For personal use. Only reproduce with permission from The Lancet Publishing Group. TRAUMA III 1988 THE LANCET ��� Vol 363 ��� June 12, 2004 ��� www.thelancet.com Advances in shock resuscitation have occurred during military conflict because of concentrated experience with patients. With other clashes around the world that will probably involve both military personnel and civilians, what further advances should we expect in the next several years? Shock resuscitation is an obligatory intervention that with refinement has changed the epidemiology of deaths from trauma (table 1). During World War I, as a result of the wound toxin hypothesis, no preoperative resuscitation was administered and many soldiers died. In World War II and the Korean conflict, due to the misconception over haemoconcentration, colloid was administered and eventually banked blood resuscitation became standard care. Early survival improved, but many casualties later died of acute renal failure. During the Vietnam conflict, with the recognition that the extracellular fluid space had to be repleted, large volume isotonic crystalloid solutions replaced colloids for initial shock resuscitation. Mortality rates and the incidence of acute renal failure decreased but adult respiratory distress syndrome emerged as a major source of morbidity and mortality. Through the 1970s and early 1980s, intensive care units (ICUs) were developed, where advanced technology and improved care allowed patients with single organ failure to survive for long periods. Patients no longer died of acute renal failure or adult respiratory distress syndrome, but developed multiple organ failure, which, at that time, was usually fatal. From the mid-1980s to the present, regional trauma systems have been implemented, in which the most severely injured patients are rapidly triaged to high volume, level 1 trauma centres where rapid intervention and Lancet 2004 363: 1988���96 Department of Surgery, University of Texas���Houston Medical School, Houston, TX, USA (Prof F A Moore MD, B A McKinley PhD) and Department of Surgery, University of Colorado���Denver Health Medical Center, Denver, CO USA (Prof E E Moore MD) Correspondence to: Dr Frederick A Moore, Department of Surgery, University of Texas���Houston Medical School, 6431 Fannin Street, Suite 4��264, Houston, TX 77030, USA (e-mail: Frederick.A.Moore@uth.tmc.edu) prompt definitive care are given.1 Consequently, the epidemiology of death from trauma has changed.2 Rates of early hospital death from blood-loss have been reduced with the introduction of damage control surgery. For example, damage control laparotomy is well established in the initial surgical care of patients with severe torso injury, and consistent results have been achieved in the USA and Europe.3 These patients, however, remain at high risk of multiple organ failure.4 Although mortality rate from multiple organ failure is decreasing, it remains the major cause of prolonged ICU stays. As we learn more about the relations between the initial traumatic shock insult, the obligatory resuscitation, and the dysfunctional inflamma- tory response that causes multiple organ failure, some The next generation in shock resuscitation Frederick A Moore, Bruce A McKinley, Ernest E Moore Search strategy For this review we drew on a continuously updated collection of references on shock resuscitation, maintained at the Trauma Research Center, University of Texas-Houston Medical School, supplemented by a PubMed search using the keywords shock, trauma, and resuscitation. Resuscitation of the severely injured patient who presents in shock has improved greatly, following focused wartime experience and insight from laboratory and clinical studies. Further benefit is probable from technologies that are being brought into clinical use, especially hypertonic saline dextran, haemoglobin-based oxygen carriers, less invasive early monitors, and medical informatics. These technologies could improve the potential of prehospital and early hospital care to pre-empt or more rapidly reverse hypoxaemia, hypovolaemia, and onset of shock. Damage control surgery and definitive interventional radiology will probably combine with more real-time detection and intervention for hypothermia, coagulopathy, and acidosis, to avoid extreme pathophysiology and the ���bloody vicious cycle���. Although now widely practised as standard of care in the USA and Europe, shock resuscitation strategies involving haemoglobin replacement and fluid volume loading to regain tissue perfusion and oxygenation vary between trauma centres. One of the difficulties is the scarcity of published evidence for or against seemingly basic intervention strategies, such as early or large- volume fluid loading. Standardised protocols for resuscitation, representing the best and most current knowledge of the clinical process, could be devised and widely implemented as interactive computerised applications among trauma centres in the USA and Europe. Prevention of injury is preferable and feasible, but early care of the severely injured patient and modulation of exaggerated systemic inflammatory response due to transfusion and other complications of traditional strategies will probably provide the next generation of improvements in shock resuscitation. Era Focus Resuscitation Outcome World War I Wound toxins None Early death World War II, Intravascular Colloids, blood ��� Early survival Korean war repletion ARF ��� death Vietnam war Intravascular Crystalloids, ��� Early survival and extracellular banked blood ��� ARF fluid repletion ARDS ��� death 1970s���80s ICUs, organ PA catheters, ��� ARF failure, metabolic endpoints of ARDS ? MOF support resuscitation ��� MOF deaths Mid-1980s Trauma centres, Rapid triage, ��� Early survival to present trauma systems damage control, ��� ARDS/MOF shock and ��� ARDS/MOF deaths trauma ICUs ARF=acute renal failure. ARDS=adult respiratory distress syndrome. MOF=multiple organ failure. Table 1: Improvements in resuscitation and the changing epidemiology of trauma deaths over time Trauma III

