus: unanimous agreement.Question 2. In addition to plasma, is there a role for other blood components and products in the resuscitation of massively bleeding patients?Question 2a.