Immediate pre-injury vital signs were recorded, the lower half of the midline incision was closed with towel clips, and then one of three primary injury mechanisms was applied as described below. As discussed above, the reasonable outcome that was being sought in this study was a,50% mortality within the first hour after injury. This mortality rate goal was set empirically, based upon the observations of severe noncompressible truncal hemorrhage in a military setting. Furthermore, we believed that an injury mechanism with either a higher or lower one-hour mortality probably would not allow us to discriminate differences in efficacy in subsequent comparisons of experimental treatments for noncompressible truncal hemorrhage. That is, if the injured subject bled too “slow” or too “fast,” then future comparison of treatment regimens might be meaningless. We believe that domestic swine are a good choice to model severe hemorrhage in humans because, by the age of 3 months, domestic swine have reasonably large size and blood volume which makes studies of severe hemorrhage practical. We have access to an inbred population of domestic swine that has been closed for.30 years, which theoretically should reduce inter-subject variability. Furthermore, domestic swine have been used for decades to model human physiology and pathophysiology, and generally have produced acceptable data for these types of studies. Although small animal models have produced some usable data in the field of hemostasis, information obtained from small animal models of hemorrhage ultimately may have limited clinical relevance because of their small organ size, blood vessel diameter, and blood volume with respect to humans. A rabbit model of noncompressible hemorrhage was described in which partial hepatectomy was combined with a systemic administration of a BEZ235 Factor X inhibitor. A fibrin sealant foam therapy was demonstrated to have hemostatic efficacy in this model compared to both placebo treatment or notreatment controls. A model of noncompressible hemorrhage involving portovenous injury in domestic swine also has been described. In this model, the investigators looped wires around the medial liver lobes through a midline laparotomy, and then transected these lobes by pulling the wires out of the abdomen after the incision had been closed. The subjects then were resuscitated with crystalloid with no limit on resuscitation volume. The one-hour mortality of this injury model was 90%, with a median survival time of 43 min. This group of investigators subsequently described a therapy for noncompressible hemorrhage consisting of an expansile polyurethane foam, which had demonstrable efficacy in their porcine model. The same investigator group described another porcine model of noncompressible hemorrhage which utilized placement of a wire around the external iliac artery via a laparotomy, and subsequent transection of the artery by wire distraction after the abdominal incision had been closed. One-hour mortality in this model was 78%, with a median survival time of 32 min. Use of a “closed” abdominal technique makes empiric sense in the design of a noncompressible injury model.