For personal use. Only reproduce with permission from The Lancet Publishing Group. Transfusion trigger Banked red blood cells have been widely available since World War II, but the indications for administration continue to be debated. Animals with haemorrhagic shock show improved survival with haemoglobin concentration maintained in the range 120���130 g/L.24,25 Animal models of isovolaemic haemodilution suggest that the optimum haemoglobin concentration for maintaining systemic oxygen delivery (DO2) is 100 g/L, but in healthy human volunteers isovolaemic haemodilution is tolerated at concentrations as low as 50 g/L.26 Until recently, clinical studies in critically ill patients concluded that 100 g/L haemoglobin was optimum for shock resuscitation,27 but more recent consensus panels have judged that a lower concentration is adequate. The American Society of Anesthesiology recommends maintaining haemoglobin at greater than 60 g/L, whereas the National Institutes of Health recommends a concentration of greater than 70 g/L.28,29 A prospective randomised controlled trial has shown that ICU patients randomised to restrictive blood transfusions (transfuse if haemoglobin concentration 70 g/L and maintain between 70 and 90 g/L) did as well and possibly better than patients who were liberally transfused (transfuse if 100 g/L and maintain between 100 and 120 g/L).30 This study was done in a select group of euvolaemic patients. Malone and colleagues report blood transfusion as a risk factor, independent of shock severity, for poor outcome in patients with trauma.31 Short-term improvement in survival is believed to be due to reduction of the adverse immunological consequences of transfusion of banked red blood cells. The concept of transfusion-related immunosuppression dates from the early 1970s with the observation that greater pretransplant transfusion was associated with improved renal allograft survival.32 Subsequently, perioperative transfusion was associated with cell-mediated immunosuppression, tumour recurrence, and postoperative infections.33���36 The capacity of banked red blood cells to provoke neutrophil cytotoxicity is a key mechanism in multiple organ failure.5,37 In epidemiological studies, early transfusion has been shown to be a strong independent risk factor for multiple organ failure.4,8 Focused clinical studies have shown that severely injured patients at high risk of multiple organ failure have circulating neutrophils that are primed for cytotoxicity within the first 6 h after injury, increased expression of adhesion receptors, activation of p38 mitogen-activated protein kinase, and delayed apoptosis.6,38,39 The precise mechanisms linking red blood cell transfusion and neutrophil priming remain to be established, but it is generally believed that passenger leucocytes accompanying red blood cells in storage are important in the generation of proinflammatory agents and progressively increase from 14 to 42 days of storage.40���42 Some investigators have implicated cytokines (tumour necrosis factor , interleukin 1, interleukin 6, interleukin 8), but our focus has been on proinflammatory lipids presumably generated from the degradation of the red blood cell membrane.43���45 Although pre-storage leucoreduction of red blood cells decreases the generation of cytokines, this process does not eliminate neutrophil priming.46 Another emerging concern regarding stored cells is that substantial shape changes and impaired deformability occur by the second week of storage and progress during further preservation.47 The effects of decreased deformability on microcirculation have been documented experimentally, and include impaired tissue access and entrapment resulting in microvascular obstruction.48 These findings might account for the observation that blood transfusion failed to improve TRAUMA III THE LANCET ��� Vol 363 ��� June 12, 2004 ��� www.thelancet.com 1989 disconcerting observations indicate that our fundamental approach to resuscitation needs to be reassessed.5���10 The purpose of this report is to briefly review traditional resuscitation, discuss its inherent flaws in view of our current understanding of the pathogenesis of multiple organ failure, and identify alternative methods of fluid resuscitation, haemorrhage control, and monitoring techniques that could be used in the near future to improve outcome. Traditional resuscitation Crystalloid versus colloid debate Since the early 1940s, when restoration of circulatory blood volume was embraced as pivotal in shock resuscitation, controversy has existed as to which fluid to use for this purpose.11 A review article from that era concluded that saline and glucose solutions were unsuitable because they were quickly lost from the intravascular space. Plasma and serum were considered the best substitutes for whole blood���and in some cases, better than blood itself.12 This was the resuscitation strategy used in World War II. Isotonic crystalloids became the standard of care in the late 1960s based on the laboratory work of Moyer, Shires, Moss, and others,13���15 which showed that: (1) large-volume resuscitation with isotonic crystalloids provided the best survival (2) extracellular fluid was redistributed during shock into both the intravascular and intracellular spaces (3) optimal resuscitation corrected this extracellular fluid deficit, required infusion of isotonic crystalloid fluid, and shed blood in a ratio of three volumes of crystalloid to one of blood and (4) in severe shock, this ratio increased up to eight to one. Isotonic crystalloid fluids were used in the Vietnam conflict. For various reasons, mortality and acute renal failure decreased, but a new entity called shock lung emerged. This complication was soon described in civilian publications as adult respiratory distress syndrome, and was shown to be the major source of morbidity and mortality in the ICU. The pathogenesis was not known, but controversy existed over whether adult respiratory distress syndrome was an iatrogenic complication of resuscitation with crystalloids.16,17 Beginning in the mid- 1970s, prospective randomised controlled trials were done to compare crystalloids with colloids in resuscitation, with pulmonary dysfunction as the primary endpoint.18���20 Results conflicted because of shortcomings in study design, which included small numbers of patients, heterogeneous populations, and different resuscitation endpoints and sodium loads, and difficulty in defining pulmonary dysfunction. These data have been assessed by meta-analysis,19,20 and no difference in the incidence of pulmonary dysfunction has been found. However, when mortality is used as the endpoint and the data are subgrouped, the use of crystalloids in trauma patients is associated with improved survival. Although these data are not definitive, this finding does lend support to the current common use of crystalloids in US trauma centres. The type of crystalloid fluid used in resuscitation engenders less controversy. Normal saline and lactated Ringer���s are the two balanced salt solutions that are com- monly used. Although important theoretical differences favour lactated Ringer���s, laboratory investigators have had difficulty demonstrating substantial differences in outcome.21,22 Results of one study showed that the two solutions were equivalent in moderate haemorrhagic shock, but that with massive haemorrhage normal saline was associated with greater physiological derangement and greater mortality rate, compared with Ringer���s.23