Background : An advanced hemostatic dressing is needed to augment current methods for the control of life-threatening hemorrhage. A systematic approach to the study of dressings is described. We studied the effects of nine hemostatic dressings on blood loss using a model of severe venous hemorrhage and hepatic injury in swine.
Methods : Swine were treated using one of nine hemostatic dressings. Dressings used the following primary active ingredients: microfibrillar collagen, oxidized cellulose, thrombin, fibrinogen, propyl gallate, aluminum sulfate, and fully acetylated poly-N-acetyl glucosamine. Standardized liver injuries were induced, dressings were applied, and resuscitation was initiated. Blood loss, hemostasis, and 60-minute survival were quantified.
Results : The American Red Cross hemostatic dressing (fibrinogen and thrombin) reduced (p < 0.01) posttreatment blood loss (366 mL; 95% confidence interval, 175-762 mL) and increased (p < 0.05) the percentage of animals in which hemostasis was attained (73%), compared with gauze controls (2,973 mL; 95% confidence interval, 1,414-6,102 mL and 0%, respectively). No other dressing was effective. The number of vessels lacerated was positively related to pretreatment blood loss and negatively related to hemostasis.
Conclusion : The hemorrhage model allowed differentiation among topical hemostatic agents for severe hemorrhage. The American Red Cross hemostatic dressing was effective and warrants further development.
The development of improved methods for hemorrhage control is a major emphasis within the U.S. Army Combat Casualty Care Research Program. In combat, casualties may not be evacuated to advanced care for hours or days, depending on the tactical situation. Hemorrhage before evacuation accounts for 49% of overall battle deaths, whereas hemorrhage after evacuation accounts for just 1% of overall battle deaths from wounds. 1 There is a population of combat casualties that are potentially salvageable with improved methods, drugs, or devices for emergency treatment. 2,3 Among those that died from combat wounds in Korea, Vietnam, and Somalia, 7% to 14% died from extremity hemorrhage. 4,5 There is great potential for an advanced hemostatic dressing (HD) to reduce overall combat deaths from wounds. In this article, we present our approach to animal model selection and results of a study of nine potential HDs for treatment of severe venous hemorrhage.
An advanced HD development integrated product team was assembled to identify requirements, select animal models, and plan the research and development program. Although diffuse bleeding in coagulopathic patients can be a lethal problem, 6 the more important problem is bleeding from larger structures. 7,8 These deaths occur rapidly. 1,9-12 The requirement for an HD was the rapid control of life-threatening hemorrhage before evacuation. The primary HD patient was identified as a young, healthy service member with a major, life-threatening, actively bleeding, vascular wound that is accessible to the medic. This patient will have a normal coagulation system at the time of wounding. Primary HD use will be medic or buddy aid.
It was important that model selection criteria ensure relevance to hemorrhage severity and coagulation status of the target patient, and no artificial bias in favor of or against any specific product or product mode of action. A range of animal models of uncontrolled hemorrhage was reviewed and considered. The principal differences among models related to the method of hemorrhage induction, source of hemorrhage, and model severity.
Hemorrhage induction methods include variations of incision, penetration, transection, or crushing. Skin 13,14 or mucosa can be incised. 14 Vessels may be incised 15-17 or torn, 18 or a portion of the vessel wall may be removed with a punch. 19 Organs may be incised 20-22 or portions of organs may be excised 20,23-27 or removed by a punch. 28 Another approach is to incise and strip capsules from sections of organs. 20,21 Methods involving penetration or puncture of a vessel may penetrate one 29,30 or both sides. 31 Organs can be sharply penetrated 32-36 or stabbed. 37 Transection may involve transection of an individual vessel, 38 a portion of the tail, 39 or the cuticle. 13 Crush injuries often involve solid organs. 28 Models of blunt abdominal trauma 40 and extremity gunshot 41 have been reported. These models involve complex injuries that include combinations of the types of defects discussed above. Quantitation of hemorrhage was also a factor. Most models directly measure blood loss, 15,18,23,25-27,32-35,40-42 whereas others use indirect methods. 13,30 Bleeding time may be determined by direct observation and continuous timing 14,25 or expressed as the required number of dressing compressions or applications. 20-22,29,42
Other important considerations pertain to the size of the defect and the severity of the hemorrhage. The geometric area of the vascular defect and the pressure gradient across the vessel wall are the main determinants of blood loss. 43 Vessel diameter is vessel and species specific. The tension within the vessel wall is related to the pressure pushing against the vessel wall and the radius of the vessel, according to the law of Laplace: Tension = Pressure × Radius. 44 For an HD to seal a damaged vessel, the clot or other seal formed must hold under the tension characteristic of the given vessel wall. The physiologic severity of the hemorrhagic insult can be determined on the basis of lethality in the absence of treatment. Severity can also be assessed from the standpoint of difficulty to control bleeding. Flowing hemorrhage from a large defect in a large vessel may be considered severe, whereas capillary bleeding that is temporarily controlled using vascular clamping before application of the experimental hemostat may be considered mild. The status of the hemostatic mechanism differs among animal models. Hemorrhage models often use anticoagulants or other methods to achieve adequate hemorrhage or investigate coagulopathic states. 13,14,22,23,25,28,29,33,35
Models of arterial, large venous, and diffuse bleeding were required. To address venous bleeding, we selected a model of severe venous hemorrhage and liver injury in swine. 32,34,36 Although this model includes large incisions through liver parenchyma, the major hemorrhage is vascular. Intra-abdominal hemorrhage is not an exact match with the accessible hemorrhage of the primary patient. This limitation was seriously considered in model selection. The model was selected on the basis of several desirable characteristics, including large-diameter veins, ability to apply HD in the presence of free-flowing hemorrhage, ease of blood loss quantitation, ability to determine time to hemostasis, ease of instrumentation, potential for lethality, no requirement for anticoagulation, and reproducibility.
As a part of the program to identify an advanced HD for military use, a request for proposals for HDs was published in February 1999. Nine HDs were submitted, including both commercially produced and experimental HDs. We examined the effects of these HDs on blood loss, hemostasis, and short-term survival. This study was completed in early 2000.